EDITOR IN CHIEF
William C. Roberts,
ASSOCIATE EDITORS
Vincent E. Friedewald Paul A. Grayburn
MD
Baylor Heart & Vascular Institute Baylor University Medical Center Wadley Tower No. 457 3600 Gaston Avenue Dallas, Texas 75246 (214)826-8252 Fax: (214)826-2855
ASSISTANT EDITORS
Robert C. Kowal Jeffrey M. Schussler
EDITORIAL BOARD CARDIOVASCULAR MEDICINE In Adults
Antonio Abbate J. Dawn Abbott George S. Abela Jamil Aboulhosn Joseph S. Alpert Martin A. Alpert Ezra A. Amsterdam Jeffrey L. Anderson Evan Appelbaum Richard W. Asinger Pablo Avanzas Gary John Balady Eric Bates Jeroen J. Bax George A. Beller William E. Boden Monty M. Bodenheimer Robert O. Bonow Jeffrey S. Borer Martial G. Bourassa Eugene Braunwald Jeffrey A. Brinker David L. Brown Michael E. Cain Richard O. Cannon III Bernard R. Chaitman Kanu Chatterjee John S. Child Robert J. Cody Lawrence S. Cohen Marc Cohen C. Richard Conti Michael H. Crawford James P. Daubert Gregory J. Dehmer James A. de Lemos Anthony N. DeMaria Pablo Denes George A. Diamond John P. DiMarco Allen Dollar Michael J. Domanski Gerald Dorros Uri Elkayam Kenneth A. Ellenbogen Myrvin H. Ellestad Stephen G. Ellis Toby R. Engel Andrew E. Epstein N. A. Mark Estes, III Michael Ezekowitz
Rodney H. Falk John A. Farmer David P. Faxon Ted Feldman Jack Ferlinz Jerome L. Fleg Gerald F. Fletcher Joseph A. Franciosa Gary S. Francis W. Bruce Fye William H. Gaasch Julius M. Gardin Bernard J. Gersh S. David Gertz Mihai Gheorghiade D. Luke Glancy Stephen P. Glasser Michael R. Gold Samuel Z. Goldhaber Robert E. Goldstein Sidney Goldstein Steven A. Goldstein J. Anthony Gomes Antonio M. Gotto, Jr. K. Lance Gould Donald C. Harrison Philip D. Henry L. David Hillis David R. Holmes, Jr. Mun K. Hong William G. Hundley Ami S. Iskandrian Allan S. Jaffe Hani Jneid Greg L. Kaluza Joel S. Karliner John A. Kastor Sanjiv Kaul Ellen C. Keeley Kenneth M. Kent Richard E. Kerber Dean J. Kereiakes Morton J. Kern Spencer B. King III Robert E. Kleiger George J. Klein Lloyd W. Klein Paul Kligfield Robert A. Kloner John B. Kostis Charles Landau Richard L. Lange Carl J. Lavie Carl V. Leier B. T. Liang
Joseph Lindsay, Jr. Gregory Y.H. Lip Francisco Lopez-Jimenez Joseph Loscalzo G.B. John Mancini Francis E. Marchlinski Frank I. Marcus Barry J. Maron Martin S. Maron Randolph P. Martin Attilo Maseri Charles Maynard Michael D. McGoon Darren K. McGuire Raymond G. McKay Jawahar L. Mehta Bernard Meier Franz H. Messerli Eric L. Michelson Richard V. Milani Alan B. Miller Wayne L. Miller Gary S. Mintz Fred Morady Lori Mosca Arthur J. Moss James E. Muller Gerald B. Naccarelli Navin C. Nanda Robert A. O’Rourke Erik Magnus Ohman Richard L. Page Sebastian T. Palmeri Seung-Yung Park Eugene R. Passamani Alan S. Pearlman Carl J. Pepine Joseph K. Perloff Bertram Pitt Philip J. Podrid Don Poldermans Arshed A. Quyyumi Charles E. Rackley C. Venkata Ram Nathaniel Reichek Robert Roberts Jennifer G. Robinson Lynda E. Rosenfeld Melvin M. Scheinman David J. Schneider John S. Schroeder Patrick Washington Serruys Pravin M. Shah Prediman K. Shah Jamshid Shirani
Robert J. Siegel Marc A. Silver Ross J. Simpson, Jr. Steven N. Singh Burton E. Sobel John C. Somberg David H. Spodick Lynne W. Stevenson Gregory W. Stone John R. Stratton Jonathan M. Tobis Eric J. Topol Byron F. Vandenberg Hector O. Ventura George W. Vetrovec Robert A. Vogel Ron Waksman David D. Waters Nanette K. Wenger Robert Wilensky James T. Willerson Clyde W. Yancy Barry L. Zaret Douglas P. Zipes In Infants and Children
Hugh D. Allen Bruce S. Alpert Stanley J. Goldberg Warren G. Guntheroth Howard P. Gutgesell John D. Kugler James E. Lock John W. Moore Lowell W. Perry David J. Sahn Richard M. Schieken CARDIOVASCULAR SURGERY
Eugene H. Blackstone Lawrence I. Bonchek Lawrence H. Cohn John A. Elefteriades Hartzel V. Schaff RELATED SPECIALISTS
L. Maximilian Buja Jean-Pierre Despres Michael Emmett Giovanni Filardo Barry A. Franklin Charles B. Higgins Jeffrey E. Saffitz Renu Virmani
A3
THE AMERICAN JOURNAL OF CARDIOLOGY姞 VOL. 107, NO. 5 MARCH 1, 2011
CONTENTS Coronary Artery Disease Usefulness of Comprehensive Cardiothoracic Computed Tomography in the Evaluation of Acute Undifferentiated Chest Discomfort in the Emergency Department (CAPTURE) ....................................643 Ian S. Rogers, Dahlia Banerji, Emily L. Siegel, Quynh A. Truong, Brian B. Ghoshhajra, Thomas Irlbeck, Suhny Abbara, Rajiv Gupta, Ricardo J. Benenstein, Garry Choy, Laura L. Avery, Robert A. Novelline, Fabian Bamberg, Thomas J. Brady, John T. Nagurney, and Udo Hoffmann
Comparison of Mortality Rates in Women Versus Men Presenting With ST-Segment Elevation Myocardial Infarction ......................................651 Fabrizio D’Ascenzo, Anna Gonella, Giorgio Quadri, Giada Longo, Giuseppe Biondi-Zoccai, Claudio Moretti, Pierluigi Omede`, Filippo Sciuto, Fiorenzo Gaita, and Imad Sheiban
Global Variability in Angina Pectoris and Its Association With Body Mass Index and Poverty ..........................................................655 Longjian Liu, Jixiang Ma, Xiaoyan Yin, Ellie Kelepouris, and Howard J. Eisen
Igor Balevski, Mojca Cizek Sajko, Vojko Kanic, and Marko Noc
Identifying Patients at Risk for Premature Discontinuation of Thienopyridine After Coronary Stent Implantation ...........................................685 Alexandre S. Quadros, Dulce I. Welter, Fernanda O. Camozzatto, Áurea Chaves, Rajendra H. Mehta, Carlos A. Gottschall, and Renato D. Lopes
Preventive Cardiology Therapeutic Benefit of Preventive Telehealth Counseling in the Community Outreach Heart Health and Risk Reduction Trial ........................690 Robert P. Nolan, Ross E.G. Upshur, Hazel Lynn, Thomas Crichton, Ellen Rukholm, Donna E. Stewart, David A. Alter, Caroline Chessex, Paula J. Harvey, Sherry L. Grace, Louise Picard, Isabelle Michel, Jan Angus, Kim Corace, Susan M. Barry-Bianchi, and Maggie H. Chen
Arrhythmias and Conduction Disturbances
Meta-Analysis of B-Type Natriuretic Peptide’s Ability to Identify Stress Induced Myocardial Ischemia ..662 M. Adnan Nadir, Miles D. Witham, Benjamin R. Szwejkowski, and Allan D. Struthers
Temporal Trends (over 30 Years), Clinical Characteristics, Outcomes, and Gender in Patients <50 Years of Age Having Percutaneous Coronary Intervention ....................................................668 Farhan J. Khawaja, Charanjit S. Rihal, Ryan J. Lennon, David R. Holmes, and Abhiram Prasad
Effect of Insurance Type on Adverse Cardiac Events After Percutaneous Coronary Intervention ......... 675 Michael A. Gaglia, Jr., Rebecca Torguson, Zhenyi Xue, Manuel A. Gonzalez, Itsik Ben-Dor, Gabriel Maluenda, Michael Mahmoudi, Gabriel Sardi, Kohei Wakabayashi, Kimberly Kaneshige, William O. Suddath, Kenneth M. Kent, Lowell F. Satler, Augusto D. Pichard, and Ron Waksman
A4 THE AMERICAN JOURNAL OF CARDIOLOGY姞
Catheter Aspiration in ST-Elevation Myocardial Infarction and Different Extent of Coronary Thrombus .......................................................681
VOL. 107
Patterns of Ventricular Tachyarrhythmias Associated With Training, Deconditioning and Retraining in Elite Athletes Without Cardiovascular Abnormalities .................................................697 Alessandro Biffi, Barry J. Maron, Franco Culasso, Luisa Verdile, Fredrick Fernando, Barbara Di Giacinto, Fernando M. Di Paolo, Antonio Spataro, Pietro Delise, and Antonio Pelliccia
Safety of Lower Activated Clotting Times During Atrial Fibrillation Ablation Using Open Irrigated Tip Catheters and a Single Transseptal Puncture .....704 Roger A. Winkle, R. Hardwin Mead, Gregory Engel, and Rob A. Patrawala
Sleep-Disordered Breathing in Patients With the Brugada Syndrome .........................................709 Paula G. Macedo, Josep Brugada, Pavel Leinveber, Begoña Benito, Irma Molina, Fatima Sert-Kuniyoshi, Taro Adachi, Jan Bukartyk, Christelle van der Walt, Tomas Konecny, Shantal Maharaj, Tomas Kara, Josep Montserrat, and Virend Somers
MARCH 1, 2011
Frequency of Cardiac Events at Four Years Among Initially Asymptomatic Filipinos With the Brugada Type 1 Electrocardiographic Pattern .................714
Incidence, Epidemiology, and Prognosis of Residual Pulmonary Hypertension After Mitral Valve Repair for Degenerative Mitral Regurgitation ...............755
Giselle Gervacio Domingo, Gabriel Jocson, and Antonio Dans
Andrew B. Goldstone, Joanna Chikwe, Sean P. Pinney, Anelechi C. Anyanwu, Samuel A. Funt, Antonio Polanco, and David H. Adams
Systemic Hypertension Differential Effect of Elevated Blood Pressure on Left Ventricular Geometry Types in Black and White Young Adults in a Community (from the Bogalusa Heart Study) ...................................................717 Jian Wang, Wei Chen, Litao Ruan, Ahmet Toprak, Sathanur R. Srinivasan, and Gerald S. Berenson
Heart Failure Predictive Value of Depressive Symptoms and BType Natriuretic Peptide for New-Onset Heart Failure and Mortality .......................................723 Krista C. van den Broek, Christopher R. deFilippi, Robert H. Christenson, Stephen L. Seliger, John S. Gottdiener, and Willem J. Kop
Effect and Clinical Prediction of Worsening Renal Function in Acute Decompensated Heart Failure ...........................................................730 Tobias Breidthardt, Thenral Socrates, Markus Noveanu, Theresia Klima, Corinna Heinisch, Tobias Reichlin, Mihael Potocki, Albina Nowak, Christopher Tschung, Nisha Arenja, Roland Bingisser, and Christian Mueller
Value of the Surface Electrocardiogram in Detecting Right Ventricular Dilatation in the Presence of Left Bundle Branch Block .......................................736 Rutger J. Van Bommel, Nina Ajmone Marsan, Victoria Delgado, Eva P.M. van Rijnsoever, Martin J. Schalij, Jeroen J. Bax, and Hein J. Wellens
Valvular Heart Disease Relation of Aortic Valve Weight to Severity of Aortic Stenosis .........................................................741 Renato Razzolini, Susy Longhi, Giuseppe Tarantini, Stefania Rizzo, Massimo Napodano, Elena Abate, Chiara Fraccaro, Gaetano Thiene, Sabino Iliceto, Gino Gerosa, and Cristina Basso
Incidence, Predictors, and Outcome of Conduction Disorders After Transcatheter Self-Expandable Aortic Valve Implantation ................................747 Chiara Fraccaro, Gianfranco Buja, Giuseppe Tarantini, Valeria Gasparetto, Loira Leoni, Renato Razzolini, Domenico Corrado, Raffaele Bonato, Cristina Basso, Gaetano Thiene, Gino Gerosa, Giambattista Isabella, Sabino Iliceto, and Massimo Napodano A8 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 107
Congenital Heart Disease Spontaneous Rupture of Atrioventricular Valve Tensor Apparatus as Late Manifestation of Anti-Ro/ SSA Antibody-Mediated Cardiac Disease .......... 761 Bettina F. Cuneo, Deborah Fruitman, D. Woodrow Benson, Bo-Yee Ngan, Michael R. Liske, Marie Wahren-Herlineus, S. Yen Ho, and Edgar Jaeggi
Cardiac Magnetic Resonance Imaging and the Assessment of Ebstein Anomaly in Adults .......... 767 Sergey Yalonetsky, Daniel Tobler, Matthias Greutmann, Andrew M. Crean, Bernd J. Wintersperger, Elsie T. Nguyen, Erwin N. Oechslin, Candice K. Silversides, and Rachel M. Wald
Miscellaneous Prognosis Based on Creatine Kinase Isoenzyme MB, Cardiac Troponin I, and Right Ventricular Size in Stable Patients With Acute Pulmonary Embolism .......................................................774 Paul D. Stein, Muhammad Janjua, Fadi Matta, Pramod K. Pathak, Fadel Jaweesh, Ahmad Alrifai, and Haroon L. Chughtai
Usefulness of Postexercise Ankle-Brachial Index to Predict All-Cause Mortality ...............................778 Mobeen A. Sheikh, Deepak L. Bhatt, Jianbo Li, Songhua Lin, and John R. Bartholomew
Comparison of Central Artery Elasticity in Swimmers, Runners, and the Sedentary ............783 Nantinee Nualnim, Jill N. Barnes, Takashi Tarumi, Christopher P. Renzi, and Hirofumi Tanaka
Case Report Carcinoid Heart Disease Without the Carcinoid Syndrome but With Quadrivalvular Regurgitation and Unsuccessful Operative Intervention ........... 788 William Clifford Roberts, Cyril Abie Varughese, Jong Mi Ko, Paul A. Grayburn, Robert Frederick Hebeler, Jr., and Elizabeth C. Burton
Disappearance of Angina Pectoris by LipidLowering in Type III Hyperlipoproteinemia ........ 793 Eun Jeung Cho, Yun Joo Min, Min Seok Oh, Jee Eun Kwon, Jeung Eun Kim, and Chee Jeong Kim MARCH 1, 2011
Readers’ Comments Authors’ Reply ................................................797 Paul Dudley White on Echocardiography. “Nobody Is Perfect” ......................................................797 Usefulness of Serial of B-Type Natriuretic Peptide and Troponin-I Levels to Predict Left Ventricular Remodeling After Primary Coronary Angioplasty ....................................................797
Instructions to Authors can be found at the AJC website: www.AJConline.org Classifieds on page A25
Full Text: www.ajconline.org
CONTENTS
A11
Usefulness of Comprehensive Cardiothoracic Computed Tomography in the Evaluation of Acute Undifferentiated Chest Discomfort in the Emergency Department (CAPTURE) Ian S. Rogers, MD, MPHa,b, Dahlia Banerji, MDa, Emily L. Siegel, BAa, Quynh A. Truong, MD, MPHa, Brian B. Ghoshhajra, MD, MBAa, Thomas Irlbeck, BSa, Suhny Abbara, MDa, Rajiv Gupta, MD, PhDa, Ricardo J. Benenstein, MDa, Garry Choy, MD, MSca, Laura L. Avery, MDc, Robert A. Novelline, MDc, Fabian Bamberg, MD, MPHa, Thomas J. Brady, MDa, John T. Nagurney, MD, MPHd, and Udo Hoffmann, MD, MPHa,* Newer cardiac computed tomographic (CT) technology has permitted comprehensive cardiothoracic evaluations for coronary artery disease, pulmonary embolism, and aortic dissection within a single breath hold, independent of the heart rate. We conducted a randomized diagnostic trial to compare the efficiency of a comprehensive cardiothoracic CT examination in the evaluation of patients presenting to the emergency department with undifferentiated acute chest discomfort or dyspnea. We randomized the emergency department patients clinically scheduled to undergo a dedicated CT protocol to assess coronary artery disease, pulmonary embolism, or aortic dissection to either the planned dedicated CT protocol or a comprehensive cardiothoracic CT protocol. All CT examinations were performed using a 64-slice dual source CT scanner. The CT results were immediately communicated to the emergency department providers, who directed further management at their discretion. The subjects were then followed for the remainder of their hospitalization and for 30 days after hospitalization. Overall, 59 patients (mean age 51.2 ⴞ 11.4 years, 72.9% men) were randomized to either dedicated (n ⴝ 30) or comprehensive (n ⴝ 29) CT scanning. No significant difference was found in the median length of stay (7.6 vs 8.2 hours, p ⴝ 0.79), rate of hospital discharge without additional imaging (70% vs 69%, p ⴝ 0.99), median interval to exclusion of an acute event (5.2 vs 6.5 hours, p ⴝ 0.64), costs of care (p ⴝ 0.16), or the number of revisits (p ⴝ 0.13) between the dedicated and comprehensive arms, respectively. In addition, radiation exposure (11.3 mSv vs 12.8 mSv, p ⴝ 0.16) and the frequency of incidental findings requiring follow-up (24.1% vs 33.3%, p ⴝ 0.57) were similar between the 2 arms. Comprehensive cardiothoracic CT scanning was feasible, with a similar diagnostic yield to dedicated protocols. However, it did not reduce the length of stay, rate of subsequent testing, or costs. In conclusion, although this “triple rule out” protocol might be helpful in the evaluation of select patients, these findings suggest that it should not be used routinely with the expectation that it will improve efficiency or reduce resource use. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:643– 650) Contrast-enhanced computed tomographic (CT) angiography has become a standard procedure in the evaluation of pulmonary embolus (PE)1 and aortic dissection (AD).2 Cara
Cardiac MR PET CT Program, Division of Cardiology and Department of Radiology, cDivision of Emergency Imaging, and dDepartment of Emergency Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA; and bDivision of Cardiovascular Medicine, Stanford University, Falk Cardiovascular Research Center, Stanford, CA. Manuscript received July 22, 2010; manuscript received and accepted October 26, 2010. This trial was investigator initiated and supported by an unrestricted grant from Bracco Diagnostics, Inc, Princeton, New Jersey. Drs. Rogers, Truong, and Ghoshhajra received salary support from grant T32HL076136 from the National Institutes of Health, Bethesda, Maryland. The investigators maintained full control of the data and statistical analysis. *Corresponding author: Tel: (617) 725-1266; fax: (617) 724-4152. E-mail address:
[email protected] (U. Hoffmann). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.039
diac computed tomography has proved to be an effective tool to rule out coronary artery disease (CAD), with a sensitivity of 93% to 99% and negative predictive value of 95% to 99%.3– 6 Recent data have suggested the potential to improve the efficiency of the treatment of patients with acute chest pain and a low-to-intermediate probability of acute coronary syndrome. These data have suggested that the noninvasive detection of CAD and left ventricular function using computed tomography might significantly improve the diagnosis and treatment of patients with suspicion of acute coronary syndrome.7 About 20% of patients with acute chest pain present to the emergency department (ED) with undifferentiated acute chest pain and often require multiple examinations to exclude PE and/or AD, in addition to excluding obstructive CAD.8 The results of several recent studies have suggested that high temporal resolution, dualsource CT technology permits simultaneous assessment of CAD, PE, and AD, independent of the heart rate, in a single www.ajconline.org
644
The American Journal of Cardiology (www.ajconline.org)
scan.9 –11 However, it is unknown whether providing such a comprehensive assessment would lead to improved efficiency in the treatment of these patients. Thus, our goal was to determine whether providing a comprehensive cardiothoracic CT examination would result in significant improvement in the efficiency of treating patients presenting to the ED with undifferentiated acute chest discomfort or dyspnea in the setting of a tertiary academic hospital. Methods Our subject population consisted of patients who presented to the ED of a tertiary academic center with undifferentiated acute chest discomfort and/or shortness of breath with a component of chest discomfort of ⱖ5 minutes’ duration within the previous 24 hours. Specifically, we included subjects if the assessing ED attending physician had independently decided that the patient’s care plan should include a cardiac, PE, or AD CT examination after the standard initial clinical evaluation to rule out acute coronary syndrome, PE, or AD. The potential subjects were ⱖ30 years old and in sinus rhythm. The women were required to be of nonchildbearing potential or to have had negative findings from a pregnancy test. The exclusion criteria included positive cardiac biomarkers, changes on the electrocardiogram diagnostic of myocardial ischemia, a known history of CAD (e.g., previous myocardial infarction, previous coronary stent placement, and/or coronary artery bypass graft surgery), a known history of thoracic aortic disease (i.e., thoracic aortic aneurysm ⬎5 cm in diameter, history of AD, and/or a history of thoracic aortic aneurysm repair), a known history of PE within the previous 6 months, and cardiopulmonary instability (e.g., heart rate ⬎100 beats/min, systolic blood pressure ⬍105 mm Hg, or oxygen saturation ⬍90%). The potential subjects were not eligible if they had a serum creatinine clearance of ⬍60 ml/min by Cockcroft-Gault, had a known allergy to iodinated contrast agents, were receiving metformin therapy, or were unable or unwilling to discontinue therapy for 48 hours after the CT evaluation. Nitroglycerin was not administered if the subject had used a PDE-5 inhibitor (i.e., sildenafil, tadalafil, or vardenafil) within the previous 72 hours. Our trial was designed as a prospective, randomized, controlled diagnostic trial to assess the relative efficiency of the comprehensive cardiothoracic CT protocol. We screened all patients who had presented with a chief complaint of chest discomfort and/or shortness of breath to the ED on weekdays from 8 A.M. to 6 P.M. The eligible patients willing to participate in the present trial were randomized to receive either the dedicated cardiac, pulmonary, or aortic CT scan initially scheduled as a part of their clinical care to exclude CAD, PE, or AD or to receive a comprehensive cardiothoracic CT scan—a single CT scan designed to exclude CAD, PE, and AD. Before recruitment, the randomized study group allocations were placed into sealed and numbered opaque envelopes. Accordingly, the treating ED physicians, study physicians, and subjects had no knowledge of the study arm into which the subject would be assigned before randomization. After consent, each subject opened the appropriate, sealed
envelope in the presence of the study physicians to reveal the study arm into which the subject was assigned. After randomization, the treating ED physicians, study physicians, and subjects had full knowledge of the group assignment. The CT scans were completed immediately after patient consent and randomization. We prospectively collected data on the demographics, risk factor profile, and clinical course for all patients. The presence of risk factors was established from patient interview and a review of the available medical records. The subjects were then followed for the remainder of their hospitalization to determine the length of stay, costs incurred, and any additional diagnostic testing thought necessary to achieve a diagnosis for the presenting symptoms. The medical records were reviewed to obtain the results of all diagnostic tests performed during the index hospitalization and for a period of 30 days after discharge. The institutional review board of Massachusetts General Hospital approved the study protocol, and all patients provided written informed consent. Electrocardiographic-gated comprehensive computed tomography was performed using dual-source computed tomography (Definition, Siemens Medical Solutions, Forchheim, Germany). To determine the contrast volume and scan trigger time, a 20-mL intravenous bolus of contrast agent (iopamidol 370 mg iodine/ml, Isovue 370, Bracco Diagnostics, Princeton, New Jersey) was injected intravenously at a rate of 5 ml/s for opacification of the coronary artery lumen followed by saline injected intravenously at the same rate. Sequential scans were obtained at the level of the right pulmonary artery and aortic root at 2-second intervals beginning 4 seconds after contrast administration. The peak opacification time for the right pulmonary artery and aortic root was recorded. The interval to peak opacification in the aortic root served as the trigger time for initiating scan acquisition. To determine the total contrast volume to be infused for CT angiography, the difference between the interval to peak opacification in the right pulmonary artery and the interval to peak opacification in the aorta was added to the time required to cover the scan length, and this value was multiplied by the infusion rate. Retrospective image acquisition was performed during a single breath hold in inspiration in a caudal– cranial direction from the diaphragm to the lung apices. The imaging parameters included a slice collimation of 2 ⫻ 64 ⫻ 0.6 mm, a gantry rotation time of 330 ms, a temporal resolution of 83 ms, a tube voltage of 100 to 120 kVp, and a maximal tube current of 400 mA. To lower the radiation exposure, we used a 100 kVp for subjects with a body mass index ⬍30 kg/m2 and a 120 kVp for subjects with a body mass index ⱖ30 kg/m2. Aggressive tube current modulation, attenuation adaptation, and adaptive pitch selection were used in all patients. Imaging in the dedicated arm was also performed using dual-source computed tomography (Definition, Siemens Medical Solutions). Cardiac CT scans were performed according to the standard protocol using retrospective electrocardiographic gating, field of view from the carina to the diaphragm, test contrast bolus followed by sufficient contrast with cover of the scan length as triggered by the aortic root, and a maximal tube current of 400 mA and 100 kVp
Coronary Artery Disease/Comprehensive CT for Undifferentiated Chest Discomfort
645
Table 1 Clinical characteristics Variable Age (years) Men Body mass index (kg/m2) White Primary suspicion Acute coronary syndrome Pulmonary embolus Aortic dissection Average Thrombolysis In Myocardial Infarction score Hypertension Hyperlipidemia Diabetes mellitus Current or past smoker Family history of coronary artery disease
Overall (n ⫽ 59)
Comprehensive (n ⫽ 29)
Dedicated (n ⫽ 30)
p Value
51.2 ⫾ 11.4 43 (73%) 28.7 ⫾ 5.3 52 (88%)
49.7 ⫾ 12.2 23 (79%) 28.7 ⫾ 6.0 27 (93%)
52.6 ⫾ 10.6 20 (67%) 28.7 ⫾ 4.7 25 (83%)
0.34 0.38 0.99 0.42 0.93
30 (51%) 24 (41%) 5 (9%) 0.9 ⫾ 0.9 16 (27%) 13 (22%) 6 (10%) 12 (20%) 26 (44%)
14 (48%) 12 (41%) 3 (10%) 1.1 ⫾ 1.1 13 (45%) 8 (28%) 3 (10%) 6 (21%) 13 (45%)
16 (53%) 12 (40%) 2 (7%) 0.7 ⫾ 0.8 3 (10%) 5 (17%) 3 (10%) 6 (20%) 13 (4%)
for body mass index ⬍30 kg/m2 and 120 kVp for a body mass index ⱖ30 kg/m2, tube current modulation, attenuation adaptation, and adaptive pitch selection. Dedicated helical CT evaluation of the pulmonary arteries and dedicated CT evaluation of the aorta were performed according to the department protocol without electrocardiographic gating. Dedicated CT examination of the pulmonary arteries (240 mA, 120 kVp) covered a field of view from the lung apices to the diaphragm and used bolus tracking software in the right ventricle to trigger delivery of a fixed bolus of contrast according to weight (ⱕ100 lb, 110 ml; 101 to 160 lb, 120 ml, and ⬎160 lb, 130 ml). Lower extremity venography was acquired from the tibial plateaus to the top of the iliac crests approximately 180 seconds after the start of the initial infusion. Dedicated CT examination of the aorta (250 mA, 120 kVp) covered a field of view from the lung apices to the iliac bifurcation. A noncontrast-enhanced scan to evaluate for intramural hematoma was followed by a contrast-enhanced scan that used bolus tracking software in the descending thoracic aorta to trigger delivery of a fixed bolus (90 ml) of contrast. The CT data sets were postprocessed and interpreted in real-time by cardiac imaging and emergency radiology physicians using a dedicated workstation immediately after the examination. Dedicated cardiac CT examinations and comprehensive CT examinations were qualitatively evaluated by Core Cardiology Training Symposium (COCATS) level III-trained physicians from the cardiac imaging service for the presence of coronary atherosclerotic plaque,12 luminal narrowing, and global and regional left ventricular function. All CT data sets were assessed for the presence of extracardiac findings, such as pneumonia, pneumothorax, and rib fracture and other cardiac pathologic features. The results were relayed to the emergency medicine attending physician caring for the patient immediately after the final interpretation. Although the subsequent patient treatment was at the full discretion of the ED attending physician caring for the patient, the ED physicians were offered nonbinding recommendations for the management of CAD as determined from the previous findings7,13 at the reporting of the CT findings by the cardiac imaging attending physicians: (1) no CAD and normal left ventricular function, discharge
0.10 0.003 0.36 1.00 1.00 1.00
without additional cardiopulmonary diagnostic testing; (2) CAD with ⬍25% luminal narrowing and normal left ventricular function, discharge without additional cardiopulmonary diagnostic testing if the findings from repeated electrocardiography and cardiac enzyme testing were negative at 6 hours; (3) CAD with ⬍25% narrowing and global or regional left ventricular dysfunction or CAD with 25% to 75% narrowing, myocardial perfusion imaging if the findings from repeated electrocardiography and cardiac enzyme testing were negative at 6 hours; and (4) CAD with ⬎75% narrowing, cardiology consultation for consideration of myocardial perfusion imaging or invasive angiography. The primary end point of the present study was the length of hospital stay, defined as the interval from presentation to the ED to discharge from the ED or inpatient hospital unit, whichever occurred later. The present study was powered to detect a difference of 11.6 hours, which was previously reported for computed tomography versus nuclear imaging,13 with 90% power at a significance level of 0.05. Although no previous studies we are aware of have compared the length of stay between these CT strategies, we chose this length of stay difference as a surrogate, given the similarities in the diagnostic strategies. The secondary end points included the interval to exclusion of an acute event, the rate of hospital discharge without additional imaging, the index hospitalization cost, radiation exposure, rate of incidental findings requiring follow-up, and the occurrence of major cardiovascular adverse events or follow-up visits within 30 days of follow-up. The interval to the exclusion of an acute event was defined as the interval from presentation to the ED to the performance of the last diagnostic test clinically necessary to exclude acute coronary syndrome, PE, and/or thoracic AD. The rate of hospital discharge without additional imaging was defined as discharge that occurred when no additional diagnostic imaging test (e.g., single photon emission computed tomography, transthoracic echocardiography, or magnetic resonance imaging) was performed after the CT scan. The cost of the index hospitalization was determined from the ED and inpatient care costs, including laboratory and imaging tests. Acute coronary syndrome was defined as either an acute myocardial infarction (ST-segment elevation myocardial in-
646
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Three-dimensional volume-rendered and multiplanar reconstructions of comprehensive CT study of 32-year-old man who presented with chest pain that radiated to his back showing no evidence of CAD, PE, or AD. Table 2 Clinical end points Variable Length of stay (hours: minutes) Median Interquartile range Interval from emergency department presentation to computed tomography (hours: minutes) Median Interquartile range Interval to exclusion of acute event (hours: min) Median Interquartile range Direct hospital discharge (n) Cost of stay ($) Median Interquartile range
Overall (n ⫽ 59)
Comprehensive (n ⫽ 29)
Dedicated (n ⫽ 30)
8:12 5:19, 30: 05
7:38 5:23, 28:14
p Value 0.79
8:02 5:19, 29:20
0.56 3:34 2:50, 4:38
3:25 2:50, 4:26
3:42 3:08, 4:46
6:12 4:11, 20:26 41 (69.5%)
6:29 4:17, 18:19 20 (69%)
5:15 4:05, 20:26 21 (70%)
0.64
1809 1424, 4877
farction or non–ST-segment elevation myocardial infarction) or unstable angina according to the clinical discharge diagnosis.1,14 The presence of PE and AD was established or excluded according to the CT findings. A follow-up telephone interview using a standardized questionnaire was
1898 1554, 5116
0.99 0.16
1724 1277, 4208
conducted 1 month after enrollment to determine the occurrence of events (i.e., death from all causes, myocardial infarction, and coronary revascularization). Any potential major cardiovascular adverse events, such as a report of recurrent symptoms resulting in medical consultation, diag-
Coronary Artery Disease/Comprehensive CT for Undifferentiated Chest Discomfort
647
Figure 2. Summary of additional testing performed after dedicated and comprehensive CT scans performed as part of present trial. No significant difference was found in additional testing to exclude acute events between the dedicated and comprehensive CT arms. Arrows pointing exclusively to “Home” indicate patients discharged without additional diagnostic testing. CE ⫽ second set of cardiac enzymes; ICA ⫽ invasive coronary angiography; SPECT ⫽ single photon emission computed tomography (exercise or pharmacologic cardiac SPECT); TTE ⫽ transthoracic echocardiography. Table 3 Scan characteristics Variable
Overall (n ⫽ 59)
Comprehensive (n ⫽ 29)
Dedicated (n ⫽ 30)
p Value
Diagnostic image quality Radiation exposure (mSv) Median Interquartile range Contrast volume (ml) Median Interquartile range Incidental findings Requiring follow-up Changing patient treatment
58 (98%)
28 (97%)
30 (100%)
0.49 0.16
12.6 9.2, 22
12.8 10.7, 23.1
11.3 5.7, 22
113 95,145 37 (63%) 17 (29%) 0
145 125, 155 13 (45%) 7 (24%) 0
95 85, 110 24 (80%) 10 (33%) 0
⬍0.0001
nostic testing, or hospital admission, were subsequently validated by a review of the medical records by the outcome panel, whenever available. The descriptive statistics were expressed as the mean ⫾ SD or median with the interquartile range for continuous variables, depending on normality, and as the frequency and proportions for nominal variables. An analysis of the baseline characteristics of the groups was conducted using t tests for normal continuous variables, the Wilcoxon rank sum test for the non-normal continuous variables, and Fisher’s exact tests for binary variables. The length of stay, cost of hospitalization, and scan data are presented as the median and interquartile range, and the differences between the groups were assessed using Wilcoxon rank sum tests. All analyses were performed using SAS, version 9.1 (SAS Institute, Cary, North Carolina). A 2-tailed p value of ⬍0.05 was considered statistically significant.
0.007 0.57 —
Results A total of 59 patients agreed to participate and were enrolled during a cumulative 9-month enrollment period, 30 of whom were randomized to undergo a dedicated CT evaluation and 29 to undergo a comprehensive CT evaluation. The baseline characteristics of the enrolled subjects are listed in Table 1. The most common exclusion criteria were a known history of CAD (45%) and serum creatinine clearance of ⬍60 ml/min (22%). Sample images from a comprehensive scan are displayed in Figure 1. The mean age of all subjects was 51 years; no difference was found in the age between subjects undergoing a dedicated CT scan (52.6 ⫾ 10.6 years) and those undergoing a comprehensive CT scan (49.7 ⫾ 12.2 years; p ⫽ 0.34). The subjects were primarily white (88%) and men (73%); however, no significant difference was found in gender composition (67% vs 79%, p ⫽ 0.38) or race (93% vs 83%, p ⫽ 0.42) between the 2 groups.
648
The American Journal of Cardiology (www.ajconline.org)
Figure 3. Summary of CAD detected on dedicated coronary or comprehensive CT scans performed as part of trial with additional testing ordered by treating clinicians because of the findings. Arrows pointing exclusively to “Home” indicate patients discharged without additional diagnostic testing. CE ⫽ second set of cardiac enzymes; ICA ⫽ invasive coronary angiography; SPECT ⫽ single photon emission computed tomography (exercise or pharmacologic cardiac SPECT); TTE ⫽ transthoracic echocardiography.
Overall, the primary clinical suspicion for which a CT scan had been considered before recruitment was 50.8% acute coronary syndrome, 40.7% PE, and 8.5% AD. No difference was found between the dedicated and comprehensive groups (acute coronary syndrome 53.3% vs 48.3%; PE 40.0% vs 41.4%; AD 6.7% vs 10.3%, p ⫽ 0.93). acute coronary syndrome, PE, and AD were ruled out in all but 1 patient in the dedicated CT arm, who was diagnosed with PE according to the CT findings. As listed in Table 2, no significant difference was found in the primary end point, the median length of stay (7.6 vs 8.2 hours, p ⫽ 0.79), or in the secondary end points, the rate of hospital discharge without additional imaging (70% vs 69%, p ⫽ 0.99), interval to the exclusion of the acute event (5.2 vs 6.5 hours, p ⫽ 0.64), or hospitalization costs (p ⫽ 0.16), between the dedicated and comprehensive CT arms, respectively. Additional testing was ordered by the treating physicians for 47% of all subjects in the dedicated CT arm to exclude an acute event versus 52% of subjects in the comprehensive CT arm, a difference that was not statistically significant (p ⫽ 0.61). A summary of additional testing is shown in Figure 2.
All but one comprehensive CT scans and all dedicated CT scans had clinically diagnostic image quality. The comprehensive CT scan with suboptimal imaging quality was limited secondary to the body mass index of the subject (body mass index 42 kg/m2), which was not an exclusion criterion for the present trial. As listed in Table 3, the median radiation exposure was similar between the 2 arms (11.3 vs 12.8 mSv, p ⫽ 0.16); however, the subjects undergoing dedicated CT studies had received less contrast than those undergoing comprehensive CT studies (95 vs 145 ml, respectively; p ⬍0.0001). No important adverse events or side effects occurred from any of the CT protocols. One major cardiovascular adverse events occurred during the 1-month follow-up period in a subject in the dedicated CT arm who had had negative findings for PE on a dedicated CT scan as a part of the present trial but subsequently experienced a myocardial infarction. No major cardiovascular adverse events were observed in the comprehensive CT arm. A similar number of subjects in the comprehensive and dedicated CT arms (20.7% vs 33.3%, p ⫽ 0.39) reported recurrent chest discomfort or dyspnea during follow-up. Fewer subjects in the comprehensive CT
Coronary Artery Disease/Comprehensive CT for Undifferentiated Chest Discomfort
group reported returning to the ED or their physician’s office for evaluation (13.8% vs 33.3%), although this difference was not statistically significant (p ⫽ 0.13). CAD was newly diagnosed in 19 (42%) of 45 subjects who had had their coronary arteries assessed with either a dedicated cardiac or comprehensive CT scan. Of the 29 subjects who had undergone comprehensive CT evaluation, 12 had had CAD diagnosed from the comprehensive CT findings. Of the 12, 5 were thought to have 1% to 24% luminal narrowing, 6 were thought to have 25% to 75% narrowing, and 1 was thought to have ⬎75%. A summary of the evaluation of newly diagnosed CAD is provided in Figure 3. Overall, ED physicians followed the suggested recommendation for 74% of the subjects with newly diagnosed CAD. In the remaining 26% of subjects, the ED physicians chose less additional testing than was suggested from the imaging findings. Discussion We performed a randomized diagnostic trial to determine whether providing a comprehensive cardiothoracic CT scan would result in changes in the efficiency of treating low-risk patients presenting to the ED with undifferentiated acute chest discomfort or dyspnea in the setting of a tertiary academic hospital. In our randomized comparison of dedicated standard CT evaluations to exclude CAD, PE, or AD and comprehensive cardiothoracic CT evaluations, we found no differences in the primary end point, median length of stay, or in the secondary end points, such as the rates of hospital discharge without additional imaging, the interval to exclusion of an acute event, or the hospitalization costs in this patient population. Our trial design was determined from various studies, including several case series that established the feasibility of achieving diagnostic image quality with moderate radiation using a comprehensive cardiothoracic electrocardiographic-gated CT protocol,9 –11,15 studies that reported that health care use could be improved (length of stay, costs) with cardiac CT, and studies that found that ED patients with acute chest pain often undergo evaluation for ⬎1 life-threatening condition (i.e., PE, CAD, or AD). That we did not detect significant differences in the length of stay or other measures of efficiency of patient management was most likely related to the very low rate of acute coronary syndrome, PE, and AD (ruled out for 58 of 59 patients) and the additional evaluation of newly discovered CAD in the comprehensive protocol. Overall, the number of patients in whom additional testing for a primary differential diagnosis was spared by the comprehensive protocol was largely offset by the number of patients in which newly diagnosed nonobstructive CAD was incidentally discovered and subsequently evaluated. Although a considerable amount of subjects in both study arms had other incidentally discovered findings, none of these incidental findings were important enough to warrant a beneficial change in their index care plan. No serious adverse events occurred during the trial, and only 1 clinical event occurred during follow-up, a cardiac arrest that required percutaneous coronary intervention for an occluded marginal branch in 1 patient without a known
649
history of CAD who had undergone a dedicated PE CT scan as part of the present trial for the evaluation of dyspnea with an elevated D-dimer assay. Although this patient probably epitomizes the potential value of a comprehensive protocol because the CAD would have likely been detected on a comprehensive CT scan, the low number of events precluded any meaningful conclusions. Although all patients had a clinical indication for computed tomography, the low event rate also suggested that these examinations might be overused in low-risk patients, arguably some of whom might not have needed any imaging at all. This fact is well known from recent studies, documenting the decreasing prevalence of findings even in patients with specific suspicion for one of the differential diagnoses (PE, AD, acute coronary syndrome). Dedicated CT examinations for PE have an event rate as low as 2%, and the event rates for acute coronary syndrome have been 0%, even in a study of dedicated acute coronary syndrome suspicion.13 Our results have shown that the effective use of aggressive dose-reduction strategies by trained physicians directly involved in CT scanning will lead to a radiation exposure that was similar between the 2 arms while maintaining a similar image quality, despite preclusion of intravenous  blockade. Our data suggest a reduction in the repeat visits for the presenting symptom might be a potential benefit of a comprehensive cardiothoracic CT scan, because fewer subjects in the comprehensive CT group reported returning to the ED or their physician’s office than in the dedicated CT group (13.8% vs 33.3%, p ⫽ 0.13). However, additional studies are warranted to confirm this observation. The most important limitations to the present study were the relatively small sample size and the short term followup. Although additional cardiac testing was performed in patients with newly diagnosed CAD that would not have been identified with a dedicated CT evaluation for PE or AD, the longer term implications of establishing the diagnosis of CAD could not be assessed. Perhaps these patients will be more likely to have medical therapy for CAD initiated at an earlier age, reducing longer term events; however, they might be more likely to be referred for invasive angiography in future years. Moreover, our short-term follow-up period might not reflect the follow-up care at another institution that was not declared to our team, and the present trial was not powered to determine differences in major cardiovascular adverse events. The radiation doses administered in our trial might not be generalizable to other institutions or other CT vendors, particularly owing to differences in scanner type and aggressiveness with dose reduction. Physicians and technologists should be vigilant in selecting the lowest tube current and voltage possible for diagnostic image quality for each patient and using the dose reduction techniques available from their vendor, such as attenuation adaptation, adaptive pitch selection, and iterative reconstruction. Newer studies have demonstrated that heart rate control can facilitate techniques to lower radiation exposure to ⬍1 mSv16; however, these techniques have not been validated for dedicated PE or AD protocols. Although no difference was found in the radiation exposure between the 2 arms, variation in the exposure was present within the dedicated arm, with AD scans having the greatest exposure, given the acquisition of
650
The American Journal of Cardiology (www.ajconline.org)
a noncontrast-enhanced scan followed by a contrast-enhanced scan. In addition, the routine department protocol was used for the dedicated AD image acquisition that was not specifically altered to lower radiation. Prospective triggering can significantly lower exposure; however, this frequently requires -blocker administration to achieve a low and steady heart rate, which some believe is relatively contraindicated if pulmonary embolus is in the differential diagnosis. Not surprisingly, the comprehensive protocol required a median 50-ml greater contrast than the dedicated protocols to opacify all the cardiothoracic vasculature. 1. Kruip MJ, Söhne M, Nijkeuter M, Kwakkel-Van Erp HM, Tick LW, Halkes SJ, Prins MH, Kramer MH, Huisman MV, Büller HR, Leebeek FW. Christopher Study Investigators. A simple diagnostic strategy in hospitalized patients with clinically suspected pulmonary embolism. J Intern Med 2006;260:459 – 466. 2. Yoshida S, Akiba H, Tamakawa M, Yama N, Hareyama M, Morishita K, Abe T. Thoracic involvement of type A aortic dissection and intramural hematoma: diagnostic accuracy— comparison of emergency helical CT and surgical findings. Radiology 2003;228:430 – 435. 3. Hoffmann U, Moselewski F, Cury RC, Ferencik M, Jang IK, Diaz LJ, Abbara S, Brady TJ, Achenbach S. Predictive value of 16-slice multidetector spiral computed tomography to detect significant obstructive coronary artery disease in patients at high risk for coronary artery disease: patient-versus segment-based analysis. Circulation 2004;110: 2638 –2643. 4. Leschka S, Alkadhi H, Plass A, Desbiolles L, Grünenfelder J, Marincek B, Wildermuth S. Accuracy of MSCT coronary angiography with 64-slice technology: first experience. Eur Heart J 2005;26:1482– 1487. 5. Mollet NR, Cademartiri F, van Mieghem CA, Runza G, McFadden EP, Baks T, Serruys PW, Krestin GP, de Feyter PJ. High-resolution spiral computed tomography coronary angiography in patients referred for diagnostic conventional coronary angiography. Circulation 2005;112: 2318 –2323. 6. Ropers D, Rixe J, Anders K, Küttner A, Baum U, Bautz W, Daniel WG, Achenbach S. Usefulness of multidetector row spiral computed tomography with 64- ⫻ 0.6-mm collimation and 330-ms rotation for the noninvasive detection of significant coronary artery stenoses. Am J Cardiol 2006;97:343–348. 7. Seneviratne SK, Truong QA, Bamberg F, Rogers IS, Shapiro MD, Schlett CL, Chae CU, Cury R, Abbara S, Brady TJ, Nagurney JT, Hoffmann U. Incremental diagnostic value of regional left ventricular function over coronary assessment by cardiac computed tomography for the detection of acute coronary syndrome in patients with acute
8.
9.
10.
11.
12.
13.
14.
15.
16.
chest pain—From The ROMICAT Trial. Circ Cardiovasc Imaging 2010;3:375–383. Rogg JG, Neve JW, Huang C, Brown D, Jang IK, Chang Y, Marill K, Parry B, Hoffmann U, Nagurney JT. The triple work-up for emergency department patients with acute chest pain: how often does it occur? J Emerg Med Epub 2008 Sept 12. White CS, Kuo D, Kelemen M, Jain V, Musk A, Zaidi E, Read K, Sliker C, Prasad R. Chest pain evaluation in the emergency department: can MDCT provide a comprehensive evaluation? AJR Am J Roentgenol 2005;185:533–540. Raptopoulos VD, Boiselle PB, Michailidis N, Handwerker J, Sabir A, Edlow JA, Pedrosa I, Kruskal JB. MDCT angiography of acute chest pain: evaluation of ECG-gated and nongated techniques. AJR Am J Roentgenol 2006;186:S346 –S356. Savino G, Herzog C, Costello P, Schoepf UJ. 64 Slice cardiovascular CT in the emergency department: concepts and first experiences. Radiol Med 2006;111:481– 496. Hoffmann U, Bamberg F, Chae CU, Nichols JH, Rogers IS, Seneviratne SK, Truong QA, Cury RC, Abbara S, Shapiro MD, Moloo J, Butler J, Ferencik M, Lee H, Jang IK, Parry BA, Brown DF, Udelson JE, Achenbach S, Brady TJ, Nagurney JT. Coronary computed tomography angiography for early triage of patients with acute chest pain: the ROMICAT (Rule Out Myocardial Infarction using Computer Assisted Tomography) trial. J Am Coll Cardiol 2009;53:1642–1650. Goldstein JA, Gallagher MJ, O’Neill WW, Ross MA, O’Neil BJ, Raff GL. A randomized controlled trial of multi-slice coronary computed tomography for evaluation of acute chest pain. J Am Coll Cardiol 2007;49:863– 871. Kushner FG, Hand M, Smith SC Jr, King SB III, Anderson JL, Antman EM, Bailey SR, Bates ER, Blankenship JC, Casey DE Jr, Green LA, Hochman JS, Jacobs AK, Krumholz HM, Morrison DA, Ornato JP, Pearle DL, Peterson ED, Sloan MA, Whitlow PL, Williams DO. Focused updates: ACC/AHA Guidelines for the Management of Patients with ST-Elevation myocardial infarction (updating the 2004 Guideline and 2007 focused update) and ACC/AHA/SCAI Guidelines on Percutaneous Coronary Intervention (updating the 2005 Guideline and 2007 focused update): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2009;120:2271–2306. Schertler T, Scheffel H, Frauenfelder T, Desbiolles L, Leschka S, Stolzmann P, Seifert B, Flohr TG, Marincek B, Alkadhi H. Dualsource computed tomography in patients with acute chest pain: feasibility and image quality. Eur Radiol 2007;17:3179 –3188. Achenbach S, Marwan M, Ropers D, Schepis T, Pflederer T, Anders K, Kuettner A, Daniel WG, Uder M, Lell MM. Coronary computed tomography angiography with a consistent dose below 1 mSv using prospectively electrocardiogram-triggered high-pitch spiral acquisition. Eur Heart J 2010;3:340 –346.
Comparison of Mortality Rates in Women Versus Men Presenting With ST-Segment Elevation Myocardial Infarction Fabrizio D’Ascenzo, MD*,†, Anna Gonella, MD†, Giorgio Quadri, MD, Giada Longo, MD, Giuseppe Biondi-Zoccai, MD, Claudio Moretti, PhD, Pierluigi Omedè, MD, Filippo Sciuto, MD, Fiorenzo Gaita, MD, and Imad Sheiban, MD Women who present with coronary artery disease have different characteristics, undergo different treatment, and have a different prognosis than men. The increasing use of coronary stenting has improved the outcome of percutaneous coronary intervention (PCI). However, little is known about the outcomes for men versus women after PCI, especially for those presenting with a diagnosis of acute coronary syndrome. Thus, we compared the baseline features, management, and long-term outlook of men versus women undergoing PCI. All consecutive patients who had undergone PCI with stents at our center from July 1, 2002 to June 30, 2004 were identified retrospectively. The primary end point was the long-term rate of major adverse cardiac events (i.e., death, infarction, and repeat revascularization). The secondary end points were the individual components of the major adverse cardiac events and stent thrombosis. A total of 833 patients were included, 210 women (25.2%) and 623 men (75.8%). The women were significantly older (70.9 vs 63 years, p <0.001) and more often had diabetes mellitus (36.2% vs 21.0%, p <0.001) and hypertension (82.3% vs 73.7%, p ⴝ 0.006). The number of drug-eluting stents and the length were significantly lower in the female patients. The incidence of major adverse cardiac events after a median follow-up of 60 months was similar for both women and men (38.8% vs 46.4%, p ⴝ 0.075), with a trend toward greater mortality rate for women (21.2% vs 15.4%, p ⴝ 0.090). All other end points occurred with similar frequencies. Only in the subgroup of ST-segment elevation myocardial infarction were the rates of death significantly greater for the women than for the men (20.0% vs 8.1%; p ⴝ 0.029). In conclusion, very long-term follow-up of women undergoing PCI with coronary artery stenting resulted in similar rates of cardiac event compared to those of men, but greater care should be given to women presenting with ST-segment elevation myocardial infarction. Also, despite their greater baseline risk profile, women were significantly less likely to have received effective treatment, the use of including drug-eluting stents. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:651– 654) The aim of the present study was to compare the clinical outcomes of percutaneous coronary intervention (PCI) in male and female patients in the short and very long term according to their admission diagnosis. Methods The present retrospective study included all consecutive patients who had undergone percutaneous transluminal coronary angioplasty at our center from July 2002 to December 2004. These patients were divided into 2 cohorts according to their gender. All patients had provided written informed consent for the procedure, and ethical approval was waived, given the retrospective, observational design. Division of Cardiology, University of Turin, Turin, Italy. Manuscript received August 5, 2010; manuscript received and accepted October 26, 2010. *Corresponding authors: Tel: (39) 33-3399-2707; fax: (39) 01-16336769. E-mail address:
[email protected] (F. D’Ascenzo). †
Drs. D’Ascenzo and Gonella contributed equally to this report.
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.038
All patients had been pretreated with aspirin 100 mg/day and clopidogrel 75 mg/day or ticlopidine 250 mg twice daily for ⱖ3 days before the procedure. Loading doses of 300 to 600 mg of clopidogrel and 250 to 500 mg of aspirin were given to patients who had not been pretreated. At the start of the procedure, unfractionated heparin was administered at a dose of 70 to 100 UI/kg to achieve an activated clotting time of ⱖ250 seconds. The use of glycoprotein IIb/IIIa was left to the discretion of the operators. Coronary vasodilators (nitroglycerin) were routinely used both before and after the procedure. Coronary angioplasty and stent implantation were performed according to the current practice and technical guidelines. The primary end point was the rate of major adverse cardiac events (MACE), defined as a nonhierarchical composite of all causes of death, nonfatal acute myocardial infarction, and target vessel revascularization. The secondary end points were the rates of the single components of MACE (i.e., death, acute myocardial infarction, percutaneous or surgical revascularization on the procedure target vessels or additional procedures such as nontarget vessel revascularization, stent thrombosis, and stroke). Moreover, we evaluated death, myocardial infarction, stroke, and stent www.ajconline.org
652
The American Journal of Cardiology (www.ajconline.org)
Table 1 Baseline characteristics Variable Mean age (years) Type 2 diabetes mellitus Type 1 diabetes mellitus Hyperlipidemia* Hypertension† Smoker Current Former Never Previous myocardial infarction Renal failure Left ventricular ejection fraction ⱕ35% Heart failure Peripheral artery disease Indication Silent myocardial ischemia Stable coronary artery disease Unstable coronary artery disease Non–ST-segment elevation myocardial infarction ST-segment elevation myocardial infarction Other Primary percutaneous coronary intervention
Table 2 Angiographic features Women (n ⫽ 210)
Men (n ⫽ 623)
p Value
Variable
70 ⫾ 9 36.2% 6.0% 56.6% 82.3%
63 ⫾ 8 21.0% 2.6% 59.3% 73.7%
⬍0.0001 0.001 0.030 0.546 0.006
82.6% 9.4% 8.0% 21.8% 11.1% 10.9%
50.8% 32.7% 16.5% 22.3% 11.5% 8.3%
⬍0.001
6.6% 9.5%
4.9% 10.5%
0.381 0.704
5.6% 4.5% 43.4%
8.7% 12.6% 32.1%
0.005 0.03 ⬍0.001
15.7%
16.7%
0.972
27.8%
28.8%
0.8778
3.0% 19.7%
1.1% 20.3%
0.252 0.843
Chronic total occlusion Coronary perforation Unprotected left main coronary artery Bifurcation Thrombosis Vessels ⬍2.75 mm American College of Cardiology/American Heart Association type C lesion Total number of diseased vessels Total number of treated vessels Total number of target lesions Total number of stents Total length of stent (mm) Total number of bare metal stents Total bare metal stent length (mm) Total number of drugeluting stents Total drug-eluting stent length (mm) Use of both bare metal stents and drug-eluting stents Use of paclitaxel-eluting stents only Use of sirolimus-eluting stents only Intra-aortic balloon pump Intravenous glycoprotein IIb/IIIa inhibitors Dual antiplatelet therapy at discharge Length of dual antiplatelet therapy (months)
0.894 0.912 0.355
* Defined as total cholesterol ⬎200 mg/dl, low-density lipoprotein cholesterol ⬎130 mg/dl, or triglycerides ⬎175 mg/dl. † Defined as blood pressure ⬎140/90 mm Hg for overall population, ⬎130/80 mm Hg for people with diabetes mellitus, or the use of antihypertensive drugs.
thrombosis according to the admission diagnosis, stratified as stable angina or stable ischemia, unstable angina, or non–ST-segment elevation myocardial infarction (nonSTEMI) and STEMI. Myocardial infarction was defined as Q-wave or non–Qwave myocardial infarction with elevation of total creatine kinase-MB 2 times greater than the upper limit of normal. Target vessel revascularization was defined as any intervention, surgical or percutaneous, to treat a luminal stenosis occurring in the same coronary vessel treated at the index procedure. Stent thrombosis was adjudicated according to the Academic Research Consortium definitions as definite, probable or possible. Stroke was defined as any ischemic neurologic event extended ⬎24 hours with irreversible neurologic injury or permanent disability. To assess all procedural and in-hospital outcomes, we consulted our institutional electronic database and individual patient charts. We recorded the long-term outcomes of those with ⱖ3 years of follow-up, determined by telephone interviews, ambulatory visits, or formal query of the primary care physician. The continuous variables are expressed as the mean ⫾ SD and were compared using analysis of variance. The categorical variables are presented as the counts and percentage and were compared using the chi-square test, with
Women (n ⫽ 210)
Men (n ⫽ 623)
p Value
19.8% 0% 8.4%
22.3% 0.6% 7.4%
0.460 0.263 0.630
15.8% 9.8% 19.9% 10.4%
14.1% 12.6% 19.2% 15.2%
0.599 0.319 0.972 0.110
1.94 ⫾ 0.813
2.02 ⫾ 0.813
0.789
1.70 ⫾ 0.755
1.70 ⫾ 0.737
0.530
2.27 ⫾ 1.383
2.42 ⫾ 1.377
0.372
2.30 ⫾ 1.628 32.48 ⫾ 24.07 1.95 ⫾ 1.747
2.39 ⫾ 1.476 34.54 ⫾ 23.60 1.82 ⫾ 1.51
0.678 0.788 0.242
27.17 ⫾ 24.66
24.64 ⫾ 21.89
0.268
0.35 ⫾ 0.948
0.58 ⫾ 1.143
⬍0.001
8.90 ⫾ 14.531
13.87 ⫾ 23.090
0.002
8.1%
12.3%
0.092
2.4%
5.3%
0.082
7.6%
11.0%
0.162
2.0% 15.2%
1.1% 23.7%
0.323 0.010
90.5%
90.3%
0.960
1.81 ⫾ 1.54
1.78 ⫾ 1.64
0.837
the results reported as the hazard ratios and 95% confidence intervals. Statistical significance was set at the 2-tailed 0.05 level. A multivariate-adjusted Cox proportional hazard analysis was performed that included all variables with a statistically significant difference (p ⬍0.05) on univariate analysis Computations were performed using the Statistical Package for Social Sciences, version 11.0 (SPSS, Chicago, Illinois). Results A total of 833 patients were selected, 210 women and 623 men. The clinical, angiographic, and procedural characteristics are summarized in Tables 1 and 2. Although the women were older and had greater rates of the most important cardiovascular risk factors, such as diabetes mellitus and hypertension, the men had more frequently reported a
Coronary Artery Disease/Comparison of Mortality Rates Table 3 Long-term outcomes Variable Follow-up (months) Major adverse cardiac events Death from any cause Noncardiac death Cardiac nonischemic death Vascular noncardiac death Nonsudden ischemic death Sudden death Myocardial infarction Repeat percutaneous transluminal coronary angioplasty/target vessel revascularization Repeat percutaneous transluminal coronary angioplasty/nontarget vessel revascularization Coronary artery bypass grafting Stroke Defined stent thrombosis Defined and probable stent thrombosis Defined and probable and possible stent thrombosis
Women (n ⫽ 210)
Men (n ⫽ 623)
p Value
53 ⫾ 16 42.3% 21.2% 23.3% 4.7% 11.6% 16.3% 7.0% 7.9% 19.2%
58 ⫾ 17 48.8% 15.4% 15.3% 3.8% 5.5% 6.0% 9.8% 9.0% 17.6%
0.08 0.090 0.090 0.0497 0.2395 0.3544 0.2774 0.2020 0.952 0.652
12.6%
14.1%
0.626
4.0% 4.0% 0.65% 1.3%
2.7% 3.1% 1.4% 1.8%
0.435 0.603 0.4831 0.4352
1.3%
2.0%
0.921
Table 4 Interaction between admission diagnosis and long-term outcomes Variable Stable angina/silent ischemia Death from any cause Myocardial infarction Stroke Defined and probable and possible stent thrombosis Unstable angina/non–ST-segment elevation myocardial infarction Death from any cause Myocardial infarction Stroke Defined and probable and possible stent thrombosis ST-segment elevation myocardial infarction Death from any cause Myocardial infarction Stroke Defined and probable and possible stent thrombosis
Women
Men
p Value
37 22.2% 0% 0% 0%
134 16.0% 5.8% 4.2% 0.8%
0.508 0.293 0.376 0.696
155 20.5% 9.5% 4.8% 1.2% 59 20.0% 5.0% 5.0% 2.5%
307 16.6% 10.5% 3.7% 2.8%
0.421 0.794 0.639 0.397
181 8.1% 5.4% 1.3% 1.3%
0.029 0.926 0.154 0.603
previous surgical or percutaneous revascularization. Also, the admission diagnosis showed significant differences, with greater rates of unstable coronary disease in the women, with the men more often referred to the hemodynamic laboratory for silent ischemia and stable angina. During coronary angiography and angioplasty, chronic total occlusion and small vessels were more frequently found in the men, and the number and length of the drugeluting stents were inferior in the women. Moreover, the women were less likely to receive intravenous glycoprotein IIb/IIIa inhibitors.
653
The long-term outcomes are reported in Table 3. The rates of both MACE and the single components of MACE and the other secondary outcomes did not differ between the 2 groups. As reported in Table 4, the outcomes of patients stratified by their admission diagnosis were not significantly different, apart from those presenting with a diagnosis of STEMI. The rate of death for those presenting with STEMI was significantly greater for the women, and this difference persisted after adjustment for confounders. Discussion The most important findings of our long-term observational study were first that the men were more likely to report a previous cardiac event and women presented with greater rates of cardiovascular risk factors. Second, the long-term outcomes between the female and male patients were not statistically significant. Finally, great attention should be given to women presenting with a diagnosis of STEMI. In our registry, the women had more risk factors than did the men. In contrast, the men had had a greater number of previous cardiologic events. Also, as reported by most of the studies investigating gender differences,1–7 the women, as a group, were older with more co-morbidities than the men, including hypertension and diabetes. In contrast, the male patients more frequently had a history of PCI or coronary artery bypass grafting. Moreover, women have generally developed coronary artery disease 4 to 10 years later than men, probably because of the protective role of endogenous estrogen.7,8 In our patients, the long-term outcomes showed similar rates of events between the men and women undergoing PCI, despite the baseline differences. The association between gender and mortality among the patients with cardiovascular disease has been a major topic of study in the past several decades. Despite the increased attention, this relation is poorly understood, and we found contrasting results in the published data. For example, in the studies by Berger et al5 and Singh et al,9 the mortality rates were similar. However, in the recent report by Movahed et al,10 greater rates of mortality were demonstrated in the women. These different findings could have been related to the different years of selection and follow-up length. Another important finding was the greater rates of mortality for women presenting with a diagnosis of STEMI. This worse outcome could have been related to the observations that the female patients more often presented to emergency department with atypical symptoms and, thus, longer door-to-needle times resulted compared to those with men.11 Moreover, a referral bias for women involving both the clinical presentation and the initial management has been reported, the so-called Yentl syndrome.12 In our study also, as reported by other large European registries,13,14 women were less likely to undergo reperfusion or to receive adjunctive therapies. Thus, it could be argued that the low use of drug-eluting stents and anti-glycoprotein IIb/IIIa agents in the female patients contributed to their worse outcomes. The results of the present study have suggested that greater attention to the treatment and control of risk factors
654
The American Journal of Cardiology (www.ajconline.org)
could improve the outcomes in women. Moreover, a physician and public awareness program could be important for prompt diagnosis and appropriate unbiased treatment.2 The main limitation of our study was the observational design, because, despite our use of appropriate statistical adjustments, differences in the baseline characteristics or the selection criteria that might not have been recorded, could remain. The nonrandomized design did not allow us to obtain a definite answer to the gender differences in coronary artery disease. Thus, our study was only hypothesis generating. The smaller number of patients with drugeluting stents was another potential confounder, together with the lack of information about the baseline features. 1. Chang WC, Kaul P, Westerhout CM, Graham MM, Fu Y, Chowdhury T, Armstrong PW. Impact of sex on long-term mortality from acute myocardial infarction vs unstable angina. Arch Intern Med 2003;163: 2476 –2484. 2. El-Menyar A, Zubaid M, Rashed W, Almahmeed W, Al-Lawati J, Sulaiman K. Comparison of men and women with acute coronary syndrome in six Middle Eastern countries. Am J Cardiol 2009;104: 1018 –1022. 3. Berger JS, Elliott L, Gallup D, Roe M, Granger CB, Armstrong PW. Sex differences in mortality following acute coronary syndromes. JAMA 2009;302:874 – 882. 4. Akhter N, Milford-Beland S, Roe MT, Piana RN, Kao J, Shroff A. Gender differences among patients with acute coronary syndromes undergoing percutaneous coronary intervention in the American College of Cardiology-National Cardiovascular Data Registry (ACCNCDR). Am Heart J 2009;157:141–148. 5. Berger JS, Sanborn TA, Sherman W, Brown DL. Influence of sex on in-hospital outcomes and long-term survival after contemporary percutaneous coronary intervention. Am Heart J 2006;151:1026 –1031. 6. Ortolani P, Solinas E, Guastaroba P, Casella G, Manari A, Piovaccari G, Balducelli M, Tondi S, Percoco G, Tarantino F, Passerini F, Rossi
7.
8.
9.
10.
11.
12. 13.
14.
R, Vignali L, De Palma R, Grilli R, Marzocchi A. Long-term clinical outcomes after drug eluting stent implantation in women with de novo coronary lesions: results from the REAL (REgistro Regionale AngiopLastiche Emilia-Romagna) multicenter registry. Int J Cardiol Epub 2010 Jan 6. Mehilli J, Kastrati A, Dirschinger J, Bollwein H, Neumann FJ, Schömig A. Differences in prognostic factors and outcomes between women and men undergoing coronary artery stenting. JAMA 2000; 284:1799 –1805. Williams JK, Adams MR, Klopfenstein HS. Estrogen modulates responses of atherosclerotic coronary arteries. Circulation 1990;81: 1680 –1687. Singh M, Rihal CS, Gersh BJ, Roger VL, Bell MR, Lennon RJ, Lerman A, Holmes DR Jr. Mortality differences between men and women after percutaneous coronary interventions: a 25-year, singlecenter experience. J Am Coll Cardiol 2008;51:2313–2320. Movahed MR, Hashemzadeh M, Jamal MM, Ramaraj R. Decreasing in-hospital mortality of patients undergoing percutaneous coronary intervention with persistent higher mortality rates in women and minorities in the United States. J Invasive Cardiol 2010;22:58 – 60. Onuma Y, Kukreja N, Daemen J, Garcia-Garcia HM, Gonzalo N, Cheng JM, van Twisk PH, van Domburg R, Serruys PW; Interventional Cardiologists of Thoraxcenter. Impact of sex on 3-year outcome after percutaneous coronary intervention using bare-metal and drugeluting stents in previously untreated coronary artery disease insights from the RESEARCH (Rapamycin-Eluting Stent Evaluated at Rotterdam Cardiology Hospital) and T-SEARCH (Taxus-Stent Evaluated at Rotterdam Cardiology Hospital) registries. J Am Coll Cardiol Interv 2009;2:603– 610. Healy B. The Yentl syndrome. N Engl J Med 1991;325:274 –276. Mehilli J, Kastrati A, Dirschinger J, Bollwein H, Neumann FJ, Schömig A. Differences in prognostic factors and outcomes between women and men undergoing coronary artery stenting. JAMA 2000; 284:1799 –1805. Williams JK, Adams MR, Klopfenstein HS. Estrogen modulates responses of atherosclerotic coronary arteries. Circulation 1990;81:1680 – 1687.
Global Variability in Angina Pectoris and Its Association With Body Mass Index and Poverty Longjian Liu, MD, MSc, PhDa,*, Jixiang Ma, PhDb, Xiaoyan Yin, MDc, Ellie Kelepouris, MDd, and Howard J. Eisen, MDd In the absence of a previous global comparison, we examined the variability in the prevalence of angina across 52 countries and its association with body weight and the poverty index using data from the World Health Organization-World Health Survey. The participants with angina were defined as those who had positive results using a Rose angina questionnaire and/or self-report of a physician diagnosis of angina. The body mass index (BMI) was determined as the weight in kilograms divided by the square of the height in meters. The poverty index (a standard score of socioeconomic status for a given country) was extracted from the United Nations’ statistics. The associations of angina with the BMI and poverty index were analyzed cross-sectionally using univariate and multivariate analyses. The results showed that the total participants (n ⴝ 210,787) had an average age of 40.64 years. The prevalence of angina ranged from 2.44% in Tunisia to 23.89% in Chad. Those participants with a BMI of <18.5 kg/m2 (underweight), 25 to 29 kg/m2 (overweight), or BMI >30 kg/m2 (obese) had a significantly greater risk of having angina compared to those with a normal BMI (>18.5 but <25 k/m2). The odds ratios of overweight and obese for angina remained significant in the multilevel models, in which the influence of the countrylevel poverty status was considered. A tendency was seen for underweight status and a poverty index >14.65% to be associated with the risk of having angina, although these associations were not statistically significant in the multilevel models. In conclusion, significant variations were found in the anginal rates across 52 countries worldwide. An increased BMI was significantly associated with the odds of having angina. Published by Elsevier Inc. (Am J Cardiol 2011;107:655– 661)
Evidence of the global variability in angina pectoris (AP) rates has been limited. In the present study, we examined the prevalence of AP across 52 countries using data from the World Health Organization-World Health Survey (WHS) and to examine the associations between the body mass index (BMI) and odds of having AP.1–3 However, because huge differences exist in the socioeconomic development among different countries, we applied multilevel analysis techniques to test the BMI–AP association (individual level), taking into account the influence of the country-level socioeconomic status (SES), assessed using the poverty index, on the AP rates. The advantage of using a multilevel analysis is
a Department of Epidemiology and Biostatistics, Drexel University School of Public Health, Philadelphia, Pennsylvania; bInstitution of NonCommunicable Disease Prevention, Shandong Center for Disease Control and Prevention, Jinan, China; cDepartment of Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania; dDepartment of Medicine, Drexel University College of Medicine, Philadelphia, Pennsylvania. Manuscript received July 5, 2010; manuscript received and accepted October 11, 2010. The reviews expressed in the present study are the authors’ and do not necessarily reflect those of the WHO. Dr. Liu conceived and contributed the analytic study design, conducted the analysis, and drafted and finalized the paper; Drs. Ma, Yin, Kelepouris, and Eisen gave critical reviews of the analytic design and editing of the report. *Corresponding author: Tel: (215) 762-1370; fax: (215) 762-1174. E-mail address:
[email protected] (L. Liu).
0002-9149/11/$ – see front matter Published by Elsevier Inc. doi:10.1016/j.amjcard.2010.10.040
that it takes the hierarchical structure of the data (e.g., both the individual and the country level effects on health) into account by specifying random effects at each analysis level. Thus, this approach of multilevel modeling will result in a more conservative inference for the associations between the predictors and outcomes.4 –10 Methods The present analysis was conducted cross-sectionally using individual data from the WHS and country-level data on SES from the United Nations’ health statistics.11,12 The study design and methods of the WHS have been previously described in detail.11,13–16 In brief, the WHS, with a cross-sectional study design, was administered in 70 countries in 2002 to 2003. The countries included in the survey comprised a large proportion of the world’s population and geographically represented the 6 World Health Organization regions. These included high-, middle-, and low-income countries. The target population included adult, men and women, aged ⱖ18 years, living in private households, who were not out of the country during the survey period. In the WHS, randomized national samples, using a multistage cluster design, were obtained in all countries, except for the 6 countries of China, Comoros, the Republic of Congo, Côte d’Ivoire, India, and the Russian Federation. A Kish table was used to select the respondents within each household.11,12 Face-to-face interviews were conducted by lay people with at least a high school education. The interwww.ajconline.org
656
The American Journal of Cardiology (www.ajconline.org)
Table 1 Prevalence of angina pectoris from the World Health Survey (WHS) (2002–2003) Country Bangladesh Bosnia Herzegovina Brazil Burkina Faso Chad China Comoros The Congo Côte d’Ivoire Croatia Czech Republic Dominican Republic Ecuador Estonia Ethiopia Georgia Ghana Guatemala Hungary India Kazakhstan Kenya Lao People’s Democratic Republic Latvia Mali Malawi Malaysia Mauritania Mauritius Morocco Myanmar Namibia Nepal Pakistan Paraguay Philippines Russian Federation Senegal Slovakia Slovenia South Africa Spain Sri Lanka Swaziland Tunisia Turkey Ukraine United Arab Emirates Uruguay Viet Nam Zambia Zimbabwe
Initial Sample Size
Final Sample Size (%)
Mean Age ⫾ SEM
Crude Rate (SEP)
Age-Adjusted Rate (SEP)
5,942 1,031 5,000 4,948 4,875 3,994 1,836 3,077 3,251 993 949 5,027 5,677 1,021 5,090 2,950 4,165 4,890 1,419 10,692 4,499 4,640 4,989 929 5,209 5,551 6,145 3,907 3,968 5,000 6,045 4,379 8,822 6,502 5,288 10,083 4,427 3,465 687 2,535 2,629 6,373 6,805 3,121 5,203 11,481 1,183 2,860 2,996 4,174 4,166 4,292
5,509 (92.71) 1,012 (98.16) 4,992 (99.84) 4,798 (96.97) 4,496 (92.23) 3,990 (99.90) 1,752 (95.42) 2,133 (69.32) 3,089 (95.02) 979 (98.59) 915 (96.42) 4,497 (89.46) 4,403 (77.56) 1,006 (98.53) 4,651 (91.38) 2,731 (92.58) 3,898 (93.59) 4,689 (95.89) 1,416 (99.79) 9,333 (87.29) 4,484 (99.67) 4,322 (93.15) 4,815 (96.51) 853 (91.82) 3,481 (66.83) 4,582 (82.54) 5,999 (97.62) 3,620 (92.65) 3,880 (97.78) 4,471 (89.42) 5,886 (97.37) 3,952 (90.25) 8,606 (97.55) 6,033 (92.79) 5,106 (96.56) 9,913 (98.31) 4,360 (98.49) 2,699 (77.89) 579 (84.28) 1,783 (70.34) 2,310 (87.87) 6,258 (98.20) 6,509 (95.65) 2,013 (64.50) 4,991 (95.93) 11,184 (97.41) 1,152 (97.38) 2,474 (86.50) 2,969 (99.10) 3,485 (83.49) 3,791 (91.00) 3,938 (91.75)
38.83 ⫾ 0.30 43.50 ⫾ 0.58 41.06 ⫾ 0.34 36.81 ⫾ 0.42 36.45 ⫾ 0.33 45.07 ⫾ 0.76 40.74 ⫾ 0.53 35.32 ⫾ 0.57 34.67 ⫾ 0.42 49.46 ⫾ 0.67 46.48 ⫾ 1.17 39.99 ⫾ 0.41 40.65 ⫾ 0.42 49.71 ⫾ 0.46 35.58 ⫾ 0.31 45.87 ⫾ 0.50 40.05 ⫾ 0.40 39.99 ⫾ 0.24 49.78 ⫾ 0.57 38.35 ⫾ 0.32 42.56 ⫾ 1.17 35.34 ⫾ 0.44 38.21 ⫾ 0.28 50.48 ⫾ 0.92 41.93 ⫾ 0.39 35.30 ⫾ 0.48 39.87 ⫾ 0.37 37.45 ⫾ 0.39 41.18 ⫾ 0.35 40.08 ⫾ 0.55 40.62 ⫾ 0.29 37.34 ⫾ 0.48 38.79 ⫾ 0.24 37.56 ⫾ 0.79 38.95 ⫾ 0.30 38.87 ⫾ 0.25 51.44 ⫾ 0.70 37.93 ⫾ 0.41 47.25 ⫾ 0.75 34.34 ⫾ 0.88 37.56 ⫾ 0.48 50.35 ⫾ 0.34 41.40 ⫾ 0.47 38.88 ⫾ 0.64 40.11 ⫾ 0.32 41.27 ⫾ 0.25 36.68 ⫾ 0.48 46.30 ⫾ 0.63 44.50 ⫾ 0.53 39.19 ⫾ 0.48 35.41 ⫾ 0.00 37.04 ⫾ 0.68
8.57 (0.73) 7.32 (0.88) 7.83 (0.45) 14.41 (1.12) 22.79 (1.75) 3.65 (0.71) 5.49 (0.69) 9.66 (1.58) 8.57 (0.69) 8.78 (1.08) 8.20 (1.35) 4.28 (0.49) 5.28 (0.65) 18.97 (1.28) 14.81 (1.30) 15.69 (1.07) 5.77 (0.45) 4.91 (0.32) 18.90 (1.24) 9.12 (0.85) 10.67 (1.56) 3.29 (0.40) 5.16 (0.40) 23.81 (1.77) 10.27 (0.78) 14.38 (1.24) 4.35 (0.33) 9.31 (0.79) 4.97 (0.45) 5.84 (0.63) 3.02 (0.48) 8.10 (0.69) 6.10 (0.37) 3.19 (0.40) 6.60 (0.40) 6.40 (0.41) 30.04 (1.79) 9.88 (0.83) 20.03 (1.66) 4.20 (0.81) 5.70 (0.65) 5.14 (0.40) 4.15 (0.91) 12.88 (2.09) 2.16 (0.24) 6.38 (0.36) 3.70 (0.67) 21.31 (1.72) 7.58 (0.82) 4.94 (0.60) 3.27 (0.00) 6.94 (0.69)
9.14 (0.42) 7.49 (0.73) 7.77 (0.39) 14.99 (0.57) 23.89 (0.74) 3.31 (0.27) 5.62 (0.56) 11.34 (0.73) 8.67 (0.55) 9.94 (0.99) 6.78 (0.89) 3.84 (0.29) 4.88 (0.33) 14.15 (1.01) 15.82 (0.60) 13.32 (0.60) 6.05 (0.39) 5.00 (0.33) 13.63 (0.84) 11.03 (0.34) 12.62 (0.57) 4.52 (0.33) 5.74 (0.35) 15.73 (1.09) 9.23 (0.51) 13.19 (0.56) 4.92 (0.29) 9.34 (0.50) 4.99 (0.36) 6.39 (0.36) 3.03 (0.22) 9.20 (0.51) 6.06 (0.27) 3.75 (0.26) 7.42 (0.39) 6.62 (0.27) 17.66 (0.51) 9.35 (0.58) 18.72 (1.74) 7.56 (0.74) 8.10 (0.64) 3.26 (0.18) 4.57 (0.26) 9.83 (0.69) 2.44 (0.22) 6.46 (0.24) 5.04 (0.83) 16.33 (0.66) 7.21 (0.47) 4.78 (0.38) 4.04 (0.34) 7.63 (0.45)
SEP, standard error of proportion.
viewers had attended a weeklong training course using a standard protocol manual and participated in field interviews that were reviewed by the supervisors before actual data collection. In addition to the standard questions asked of all participants regarding social demographic variables, general health status, and chronic health condition, each country was able to select additional questionnaire modules
of specific research interest. Of the 70 countries, 52 countries administered the dietary behavior module. These countries represented 6 regions, with 18 countries from the African region, 6 from the Americas, 4 from the Eastern Mediterranean region, 14 from the European region, 5 from the South-East Asia region, and 5 from the Western Pacific region. The present analysis included data from these 52
Coronary Artery Disease/Global Variability in Angina Pectoris
657
Figure 1. Angina rate (%) in 52 countries, from WHS (2002 to 2003).
Table 2 Human poverty index (HPI) in 52 countries by regions World Health Organization region
African region Region of the Americas Eastern Mediterranean region European region South-East Asia region Western Pacific region
n
18 6 4 14 5 5
HPI (%) Mean ⫾ SEM
Range
34.03 ⫾ 3.02 9.80 ⫾ 2.24 19.05 ⫾ 7.67 5.03 ⫾ 1.10 26.68 ⫾ 3.58 13.86 ⫾ 4.39
9.50–54.50 3.00–19.70 4.00–33.40 0.20–12.50 16.80–36.10 6.10–30.70
Regions defined according to World Health Organization criteria.
countries, because they had completed data on the BMI and dietary habits. In the WHS, all questionnaires were translated and backtranslated using a standard World Health Organization protocol. The quality of translations was independently verified by bilingual experts before field implementation. All respondents provided informed consent, and the study was cleared by the ethics review committees at each site.11 Five chronic conditions were included in the WHS and defined according to the participants’ self-report medical history of a physician diagnosis, which included questions related to whether the respondent had ever been diagnosed with AP, diabetes mellitus, depression, asthma, or arthritis, and whether the respondent had ever received, or was currently receiving, therapy for these conditions. The classification of AP was also examined and defined according to the World Health Organization Rose questionnaires. The questionnaire is a validated instrument to assess the symptoms of typical AP in the general population. It is highly specific compared to physician-diagnosed AP and has been
Figure 2. Human poverty index (%) and angina rate (%) in 52 countries, from WHS (2002 to 2003).
associated with coronary artery calcification and the risk of coronary events.2,3,17–22 To examine whether cross-country difference in SES played a role in the associations between BMI and AP, we applied a multilevel analysis technique using country level data on the human poverty index (HPI). The HPI is an indication of the standard of living in a country that was developed by the United Nations. The calculation of the HPI has been previously described in detail by the United Nations.12 In brief, the HPI measures the deprivations in the 3 basic dimensions of human development: (1) early life expectancy, as measured by the probability at birth of not surviving to age 40 years; (2) adult illiteracy rate; and (3) the standard of living defined as the lack of access in overall economic provisioning, as measured by the percentage of the population not using an improved water source and the percentage of children who are underweight for age. A sum
658
The American Journal of Cardiology (www.ajconline.org)
Table 3 Age and other predictors and chronic conditions by body mass index (BMI), from World Health Survey (WHS) (2002–2003) Individual Level
Mean age (years) Age ⱖ50 years Men (%) Married (%) Educational level No formal education Elementary Middle school High school or more Current smoker (%) Vegetable/fruit intake ⱖ5 servings/day Medical conditions (%) Angina pectoris Diabetes Arthritis Asthma Depression
Body Mass Index Category*
p Value
Underweight
Normal Weight
Overweight
Obesity
40.02 (0.20) 29.10 (0.45) 42.09 (0.50) 67.32 (0.56)
38.61 (0.16) 24.51 (0.45) 52.70 (0.59) 64.43 (0.60)
45.74 (0.26) 40.07 (0.79) 49.67 (0.74) 69.00 (0.80)
48.98 (0.36) 46.85 (1.14) 36.68 (1.03) 65.79 (1.32)
25.59 (0.69) 30.69 (0.56) 33.48 (0.61) 10.25 (0.53) 24.30 (0.62) 2.09 (0.34)
17.22 (0.89) 24.32 (0.52) 40.59 (0.58) 17.87 (0.53) 25.33 (0.73) 2.53 (0.41)
10.54 (0.73) 24.99 (0.71) 41.14 (0.75) 23.33 (0.79) 20.77 (0.59) 1.81 (0.19)
10.40 (0.81) 26.53 (0.99) 41.40 (1.18) 21.67 (1.21) 14.61 (0.79) 2.37 (0.44)
9.61 (0.39) 2.70 (0.16) 16.79 (0.44) 5.75 (0.24) 7.99 (0.38)
7.32 (0.30) 2.44 (0.14) 13.48 (0.44) 5.52 (0.22) 8.57 (0.38)
12.34 (0.57) 6.12 (0.35) 17.44 (0.51) 6.87 (0.47) 10.23 (0.52)
17.63 (0.95) 11.47 (1.07) 25.49 (1.00) 8.02 (0.52) 11.60 (0.70)
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.134 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
Data are presented as the rate (standard error of proportion). * Underweight, BMI ⬍18.5 kg/m2; normal weight, BMI 18.5–24 kg/m2; overweight, BMI 25–29 kg/m2; and obesity, BMI ⱖ30 kg/m2.
score is added to produce the HPI. Greater HPI levels indicate poorer SES status. In the present study, the HPI represented each country’s SES for 1998 to 2002.12 To test the research questions, several univariate and multivariate analyses were conducted. In the first group analysis, the overall and age-specific prevalence of AP for each country were calculated. Next, the age-adjusted prevalence of AP was estimated using the 1999 World Health Organization’s World Standard Population.23 The standard error of the prevalence was estimated using Keyfitz’s method.24 The mean, standard error of the mean, and range of HPI have been described according to the 6 World Health Organization regions. In the second group analysis, differences in age, gender, marital status, education, smoking status, vegetable and/or fruit intake, and participants’ self-reported history of a physician diagnosis of the chronic conditions of AP, diabetes mellitus, arthritis, asthma, and depression were stratified by 4 BMI groups. The BMI groups were defined according to the World Health Organization criteria of underweight (BMI ⬍18.5 kg/ m2), normal weight (BMI 18.5 to 24 kg/m2), overweight (BMI 25 to 29 kg/m2), and obesity (BMI ⱖ30 kg/m2). In the third group analysis, multiple logistic regression models were used to examine the associations of AP with BMI and HPI separately.25 Odds ratios and 95% confidence intervals (CIs) of the BMI and HPI for the prevalence of AP were estimated, with adjustments for age and gender. The participants with a normal BMI were the reference group. We dichotomized HPI (⬍14.65% vs ⱖ14.65%) according to the 50% outpoint on the distribution of HPI across the 52 countries, because the distribution of this variable was quite wide. In the fourth group analysis, multilevel analysis techniques were applied to account for the influence of HPI on the association between BMI and AP.6,7 In these analyses, generalized linear mixed models were conducted using the SAS GLIMMIX procedure.6,7 Three sets of models were
tested. Model 1 was adjusted for age (per 10-year increase) and gender (male vs female). Model 2 was adjusted for age, gender, marital status (married vs unmarried), educational level (no formal education, elementary, middle school, vs high school or more), current smoking status (yes vs no), and vegetable and/or fruit intake (ⱖ5 vs ⬍5 servings/day). Finally, model 3 was adjusted for all covariates in model 2 plus the participants’ self-reported chronic condition of a physician diagnosis of diabetes mellitus, asthma, arthritis, or depression. The interaction effects of BMI with age, gender, and HPI on the odds of AP were tested. Finally, sensitivity analyses were conducted to test the robustness of the models. In particular, we wanted to assess the effects of the country level variable HPI on the association between BMI and AP. We reran our models with exclusions of the extreme values (lowest and highest) of HPI. Because little change occurred in the overall results, we have presented our findings using the total sample size. To account for the complex sampling design of the WHS, we used SAS software, version 9.1 (procedures for complex survey data) to adjust for sampling weights (SAS Institute, Cary, North Carolina).26 The level of significance was set at a 2-sided test at p ⱕ0.05. Results A total of 210,787 participants were included, with an average age of 40.64 years. The age range was 34.34 years (Slovenia) to 51.44 years (Russian) in the 52 countries of the WHS. The response rates ranged from 64.5% (Swaziland) to 99.9% (China). Table 1 lists the age-adjusted prevalence of AP, which ranged from 2.44% in Tunisia to 23.89% in Chad. Figure 1 depicts the significant variability in the prevalence of AP across the 52 countries. The data in Table 2 show that the African region had the greatest HPI (34.03), followed by the South-East Asia region, Eastern Mediterranean region, and Western Pa-
Coronary Artery Disease/Global Variability in Angina Pectoris
659
Table 4 Multivariate analyses of associations of age, poverty index, and body mass index with prevalent angina pectoris: findings from the World Health Survey, 2002 to 2003 Variable
Single-Level Models OR (95% CI)
Country level Age (per 10 years) Gender (female vs male) Poverty index ⱖ14.65% Individual level Age (per 10 years) Gender (female vs. male) Body mass index (kg/m2) ⬍18.5 18.5–24 25–29 ⱖ30 Interaction terms Age* body mass index Gender* body mass index Poverty index* body mass index
Multilevel Models
OR (95% CI)
Model 1 (OR [95% CI])
Model 2 (OR [95% CI])
Model 3 (OR [95% CI])
1.29 (0.91–1.81)
1.25 (0.89–1.77)
1.31 (0.92–1.87)
1.37 (1.36–1.39)* 1.44 (1.40–1.49)*
1.39 (1.38–1.40)* 1.46 (1.41–1.51)*
1.38 (1.37–1.40)* 1.42 (1.38–1.48)*
1.26 (1.25–1.28)* 1.25 (1.21–1.30)*
1.20 (1.16–1.25)* 1 1.27 (1.22–1.33)* 1.67 (1.57–1.76)*
1.01 (0.97–1.06) 1 1.21 (1.16–1.27)* 1.56 (1.47–1.65)*
1.01 (0.96–1.05) 1 1.21 (1.15–1.27)* 1.55 (1.46–1.64)*
1.02 (0.97–1.08) 1 1.17 (1.11–1.22)* 1.35 (1.26–1.43)*
1.21 (1.08–1.35)* 1.06 (0.94–1.20) 0.77 (0.68–0.85)*
1.20 (1.08–1.35)* 1.07 (0.94–1.20) 0.77 (0.69–0.87)*
1.08 (0.97–1.22) 1.00 (0.88–1.13) 0.83 (0.73–0.93)*
1.39 (1.38–1.41)* 1.48 (1.43–1.53)* 1.14 (1.10–1.18)*
Single-level model, adjusted for age and gender; odds ratios estimated using multivariate logistic regression models. Multilevel models conducted using generalized linear mixed model (GLIMM). Model 1, adjusted for age and gender; model 2, adjusted for age, gender, marital status, smoking status, and fruit and/or vegetable intake; model 3, adjusted for all covariates in model 2, and medical conditions of diabetes, asthma, arthritis, and depression. * p ⬍0.01. OR, odds ratio.
cific region, region of the Americas (9.80), and the European region. Figure 2 shows the tendency of a U-shaped association between the HPI and AP rates across the 52 countries. The differences in HPI explained 26% of the variance in the prevalence of AP worldwide (R2 ⫽ 0.26). The data in Table 3 show that the participants with a BMI ⱖ30 kg/m2 had the greatest mean age, followed by those with a BMI of 25 to 29, ⬍18.5, and 18.5 to 24 kg/m2 (p ⬍0.001). Significant differences in the proportions of gender, marital status, educational level, and smoking status were also observed in the different BMI groups. The participants with a BMI of ⱖ30 kg/m2 had the greatest AP prevalence (17.63%), followed by those with a BMI of 25 to 29 (12.34%), ⬍18.5 (9.61%), and 18.5 to 24 kg/m2 (7.23%; p ⬍0.001). They also had the greatest rates of diabetes mellitus, arthritis, asthma, and depression. In the analysis of the single-level logistic regression models, countries with an HPI of ⱖ14.65% (SES at the median or poorer level among the 52 countries) had a high risk of having AP compared to those with an HPI of ⬍14.65% (odds ratio 1.14; Table 4). The participants who were underweight, overweight, or obese had significantly greater odds of AP compared to those with a normal BMI. The corresponding odds ratios were 1.20 (95% CI 1.16 to 1.25), 1.27 (95% CI 1.22 to 1.33), and 1.67 (95% CI 1.57 to 1.76). In the multilevel models, which took the hierarchical structure of the data into consideration, a tendency was seen for an HPI of ⱖ14.65% to be associated with greater odds of AP (Table 4), although these odds ratios and 95% CIs were not statistically significant in models 1, 2, or 3. The
odds ratios for the overweight and obese categories for AP remained significant. A significantly interactive effect was seen between HPI and BMI on the presence of AP (model 3, p ⬍0.01; Table 4). Therefore, to test the degrees of the associations between BMI and AP, as affected by HPI, we repeated the multilevel analysis, stratifying for HPI (⬍14.65% vs ⱖ14.65%). The results showed that the BMI–AP associations were greater in those with an HPI ⬍14.65% than in those with an HPI of ⱖ14.65%. Among those with an HPI ⬍14.65%, the odds ratio of the underweight, overweight, and obese categories for AP was 1.03 (95% CI 0.96 to 1.11), 1.28 (95% CI 1.20 to 1.36), and 1.71 (95% CI 1.58 to 1.85), respectively, compared to the normal BMI category The corresponding odds ratios for those with an HPI ⱖ14.65% were 0.98 (95% CI 0.93 to 1.04), 1.11 (95% CI 1.03 to 1.19), and 1.33 (95% CI 1.21 to 1.46). Discussion The present study has extended the previous studies using multiple data sources from the World Health Organization and the United Nations and examined the complex associations of BMI, SES, and AP. The present study was the first to describe the significant variability in the prevalence of AP across 52 countries worldwide. These findings indicate that increased body weight (i.e., overweight and obesity) is significantly associated with an increased relative risk of developing AP in countries with either greater or lower SES status. AP is a major concern, because it is causes disability and is a marker for the potentially severe manifestations of
660
The American Journal of Cardiology (www.ajconline.org)
coronary heart disease. In the WHS, the classification of AP included self-report of a physician diagnosis of AP and undiagnosed AP (assessed using the Rose questionnaire). This classification approach has enhanced the evaluation of the burden of AP.2,3,17,22 The results from the present study have indicated that eastern European and southeast Asia people have a much greater prevalence of AP than the other regions in the WHS samples. The increasing rates of overweight and obesity pose a serious public health problem worldwide, not only because of the association with the risks of cardiovascular disease, diabetes mellitus, and several forms of cancer, but also because of the association with human poverty status.27–32 In developed countries, such as in the United States, data have suggested that poverty is associated with greater obesity rates. In contrast, in many developing countries, greater rates of obesity have been found in the higher income groups because of economic growth and improved standards of living. One explanation for these observations has been that the low-income groups in the United States and the high-income groups in the developing countries are better able to afford, or have greater access to, energydense, but nutrient-poor, foods.10,12,27,33 The different phenotypes of overweight and obesity by poverty status have challenged the current strategies of cardiovascular disease and diabetes mellitus prevention. Our present analysis has confirmed the BMI–AP association by SES status. The present study had several limitations that should be remembered when interpreting the results. First, the analysis was of data from the WHS and the United Nations’ HPI reports, which were conducted cross-sectionally. Therefore, the associations among BMI, AP, and HPI are not necessities of a causal association. Second, the WHS aimed to describe an overall burden of chronic conditions. Information on detailed diagnoses of the different forms of cardiovascular disease was not available. Thus, we were unable to describe the distributions of the different forms of cardiovascular disease. Finally, in the multilevel analysis, the country level indicator of HPI was the only contextual factor and was a relatively broad conceptual variable. The nonsignificant associations between HPI and AP observed might have resulted from the nonlinear relation between HPI and AP (Figure 2). Although detailed comparable data of the contextual factors would be difficult to obtain, additional cross-national studies are needed. The strengths of the present study included that our results have added new evidence of the AP prevalence for such a large diversity of countries. Second, our results, using combined individual-level and country-level data, have extended the application of the WHS. This provided an opportunity to examine the associations between BMI and AP using a multilevel modeling technique. The multilevel (or hierarchical) models have the advantage of allowing the analysis of nonindependent (i.e., clustered) data that arise when studying topics such as the participants’ SES nested within counties, regions, or countries. Traditional multivariate analysis models are not well-suited for the analysis of these types of clustered data, given the violation of the assumption of independence.7,9,10 Third, the present study focused on the most modifiable risk factors at the individual level (BMI) and country level (HPI). The results have
enhanced the importance of the control of risk factors and the prevention of disease through a multilevel effort. Finally, the study addressed the global issues of increased body weight in relation to AP across nations with different levels of SES. Acknowledgment: The present study used data from the World Health Organization-World Health Surveys (WHS)/ Multi-Country Survey Study, and the United Nations. The authors appreciate the World Health Organization WHS Consortium and the many participants from each country worldwide. The authors thank Nancy K. Smith, MS, Judi Wong, MPH, and Anjana Patel, MPH, for their careful proofreading and editing of the report. 1. Gibbons RJ, Abrams J, Chatterjee K, Daley J, Deedwania PC, Douglas JS, Ferguson TB Jr, Fihn SD, Fraker TD Jr, Gardin JM, O’Rourke RA, Pasternak RC, Williams SV, American College of Cardiology, American Heart Association Task Force on Practice Guidelines (Committee on the Management of Patients With Chronic Stable Angina). ACC/ AHA 2002 guideline update for the management of patients with chronic stable angina—summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on the Management of Patients With Chronic Stable Angina). J Am Coll Cardiol 2003;41:159 –168. 2. Lam TH, Liu LJ, Janus ED, Lau CP, Hedley AJ. Fibrinogen, angina and coronary heart disease in a Chinese population. Atherosclerosis 2000;149:443– 449. 3. Cook DG, Shaper AG, MacFarlane PW. Using the WHO (Rose) angina questionnaire in cardiovascular epidemiology. Int J Epidemiol 1989;18:607– 613. 4. Hope MAD, Shannon ED. A comparison of two procedures to fit multi-level data: PROC GLM versus PROC MIXED. Available at: http://www2.sas.com/proceedings/sugi30/200-30.pdf. Accessed March 16, 2010. 5. Flom PL, McMahon JM, Pouget ER. Using PROC NLMIXED and PROC GLMMIX to analyze dyadic data with a dichotomous dependent variable. Available at: http://www2.sas.com/proceedings/forum2007/ 179-2007.pdf. Accessed January 18, 2010. 6. Dai J, Li Z, Rocke D. Hierarchical logistic regression modeling with SAS GLIMMIX. Available at: http://www.lexjansen.com/wuss/2006/ Analytics/ANL-Dai.pdf. Accessed January 18, 2010. 7. Li J, Alterman T, Deddens JA. Analysis of large hierarchical data with multilevel logistic modeling using PROC GLIMMIX. Available at: http://www2.sas.com/proceedings/sugi31/151-31.pdf. Accessed January 16, 2010. 8. Diez-Roux AV, Nieto FJ, Muntaner C, Tyroler HA, Comstock GW, Shahar E, Cooper LS, Watson RL, Szklo M. Neighborhood environments and coronary heart disease: a multilevel analysis. Am J Epidemiol 1997;146:48 – 63. 9. Jia H, Moriarty DG, Kanarek N. County-level social environment determinants of health-related quality of life among US adults: a multilevel analysis. J Community Health 2009;34:430 – 439. 10. Subramanian SV, Chen JT, Rehkopf DH, Waterman PD, Krieger N. Racial disparities in context: a multilevel analysis of neighborhood variations in poverty and excess mortality among black populations in Massachusetts. Am J Public Health 2005;95:260 –265. 11. World Health Organization. World Health Survey 2010. Available at: http://www.who.int/healthinfo/survey/en/index.html. Accessed May 18, 2010. 12. United Nations Data. Human Poverty Index. Available at: http:// data.un.org/DocumentData.aspx?q⫽hdi&id⫽187. Accessed May 6, 2010. 13. Subramanian SV, Huijts T, Avendano M. Self-reported health assessments in the 2002 World Health Survey: how do they correlate with education? Bull World Health Organization 2010;88:131–138. 14. Hall JN, Moore S, Harper SB, Lynch JW. Global variability in fruit and vegetable consumption. Am J Prev Med 2009;36:402– 409.e5.
Coronary Artery Disease/Global Variability in Angina Pectoris 15. Guthold R, Ono T, Strong KL, Chatterji S, Morabia A. Worldwide variability in physical inactivity a 51-country survey. Am J Prev Med 2008;34:486 – 494. 16. Moussavi S, Chatterji S, Verdes E, Tandon A, Patel V, Ustun B. Depression, chronic diseases, and decrements in health: results from the World Health Surveys. Lancet 2007;370:851– 858. 17. Rose G, McCartney P, Reid DD. Self-administration of a questionnaire on chest pain and intermittent claudication. Br J Prev Soc Med 1977;31:42–48. 18. Lampe FC, Walker M, Lennon LT, Whincup PH, Ebrahim S. Validity of a self-reported history of doctor-diagnosed angina. J Clin Epidemiol 1999;52:73– 81. 19. Cosin J, Asin E, Marrugat J, Elosua R, Aros F, de los Reyes M, Castro-Beiras A, Cabades A, Diago JL, Lopez-Bescos L, Vila J; PANES Study Group. Prevalence of angina pectoris in Spain. Eur J Epidemiol 1999;15:323–330. 20. Biloglav Z, Ivankovic D, Campbell H, Rudan I. Performance of WHO Angina Questionnaire in measuring burden of coronary heart disease in human isolate populations. Coll Antropol 2004;28:205–213. 21. Hemingway H, Langenberg C, Damant J, Frost C, Pyorala K, BarrettConnor E. Prevalence of angina in women versus men: a systematic review and meta-analysis of international variations across 31 countries. Circulation 2008;117:1526 –1536. 22. Lam TH, Liu LJ, Janus ED, Lam KSL, Hedley AJ; Hong Kong Cardiovascular Risk Fact. Fibrinogen, other cardiovascular risk factors and diabetes mellitus in Hong Kong: a community with high prevalence of type 2 diabetes mellitus and impaired glucose tolerance. Diabet Med 2000;17:798 – 806. 23. World Health Organization. World standard population. Available at: http://www.who.int/research/en/. Accessed January 6, 2010. 24. Keyfitz N. 3. Sampling variance of standardized mortality rates. Hum Biol 1966;38:309 –317. 25. Tabachnick BG, Fidell LS. Using Multivariate Statistics, 5th ed. Boston: Allyn & Bacon; 2007.
661
26. SAS Institute. SAS/STAT User’s Guide, version 9.1. Cary, North Carolina: SAS Institute; 2005. 27. Drewnowski A, Specter SE. Poverty and obesity: the role of energy density and energy costs. Am J Clin Nutr 2004;79:6 –16. 28. Liu L, Sullivan DH. Relationship between social demographic factors and survival within one year of hospital discharge in a cohort of elderly male patients. J Epidemiol 2003;13:203–210. 29. Liu L, Liu L, Ding Y, Huang Z, He B, Sun S, Zhao G, Zhang H, Miki T, Mizushima S, Ikeda K, Nara Y, Yamori Y. Ethnic and environmental differences in various markers of dietary intake and blood pressure among Chinese Han and three other minority peoples of China: results from the WHO Cardiovascular Diseases and Alimentary Comparison (CARDIAC) study. Hypertens Res 2001;24:315–322. 30. Liu L, Choudhury SR, Okayama A, Hayakawa T, Kita Y, Ueshima H. Changes in body mass index and its relationships to other cardiovascular risk factors among Japanese population: results from the 1980 and 1990 national cardiovascular surveys in Japan. J Epidemiol 1999; 9:163–174. 31. Yamori Y, Liu L, Mizushima S, Ikeda K, Nara Y; CARDIAC Study Group. Male cardiovascular mortality and dietary markers in 25 population samples of 16 countries. J Hypertens 2006;24:1499 –1505. 32. Yamori Y, Liu L, Ikeda K, Miura A, Mizushima S, Miki T, Nara Y; WHO-Cardiovascular Disease and Alimentary Comparison (CARDIAC) Study Group. Distribution of twenty-four hour urinary taurine excretion and association with ischemic heart disease mortality in 24 populations of 16 countries: results from the WHO-CARDIAC Study. Hypertens Res 2001;24:453– 457. 33. Liu L, Ikeda K, Chen M, Yin W, Mizushima S, Miki T, Nara Y, Yamori Y. Obesity, emerging risk in China: trend of increasing prevalence of obesity and its association with hypertensions and hypercholesterolaemia among the Chinese. Clin Exp Pharmacol Physiol 2004; 31(suppl):S8 –S10.
Meta-Analysis of B-Type Natriuretic Peptide’s Ability to Identify Stress Induced Myocardial Ischemia M. Adnan Nadir, MBBSa,*, Miles D. Witham, PhDb, Benjamin R. Szwejkowski, MBChBa, and Allan D. Struthers, MDa Studies in victims of sudden cardiac death and those surviving a cardiac arrest have confirmed that extent of coronary artery disease is similar in those with and without angina, suggesting that it is the presence of myocardial ischemia rather than associated symptoms that determine the prognosis. Experimental models show that hypoxic myocardial tissue results in production of extra B-type natriuretic peptide (BNP), suggesting that BNP could potentially serve as a biomarker of myocardial ischemia. We performed a meta-analysis of the studies that link BNP to inducible myocardial ischemia as indicated by noninvasive stress tests. Values of true positive, false positive, true negative, and false negative were calculated from the reported sensitivity, specificity, disease prevalence, and total number of patients studied. Sixteen studies reporting data on 2,784 patients across 14 study populations were included in the final analysis. Mean age of participants was 55 to 69 years and 55% to 90% were men. Pooled sensitivity and specificity of BNP for detection of stress-induced myocardial ischemia were 71% (95% confidence interval [CI] 68 to 74) and 52% (95% CI 52 to 54), respectively. Pooled diagnostic odds ratio was 3.5 (95% CI 2.46 to 5.04) and summary receiver operating characteristic curve revealed an area under the curve of 0.71 ⴞ 0.02 (mean ⴞ SE). In conclusion, this meta-analysis suggests that an increased BNP level can identify inducible ischemia as detected by standard noninvasive stress tests. This raises the possibility of a whole new role for BNP in the diagnosis and management of myocardial ischemia. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:662– 667) B-type natriuretic peptide (BNP) has been studied extensively in the diagnosis and exclusion of left ventricular (LV) systolic dysfunction (SD).1 The high negative predictive value of BNP in this particular scenario has led to it being used in day-to-day clinical practice as a “rule-out test” to exclude the presence of LVSD. However, BNP also has the potential to be useful in other patient populations that do not have heart failure or LVSD, albeit at much lower levels than those used to identify LVSD.2 Increased BNP values (typically 100 ⱖpg/ml) are used to detect LVSD but BNP values in the lower range (10 to 20 pg/mL) are of independent prognostic value over and above a wide range of echocardiographic abnormalities including LV dysfunction and LV hypertrophy.3 The question that naturally arises is why BNP identifies a poor prognosis in so many populations irrespective of LV function and other echocardiographic abnormalities. One possible answer is that BNP identifies myocardial ischemia itself.4 Studies have shown that transcription and release of BNP are affected by oxygen tension, and hypoxia as a trigger for the release of BNP has been demonstrated in experimental models.5 Thus, it seems that repeated episodes of ischemia, silent or symptomatic, a
Department of Clinical Pharmacology and bAgeing and Health, Centre for Cardiovascular and Lung Biology, Ninewells Hospital and Medical School, University of Dundee, Dundee, United Kingdom. Manuscript received September 6, 2010; revised manuscript received and accepted October 19, 2010. *Corresponding author: Tel: 44-1382-632-180; fax: 44-1382-644-972. E-mail address:
[email protected] (M.A. Nadir). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.043
may stimulate BNP formation and that BNP can be used to identify those patients who have underlying myocardial ischemia.6 Indeed, many small clinical studies have suggested that this is true. To examine this question, we therefore performed a meta-analysis of the studies that link BNP to inducible myocardial ischemia as indicated by noninvasive stress tests. Methods We conducted a meta-analysis based on a prespecified protocol that was devised according to guidelines of the Cochrane Collaboration. Studies were eligible if they compared BNP or N-terminal pro-BNP (NT–pro-BNP) to evidence of inducible ischemia from a noninvasive stress test (stress echocardiography or myocardial perfusion imaging with single-photon emission computed tomography) in patients with known or suspected coronary artery disease (CAD). The other minimum requirement to be included in our meta-analysis was the availability of enough information to complete a 2 ⫻ 2 contingency table. Studies using BNP as a prognostic rather than a diagnostic marker, those testing BNP in acute coronary syndromes, and studies where the reference test was not a noninvasive stress test for myocardial ischemia were excluded. We did not include studies where treadmill electrocardiography alone was used as a reference standard because of its known poor sensitivity and specificity to diagnose inducible myocardial ischemia. On occasions when several published studies reported data from a single patient group, we selected the most complete www.ajconline.org
Coronary Artery Disease/BNP and Stress-Induced Myocardial Ischemia
or most recent version. When a study reported numerous cutoffs, we used the one with best combination of the reported sensitivity and specificity. We searched MEDLINE, EMBASE, the Cochrane Database, and CINAHL from the earliest date in 1966 to September 2009. Gray literature was sought using the Google search engine. In addition, we hand-searched references listed in the original studies and previous review articles to identify other potentially eligible studies. No language restrictions were imposed. Search terms used included “natriuretic peptides,” “BNP,” “amino terminal pro brain natriuretic peptide” and “NT–pro-BNP” combined with “myocardial ischemia/ischemia,” “coronary artery disease,” and “ischemic/ischemic heart disease.” When necessary we also attempted to contact study authors for further information to complete data abstraction from the potentially eligible studies. Two reviewers (M.A.N. and M.D.W.) independently examined titles and abstracts of all potentially relevant studies. Full articles were obtained from all citations meeting prespecified selection criteria. Data were abstracted by 2 reviewers (M.A.N. and B.R.S.) independently using a standard ProForma and discrepancies were resolved by consensus. We abstracted data on study characteristics, methods, and final results from each selected article. Study characteristics included study settings and type of population examined (including age, gender, and reported disease prevalence in study population), year of publication, assay type and cutoff of index test, reference test standards, and source of funding (industry vs nonindustry). Quality was assessed by the QUADAS tool with particular emphasis on blinding procedures during interpretation of index and reference tests for each study. QUADAS is an evidence-based tool used for quality assessment of diagnostic accuracy studies. It consists of 14 items phrased as questions, each of which should be scored a “yes”, “no,” or “unclear,” that examine bias in a study.7 Published data suggest that a study with a score 7 to ⱖ10 is considered high quality. For this analysis we used a minimum score of 10 as an indicator of good quality.7 For all studies included in the review a 2 ⫻ 2 contingency table was constructed. Values of true positives, false positives, true negatives, and false negatives were deduced from reported sensitivity, specificity, disease prevalence, and total number of patients studied. Where the published information was insufficient or unclear to construct a 2 ⫻ 2 table we requested missing data from the respective corresponding author. We used Meta-Disc 1.4 (Clinical Biostatistics, Ramón y Cajal Hospital, Madrid, Spain) to perform all statistical analyses. This is a publically accessible statistical software package designed purposefully for meta-analysis of diagnostic accuracy and has previously been used in several meta-analyses.8 Pooled sensitivities, specificities, positive and negative likelihood ratios, and diagnostic odds ratios were calculated using the DerSimonian-Liard method. We calculated the I2 statistic to assess heterogeneity between studies. An I2 value ⬎25% was considered to represent significant heterogeneity. Although random and fixed effect models were used, owing to significant heterogeneity we reported only results from random effect models. A summary receiver operating characteristic curve9 was constructed from pooled data for studies included in the
663
Figure 1. Study selection flow chart.
final analysis. Subgroup analyses were performed to explore the significant heterogeneity observed. First, we compared studies with BNP to those with NT–pro-BNP. Second, we reanalyzed the data by including results from studies that explicitly excluded patients with decreased ejection fraction or known heart failure. Third, we pooled data separately for studies using BNP levels after stress. Random effects metaregression (inverse variance weights) was used to explore sources of variation. Mean age of study population, disease prevalence, exclusion of heart failure, timing of BNP measurement, and blinding of test results were considered variables. We also investigated whether type of reference test and type (BNP or NT–pro-BNP) and timing (before or after stress) of NP measured influenced the results. Results The initial search strategy identified 832 citations. Examination of titles and abstracts of these citations identified 33 articles for detailed evaluation. After detailed evaluation 20 studies met inclusion criteria (Figure 1). Reasons for exclusion included no reference test or the reference test not being a stress test. Where not enough information was available to complete 2 ⫻ 2 tables, a request was made to the corresponding study author to provide the missing data. We were unable to obtain further data on 4 studies and hence could not include these in the final analysis. Sixteen studies reporting data on 2,784 patients across 14 study populations were included in the final analysis. Table 1 presents key characteristics of the included studies. All studies reported a prospective design and all studies except 1 declared blind interpretation of test results.10 All but 4 studies11–14 reported consecutive enrollment. The reference test was applied equally in all study participants. In 11 studies11,13–22 myocardial perfusion imaging with singlephoton emission computed tomography (using exercise or dipyridamole as the stressor) was the reference test to confirm stress-induced ischemia, whereas in 3 studies10,12,23 stress echocardiography (using exercise or dobutamine as the stressor) provided the reference standard. The QUADAS score was 10 to 14 (mean score 11). All studies scored ⱖ10 indicating that all studies included in the meta-analysis were of good quality. Study populations and settings were heterogenous (Table 1). Although most studies included patients with and without known CAD, 4 studies12,14,17 restricted recruitment to pa-
664
The American Journal of Cardiology (www.ajconline.org)
Table 1 Characteristics of studies included in meta-analysis Study
Year Published
Study Population
Total Number of Subjects
Mean Age
Men (%)
Prevalence of Ischemia (%)
Sabatine et al13
2004
112
63
78%
50%
Palumbo et al14
2004
100
59
80%
70%
Foote et al19
2004
74
59
85%
54%
Asada et al23
2004
317
68
55%
9%
Win et al24
2005
60
58
80%
16%
Zaid et al11
2006
203
59
66%
62%
Conen et al20
2006
62
67
100%
24%
Staub et al15,16
2006
260
63
70%
48%
Wong et al22
2006
56
68
69%
55%
Vanzetto et al17
2007
102
61
88%
55%
Karabinos et al10 Kurz et al21
2007 2007
128 86
55 68
72% 58%
28% 28%
Cosson et al18
2009
323
59
57%
33%
Heart and Soul Study12,25
2009
patients undergoing exercise MPI-SPECT for evaluation of chest pain or ischemia patients with known CAD referred for exercise MPI-SPECT in tertiary center for evaluation of ischemia patients with known CAD referred for exercise MPI-SPECT in tertiary center patients undergoing dobutamine echocardiography for evaluation of ischemia patient undergoing exercise MPI-SPECT for evaluation of chest pain or ischemia patient undergoing exercise MPI-SPECT for evaluation of chest pain consecutive patients referred for exercise echocardiography with rapid access to angiography consecutive patients referred for exercise MPISPECT for evaluation of chest pain or ischemia asymptomatic patients with stroke who underwent dipyridamole MPI-SPECT patients with known CAD referred for exercise MPI-SPECT in tertiary center patient undergoing dobutamine echocardiography patients undergoing exercise or dipyridamole MPI-SPECT asymptomatic diabetics undergoing exercise or dipyridamole MPI-SPECT patients with stable CAD undergoing exercise echocardiography for evaluation of ischemia
901
69
90%
25%
CAD ⫽ coronary artery disease; MPI-SPECT ⫽ myocardial perfusion imaging with single-photon emission computed tomography.
tients with known CAD only. Two studies were performed in asymptomatic but high-risk patients (stroke survivors22 and patients with diabetes and ⱖ1 other cardiovascular risk factor18). Mean age of participants was 55 to 69 years; 55% to 100% of study subjects were men. Prevalence of inducible myocardial ischemia was 9% to 70%. Five studies reported data on BNP alone,11,14,22–24 whereas 4 studies measured NT–pro-BNP only.10,17,18,21 In 5 studies data on BNP and NT–pro-BNP were available.13,15,16,25 Timing of measurement varied at baseline, peak stress, and after stress. Similarly, a wide range of threshold levels was reported (Table 2). Pooled sensitivity and specificity of BNP for detection of stress induced myocardial ischemia were 71% (95% confidence interval [CI] 68 to 74) and 52% (95% CI 50 to 54), respectively. Pooled diagnostic odds ratio was 3.5 (95% CI 2.46 to 5.04) and summary receiver operating characteristic revealed an area under the curve of 0.71 ⫾ 0.02 (Figures 2 to 5). Interestingly, a study by Conen et al20 reported very poor discrimination by BNP and NT– pro-BNP in suspected myocardial ischemia. However, this study included only 62 patients and the investigators concluded that major differences in baseline characteristics including age and higher incidence of kidney disease in patients without ischemia outweighed myocardial ischemia as a trigger for BNP and NT–pro-BNP secretion.
Six studies excluded patients with known heart failure or decreased ejection fraction.11,14,17–19,24 Pooled sensitivity (74%, 95% CI 69 to 78) in this group was marginally better with a decrease in measurements of heterogeneity in this model compared to the model that included all studies. NT–pro-BNP had better pooled sensitivity (74%, 95% CI 70 to 77, vs 67%, 95% CI 63 to 70), whereas pooled specificity (53%, 95% CI 50 to 56, vs 60%, 95% CI 57 to 63) and pooled diagnostic odds ratios (4.3 vs 5.1) appeared better in studies using BNP. Pooled sensitivity was better for studies using BNP levels at rest (73%, 95% CI 70 to 77, vs 59%, 95% CI 56 to 62), whereas pooled specificity (51%, 95% CI 54 to 59, vs 59%, 95% CI 56 to 62) was better for studies reporting measurements after stress. None of the following factors explained the demonstrated heterogeneity in results: mean age of each study population, disease prevalence, known CAD, exclusion of heart failure, type and timing of NP tested, type of reference test, and blinding of test results. Discussion Our results suggest that a BNP or NT–pro-BNP measurement can identify inducible myocardial ischemia as detected by myocardial perfusion imaging or stress echo-
Coronary Artery Disease/BNP and Stress-Induced Myocardial Ischemia
665
Table 2 B-type natriuretic peptide and N-terminal pro–B-type natriuretic peptide measurements Study
Sabatine et al
Assay Manufacturer
13
Biosite Diagnostics, Elecys, Roche Diagnostics
Palumbo et al14
Shionoria-BNP, CIS Bio International-Shering
Foote et al19
Biosite Diagnostics, Elecys, Roche Diagnostics
Asada et al23
Shionoria-BNP, Shionogi and Co.
Win et al24
Biosite Diagnostics
Zaid et al11
AxSYM, Axis Shield Diagnostics
Conen et al20
Biosite Diagnostics, Elecys, Roche Diagnostics
Staub et al15,16
AxSYM, Abbot Laboratories, Elecys, Roche Diagnostics
Wong et al22
RIA, Bachem, Ltd.
Vanzetto et al17
Elecys, Roche Diagnostics
Karabinos et al10
Elecys, Roche Diagnostics
Kurz et al21
Elecys, Roche Diagnostics
Cosson et al18
Dade Behring
Heart and Soul Study12,25
Biosite Diagnostics, Elecys, Roche Diagnostics
Thresholds (pg/ml) Timing
BNP
NT-pro-BNP
baseline after stress change* baseline after stress change baseline after stress change baseline after stress change baseline after stress change baseline after stress change baseline after stress change baseline after stress change baseline after stress change baseline after stress change baseline after stress change baseline after stress change baseline after stress change baseline after stress change
— 80 10 15 — — — — 12 18 — — — — 10% 36 48 7 18 26 — 90 110 22 11 — — — — — — — — — — — 15 — — 18 — —
— — — — — — — — 8 — — — — — — — — — 59 66 — 229 247 19 — — — 104 — — 153 172 — — 125 — — — — 100 — —
* “Change” refers to difference in peptide levels from baseline to after stress.
Figure 2. Pooled sensitivity.
Figure 3. Pooled specificity.
666
The American Journal of Cardiology (www.ajconline.org)
Figure 4. Pooled diagnostic odds ratio.
Figure 5. Summary receiver operating characteristic (SROC) curve.
cardiography, even in patients with LVSD. An increased BNP measurement had pooled sensitivity of 71% with an area under the curve of 0.71. Pooled sensitivity was even better for those studies that excluded patients with LVSD and for studies using pretest BNP levels in contrast to poststress BNP measurements. However, pooled specificity was low, perhaps because many studies did not exclude patients with abnormalities that are well known to increase BNP levels, e.g., LV dysfunction, LV hypertrophy, or renal impairment. There was a significant amount of heterogeneity that could not be explained, which limits the applicability of our findings in clinical practice. The studies we evaluated used different thresholds for BNP and NT–proBNP assays to represent an abnormal value. Because of the limited number of studies, our meta-analysis cannot provide clear guidance on the level of BNP or NT–pro-BNP that may be considered abnormal. Experimental data have confirmed that myocardial stretch is not the only trigger for release of BNP. Transcription and release of BNP are also affected by oxygen tension, and hypoxia as a trigger for the release of BNP has been demonstrated in animal models.4 Release of extra BNP
triggered by hypoxia is independent of any increase in end-diastolic pressure and is thought to result from increased BNP gene expression in hypoxic myocardial tissue irrespective of hemodynamic considerations.5,26 In line with experimental studies, published clinical data have suggested that BNP is more closely related to number of stenosed coronary arteries than to intracardiac pressure.27 An increased BNP level in affluent coronary circulation has also been demonstrated in humans within minutes of myocardial ischemia induced by balloon inflation in coronary arteries.28 Because BNP at rest is higher in those with ischemia it seems that that repeated episodes of ischemia, silent or symptomatic, may result in increased baseline production and release BNP. However, so far it remains unclear whether it is rapid upregulation of gene expression alone that is primarily responsible for ischemia-induced increased BNP levels or whether a storage granule-based pathway is also involved. Because BNP at rest is higher in those with ischemia, the first logical clinical use could be as a screening test for silent inducible myocardial ischemia or CAD. Those identified by BNP could then be selected for further more definitive tests for ischemia, e.g., exercise testing or myocardial perfusion scans. Two distinct populations in which such screening for silent myocardial ischemia would be useful are diabetics and those with noncardiac vascular disease, e.g., cerebrovascular disease. Diabetics and patients with cerebrovascular disease have high cardiac mortality rates. In a population with angiographic evidence of CAD, those patients who have angina and inducible ischemia are at higher risk of future events compared to those who have angina but no inducible myocardial ischemia.29 Also, Takase et al30 showed that in patients with stable angina BNP spontaneously increased in those who subsequently had a recurrence of angina, whereas BNP decreased in those with no recurrence (p ⬍0.0001). Hence, BNP could perhaps be used to guide the intensity of anti-ischemic and risk factor modification therapy in selected patients because those with high BNP are more likely to have inducible ischemia, future angina, and adverse cardiac events. This meta-analysis has several strengths. We undertook a comprehensive search, conducted eligibility decisions and data abstraction in duplicate, and, where possible, obtained data from study authors. Although the notion of using BNP as biomarker of ischemia is promising, there are several limitations to our systematic review. First, most of these studies were performed in different populations with varying pretest probability and likelihood of a positive stress test result. Second, most studies did not fully exclude patients with heart failure. Third, a wide range of BNP cut-off limits was reported, thus limiting its use in clinical practice. Lack of a clear diagnostic threshold may also have been compounded by issues of standardization and biases in BNP and NT–pro-BNP assays used. To establish whether there is a single threshold or a few important BNP or NT–pro-BNP thresholds (e.g., age dependent) requires evaluation in prespecified groups of large number of patients, e.g., suspected versus known CAD. Fourth, most studies were done in symptomatic subjects and we have only a limited amount of information on asymptomatic subjects, although the data that do exist are certainly encouraging.
Coronary Artery Disease/BNP and Stress-Induced Myocardial Ischemia
Acknowledgment: We thank Hisako Tsuji, MD, Demosthenes Katritsis, MD, Mary Whooley, MD, and Christian Mueller, MD for providing us additional data from their studies.10,12,20,23,25 1. Maisel A, Mueller C, Adams K Jr, Anker SD, Aspromonte N, Cleland JG, Cohen-Solal A, Dahlstrom U, DeMaria A, Di Somma S, Filippatos GS, Fonarow GC, Jourdain P, Komajda M, Liu PP, McDonagh T, McDonald K, Mebazaa A, Nieminen MS, Peacock WF, Tubaro M, Valle R, Vanderhyden M, Yancy CW, Zannad F, Braunwald E. State of the art: using natriuretic peptide levels in clinical practice. Eur J Heart Fail 2008;10:824 – 839. 2. McKie PM, Rodeheffer RJ, Cataliotti A, Martin FL, Urban LH, Mahoney DW, Jacobsen SJ, Redfield MM, Burnett JC Jr. Amino-terminal pro-B-type natriuretic peptide and B-type natriuretic peptide: biomarkers for mortality in a large community-based cohort free of heart failure. Hypertension 2006;47:874 – 880. 3. Struthers A, Lang C. The potential to improve primary prevention in the future by using BNP/N-BNP as an indicator of silent “pancardiac” target organ damage: BNP/N-BNP could become for the heart what microalbuminuria is for the kidney. Eur Heart J 2007;28:1678 –1682. 4. D’Souza SP, Baxter GF. B-type natriuretic peptide: a good omen in myocardial ischaemia? Heart 2003;89:707–709. 5. Goetze JP, Christoffersen C, Perko M, Arendrup H, Rehfeld JF, Kastrup J, Nielsen LB. Increased cardiac BNP expression associated with myocardial ischemia. FASEB J 2003;17:1105–1107. 6. Struthers AD, Davies J. B-type natriuretic peptide: a simple new test to identify coronary artery disease? QJM 2005;98:765–769. 7. Whiting PF, Weswood ME, Rutjes AW, Reitsma JB, Bossuyt PN, Kleijnen J. Evaluation of QUADAS, a tool for the quality assessment of diagnostic accuracy studies. BMC Med Res Methodol 2006;6:9. 8. Zamora J, Abraira V, Muriel A, Khan K, Coomarasamy A. Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol 2006;6:31. 9. Matthew DM. Validation of the summary ROC for diagnostic test Metaanalysis: A Monte Carlo simulation. Acad Radiol 2003;10:25–31. 10. Karabinos I, Karvouni E, Chiotinis N, Papadopoulos A, Simeonidis P, Tsolas O, Katritsis D. Acute changes in N-terminal pro-brain natriuretic peptide induced by dobutamine stress echocardiography. Eur J Echocardiogr 2007;8:265–274. 11. Zaid G, Tanchilevitch A, Rivlin E, Gropper R, Rosenschein U, Lanir A, Goldhammer E. Diagnostic accuracy of serum B-type natriuretic peptide for myocardial ischemia detection during exercise testing with SPECT perfusion imaging. Int J Cardiol 2007;117:157–164. 12. Bibbins-Domingo K, Ansari M, Schiller NB, Massie B, Whooley MA. B-type natriuretic peptide and ischemia in patients with stable coronary disease: data from the Heart and Soul study. Circulation 2003; 108:2987–2992. 13. Sabatine MS, Morrow DA, de Lemos JA, Omland T, Desai MY, Tanasijevic M, Hall C, McCabe CH, Braunwald E. Acute changes in circulating natriuretic peptide levels in relation to myocardial ischemia. J Am Coll Cardiol 2004;44:1988 –1995. 14. Palumbo B, Siepi D, Lupattelli G, Sinzinger H, Fiorucci G, Anniboletti PF, Latini RA, Mannarino E, Palumbo R. Usefulness of brain natriuretic peptide levels to discriminate patients with stable angina pectoris without and with electrocardiographic myocardial ischemia and patients with healed myocardial infarction. Am J Cardiol 2004;94: 780 –783.
667
15. Staub D, Jonas N, Zellweger MJ, Nusbaumer C, Wild D, Pfisterer ME, Mueller-Brand J, Perruchoud AP, Mueller C. Use of N-terminal proB-type natriuretic peptide to detect myocardial ischemia. Am J Med 2005;118(suppl):1287.e9 –1287.e16. 16. Staub D, Nusbaumer C, Zellweger MJ, Jonas N, Wild D, Pfisterer ME, Mueller-Brand J, Perruchoud AP, Mueller C. Use of B-type natriuretic peptide in the detection of myocardial ischemia. Am Heart J 2006; 151:1223–1230. 17. Vanzetto G, Jacon P, Calizzano A, Neuder Y, Faure P, Fagret D, Machecourt J. N-terminal pro-brain natriuretic peptide predicts myocardial ischemia and is related to postischemic left-ventricular dysfunction in patients with stable coronary artery disease. J Nucl Cardiol 2007;14:835– 842. 18. Cosson E, Nguyen MT, Pham I, Pontet M, Nitenberg A, Valensi P. N-terminal pro-B-type natriuretic peptide: an independent marker for coronary artery disease in asymptomatic diabetic patients. Diabet Med 2009;26:872– 879. 19. Foote RS, Pearlman JD, Siegel AH, Yeo K-TJ. Detection of exerciseinduced ischemia by changes in B-type natriuretic peptides. J Am Coll Cardiol 2004;44:1980 –1987. 20. Conen D, Jander N, Trenk D, Neumann F-J, Mueller C. The use of B-type natriuretic peptides in the detection of myocardial ischemia in settings with rapid access to coronary angiography. Int J Cardiol 2007;119:416 – 418. 21. Kurz K, Voelker R, Zdunek D, Wergeland R, Hess G, Ivandic B, Katus H, Giannitsis E. Effect of stress-induced reversible ischemia on serum concentrations of ischemia-modified albumin, natriuretic peptides and placental growth factor. Clin Res Cardiol 2007;96:152–159. 22. Wong KY, McSwiggan S, Kennedy NS, MacWalter RS, Struthers AD. B-type natriuretic peptide identifies silent myocardial ischaemia in stroke survivors. Heart 2006;92:487– 489. 23. Asada J, Tsuji H, Iwasaka T, Thomas JD, Lauer MS. Usefulness of plasma brain natriuretic peptide levels in predicting dobutamine-induced myocardial ischemia. Am J Cardiol 2004;93:702–704. 24. Win HK, Chang SM, Raizner M, Shah G, Al Basky F, Desai U, Plana JC, Mahmarian JJ, Quinones MA, Zoghbi WA. Percent change in B-type natriuretic peptide levels during treadmill exercise as a screening test for exercise-induced myocardial ischemia. Am Heart J 2005; 150:695–700. 25. Singh HS, Bibbins-Domingo K, Ali S, Wu AH, Schiller NB, Whooley MA. N-terminal pro-B-type natriuretic peptide and inducible ischemia in the Heart and Soul Study. Clin Cardiol 2009;32:447– 453. 26. Goetze JP, Gore A, Moller CH, Steinbruchel DA, Rehfeld JF, Nielsen LB. Acute myocardial hypoxia increases BNP gene expression. FASEB J 2004;18:1928 –1930. 27. Davidson NC, Pringle SD, Pringle TH, McNeill GP, Struthers AD. Right coronary artery stenosis is associated with impaired cardiac endocrine function during exercise. Eur Heart J 1997;18:1749 –1754. 28. Pascual-Figal DA, Antolinos MJ, Bayes-Genis A, Casas T, Nicolas F, Valdes M. B-type natriuretic peptide release in the coronary effluent after acute transient ischaemia in humans. Heart 2007;93:1077–1080. 29. Bibbins-Domingo K, Gupta R, Na B, Wu AH, Schiller NB, Whooley MA. N-terminal fragment of the prohormone brain-type natriuretic peptide (NT-proBNP), cardiovascular events, and mortality in patients with stable coronary heart disease. JAMA 2007;297:169 –176. 30. Takase H, Toriyama T, Sugiura T, Ueda R, Dohi Y. Brain natriuretic peptide in the prediction of recurrence of angina pectoris. Eur J Clin Invest 2004;34:79 – 84.
Temporal Trends (over 30 Years), Clinical Characteristics, Outcomes, and Gender in Patients <50 Years of Age Having Percutaneous Coronary Intervention Farhan J. Khawaja, MDa, Charanjit S. Rihal, MDa, Ryan J. Lennon, MSb, David R. Holmes, MDa, and Abhiram Prasad, MDa,* Little is known regarding temporal trends in characteristics and outcomes of young (<50 years) patients who develop symptomatic premature coronary artery disease (CAD). The aim of this study was to describe temporal trends in clinical characteristics and outcomes and gender differences in patients with premature CAD undergoing percutaneous coronary intervention (PCI) over 3 decades. A retrospective analysis of 2,922 consecutive patients <50 years of age undergoing PCI from 1980 through 2007 was conducted. Baseline characteristics and in-hospital and long-term outcomes were compared by decade. Gender differences and predictors of mortality were analyzed in the most recent cohort. Although most patients were men (80%), there was an increasing proportion of women over time. An increasing prevalence of diabetes mellitus (10% in 1980 to 1989, 16% in 1990 to 1999, 20% in 2000 to 2007, p <0.001), hypertension (29%, 41%, 57%, p <0.001), and hyperlipidemia (39%, 55%, 73%, p <0.001) coincided with increasing body mass index (28.2 ⴞ 4.6, 29.9 ⴞ 5.8, 30.9 ⴞ 6.7 kg/m2, p <0.001). The proportion of smokers decreased (84%, 76%, 74%, p <0.001). In-hospital mortality (1.0%, 0.8%, 0.9%, p ⴝ 0.93) and long-term mortality at 5 years (6%, 6%, 7%, p ⴝ 0.97) did not change over time. In contemporary PCI practice, women with premature CAD were more likely to have diabetes mellitus (25% vs 19%, p ⴝ 0.02), single-vessel disease (56% vs 41%, p <0.001), and a bleeding complication. In conclusion, there is an increasing burden of cardiovascular risk factors, related mostly to obesity, in patients with premature CAD requiring PCI. Long-term morbidity or mortality in these patients has not improved over the previous 3 decades. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:668 – 674) A small proportion of patients manifest symptoms of coronary artery disease (CAD) at a relatively young age and are often referred to as having premature CAD. Previous studies have demonstrated that patients with premature CAD are predominantly men and have a high prevalence of conventional cardiovascular risk factors and a family history of premature CAD.1– 4 There is a paucity of data on temporal trends in the cardiovascular risk factor profile and outcomes in this patient population.5–7 We hypothesized that there is an increasing burden of cardiovascular risk factors in patients with premature CAD due to adverse lifestyle trends and that their long-term outcomes have not improved. Thus, the aim of this study was to describe trends in clinical characteristics and outcomes of patients with premature CAD undergoing percutaneous coronary intervention (PCI) over the previous 3 decades. Methods Since 1979, patients undergoing percutaneous coronary revascularization at the Mayo Clinic in Rochester, Minne-
a
Division of Cardiovascular Diseases, Department of Medicine, and Division of Biomedical Statistics and Informatics, Mayo Clinic, Rochester, Minnesota. *Corresponding author: Tel: 507-538-6325; fax: 507-255-2550. E-mail address:
[email protected] (A. Prasad).
b
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.044
sota have been prospectively followed in our registry, which includes demographic, clinical, angiographic, and procedural data. Immediate postprocedure and in-hospital events are recorded and each patient is surveyed by telephone using a standardized questionnaire at 6 months, 1 year, and then annually after the procedure by trained data technicians. The inclusion criterion was an age ⱕ50 years in patients undergoing PCI. Patients were excluded if they declined authorization allowing use of their medical records for research, as required by Minnesota statute. Of patients with ⬎1 qualifying procedure, only the earliest procedure was used. We identified 2,922 patients who met these criteria. Patients were grouped into 3 eras according to time of PCI: group 1, January 1980 to December 1989 (n ⫽ 523); group 2, January 1990 through December 1999 (n ⫽ 1,265); and group 3, January 2000 through December 2007 (n ⫽ 1,134). For comparison, data regarding clinical characteristics and in-hospital outcomes are provided for patients ⬎50 years old for the most recent period. The study was approved by the Mayo Clinic institutional review board. Hypertension was defined as a documented history of hypertension treated with medication. Hyperlipidemia was defined as any history of total cholesterol measurement ⬎240 mg/dl or treatment with a lipid-lowering agent. Diabetes mellitus was defined as a documented history of diabetes treated with medication or diet. Multivessel CAD required ⱖ70% diameter stenosis in 1 major artery and www.ajconline.org
Coronary Artery Disease/Premature Coronary Artery Disease
669
Table 1 Baseline clinical characteristics Variable
Age (years) Men Diabetes mellitus Hypertension Hyperlipidemia Smoking status Never Former Current Obesity Body mass index (kg/m2) Most recent myocardial infarction (days) ⬍1 1–7 ⬎7 Never Mayo Clinic Risk Score predicted risk (%) Preprocedure shock Previous percutaneous coronary intervention Previous coronary bypass surgery Peripheral vascular disease Cerebrovascular disease Moderate/severe renal disease
1980–1989 (n ⫽ 523)
1990–1999 (n ⫽ 1,265)
2000–2007 (n ⫽ 1,134)
p Value (age ⱕ50)
44.8 ⫾ 4.7 438 (84%) 52 (10%) 151 (29%) 164 (39%)
45.1 ⫾ 4.8 1,031 (82%) 200 (16%) 506 (41%) 600 (55%)
45.5 ⫾ 4.8 880 (78%) 228 (20%) 583 (57%) 711 (73%)
0.007 0.002 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
79 (15%) 153 (29%) 289 (55%) 166 (32%) 28.2 ⫾ 4.6
284 (23%) 330 (26%) 641 (51%) 521 (43%) 29.9 ⫾ 5.8
290 (26%) 255 (23%) 569 (51%) 550 (49%) 30.9 ⫾ 6.7
110 (21%) 76 (15%) 125 (24%) 212 (41%) 0.6 (0.4–0.8)
260 (21%) 219 (17%) 270 (21%) 508 (40%) 0.8 (0.6–1.5)
387 (34%) 188 (17%) 133 (12%) 415 (37%) 1.1 (0.8–1.5)
36 (3%) 109 (9%) 117 (9%) 33 (4%) 21 (2%) 29 (3%)
0 (0%) 36 (7%)
⬍0.001 ⬍0.001 ⬍0.001
2000–2007 (age ⬎50) (n ⫽ 9,479) 69.5 ⫾ 10.0 6,539 (69%) 2,468 (26%) 6,927 (77%) 7,120 (82%) 3,402 (37%) 4,459 (49%) 1,329 (14%) 3,875 (41%) 29.6 ⫾ 5.7
p Value (age ⱕ50 vs ⬎50, 2000–2007) ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
⬍0.001 ⬍0.001 ⬍0.001
⬍0.001
1,912 (21%) 1,465 (16%) 1,890 (20%) 4,039 (43%) 2.1 (1.5–3.9)
⬍0.001
48 (4%) 136 (12%)
0.21 ⬍0.001
396 (4%) 2,440 (24%)
0.90 ⬍0.001
73 (6%) 41 (4%) 38 (3%) 42 (4%)
0.33 0.93 0.19 0.60
2,029 (21%) 1,020 (11%) 1,139 (12%) 353 (4%)
⬍0.001 ⬍0.001 ⬍0.001 0.96
Values are presented as mean ⫾ SD, number of patients (percentage), or median (quartiles 1 to 3).
ⱖ50% stenosis in a second major vessel. Patients with ⱖ50% diameter stenosis in the left main coronary artery were considered to have 2-vessel disease if there was right dominance and 3-vessel disease if there was left dominance. Major adverse cardiovascular events (MACEs) were defined as ⱖ1 of the following: (1) in-hospital death; (2) Q-wave myocardial infarction (MI); (3) urgent or emergency coronary artery bypass grafting (CABG) during the index hospitalization; and (4) cerebrovascular accident defined as transient ischemic attack or stroke. The Mayo Clinic Risk Score estimates the risk of MACE based on 8 clinical and angiographic variables.8 MI was diagnosed in the presence of 2 of the following 3 criteria: (1) typical chest pain for ⱖ20 minutes; (2) increase of serum creatinine kinase levels (or the MB fraction) ⬎2 times normal; and (3) a new Q wave on electrocardiogram. In-hospital deaths included all deaths during the index hospital admission. Severe renal dysfunction was defined as a creatinine level ⬎3.0 mg/dl or a history of dialysis or renal transplantation. Procedural success was defined as a decrease of residual luminal diameter stenosis to ⱕ20% in ⱖ1 treated lesion without in-hospital death, Q-wave MI, or CABG to allow comparison among the different eras. Long-term outcomes included all-cause mortality and the combined end point of death or any MI and target lesion revascularization. Target lesion revascularization was defined as any attempted percutaneous or surgical revascularization of the target lesion after the initial procedure.
Figure 1. Frequency distribution of ⱖ1 cardiovascular risk factor (diabetes mellitus, smoking, hypertension, hyperlipidemia, and obesity) present in 1,134 patients ⱕ50 years old from 2000 to 2007.
Continuous data are summarized as mean ⫾ SD unless otherwise stated. Categorical data are summarized as frequency (group percentage). Trends in distributions of continuous variables across the 3 periods were tested using a linear contrast of means in association with analysis of
670
The American Journal of Cardiology (www.ajconline.org)
Table 2 Angiographic and procedural characteristics and in-hospital outcomes Variable
Single vessel coronary disease Worst lesion type A B1 B2 C Urgency of percutaneous coronary intervention Elective Urgent Emergency Number of vessels treated 1 2 3 Use of drug-eluting stents Glycoprotein IIb/IIIa inhibitors use In-hospital outcomes Death Q-wave myocardial infarction Coronary bypass surgery Target lesion revascularization Major adverse cardiovascular events Admission medications Aspirin  Blocker Angiotensin-converting enzyme inhibitors Lipid-lowering drug Discharge medications Aspirin  Blocker Angiotensin-converting enzyme inhibitors Lipid-lowering drug
1980–1989 (n ⫽ 523) 237 (47%)
1990–1999 (n ⫽ 1,265)
2000–2007 (n ⫽ 1,134)
p Value (age ⱕ50)
2000–2007 (age ⬎50) (n ⫽ 9,479)
p Value (age ⱕ50 vs ⬎50, 2000–2007)
501 (41%)
449 (43%)
0.38 ⬍0.001
2,590 (30%)
⬍0.001 0.50
34 (3%) 232 (23%) 376 (38%) 354 (36%)
32 (3%) 181 (19%) 289 (30%) 468 (48%)
235 (3%) 1,475 (17%) 2,835 (33%) 4,098 (47%) ⬍0.001
485 (93%) 0 (0%) 38 (7%)
626 (50%) 369 (29%) 268 (21%)
319 (28%) 458 (40%) 357 (31%)
467 (89%) 54 (10%) 2 (0%) 0 (0%) 0 (0%)
1,120 (89%) 138 (11%) 7 (1%) 0 (0%) 288 (23%)
979 (86%) 146 (13%) 9 (1%) 502 (44%) 782 (69%)
10 (0.8%) 4 (0.3%) 20 (1.6%) 45 (3.6%) 28 (2.2%)
175 (34%) 204 (39%) 0 (0%) 0 (0%)
0.075 3,219 (34%) 4,343 (46%) 1,914 (20%)
0.52
5 (1.0%) 8 (1.5%) 41 (7.8%) 53 (10.1%) 29 (5.5%)
422 (81%) 186 (36%) 0 (0%) 0 (0%)
0.056
⬍0.001 ⬍0.001
7,949 (84%) 1,400 (15%) 127 (1%) 4,006 (42%) 5,715 (60%)
0.20 ⬍0.001
10 (0.9%) 1 (0.1%) 13 (1.1%) 16 (1.4%) 19 (1.7%)
0.93 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
194 (2.0%) 22 (0.2%) 70 (0.7%) 108 (1.1%) 295 (3.1%)
0.007 0.32 0.14 0.42 0.007
915 (73%) 775 (61%) 140 (13%)
992 (88%) 895 (79%) 356 (32%)
⬍0.001 ⬍0.001 ⬍0.001
8,244 (88%) 7,035 (75%) 3,831 (41%)
0.81 0.003 ⬍0.001
220 (20%)
374 (33%)
⬍0.001
4,028 (43%)
⬍0.001
1,191 (95%) 871 (69%) 241 (19%)
1,095 (97%) 976 (87%) 643 (57%)
⬍0.001 ⬍0.001 ⬍0.001
8,912 (95%) 7,680 (82%) 5,333 (57%)
0.010 ⬍0.001 0.98
512 (41%)
729 (65%)
⬍0.001
6,271 (67%)
0.12
variance. Trends in categorical data were tested with the Armitage trend test. For comparison of patients ⱕ50 years old versus others in the current era and comparison between men and women, 2-sample t tests and Pearson chi-square tests were used. Kaplan-Meier methods are used to estimate follow-up event rates, with follow-up beginning at time of discharge (patients who died in hospital were excluded). Trends in follow-up risk were tested using a Cox proportional hazards model with the 3 groups scored as 1, 2, and 3 and modeled continuously. A multiple logistic regression model was employed to estimate the adjusted association between era and in-hospital mortality. Covariates chosen for adjustment were selected by clinical relevance. A Cox proportional hazards multiple regression model was used to investigate which variables were associated with follow-up mortality in patients ⱕ50 years old undergoing PCI in the 2000 to 2007 cohort. A backward selection variable procedure was employed on the following variables chosen for clinical relevance: MI within 1 week, preprocedure shock, history of congestive heart failure, diabetes mellitus, hyper-
tension, body mass index (linear), hyperlipidemia, ever smoked, peripheral vascular disease, history of cerebral vascular accident/transient ischemic attack, moderate/severe renal disease, chronic obstructive pulmonary disease, tumor/lymphoma/leukemia, and multivessel disease. Variables were removed from the model if they were not significant at the 0.05 level. Expected mortality was calculated based on a cohort matched by age, gender, and calendar year of birth from the Minnesota white population.9 Results Clinical characteristics of the 2,922 patients in the 3 periods are presented in Table 1. Patients with premature CAD represent 17%, 13%, and 11% (p ⬍0.001) of the entire PCI population for the 3 periods, respectively. Most patients were men. Prevalence of diabetes mellitus, hypertension, hyperlipidemia, and obesity increased over time. However, the proportion of patients who were current or former smokers decreased. Prevalence of ⱖ1 of the risk factors obesity,
Coronary Artery Disease/Premature Coronary Artery Disease
Figure 2. Kaplan–Meier curves for mortality (A) and combined end point of death or myocardial infarction (B). (A) Expected survival reflects mortality rates for a cohort matched by age, gender, and calendar year of birth from the Minnesota white population.
diabetes mellitus, smoking, hypertension, or hyperlipidemia in the 3 eras were 95%, 96%, and 97% (p ⫽ 0.14 for trend). Frequency distribution of risk factors in the most recent cohort is presented in Figure 1. Preprocedure risk for PCI as estimated by the Mayo Clinic Risk Score increased over time. Compared to patients ⬎50 years old (2000 to 2007), patients with premature CAD had fewer co-morbidities except for a higher prevalence of obesity and smoking, and they were more likely to have had a MI within the preceding 7 days. Table 2 presents angiographic and procedural characteristics in the 3 groups. Most patients had type B2 or C culprit lesions. Procedural success increased over time (31%, 71%, 94%, p ⬍0.001) despite increased lesion complexity. Compared to patients ⬎50 years old (2000 to 2007), patients with premature CAD were more likely to have single-vessel disease. In-hospital outcomes are presented in Table 2. Although there was no change in unadjusted in-hospital mortality over time, in-hospital rates of Q-wave MI, CABG, target lesion revascularization, and MACE decreased significantly. Kaplan–Meier estimates for all-cause mortality and combined end point of death or MI during follow-up are shown in Figure 2. Median follow-ups (interquartile ranges) were 21.0 years (19.1 to 24.0), 11.2 years (9.1 to 14.1), and 3.7 years (1.9 to 6.0) for the 3 groups. During follow-up,
671
there was no significant difference in unadjusted mortality rates (p ⫽ 0.97 for trend over groups). At 5 years, mortality rates were 6%, 6%, and 7%, respectively. Risk-adjusted long-term mortality (for history of diabetes, hypertension, hyperlipidemia, current smoking, body mass index, MI within 1 week, congestive heart failure, and medication score) also did not differ in the 3 groups (p ⫽ 0.21). There was no difference in the combined end point of death or MI (p ⫽ 0.44) with event rates at 5 years of 14%, 13%, and 16%, respectively. However, rates of target lesion revascularization decreased significantly (27%, 25%, 19%, p ⬍0.001). Figure 3 illustrates independent predictors of long-term mortality in the most recent cohort. Variables removed from the model (MI within previous week, peripheral vascular disease, history of cerebral vascular accident/ transient ischemic attack, preprocedure shock, body mass index, hypertension) had p values ⬎0.30 at the time of removal. There was a slight increase in the proportion of women over time such that 22% of patients were women in the most recent period. Clinical, angiographic, and procedural characteristics according to gender for the most recent cohort are presented in Table 3. Women were marginally younger and had a higher prevalence of diabetes mellitus, but otherwise clinical characteristics were similar. Single-vessel disease was more common in women. An important procedural finding was the higher frequency of bleeding complications in women with respect to hematomas, retroperitoneal bleeding, femoral bleeding, and greater need for blood transfusion. In-hospital death (women vs men 0.9% vs 0.8%, p ⫽ 0.86), MI (3.5% vs 2.4%, p ⫽ 0.31), and CABG (0.4% vs 1.4%, p ⫽ 0.20) were similar in women and men. At 5 years, total mortalities were 7% in men and 8% in women (p ⫽ 0.85), whereas incidence of death or any MI was 16% in women and men (p ⫽ 0.52) at 5 years. Discussion The major findings of the present study in 2,922 patients with symptomatic premature CAD requiring PCI are that (1) prevalence of obesity and smoking is greater than in older patients, (2) ⱖ1 conventional risk factor for atherosclerosis or obesity is present in virtually all patents, (3) there has been a 50% relative increase in the prevalence of obesity and a 100% increase in the prevalence of diabetes mellitus associated with an increase in hypertension and hyperlipidemia, (4) there has been no change in long-term outcomes despite greater use of secondary prevention, and (5) there were no differences in gender-based ischemic outcomes, but young women had a higher frequency of periprocedural bleeding complications compared to their men counterparts. In contemporary practice, 97% of patients had ⱖ1 and 79% had ⱖ2 conventional cardiovascular risk factors or obesity. Although there is great interest in identifying novel risk factors for atherosclerosis,10 particularly in those with premature CAD, our data suggest that presence of conventional risk factors such as diabetes mellitus, smoking, hypertension, and hyperlipidemia, and/or obesity may be responsible. Of these, smoking and obesity were 2 risk factors that were more common in those with premature CAD compared to older patients. This finding is consistent with
672
The American Journal of Cardiology (www.ajconline.org)
Figure 3. Hazard ratios (HRs) for independent predictors of long-term mortality in patients with premature coronary artery disease during the most recent era. CHF ⫽ congestive heart failure; CI ⫽ confidence interval; COPD ⫽ chronic obstructive pulmonary disease. Table 3 Gender-based comparisons (2000 to 2007) Variable Age (years) Diabetes mellitus Hypertension Hyperlipidemia Smoking status Never Former Current Obesity Body mass index (kg/m2) Urgency of percutaneous coronary intervention Elective Urgent Emergency Multivessel disease Number of vessels treated 1 2 3 Total number of stents placed Glycoprotein IIb/IIIa inhibitor use Bleeding complications Hematoma Pseudoaneurysm Gastrointestinal bleed Retroperitoneal bleed Femoral bleed Blood loss requiring transfusion
Women (n ⫽ 254)
Men (n ⫽ 880)
p Value
44.9 ⫾ 5.0 64 (25%) 134 (60%) 143 (67%)
45.6 ⫾ 4.7 164 (19%) 449 (56%) 568 (74%)
0.042 0.021 0.34 0.049 0.48
72 (29%) 50 (20%) 124 (50%) 124 (49%) 30.5 ⫾ 7.7
218 (25%) 205 (24%) 445 (51%) 426 (49%) 31.0 ⫾ 6.4
69 (27%) 110 (43%) 75 (30%) 104 (44%)
250 (28%) 348 (40%) 282 (32%) 497 (59%)
220 (87%) 31 (12%) 3 (1%) 1.3 ⫾ 0.9
759 (86%) 115 (13%) 6 (1%) 1.4 ⫾ 0.9
0.20
158 (62%)
624 (71%)
0.008
9 (3.5%) 2 (0.8%) 2 (0.8%) 3 (1.2%) 3 (1.2%) 12 (4.7%)
10 (1.1%) 1 (0.1%) 2 (0.2%) 0 (0.0%) 0 (0.0%) 14 (1.6%)
0.008 0.07 0.18 0.001 0.001 0.003
0.90 0.33 0.80
⬍0.001 0.90
an observation by Khot et al11 derived from a large clinical trial database from the 1990s that approximately 90% of patients ⱕ45 years old with CAD had ⱖ1 conventional cardiovascular risk factor. As observed in the general CAD population,12–14 prevalence of smoking did decrease over time. Although this is encouraging, the decrease was modest. Thus, populationbased efforts to decrease smoking must continue and targeting the young may yield benefits with in decreasing the incidence and sequels of premature CAD. Prevalence of obesity increased by approximately 50% and diabetes mellitus by 100% from 1980 through 1989 to 2000 through 2007. Trends in obesity and diabetes mellitus in patients with premature CAD mirror those in the general local15 and national12 populations. Although the epidemic of diabetes mellitus can in part be explained by the lower blood glucose threshold for diagnosing the condition, the increasing prevalence in obesity is the major factor. Thus, prevention and treatment of obesity must become a major public health priority in the prevention of premature CAD. We observed a marked increase in the prevalence of hypertension and hyperlipidemia over time in contrast to previous studies in the general population.12–14,16,17 These differences may be explained by the fact that previous studies reported trends in blood pressure and lipid levels that would have been influenced by therapeutic interventions, whereas our database records a medical history of these conditions. Moreover, previous studies investigated population-based cohorts, whereas our study was in patients with symptomatic CAD requiring PCI. In a multivariable model of predictors of long-term mortality in current practice, hyperlipidemia was seemingly protective (Figure 3), which is incongruent with conventional knowledge of the impact of hyperlipidemia on CAD.18 However, this phenomenon has been frequently
Coronary Artery Disease/Premature Coronary Artery Disease
reported in clinical trials and observational databases.19,20 To explore this apparent paradox, Wang et al21 investigated patients admitted with non–ST-segment elevation acute coronary syndromes with and without a history of hyperlipidemia. A similar apparent protective effect of a history of hyperlipidemia persisted when adjusting for baseline characteristics and previous statin use. However, a protective effect of hyperlipidemia was not observed in patients with a new diagnosis of hyperlipidemia at time of hospital admission, leading the investigators to conclude that other confounders that are not captured in observational databases likely explain the paradox. These confounders could include previous contact with the medical field before admission, which may make this population more healthy and less likely to have adverse outcomes; the pleiotropic effects of statins that cannot be fully accounted for by measuring a lipid profile; and that these patients may be taking other medications in addition to statins such as aspirin, clopidogrel,  blockers, and angiotensin-converting enzyme inhibitors before their acute coronary syndrome events that contribute to lower risk. Procedural success has dramatically improved over time, despite the increasing risk profile of the patients, as quantified by the Mayo Clinic Risk Score. This has been associated with a marked decrease in in-hospital rates of Q-wave MI, CABG, target lesion revascularization, and MACEs likely due to advances in the pharmacology and technology required for PCI in conjunction with greater operator experience. These trends are similar to our overall PCI experience,22 although in-hospital mortality and MACE rates are lower in patients with premature CAD compared to older patients due to their lower risk profile. Unlike in-hospital outcomes, there was no change in long-term adjusted mortality or combined end point of death or MI. The finding suggests that the natural history of patients with premature CAD has not been modified significantly over time. This to a large extent may be because the benefits of advances of disease management have been negated by the increasing risk profile of these patients. The observational design of our analysis does not allow accurate adjustment for differences in baseline characteristics and therapeutic interventions in the different eras. Alternatively, it is also possible that current treatment strategies are ineffective in this group of patients in modifying natural history. Independent predictors of long-term mortality included smoking and diabetes mellitus, underscoring the importance of tackling these cardiovascular risk factors to improve outcomes in premature CAD.5 In contrast, long-term need for target lesion revascularization has decreased markedly, as would be anticipated with the advent of coronary stents. Compared to men, women with premature CAD were marginally younger and more likely to have diabetes mellitus. In-hospital and long-term outcomes in men and women were similar. This is consistent with our experience in the overall PCI population.23 However, data from the New York State PCI Registry suggests that young women have higher in-hospital mortality than men. The contrasting results may be due to differences in baseline characteristics of the 2 study cohorts. Interestingly, in-hospital mortality in women in our study (0.9%) was similar to that of the New York State PCI Registry cohort (0.7%), and as such their
673
observed difference was due to lower mortality in the population of men (0.2%).24 Our study confirms that young women have a higher frequency of PCI-related bleeding complications despite having less complex CAD and lower use of glycoprotein IIb/IIIa inhibitors compared to young men, a phenomenon previously reported in women in general.25 The increased risk in women overall has been attributed to a smaller body, which may also be true for younger women.26 Data from the Early Discharge after Stenting of Coronary Arteries (EASY) trial confirms that increased risk of bleeding persists in the general population of women even with the transradial approach.27 The precise cause for increased bleeding risk in young women cannot be ascertained from our study and remains to be established. The fact that young and older women are more prone to bleeding suggests that age per se does not account for the phenomenon. This study is a retrospective analysis from a single largevolume tertiary care center, which limits its generalizability, but represents the largest analysis of temporal trends in premature CAD performed to date. Due to changing definitions of MI over time and transition from creatine kinase-MB to troponin, we could not elaborate on temporal trends of in-hospital MI. Physician and patient factors that influence referral for PCI cannot be determined from our registry and are a source of bias. The study was limited to selected patients treated with PCI, and we cannot comment on trends in outcomes in patients who are managed medically or with surgical revascularization. 1. Chesebro JH, Fuster V, Elveback LR, Frye RL. Strong family history and cigarette smoking as risk factors of coronary artery disease in young adults. Br Heart J 1982;47:78 – 83. 2. Zimmerman FH, Cameron A, Fisher LD, Ng G. Myocardial infarction in young adults: angiographic characterization, risk factors and prognosis (Coronary Artery Surgery Study Registry). J Am Coll Cardiol 1995;26:654 – 661. 3. Mukherjee D, Hsu A, Moliterno DJ, Lincoff AM, Goormastic M, Topol EJ. Risk factors for premature coronary artery disease and determinants of adverse outcomes after revascularization in patients ⬍ or ⫽40 years old. Am J Cardiol 2003;92:1465–1467. 4. Jomini V, Oppliger-Pasquali S, Wietlisbach V, Rodondi N, Jotterand V, Paccaud F, Darioli R, Nicod P, Mooser V. Contribution of major cardiovascular risk factors to familial premature coronary artery disease: the GENECARD project. J Am Coll Cardiol 2002;40:676 – 684. 5. Cole JH, Miller JI III, Sperling LS, Weintraub WS. Long-term follow-up of coronary artery disease presenting in young adults. J Am Coll Cardiol 2003;41:521–528. 6. Fournier JA, Cabezon S, Cayuela A, Ballesteros SM, Cortacero JA, Diaz De La Llera LS. Long-term prognosis of patients having acute myocardial infarction when ⱕ40 years of age. Am J Cardiol 2004;94: 989 –992. 7. Kofflard MJ, de Jaegere PP, van Domburg R, Ruygrok P, van den Brand M, Serruys PW, de Feyter PJ. Immediate and long-term clinical outcome of coronary angioplasty in patients aged 35 years or less. Br Heart J 1995;73:82– 86. 8. Singh M, Lennon RJ, Holmes DR Jr, Bell MR, Rihal CS. Correlates of procedural complications and a simple integer risk score for percutaneous coronary intervention. J Am Coll Cardiol 2002;40:387–393. 9. Therneau TM, Sicks JD, Bergstralh EJ, Offord JR. Expected Survival Based on Hazard Rates, Section of Biostatistics, Mayo Clinic, Rochester, MN. Technical Report Series No. 52. March 1994. 10. Helfand M, Buckley DI, Freeman M, Fu R, Rogers K, Fleming C, Humphrey LL. Emerging risk factors for coronary heart disease: a summary of systematic reviews conducted for the U.S. Preventive Services Task Force. Ann Intern Med 2009;151:496 –507.
674
The American Journal of Cardiology (www.ajconline.org)
11. Khot UN, Khot MB, Bajzer CT, Sapp SK, Ohman EM, Brener SJ, Ellis SG, Lincoff AM, Topol EJ. Prevalence of conventional risk factors in patients with coronary heart disease. JAMA 2003;290:898 –904. 12. Gregg EW, Cheng YJ, Cadwell BL, Imperatore G, Williams DE, Flegal KM, Narayan KM, Williamson DF. Secular trends in cardiovascular disease risk factors according to body mass index in US adults. JAMA 2005;293:1868 –1874. 13. Preis SR, Pencina MJ, Hwang SJ, D’Agostino RB Sr, Savage PJ, Levy D, Fox CS. Trends in cardiovascular disease risk factors in individuals with and without diabetes mellitus in the Framingham Heart Study. Circulation 2009;120:212–220. 14. Arnett DK, McGovern PG, Jacobs DR, Jr., Shahar E, Duval S, Blackburn H, Luepker RV. Fifteen-year trends in cardiovascular risk factors (1980 –1982 through 1995–1997): the Minnesota Heart Survey. Am J Epidemiol 2002;156:929 –935. 15. Gandhi GY, Roger VL, Bailey KR, Palumbo PJ, Ransom JE, Leibson CL. Temporal trends in prevalence of diabetes mellitus in a population-based cohort of incident myocardial infarction and impact of diabetes on survival. Mayo Clin Proc 2006;81:1034 –1040. 16. Mosterd A, D’Agostino RB, Silbershatz H, Sytkowski PA, Kannel WB, Grobbee DE, Levy D. Trends in the prevalence of hypertension, antihypertensive therapy, and left ventricular hypertrophy from 1950 to 1989. N Engl J Med 1999;340:1221–1227. 17. Kuklina EV, Yoon PW, Keenan NL. Trends in high levels of lowdensity lipoprotein cholesterol in the United States, 1999 –2006. JAMA 2009;302:2104 –2110. 18. Pekkanen J, Linn S, Heiss G, Suchindran CM, Leon A, Rifkind BM, Tyroler HA. Ten-year mortality from cardiovascular disease in relation to cholesterol level among men with and without preexisting cardiovascular disease. N Engl J Med 1990;322:1700 –1707. 19. Boersma E, Pieper KS, Steyerberg EW, Wilcox RG, Chang WC, Lee KL, Akkerhuis KM, Harrington RA, Deckers JW, Armstrong PW, Lincoff AM, Califf RM, Topol EJ, Simoons ML. Predictors of outcome in patients with acute coronary syndromes without persistent ST-segment elevation. Results from an international trial of 9461
20.
21.
22.
23.
24.
25.
26. 27.
patients. The PURSUIT Investigators. Circulation 2000;101: 2557–2567. Granger CB, Goldberg RJ, Dabbous O, Pieper KS, Eagle KA, Cannon CP, Van De Werf F, Avezum A, Goodman SG, Flather MD, Fox KA. Predictors of hospital mortality in the global registry of acute coronary events. Arch Intern Med 2003;163:2345–2353. Wang TY, Newby LK, Chen AY, Mulgund J, Roe MT, Sonel AF, Bhatt DL, DeLong ER, Ohman EM, Gibler WB, Peterson ED. Hypercholesterolemia paradox in relation to mortality in acute coronary syndrome. Clin Cardiol 2009;32(suppl):E22–E28. Singh M, Rihal CS, Gersh BJ, Lennon RJ, Prasad A, Sorajja P, Gullerud RE, Holmes DR Jr. Twenty-five-year trends in in-hospital and long-term outcome after percutaneous coronary intervention: a single-institution experience. Circulation 2007;115:2835–2841. Singh M, Rihal CS, Gersh BJ, Roger VL, Bell MR, Lennon RJ, Lerman A, Holmes DR Jr. Mortality differences between men and women after percutaneous coronary interventions. A 25-year, singlecenter experience. J Am Coll Cardiol 2008;51:2313–2320. Srinivas VS, Garg S, Negassa A, Bang JY, Monrad ES. Persistent sex difference in hospital outcome following percutaneous coronary intervention: results from the New York State reporting system. J Invasive Cardiol 2007;19:265–268. Lansky AJ, Mehran R, Cristea E, Parise H, Feit F, Ohman EM, White HD, Alexander KP, Bertrand ME, Desmet W, Hamon M, Stone GW. Impact of gender and antithrombin strategy on early and late clinical outcomes in patients with non–ST-elevation acute coronary syndromes (from the ACUITY trial). Am J Cardiol 2009;103:1196 –1203. Peterson ED, Lansky AJ, Kramer J, Anstrom K, Lanzilotta MJ. Effect of gender on the outcomes of contemporary percutaneous coronary intervention. Am J Cardiol 2001;88:359 –364. Tizon-Marcos H, Bertrand OF, Rodes-Cabau J, Larose E, Gaudreault V, Bagur R, Gleeton O, Courtis J, Roy L, Poirier P, Costerousse O, De Larochelliere R. Impact of female gender and transradial coronary stenting with maximal antiplatelet therapy on bleeding and ischemic outcomes. Am Heart J 2009;157:740 –745.
Effect of Insurance Type on Adverse Cardiac Events After Percutaneous Coronary Intervention Michael A. Gaglia, Jr., MD, MSc, Rebecca Torguson, MPH, Zhenyi Xue, MS, Manuel A. Gonzalez, MD, MPH, Itsik Ben-Dor, MD, Gabriel Maluenda, MD, Michael Mahmoudi, MD, Gabriel Sardi, MD, Kohei Wakabayashi, MD, Kimberly Kaneshige, BS, William O. Suddath, MD, Kenneth M. Kent, MD, PhD, Lowell F. Satler, MD, Augusto D. Pichard, MD, and Ron Waksman, MD* Previous studies have documented disparities in both access to invasive cardiovascular procedures and outcomes in patients with Medicaid, Medicare, or no insurance. Outcomes by insurance have yet not been examined in a percutaneous coronary intervention (PCI) population. Data from patients undergoing PCI from June 2000 to June 2009 were retrospectively analyzed. Insurance was categorized as private, Medicare, Medicaid, and uninsured, according to the primary insurance at discharge. The outcome variable of interest was major adverse cardiac events (a composite of death, Q-wave myocardial infarction, and target vessel revascularization) at 1 year. Multivariable Cox regression analysis was stratified according to age <65 and >65 years. Of the 13,573 patients who had undergone PCI, 6,653 (49.0%) had private insurance, 6,150 (45.3%) had Medicare, 486 (3.6%) had Medicaid, and 284 (2.1%) were uninsured. Of the patients <65 years old, Medicaid (hazard ratio [HR] 1.59, 95% confidence interval [CI] 1.04 to 2.43), Medicare (HR 2.18, 95% CI 1.58 to 2.99), and no insurance (HR 2.41, 95% CI 1.36 to 4.27) were associated with greater rates of adjusted major adverse cardiac events at 1 year compared with private insurance. Of the patients >65 years old, only Medicaid (HR 3.07, 95% CI 1.09 to 8.61) was associated with a greater rate of adjusted major adverse cardiac events at 1 year. In conclusion, patients with government-sponsored insurance and no insurance have worse cardiovascular outcomes than patients with private insurance after PCI at 1 year. This implies that the provision of health insurance alone might not have a dramatic effect on cardiovascular outcomes after PCI. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:675– 680) Socioeconomic status has been most commonly approximated by income level; however, it can also be conceptualized by other measures such as an individual’s education, occupation, living environment, and access to medical care. Health insurance (or the lack thereof) is an important measure of one’s access to care, although the effect of insurance type on cardiovascular outcomes has been primarily examined in the setting of acute myocardial infarction (MI). Most of these studies have shown a less utilization of invasive procedures and worse outcomes for patients with Medicaid or no insurance.1–7 Because different patterns of reimbursement, practice patterns, and health behaviors might apply to patients with no insurance or government-sponsored insurance, cardiovascular outcomes after percutaneous coronary intervention (PCI) could also vary according to a patient’s primary insurance. We, therefore, sought to examine the effect of insurance type on major adverse cardiac events (MACE) at 1 year in a diverse population undergoing PCI, hypothesizing that patients with Medicaid, Medicare, or no Washington Hospital Center, Washington, District of Columbia. Manuscript received September 9, 2010; manuscript received and accepted October 19, 2010. *Corresponding author: Tel: (202) 877-2812; fax: (202) 877-2715. E-mail address:
[email protected] (R. Waksman). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.041
insurance would have greater rates of MACE than patients with private insurance. Methods The clinical, procedural, and follow-up data for patients undergoing PCI at a single center were prospectively entered and retrospectively analyzed. Indications for PCI included stable angina pectoris, unstable angina pectoris, and acute MI. This database was then merged with the hospital billing database, which included the primary insurance type (or lack thereof) and zip code for each patient at hospital discharge. The zip code was matched with United States Census Bureau data regarding the median household income by zip code to approximate the patient’s household income.8 The present study included patients who had undergone PCI from June 2000 to June 2009. All patients had received aspirin 325 mg and clopidogrel 300 to 600 mg (at the operator’s discretion) before the procedure. Anticoagulation regimens were chosen at the operator’s discretion and included unfractionated heparin targeted to achieve an activated clotting time of 200 to 300 seconds, with or without a glycoprotein IIb/IIIa inhibitor, or bivalirudin 0.75 mg/kg, followed by an infusion of 1.75 mg/kg/hour for the duration of the procedure. After the procedure, aspirin was prescribed indefinitely, and clopiwww.ajconline.org
676
The American Journal of Cardiology (www.ajconline.org)
dogrel was prescribed for a minimum of 1 month for patients receiving bare metal stents and 6 months for patients receiving drug-eluting stents. The institutional review board at the Washington Hospital Center and MedStar Health Research Institute (Washington, DC) approved the present study. A dedicated data coordinating center performed all data management and analyses. The prespecified clinical and laboratory data during hospitalization were obtained from the hospital charts and were reviewed by independent research personnel who were unaware of the objectives of the present study. All patients routinely underwent pre- and post-PCI 12-lead electrocardiography to detect procedure-related ischemic changes and/or the presence of new pathologic Q waves. Blood samples were taken at 6 and 24 hours before and after PCI to assess the creatine kinase-MB biomarker level. If the creatine kinase-MB level was elevated to greater than the reference range (4 mg/dl), measurement was repeated every 8 hours until it had returned to less than the reference range. Clinical follow-up evaluations were conducted at 30 days, 6 months, and 1 year by telephone interview or office visits. The occurrence of major late clinical events was recorded and included death (all-cause), Q-wave MI, target vessel revascularization (TVR), and stent thrombosis. The primary end point was MACE, a composite of death from all causes, Q-wave MI, and TVR. Q-wave MI was defined as the appearance of new pathologic Q waves in the coronary distribution of the treated artery, with an increase in the creatine kinase-MB to ⱖ2 times the reference value. TVR was defined as revascularization occurring in any area along the previously treated vessel. An acute presentation was defined as presentation with non–ST-segment elevation MI or ST-segment elevation MI. Major bleeding was quantified according to the Thrombolysis In Myocardial Infarction study group definition and consisted of intracranial hemorrhage or clinically overt bleeding, with a decrease in hemoglobin of ⱖ5 g/dl or hematocrit of ⱖ15%. Primary insurance type was categorized as private, Medicare, Medicaid, or uninsured. Race was defined according to the patient’s response at admission. The patients identified themselves as African-American, Caucasian Asian, Hispanic, or Native American, and could only select one category. The patients were then defined as African-American or non-African-American for purposes of comparison. Two of us (M.A.G., R.W.) were primarily responsible for the data analysis. Continuous variables are presented as mean ⫾ SD and categorical variables as percentages. Differences in continuous variables between groups were compared using Student’s t test. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. The mean values for the different groups were compared using analysis of variance. Thirty-day and 1-year outcomes were compared using the log-rank test and are presented as Kaplan-Meier percentages. A p value ⬍0.05 was considered statistically significant. To test the independent effects of insurance type and income on the interval to a MACE, we constructed 2 multivariable Cox regression models. Given that a significant interaction was present between patient age and insurance type, we performed separate multivariable regression analyses for patients ⬍65 and ⱖ65 years old. This was in part because Medicare primarily
covers patients with end-stage renal disease in the population ⬍65 years old and covers a much broader range of patients ⱖ65 years old. Private insurance was used as the reference group for both regression analyses. A very small number of patients ⱖ65 years old were without insurance, and they were excluded from the regression analysis for patients ⱖ65 years old. We selected covariables for both models according to significant univariate p values and overall clinical relevance. The covariables in each model, in addition to insurance type, included black race, age, gender, cardiogenic shock, acute coronary syndrome, current smoking, and a history of PCI, coronary artery bypass grafting, chronic renal insufficiency, congestive heart failure, diabetes mellitus, systemic hypertension, and peripheral vascular disease. Given that certain insurance types (i.e., Medicaid and no insurance) have been tightly linked to median income, we excluded median income by zip code from the models to avoid collinearity. The covariables in the model are expressed as hazard ratios (HRs) with 95% confidence intervals (CIs). Statistical analyses were performed using SAS, version 9.1 (SAS Institute, Cary, North Carolina). Results The present study included 13,573 patients who had undergone PCI (Table 1). Overall, 6,653 (49.0%) had private insurance, 6,150 (45.3%) had Medicare, 486 (3.6%) had Medicaid, and 284 (2.1%) were uninsured at hospital discharge after PCI. Patients with Medicaid ($41,087), no insurance ($55,417), and Medicare ($59,459) all had a lower median household income compared to patients with private insurance ($62,463; p ⬍0.001 for trend.) Overall, 8,973 patients (66.1%) were white, 3,482 (25.7%) were black, 473 (3.5%) were Asian, 147 (1.1%) were Hispanic, and 51 (0.4%) were Native American. Patients with Medicare had more co-morbidities, including previous coronary artery bypass grafting or PCI, congestive heart failure, systemic hypertension, hyperlipidemia, chronic renal insufficiency, and peripheral vascular disease. Important exceptions were found, however, with diabetes mellitus more frequent in Medicaid patients, and current smoking much more common in patients with Medicaid or no insurance. Of the 13,573 patients, 51.9% were ⬍65 years old. Of those ⬍65 years old, 81.7% had private insurance, 8.7% had Medicare, 5.8% had Medicaid, and 3.8% were uninsured. Of the patients aged ⱖ65 years, 85.2% had Medicare, 13.5% had private insurance, 1.1% had Medicaid, and 0.2% were uninsured. The overall prevalence of co-morbidities and MACE rates in the population ⱖ65 years old was greater than in the population ⬍65 years old; however, the trends for differences by insurance type were similar. Patients with Medicaid (23.3%) or no insurance (29.9%) were more likely to present with STsegment elevation MI than were patients with Medicare (9.2%) or private insurance (12.3%; p ⬍0.001). A similar trend held for a presentation with non–ST-segment elevation MI, which was more common for patients with Medicaid (15.2%) and no insurance (19.4%) than for those with Medicare (7.8%) or private insurance (8.7%; p ⬍0.001). A presentation with cardiogenic shock was
Coronary Artery Disease/Insurance and PCI Outcomes
677
Table 1 Baseline characteristics by insurance type Variable Age (years) Men Race Caucasian African-American Asian Hispanic Median household income* ST-segment elevation myocardial infarction Non–ST-segment elevation myocardial infarction Shock on presentation Previous coronary artery bypass surgery Previous percutaneous coronary intervention Heart failure Diabetes mellitus Hypertension† Dyslipidemia‡ Chronic renal insufficiency Peripheral vascular disease Current smoker
Private (n ⫽ 6,653)
Medicare (n ⫽ 6,150)
Medicaid (n ⫽ 486)
Uninsured (n ⫽ 284)
p Value
57.3 ⫾ 9.4 4,891 (73.6%)
73.3 ⫾ 9.5 3,545 (57.7%)
55.1 ⫾ 10.6 262 (54.1%)
53.7 ⫾ 9.0 208 (73.2%)
⬍0.001 ⬍0.001
4,586 (68.9%) 1,489 (22.4%) 243 (3.7%) 71 (1.1%) $62463 ⫾ 20,729 815 (12.3%) 578 (8.7%) 202 (3.1%) 906 (13.7%) 1,562 (24.8%) 571 (8.9%) 2,064 (31.3%) 5,492 (82.9%) 5,883 (89.0%) 461 (7.0%) 629 (9.6%) 1,842 (27.7%)
4,143 (67.4%) 1,575 (25.6%) 198 (3.2%) 50 (0.8%) $59459 ⫾ 20,831 567 (9.2%) 480 (7.8%) 250 (4.1%) 1,499 (24.6%) 1,551 (27.9%) 1,246 (21.1%) 2,358 (38.7%) 5,523 (90.3%) 5,423 (89.3%) 1,274 (20.9%) 1,335 (22.0%) 829 (13.5%)
95 (19.5%) 324 (66.7%) 17 (3.5%) 17 (3.5%) $41087 ⫾ 15,778 113 (23.3%) 74 (15.2%) 35 (7.3%) 64 (13.2%) 107 (24.0%) 86 (18.3%) 219 (45.5%) 428 (88.2%) 406 (84.4%) 85 (17.7%) 61 (12.8%) 236 (48.6%)
149 (52.5%) 94 (33.1%) 15 (5.3%) 9 (2.2%) $55417 ⫾ 16,667 85 (29.9%) 55 (19.4%) 22 (7.8%) 22 (7.8%) 47 (17.4%) 32 (11.7%) 77 (27.4%) 211 (74.6%) 217 (77.2%) 11 (3.9%) 14 (4.9%) 164 (57.7%)
⬍0.001 ⬍0.001 0.21 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
Data are presented as mean ⫾ SD or n (%). * Median income by zip code † History of hypertension diagnosed and/or treated with medication or currently treated with diet and/or medication by a physician. ‡ Included patients with previously documented diagnosis of dyslipidemia; patients could be treated with diet or medication; a new diagnosis could be made during present hospitalization by total cholesterol level ⬎160 mg/dl; did not include elevated triglycerides. Table 2 Procedural characteristics (lesion-based) Variable Coronary vessel of intervention Right Left main Left anterior descending Left circumflex Saphenous vein graft Bare metal stent Drug-eluting stent Procedural success Intravascular ultrasound Stented length (mm) Heparin Bivalirudin Glycoprotein IIb/IIIa inhibitor
Private (n ⫽ 10,910)
Medicare (n ⫽ 10,457)
Medicaid (n ⫽ 764)
Uninsured (n ⫽ 480)
p Value
3,580 (32.8%) 80 (0.7%) 4,376 (40.1%) 2,442 (22.4%) 385 (3.5%) 6,169 (58.8%) 7,306 (72.9%) 10,441 (98.1%) 6,201 (57.9%) 19.9 ⫾ 6.7 1,054 (15.8%) 4,807 (72.3%) 801 (12.4%)
3,213 (30.7%) 251 (2.4%) 3,832 (36.6%) 2,399 (22.9%) 718 (6.9%) 5,622 (55.4%) 6,338 (65.9%) 10,046 (97.9%) 5,677 (55.4%) 19.7 ⫾ 6.8 840 (13.7%) 4,727 (76.9%) 447 (7.4%)
247 (32.3%) 1 (0.1%) 280 (36.6%) 214 (28.0%) 19 (2.5%) 379 (51.6%) 413 (56.8%) 739 (98.0%) 378 (50.1%) 20.0 ⫾ 6.2 90 (18.5%) 329 (67.7%) 58 (12.1%)
168 (35.0%) 6 (1.3%) 186 (38.8%) 113 (23.5%) 6 (1.3%) 171 (36.5%) 252 (55.3%) 461 (97.3%) 200 (42.3%) 20.6 ⫾ 6.3 50 (17.6%) 179 (63.0%) 59 (21.0%)
0.005 ⬍0.001 ⬍0.001 0.004 ⬍0.001 ⬍0.001 ⬍0.001 0.61 ⬍0.001 0.18 ⬍0.001 ⬍0.001 ⬍0.001
Data are presented as n (%) or mean ⫾ SD.
also more common in patients with Medicaid (7.3%) and no insurance (7.8%) than in those with Medicare (4.1%) or private insurance (3.1%; p ⬍0.001). The overall procedural success rate was 98.0% and did not differ according to insurance type (Table 2). Patients with Medicare were more likely to have undergone PCI of the left main artery or a saphenous vein graft. Patients with private insurance, however, were more likely to have undergone PCI of the left anterior descending artery. Patients with private insurance were also more likely to undergo intravascular ultrasound-guided PCI, although total stented length did not differ by insurance type. Patients with Medi-
care and private insurance were more likely to have received bivalirudin. In contrast, patients with Medicaid or without insurance were more likely to have received a glycoprotein IIb/IIIa inhibitor. Patients with private insurance were also more likely to have received ⱖ1 drugeluting stent (78.2%) than were those with Medicare (72.0%), no insurance (60.1%), or Medicaid (59.9%, p ⬍0.001). Patients with Medicaid (6.2%), Medicare (4.5%), or no insurance (5.1%) were more likely to die in-hospital (p ⬍0.001 for trend) than were patients with private insurance (1.5%). A similar trend for MACE at 30 days also emerged,
678
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Univariable outcomes by insurance type at 1 year (p values for analysis of variance). QWMI ⫽ Q-wave myocardial infarction.
Figure 3. Adjusted HRs for MACE among different insurance types, with private insurance as reference. (A) Patients ⬍65 years old and (B) patients ⱖ65 years old. Figure 2. Kaplan-Meier analysis for MACE (p ⬍0.001, log-rank test).
with greater rates for Medicaid (9.7%), Medicare (5.5%), and no insurance (8.1%) than for private insurance (2.2%; p ⬍0.001). These differences remained significant at 1 year, with greater rates of death for patients with Medicaid, Medicare, or no insurance compared to those with private insurance (Figure 1). The rate of TVR at 1 year was also significantly greater in patients with Medicaid (16.0%) or no insurance (14.3%), although the rates in the patients with Medicare (8.9%) or private insurance (8.1%) were similar (p ⫽ 0.003). Q-wave MI at 1 year followed a similar trend, with 3.3% of Medicaid and 1.5% of uninsured patients compared to 0.5% of Medicare and 0.2% of privately insured patients developing Q-wave MI (p ⬍0.001). Stent thrombosis at 1 year was also more common in patients with Medicaid (5.4%) and no insurance (2.6%) than in patients with Medicare (0.9%) or private insurance (0.9%; p ⬍0.001). Regarding MACE at 1 year, patients with Medicaid, Medicare, and no insurance were more likely to experience adverse events than were patients with private insurance (Figure 2). After adjustment with multivariable Cox regression analysis within the population ⬍65 years old (Figure 3), patients
with Medicaid (HR 1.59, 95% CI 1.04 to 2.43), Medicare (HR 2.18, 95% CI 1.58 to 2.99), and no insurance (HR 2.41, 95% CI 1.36 to 4.27) remained more likely than patients with private insurance to have had MACE at 1 year. Other covariables significantly associated with adjusted MACE included cardiogenic shock on presentation (HR 4.58, 95% CI 3.12 to 6.73), congestive heart failure (HR 1.49, 95% CI 1.10 to 2.02), and systemic hypertension (HR 1.81, 95% CI 1.27 to 1.59). However, after adjustment with multivariable Cox regression analysis for the population ⱖ65 years old (Figure 3), only Medicaid remained significantly associated with MACE at 1 year (HR 3.07, 95% CI 1.09 to 8.61). Medicare was not associated with MACE, and the small number of uninsured patients in this population precluded their analysis. Other covariables significantly associated with adjusted MACE included age (HR per 10 years 1.29, 95% CI 1.10 to 1.50), cardiogenic shock (HR 4.80, 95% CI 3.61 to 6.38), an acute presentation (HR 1.64, 95% CI 1.28 to 2.11), current smoking (HR 1.60, 95% CI 1.18 to 2.16), chronic renal insufficiency (HR 1.63, 95% CI 1.30 to 2.05), congestive heart failure (HR 1.61, 95% CI 1.29 to 2.01), diabetes
Coronary Artery Disease/Insurance and PCI Outcomes
mellitus (HR 1.30, 95% CI 1.06 to 1.60), and peripheral vascular disease (HR 1.52, 95% CI 1.22 to 1.90). Discussion We found that both government-sponsored insurance and no insurance, in comparison to private insurance, were associated with more adverse cardiac events in a population ⬍65 years old after PCI at 1 year. These results were also consistent for the individual outcomes of death, Q-wave MI, TVR, and stent thrombosis. Only the effect of Medicaid, however, remained significant in the population ⱖ65 years old. Previous studies have documented disparities in access to invasive cardiovascular procedures according to insurance type (in the United States), although most have focused on an acute MI population. These studies found that patients with Medicaid1,2,4 – 6 and no insurance,1– 4,6 compared to patients with private insurance, were less likely to undergo cardiovascular procedures, with only 1 study demonstrating that Medicare patients were less likely.2 Fewer studies, however, have examined cardiovascular outcomes by insurance type. A study of payer status in the National Registry of Myocardial Infarction by Sada et al4 included 17,600 patients ⬍65 years old with acute MI. Their study showed an increased rate of adjusted in-hospital mortality among Medicaid patients (odds ratio [OR] 1.55, 95% CI 1.19 to 2.01) than among private insurance patients.4 Similarly, Canto et al6 examined a more inclusive cohort of 332,221 patients in the same registry. That study, using private insurance as a reference, showed increased adjusted in-hospital mortality for patients with Medicaid (OR 1.30, 95% CI 1.14 to 1.48) and no insurance (OR 1.29, 95% CI 1.12 to 1.48). Medicare patients also appeared to have a greater rate of in-hospital mortality, although this effect was confined to the population ⬍65 years old.6 Finally, Calvin et al5 reported on 37,345 patients with non–ST-segment elevation acute coronary syndrome within the “Can Rapid Risk Stratification of Unstable Angina Patients Suppress Adverse Outcomes with Early Implementation of the ACC [American College of Cardiology] Guidelines” (CRUSADE) initiative. Their analysis showed that Medicaid patients ⬍65 years old had a greater risk of adjusted in-hospital mortality compared to those with private insurance (OR 1.33, 95% CI 1.08 to 1.63); uninsured patients were not assessed.5 The etiology of this insurance-based disparity is multifactorial. It could be that insurance status encapsulates an array of demographic factors to approximate the decidedly vague term “socioeconomic status.” In addition, and as others have observed, socioeconomic status is inherently difficult to measure and prone to misinterpretation.9 One would not conclude from these results that improving access to healthcare does not improve patients’ health. Rather, this disparity draws attention to the often oversimplified manner in which health insurance is viewed as the root of health disparities. More specifically, a multitude of factors, in concert with insurance, influence both the access to care and patient behavior. For example, a recent study by Gerber et al10 demonstrated that, in addition to individual socioeconomic status, neighborhood socioeconomic status also influenced survival after MI. Even in the setting of universal healthcare, such as in Canada, the profound influence of
679
socioeconomic status on both access to cardiac care and outcomes after acute MI remains.11 This also appeared evident in our population, which had overcome the initial barriers to access and undergone PCI. Furthermore, physician-level factors could also influence patient-level outcomes.12 For example, in the present study, both drugeluting stent and intravascular ultrasound use were less likely for patients with nonprivate insurance. To our knowledge, the present study is the first to examine the outcomes by insurance type after PCI exclusively in a broad-based and diverse population. Our analysis was also unique in that all patients were referred for PCI to a single, high-volume center, unlike previous analyses of insurance-based disparities. This eliminates any hospital-level characteristics that might have confounded the multicenter analyses.13 Nevertheless, given the retrospective nature of our analysis, it was prone to difficulties common to all nonrandomized studies. Patients with government insurance and no insurance were invariably sicker and had more co-morbidities than did patients with private insurance in our population; thus, confounding variables might have been present that were not accounted for in the final Cox regression models for MACE. Furthermore, the population ⬍65 years old covered by Medicare is very different from the population ⱖ65 years old and covered by Medicare. The number of patients without insurance was also considerably less than that for the other insurance types, which could have introduced bias. We were unable to assess the effect of the median household income by zip code on the receipt of drug-eluting stents because of the collinearity between insurance status and income. Previous studies, however, have shown that increasing income does not significantly attenuate the effect of insurance status on health outcomes.14 In addition, we did not have access to very detailed income data; thus, our use of median income by zip code might have resulted in a misclassification of the actual income.15,16 Furthermore, we defined insurance status according to the primary coverage at discharge. Given that patients might have been uninsured at PCI and then had received insurance coverage during their hospital stay, we likely underestimated the number of uninsured at PCI. 1. Wenneker MB, Weissman JS, Epstein AM. The association of payer with utilization of cardiac procedures in Massachusetts. JAMA 1990;264:1255– 1260. 2. Canto JG, Rogers WJ, Zhang Y, Roseman JM, French WJ, Gore JM, Chandra NC. The association between the on-site availability of cardiac procedures and the utilization of those services for acute myocardial infarction by payer group. The National Registry of Myocardial Infarction 2 investigators. Clin Cardiol 1999;22:519 –524. 3. Hadley J, Steinberg EP, Feder J. Comparison of uninsured and privately insured hospital patients. Condition on admission, resource use, and outcome. JAMA 1991;265:374 –379. 4. Sada MJ, French WJ, Carlisle DM, Chandra NC, Gore JM, Rogers WJ. Influence of payer on use of invasive cardiac procedures and patient outcome after myocardial infarction in the United States: Participants in the National Registry of Myocardial Infarction. J Am Coll Cardiol 1998;31:1474 –1480. 5. Calvin JE, Roe MT, Chen AY, Mehta RH, Brogan GX Jr, Delong ER, Fintel DJ, Gibler WB, Ohman EM, Smith SC Jr, Peterson ED. Insurance coverage and care of patients with non–ST-segment elevation acute coronary syndromes. Ann Intern Med 2006;145:739 –748. 6. Canto JG, Rogers WJ, French WJ, Gore JM, Chandra NC, Barron HV. Payer status and the utilization of hospital resources in acute myocar-
680
7.
8. 9. 10.
11.
The American Journal of Cardiology (www.ajconline.org) dial infarction: a report from the National Registry of Myocardial Infarction. 2. Arch Intern Med 2000;160:817– 823. Canto JG, Rogers WJ, Chandra NC, French WJ, Barron HV, Frederick PD, Maynard C, Every NR. The association of sex and payer status on management and subsequent survival in acute myocardial infarction. Arch Intern Med 2002;162:587–593. United States Census Bureau. Median household income and median owner-occupied house value by zipcode. Available at: http://factfinder. census.gov/servlet. Accessed on October 19, 2009. Braveman PA, Cubbin C, Egerter S, Chideya S, Marchi KS, Metzler M, Posner S. Socioeconomic status in health research: one size does not fit all. JAMA 2005;294:2879 –2888. Gerber Y, Benyamini Y, Goldbourt U, Drory Y; Israel Study Group on First Acute Myocardial Infarction. Neighborhood socioeconomic context and long-term survival after myocardial infarction. Circulation 2010 26;121:375–383. Alter DA, Naylor CD, Austin P, Tu JV. Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction. N Engl J Med 1999;341:1359 –1367.
12. Gaglia MA Jr, Torguson R, Xue Z, Gonzalez MA, Collins SD, BenDor I, Syed AI, Maluenda G, Delhaye C, Hanna N, Wakabayashi K, Kaneshige K, Suddath WO, Kent KM, Satler LF, Pichard AD, Waksman R. Insurance type influences the use of drug-eluting stents. J Am Coll Cardiol Interv 2010;3:773–779. 13. Halm EA, Lee C, Chassin MR. Is volume related to outcome in health care? A systematic review and methodologic critique of the literature. Ann Intern Med 2002;137:511–520. 14. Ross JS, Bradley EH, Busch SH. Use of health care services by lower-income and higher-income uninsured adults. JAMA 2006;295: 2027–2036. 15. Gornick ME, Eggers PW, Reilly TW, Mentnech RM, Fitterman LK, Kucken LE, Vladeck BC. Effects of race and income on mortality and use of services among Medicare beneficiaries. N Engl J Med 1996; 335:791–799. 16. Soobader M, LeClere FB, Hadden W, Maury B. Using aggregate geographic data to proxy individual socioeconomic status: does size matter? Am J Public Health 2001;91:632– 636.
Catheter Aspiration in ST-Elevation Myocardial Infarction and Different Extent of Coronary Thrombus Igor Balevski, MDa, Mojca Cizek Sajko, PhDa, Vojko Kanic, MDa, and Marko Noc, MD, PhDb,* Manual catheter aspiration appears to be a useful adjunct to primary percutaneous coronary intervention (PCI) in ST-elevation myocardial infarction. We investigated effects of catheter aspiration during primary PCI in patients with different extents of coronary thrombus. The study included 46 patients with no or possible thrombus (thrombus scale [TS] grades 0 to 1) and 135 patients with angiographic evidence of obvious thrombus (TS grades 2 to 5). Reference vessel diameter, which was significantly larger in the group with TS grades 2 to 5 (3.4 vs 3.2 mm, p ⴝ 0.004), was the only independent predictor of angiographically visible thrombus (odds ratio 3.3, 95% confidence interval 1.3 to 8.7, p ⴝ 0.015, per millimeter increase). Aspiration catheter was successfully advanced across the lesion in 89% of patients with TS grades 0 to 1 and 96% of those with TS grades 2 to 5 (p ⴝ 0.115). Number of aspirations varied from 1 to 5 and was significantly larger in patients with TS grades 2 to 5. Visually observable aspirate was obtained in 90% of patients with TS grades 2 to 5 and in 67% of patients with TS grades 0 to 1 (p <0.001) with more patients with TS grades 2 to 5 having aspirate >5 mm in length (49% vs 11%, p <0.001). Final Thrombolysis In Myocardial Infarction grade 3 flow (89% vs 92%), residual TS (0.2 vs 0.1), frequency of distal embolization (2% vs 6%), and early complete ST resolution (65% vs 70%) were comparable between groups with TS grades 0 to 1 and 2 to 5. In conclusion, although the amount of aspirate is larger in patients with angiographically obvious thrombus, visually observable aspirate can be obtained in most patients without definite signs of thrombus. Extent of coronary thrombus does not influence primary PCI result if manual aspiration is used. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:681– 684) Manual catheter aspiration appears to be a useful adjunct to primary percutaneous coronary intervention (PCI) in STelevation myocardial infarction (STEMI).1 Aspiration before stenting has been shown to improve the angiographic result1 and clinical outcome2 of patients with STEMI. Therefore, manual aspiration has been upgraded from class IIb level B in European STEMI guidelines of 20083 to class IIa level A in recently published European revascularization guidelines.4 In the present study, we investigated effects of catheter aspiration in patients with STEMI and different sizes of coronary thrombus determined by coronary angiography. We investigated how often an aspirate visible to the eye can be obtained despite an absence of definite angiographic signs of coronary thrombus. We also hypothesized that presence of large thrombus does not compromise results of primary PCI if manual catheter aspiration is used as part of the intervention. Methods The study protocol was approved by the National Ethical Committee (No. 23/05/09). The study enrolled consecutive patients with STEMI undergoing primary PCI with manual
a Department of Cardiology, University Clinical Center, Maribor, Slovenia; bCenter for Intensive Internal Medicine, University Medical Center, Ljubljana, Slovenia. Manuscript received August 24, 2010; revised manuscript received and accepted October 19, 2010. *Corresponding author: Tel/fax: 386-1-522-22-96. E-mail address:
[email protected] (M. Noc).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.042
catheter aspiration. Catheter aspiration was not attempted if the infarct-related artery had a diameter ⱕ2 mm and if significant tortuosity/calcification would not allow for nontraumatic passage of an aspiration catheter. All interventions were performed by 1 of 2 experienced operators. After passing a guidewire, manual aspiration with a 6 Fr guiding catheter-compatible Diver (Invatec, Brescia, Italy), Export (Medtronic, Minneapolis, Minnesota), Pronto (Vascular Solutions, Inc., Minneapolis, Minnesota), or Quickcath (Travenol Laboratories, Inc., Baxter Healthcare Corporation, Deerfield, Illinois) was performed. Predilatation using a ⱕ2-mm balloon was used only if initial passage of aspiration catheter was not possible. Number of aspiration passes was left to the discretion of the operator. In general, aspiration passes were performed until visually observable aspirate was obtained in the syringe. After aspiration, direct stenting was the preferred strategy. Sizing of a stent was performed after intracoronary administration of nitroglycerin (100 to 200 g). Predilatation after aspiration and before stenting was used if a distal vessel was not adequately visualized for safe stent placement. Epicardial coronary flow was assessed by Thrombolysis In Myocardial Infarction (TIMI) classification.5 Thrombus burden in an infarct-related artery was estimated by coronary angiography using a TIMI thrombus scale (TS).6 Based on TS grade after passing a guidewire across the culprit lesion to distal vessel, patients were categorized as having no or possible thrombus (TS grades 0 to 1) or definite thrombus (TS grades 2 to 5). The 2 groups were compared in clinical parameters, angiographic characteristics, amount www.ajconline.org
682
The American Journal of Cardiology (www.ajconline.org) Table 2 Periprocedural angiographic characteristics
Table 1 Baseline characteristics of patients Variable
Age (years) Men Hypertension Diabetes mellitus Hypercholesterolemia Smoker Onset of symptoms to coronary angiography (minutes) Antiaggregation before coronary angiography Aspirin ⫹ clopidogrel ⫹ heparin Glycoprotein Ilb/IIIa inhibitors Antiaggregation to coronary angiography (minutes)
Thrombus Scale 0–1 (n ⫽ 46)
Thrombus Scale 2–5 (n ⫽ 135)
p Value
Variable
59 ⫾ 13 36 (78%) 26 (57%) 8 (17%) 30 (65%) 18 (39%) 210 (127–304)
63 ⫾ 12 89 (66%) 87 (64%) 31 (23%) 76 (56%) 49 (36%) 168 (107–220)
0.071 0.118 0.338 0.427 0.289 0.664 0.232
Infarct-related coronary artery Left anterior descending 22 (48%) 52 (39%) Left circumflex 11 (24%) 15 (11%) Right 13 (28%) 68 (50%) Reference diameter (mm) 3.2 ⫾ 0.4 3.4 ⫾ 0.4 Predilatation before 1 (2%) 18 (13%) aspiration Passage of aspiration 41 (89%) 129 (96%) catheter Number of aspirations 1.3 ⫾ 0.5 (1–2) 1.8 ⫾ 0.9 (1–5) (range) Amount of aspirate Not visible 15 (33%) 14 (10%) ⬍5 mm 26 (57%) 55 (41%) ⬎5 mm 5 (10%) 66 (49%) Final technique Direct stenting 37 (80%) 95 (70%) Stenting with predilatation 8 (17%) 31 (23%) Only aspiration/balloon 1 (2%) 9 (7%) Final Thrombolysis In 41 (89%) 124 (92%) Myocardial Infarction grade 3 flow Residual thrombus scale 0.2 ⫾ 0.8 0.1 ⫾ 0.5 grade Distal embolization 1 (2%) 8 (6%) Early ST-segment resolution ⬎70% 30 (65%) 95 (70%) 30–70% 12 (26%) 32 (24%) ⬍30% 4 (9%) 8 (6%)
0.924
41 (89%)
121 (90%)
5 (11%)
14 (10%)
70 (49–102)
62 (34–88)
0.362
Values are presented as mean ⫾ SD, number of patients (percentage), or median (interquartile range).
of aspirated material, and final angiographic result including TIMI flow in the infarct-related artery, frequency of distal embolization, and residual TS. Early ST-segment resolution was estimated as previously reported.7 Amount of aspirate obtained by catheter aspiration was graded as not visible to the eye, ⬍5 mm in length, and ⬎5 mm in length. Numerical data are presented as mean ⫾ SD or median with interquartile range and were analyzed by unpaired t test if normally distributed or by Mann–Whitney test if skewed. Categorical data are expressed as proportion and were analyzed by chi-square test. For post hoc comparison of categorical variables, Keppel modification of Bonferroni correction was used. Multiple logistic regression was used to analyze predictors of thrombus size at first coronary angiography. A p value ⬍0.05 was considered statistically significant. Results The study included 46 patients with no or possible thrombus (TS grades 0 to 1) and 135 patients with obvious angiographic signs of thrombus (TS grades 2 to 5). The 2 groups were comparable in age, gender, and risk factors for coronary artery disease (Table 1). There were also no significant differences in time delay between onset of symptoms and urgent coronary angiography and time delay between administration of antiaggregation therapy and coronary angiography between groups. In patients with obvious thrombus (TS grades 2 to 5), the right coronary artery was more often (50% vs 28%) and the left circumflex artery less often (11% vs 24%) a culprit
Thrombus Scale 0–1 (n ⫽ 46)
Thrombus Scale 2–5 (n ⫽ 135)
p Value 0.015 0.268 0.033 0.009 0.004 0.033 0.115 ⬍0.001 ⬍0.001 ⬍0.001 0.063 ⬍0.001 0.185 0.427 0.249 0.574
0.294 0.536 0.737
artery (Table 2). Reference vessel diameter, which was significantly larger in the group with TS grades 2 to 5 (3.4 vs 3.2 mm), was the only independent predictor of angiographically visible thrombus (odds ratio 3.3, 95% confidence interval 1.3 to 8.7, p ⫽ 0.015, per millimeter increase). TIMI grade flow in the infarct-related artery at first coronary angiography and after passage of the guidewire was significantly greater in the group with TS grades 0 to 1 (Figure 1). At the same time, TS grade was higher in the group with TS grades 2 to 5. Aspiration catheter was successfully advanced across the culprit lesion in 89% of patients with TS grades 0 to 1 and 96% of those with TS grades 2 to 5 (p ⫽ 0.115; Table 2). To facilitate passage of the aspiration catheter, predilatation was required in 2% and 13%, respectively (p ⫽ 0.033). Number of aspiration passes varied from 1 to 5 and was larger in the group with TS grades 2 to 5. Visually observable aspirate was obtained in 90% of patients with TS grades 2 to 5 and 67% of those with TS grades 0 to 1 (p ⬍0.001) with more patients in the group with TS grades 2 to 5 having aspirate ⬎5 mm in length (49% vs 11%, p ⬍0.001). After catheter aspiration, direct stenting was performed in ⱖ70% of patients without significant difference between groups. Penetrations of drug-eluting stents were 27% in the group with TS grades 0 to 1 and 8% in the group with TS grades 2 to 5 (p ⫽ 0.002). Final TIMI grade 3 flow (89% vs 92%),
Coronary Artery Disease/Catheter Aspiration in STEMI
683
Figure 1. Mean Thrombolysis In Myocardial Infarction grade flow and mean Thrombolysis In Myocardial Infarction thrombus scale grades 0 to 1 (no/possible thrombus) (gray bars) and 2 to 5 (definite thrombus) (striped bars) at first coronary angiography (CAG) after passage of a guidewire and at final coronary angiography. Values are means ⫾ SDs.
residual TS (0.2 vs 0.1), frequency of distal embolization (2% vs 6%), and early complete ST-segment resolution (65% vs 70%) were comparable between the groups with TS grades 0 to 1 and 2 to 5 (Figure 1 and Table 2). Discussion Our study, which reflects real-world practice, demonstrated that size of coronary thrombus in STEMI, estimated by coronary angiography, does not influence final angiographic result if manual aspiration is used as an adjunct to primary PCI. This is in contrast to previous studies using standard techniques of PCI with balloon dilatation and stenting, in which a larger thrombotic burden unequivocally predicted worse angiographic and clinical results.8 –10 This pinpoints the importance of manual aspiration as an adjunct
to primary PCI, which is in contrast to thrombectomy devices relatively simple and cheap and may be affordable in every catheterization laboratory treating patients with STEMI. However, it is important to emphasize that although the amount of aspirate is larger in patients with angiographically obvious thrombus (TS grades 2 to 5), visually observable aspirate can be obtained in ⬎60% of patients without obvious thrombus (TS grades 0 to 1). This finding is in accord with previous investigations showing that coronary angiography is a rather insensitive method for thrombus detection11,12 and argues for systematic rather than selective use of manual aspiration. Additional myocardial salvage, also when aspirating less material, may be demonstrated only if a more sensitive method such as magnetic resonance imaging is employed.
684
The American Journal of Cardiology (www.ajconline.org)
Our study identified vessel diameter rather than the vessel as an independent predictor of definite thrombus. This seems to be logical because a larger vessel is likely to accommodate larger thrombus. Passage of a guidewire significantly increased TIMI flow and decreased TS grade in the 2 groups. This is probably related to fragmentation and distal embolization of a soft part of the thrombus. It is important to emphasize that all our patients were pretreated with ⱖ1 antiagreggation agent 60 to 70 minutes before initial coronary angiography. In contrast, decreases in TS grade and increases in TIMI flow after guidewire passage predominantly represent the added value of manual aspiration, although we did not systematically obtain these measurements immediately after aspiration. Because of the relatively small number of patients and several different aspiration catheters used, we also could not address possible differences in effectiveness between these devices. However, we believe that increasing the catheter size from 6 Fr to 7 Fr rather than some difference in design among various manufacturers is probably more important to increase aspiration capability.
4.
5. 6.
7.
8.
9. 1. Svilaas T, Vlaar PJ, van der Horst IC, Diercks GFH, de Smet BJGL, van der Heuvel AFM, Anthonio RL, Jessurun GA, Tan ES, Suurmeijer AJH, Zijlstra F. Thrombus aspiration during primary percutaneous coronary intervention. N Engl J Med 2008;358:557–567. 2. Vlaar PJ, Svilaas T, van der Horst IC, Diercks GF, Fokkema ML, de Smet BJ, van den Heuvel AF, Anthonio RL, Jessurun GA, Tan ES, Suurmeijer AJ, Zijlstra F. Cardiac death and reinfarction after 1 year in the Thrombus Aspiration during Percutaneous coronary intervention in Acute myocardial infarction Study (TAPAS): a 1-year follow-up study. Lancet 2008;371:1915–1920. 3. Van de Werf F, Bax J, Betriu A, Blomstrom-Lundqvist C, Crea F, Falk V, Filippatos G, Fox K, Huber K, Kastrati A, Rosengren A, Steg PG, Tubaro M, Verheugt F, Weidinger F, Weis M; ESC Committee for Practice Guidelines (CPG). The task force on the management of ST-segment elevation acute myocardial infarction of the European
10.
11.
12.
Society of Cardiology. Management of acute myocardial infarction in patients presenting with persistent ST-segment elevation. Eur Heart J 2008;29:2909 –2945. The Task Force on Myocardial Revascularization of the European Society of Cardiology (ESC) and the European Association for CardioThoracic Surgery (EACTS). Guidelines on myocardial revascularization. Eur Heart J 2010;20:2501–2555. The TIMI Study Group. The Thrombolysis In Myocardial Infarction (TIMI) trial: phase l findings. N Engl J Med 1985;312:932–936. Gibson CM, de Lemos JA, Murphy SA, Marble SJ, McCabe CH, Cannon CP, Antman EM, Braunwald E; for the TIMI study group. Combination therapy with abciximab reduces angiographically evident thrombus in acute myocardial infarction: a TIMI 14 substudy. Circulation 2001;103:2550 –2554. Schröder R, Dissmann R, Bruggemann T, Wegscheider K, Linderer T, Tebbe U. Extent of early ST segment elevation resolution: a simple but strong predictor of outcome in patients with acute myocardial infarction. J Am Coll Cardiol 1994;24:384 –391. Burzotta F, Trani C, Romagnoli E, Mazzari MA, Rebuzzi AG, De Vita M, Garramone B, Giannico F, Niccoli G, Biondi-Zoccai GGL, Schiavoni G, Mongiardo R, Crea F. Manual thrombus-aspiration improves myocardial reperfusion. The Randomized Evaluation of the Effect of Mechanical Reduction of Distal Embolization by Thrombus-Aspiration in Primary and Rescue Angioplasty (REMEDIA) Trial. J Am Coll Cardiol 2005;46:371–376. Yip HK, Chen MC, Chang HW, Hang CL, Hsieh YK, Fang CY, Wu CJ. Angiographic morphologic features of infarct-related arteries and timely reperfusion in acute myocardial infarction: predictors of slowflow and no-reflow phenomenon. Chest 2002;122:1322–1332. Sianos G, Papafaklis MI, Daemen J, Vaina S, van Mieghem CA, van Domburg RT, Michalis LK, Serruys PW. Angiographic stent thrombosis after routine use of drug-eluting stents in ST-segment elevation myocardial infarction. J Am Coll Cardiol 2007;50:573–583. Goldstein JA. Angiography for detection of complex and vulnerable atherosclerotic plaque. In: Naghavi M, ed. Asymptomatic Atherosclerosis: Pathophysiology, Detection, and Treatment. New York: Springer, Humana Press; 2010:455– 459. White CJ, Ramee SR, Collins TJ, Escobar AE, Karsan A, Shaw D, Jain SP, Bass TA, Heuser RR, Teirstein PS, Bonan R, Walter PD, Smalling RW. Coronary thrombi increase PTCA risk: angioscopy as a clinical tool. Circulation 1996;93:253–258.
Identifying Patients at Risk for Premature Discontinuation of Thienopyridine After Coronary Stent Implantation Alexandre S. Quadros, MD, PhDa, Dulce I. Welter, RN, MSca, Fernanda O. Camozzatto, MDa, Áurea Chaves, MD, PhDa, Rajendra H. Mehta, MD, MSb, Carlos A. Gottschall, MD, PhDa, and Renato D. Lopes, MD, PhDb,c,* We sought to identify patients at risk for premature discontinuation of thienopyridines and to develop a risk score for thienopyridine adherence after coronary stent implantation. Patients were prospectively included from December 2007 to March 2008. At 1-month follow-up, all patients were given the Morisky questionnaire and asked if they had stopped taking thienopyridines. Multivariate analysis identified predictors of thienopyridine discontinuation; points were assigned to each variable according to the odds ratios and the c-statistic of the score was calculated. Mean age of the 400 patients included was 61.0 ⴞ 10.4 years; 66 patients (16.5%) stopped thienopyridines after 1 month. Reasons for discontinuation were cost (62%), lack of information (17%), and recommendation by another doctor to stop treatment (15%). Factors associated with discontinuation included unmarried status (odds ratio 2.48, p ⴝ 0.046), lack of private health insurance (odds ratio 4.68, p ⴝ 0.041), acute coronary syndrome (odds ratio 2.31, p ⴝ 0.004), nondiabetics (odds ratio 2.20, p ⴝ 0.041), and patients who earned <2 times (odds ratio 8.23, p <0.001) and 2 to 3 times (odds ratio 4.46, p ⴝ 0.021) the minimum wage. Total risk score was 0 to 14 points and was strongly associated with thienopyridine discontinuation. For total scores of 0 to 4, 5 to 8, 9 to 12, and >13, 0%, 7%, 20%, and 37% of patients, respectively, stopped thienopyridines (c-statistic 0.76, p <0.0001). Risk score was also significantly associated with complete adherence as assessed by the Morisky questionnaire (c-statistic 0.74, p <0.001). In conclusion, we have identified patients at risk for premature discontinuation of thienopyridines using variables obtained before stent implantation and developed a risk score that accurately predicts premature thienopyridine discontinuation. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:685– 689) After implantation of coronary stents, nonadherence to dual antiplatelet therapy is the main predictor of stent thrombosis, myocardial infarction, and death.1– 4 The magnitude of this problem recently led the American Heart Association to publish an alert suggesting strategies and behavior recommendations.5 Many investigators have studied the reasons for treatment adherence problems after coronary stent implantation,6 – 8 but predictors for discontinuation of thienopyridines are not well established. Identifying these predictors and using a risk score to assess them was the objective of the present study and could be useful in improving treatment adherence. Methods All consecutive patients presenting to the cardiac catheterization laboratory for coronary stent implantation at the Institute of Cardiology of Rio Grande do Sul from 8:00 A.M. a Institute of Cardiology of Rio Grande do Sul, University Foundation of Cardiology, Porto Alegre, Rio Grande do Sul, Brazil; bDuke Clinical Research Institute, Duke University Medical Center, Durham, North Carolina; cFederal University of Sao Paulo (UNIFESP), Sao Paulo, Sao Paulo, Brazil. Manuscript received August 13, 2010; revised manuscript received and accepted October 19, 2010. *Corresponding author: Tel: 919-668-8241; fax: 919-668-7056. E-mail address:
[email protected] (R.D. Lopes).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.045
to 6:00 P.M. from November 2007 to March 2008 were considered for inclusion in this study. Those patients presenting during night or weekend hours, when physicians work on call, were not eligible. This institution is a dedicated center for cardiology, performing approximately 2,500 percutaneous coronary interventions and ⬎7,000 diagnostic catheterization procedures per year. Exclusion criteria included participation in studies involving antiplatelet treatment, unsuccessful procedures, major cardiovascular events during hospitalization, and refusal to sign informed consent. The study was approved by the institutional review board. No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, drafting and editing of the report, and its final contents. Clinical, socioeconomic, and angiographic characteristics of the patients were prospectively collected and included in a dedicated database. Hypertension was considered present in patients with a medical diagnosis or those using antihypertensive drugs. Dyslipidemia was considered present in patients with a medical diagnosis or those using antilipemic drugs. Depression was considered present in patients using antidepressives or in those reporting this diagnosis as assessed by a physician. All patients received aspirin 100 mg/day and clopidogrel 75 mg/day.6 A bolus dose of clopidogrel 300 mg was administered ⱖ6 hours before the procedure. In www.ajconline.org
686
The American Journal of Cardiology (www.ajconline.org)
Table 1 Characteristics of patients taking and not taking thienopyridines at 30 days Variable
Age (years) Men Current smoker Hypertension* Dyslipidemia† Diabetes mellitus Depression Chronic heart failure Chronic obstructive pulmonary disease Chronic renal failure Emergency procedure Acute coronary syndrome Unmarried Education (years)‡ Salary ⬍2 times minimum wage 2–3 times minimum wage ⬎3 times minimum wage Lack of private health insurance
Overall
Taking Thienopyridines
Not Taking Thienopyridines
p Value
(n ⫽ 400)
(n ⫽ 334)
(n ⫽ 66)
61 ⫾ 10 259 (65%) 113 (28%) 311 (78%) 222 (56%) 99 (25%) 44 (11%) 11 (3%) 13 (3%) 7 (2%) 90 (22%) 181 (45%) 27 (7%) 7⫾5
61 ⫾ 10 215 (64%) 95 (28%) 262 (78%) 187 (56%) 89 (27%) 36 (11%) 6 (2%) 12 (4%) 6 (2%) 65 (19%) 140 (41%) 17 (5%) 8⫾5
62 ⫾ 10 44 (67%) 18 (27%) 49 (74%) 35 (53%) 10 (15%) 8 (12%) 5 (8%) 1 (1%) 1 (1%) 25 (38%) 41 (62%) 10 (15%) 6⫾4
0.77 0.83 0.99 0.52 0.69 0.06 0.83 0.02 0.70 0.99 0.009 0.003 0.02 0.008
207 (52%) 60 (15%) 133 (33%) 329 (82%)
154 (46%) 51 (15%) 129 (39%) 265 (79%)
53 (80%) 9 (14%) 4 (6%) 64 (97%)
⬍0.001 ⬍0.001
Continuous variables are expressed as mean ⫾ SD; categorical variables are expressed as number (percentage). * Medical diagnosis or use of antihypertensive drugs. † Medical diagnosis or use of antilipemic drugs. ‡ Measured in years from primary school to after college.
urgent procedures, a clopidogrel 600-mg loading dose was administered immediately before stenting. The Institute of Cardiology has protocols for patient care after stenting procedures. At hospital discharge, patients treated with bare metal stents were instructed to maintain the use of thienopyridines for ⱖ30 days and aspirin indefinitely. Patients treated with drug-eluting stents were instructed to maintain the use of thienopyridines for ⱖ1 year and aspirin indefinitely. At hospital discharge, all patients received a prescription and a detailed explanation about the importance of 30-day adherence to dual antiplatelet therapy. Stent implantation procedures followed the protocol of the Institute of Cardiology, similar to those described in the literature.9 Clinical aspects such as type and number of stents implanted, use of other devices, and additional pharmacology were left to the discretion of the operators. All patients were interviewed 30 days after stent implantation to assess adherence to medical treatment. The main outcome of the study was discontinuation of thienopyridines before the 30-day follow-up assessment as reported by the patient at the 30-day visit. Those who discontinued their medication were questioned about the reasons. The Morisky questionnaire10 was given to all patients to assess the degree of adherence to drug therapy. This test evaluates drug adherence in an objective and standardized way, taking into account the behavior of the patient regarding use of the prescribed medicine and not just drug interruption. It is based on 4 questions that involve forgetfulness, negligence, interrupting the drug after clinical improvement, and reuse of the medicine for worsening symptoms. Adherence was classified from 0 to 4. Patients with grade 0 totally adhered to their medication; those with higher grades adhered less.
In this study, the adaptation of Goldberg et al11 was used, which allows the patient the chance to talk about the causes of discontinuation. Delay in dose taking was considered when a medication was taken ⬎2 hours from the time it should have been taken. Characteristics of patients who discontinued thienopyridine treatment were compared with those who continued use until day 30. Data were analyzed with SPSS 11.0 for Windows (SPSS, Inc., Chicago, Illinois). Continuous variables were expressed as mean ⫾ SD and categorical variables as percentage. Comparisons between groups were done using t test for continuous variables and chi-square test or Fischer’s exact test for categorical values. Sample size was calculated assuming a discontinuation rate of 20% with 3.5% variation and a 95% confidence interval and was determined to be 400 patients. A p value ⬍0.05 was considered statistically significant. Variables associated with discontinuation in univariate analysis or classified as predictors of discontinuation in previous studies were included in multiple logistic regression models. The model with better calibration was chosen according to the Hosmer–Lemeshow goodness-of-fit test.12 Based on this model, a risk score was developed for discontinuation of thienopyridine and points were assigned proportionately to the odds ratios. Continuous variables were categorized according to cutpoints that allowed the best prediction of the primary outcome. Diagnostic accuracy of the score to predict drug discontinuation was assessed through c-statistics. Diagnostic accuracy of the risk score to predict total medication adherence (defined by a Morisky score 0) was also evaluated. Developing and testing a model on the same set of patients is likely to result in biased estimates. To estimate
Coronary Artery Disease/A Risk Score for Thienopyridine Adherence
687
Table 2 Multiple logistic regression analysis of candidate variables associated with thienopyridine discontinuation (n ⫽ 400) Characteristic Unmarried Lack of health insurance Education (years)* Acute coronary syndrome Diabetes mellitus Salary ⬍2 times minimum wage 2–3 times minimum wage ⬎3 times minimum wage Constant
Beta
OR
95% CI
0.908 1.543 ⫺0.112 0.838 ⫺0.788
2.48 4.68 0.89 2.31 2.20
1.01–6.07 1.05–21 0.42–1.86 1.28–4.14 1.02–4.69
2.108 1.795 1.795 ⫺5.709
8.23 4.46 4.46
2.70–25 1.25–16 1.25–16
p Value 0.046 0.04 0.76 0.004 0.04 ⬍0.001 0.02 0.02
CI ⫽ confidence interval; OR ⫽ odds ratio. * Measured in years from primary school to postcollege. Table 3 Risk score for adherence to thienopyridine Patient Characteristic Unmarried Acute coronary syndrome Absence of diabetes mellitus Lack of private health insurance Salary ⬍2 times minimum wage 2–3 times minimum wage
Points 1 1 1 4 7 3
the impact of this, a bias-corrected c-index was generated using 1,000 bootstrapped samples by implementing the R functions Validate and Calibrate.13 A graphic representation of the calibration curve using the original dataset and biascorrected samples was also generated. Results During the study period, 405 patients were included and clinical follow-up was assessed and completed in 400 patients (99%) an average of 30 ⫾ 4 days after the procedure. Only 5 patients were lost from the follow-up period. Discontinuation of thienopyridine was reported by 66 patients (16.5%); 102 additional patients (25%) presented adherence problems, according to the Morisky questionnaire. Median time (25th, 75th percentiles) of thienopyridine discontinuation was 10 days (9, 16). Clinical characteristics of study patients are listed in Table 1. The 2 groups were similar with the exception of a trend toward a smaller percentage of diabetics and a significantly larger percentage of patients with congestive heart failure in those who discontinued thienopyridines. These patients more frequently demonstrated the following characteristics: unmarried, lack of private health insurance, and an income ⬍2 times the minimum wage ($256.00 per month). Patients who discontinued thienopyridines were also more frequently referred from the emergency department with a clinical presentation of an acute coronary syndrome and had a lower educational level. Depressive symptoms were assessed and were similar between groups. Regarding stent implantation, 3%
received drug-eluting stents and 97% received bare metal stents. This difference is due mainly to health insurance policies in Brazil, which are restrictive for drug-eluting stents. Patients who received drug-eluting stents did not report discontinuation of thienopyridines compared with a rate of 17% in those receiving bare metal stents (0 of 12 vs 66 of 388 patients, p ⫽ 0.23). The following reasons were given for discontinuing thienopyridine therapy: 41 patients said cost of medication (when data were collected, the monthly cost of clopidogrel was equivalent to $50.00), 11 patients said they had not received sufficient information about the importance of the dual antiplatelet regimen, 10 patients received advice from another physician to stop the drug, 1 patient had an allergy, 1 reported urinary bleeding, 1 had a hemorrhagic stroke, and 1 refused to accept physician advice to take the drug. Patients who discontinued thienopyridines frequently stated that they were not aware of the importance of use compared with those who maintained treatment (50% vs 26%, p ⬍0.001). There was no significant difference between patients who maintained thienopyridine use and those who discontinued use with regard to access to free supplies of the drug. Rate of 30-day stent thrombosis was 0.5%; there were 2 cases of probable stent thrombosis by the Academic Research Consortium definition (sudden death). One of these patients who died within 30 days from the procedure had discontinued thienopyridine before death, whereas the other patient was still using clopidogrel. Using the Morisky questionnaire, we found that 102 patients had problems with adherence but did not discontinue the medication. Of these, 15% had better adherence (classifications 1 and 2), although they did stop taking the medication ⱖ1 time during the month. However, 11% demonstrated serious problems regarding adherence (classifications 3 and 4) with time delays or not taking the thienopyridine several times per week. Table 2 presents the final multiple regression model for discontinuation of thienopyridine with the best calibration for this population according to Hosmer–Lemeshow goodness-of-fit test (chi-square 1.92, p ⫽ 0.96). Other variables included in multiple regression models but not retained in
688
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Thienopyridine discontinuation rates according to risk score.
Figure 2. Calibration curves tested in original sample (apparent), based on 1,000 bootstrapped samples (bias corrected), and for perfect calibration (ideal) for risk score of adherence to thienopyridine.
the final model were age, sex, chronic heart failure, and emergency procedures. The risk score developed based on results of multivariate analysis is presented in Table 3 and ranges from 0 to 14 points. Figure 1 shows 30-day thienopyridine discontinuation rates according to risk score levels (0 to 4 points ⫽ 0% of patients discontinued the drug; 5 to 8 points ⫽ 7%; 9 to 12 points ⫽ 20%; ⬎12 points ⫽ 7%, c-statistic 0.76, bias-corrected c-statistics 0.76, p ⬍0.001; Figure 2). Risk score was strongly associated with complete treatment adherence as assessed by the Morisky questionnaire (c-statistic 0.74, p ⬍0.001). Discussion In this study, independent predictors of thienopyridine discontinuation were low salary, lack of private health insurance, marital status, absence of diabetes mellitus, and clinical presentation of an acute coronary syndrome. A risk
score based on these variables showed very good diagnostic accuracy to predict drug discontinuation. This information could be useful to target strategies to increase compliance in vulnerable subgroups of patients and aid in the decision of whether to use a drug-eluting stent in patients at highest risk for thienopyridine discontinuation. Aside from the 16.5% of patients who discontinued medication, another 25% presented adherence problems according to assessment with the Morisky questionnaire. This reinforces the need for implementing strategies that aim to increase adherence by these patients, even through assessment of other therapeutic options for those with the highest nonadherence rates. To improve adherence and ensure the success of the treatment proposed, it is necessary that health professionals recognize adherence problems14 and their predictors and aim to increase awareness about this fundamental aspect of treatment. A recent study that assessed medication adherence in patients with coronary artery disease demonstrated an adherence rate of 54% after 1 year of treatment and related education level, marital status, mental health (depression/anxiety), and total number of daily medications to low adherence.15 In patients undergoing coronary stent implantation, Spertus et al6 showed that education level was the only factor associated with thienopyridine discontinuation, with a trend toward marital status, cost, and preexisting cardiovascular disease. Adherence problems can be improved by some interventions but patient involvement is essential with discussion of alternatives to improve adherence.16 Use of technologic systems is 1 strategy17 but patient education is paramount with structured reminders,18 an increase in the number of consultations or telephone calls, and simplification of the medication regimen.19,20 Lee et al21 assessed the efficiency of a pharmaceutical assistance program in older patients, demonstrating the benefit of concomitant multiple interventions such as standardized medical information, support of a pharmacist, and special time schedules. In the 6-month follow-up, a 95% adherence rate to treatment was found. Assessment of adherence based on the account of the patient might overestimate its evaluation. In contrast, its assessment in the “real world” is a favorable point because adherence in randomized clinical trials is generally better and may not represent daily practice.22,23 Although the risk score was significantly associated with thienopyridine discontinuation in this study, extrapolation of these results might be limited due to a lack of validation of the score in other groups of patients and because this study was performed in a single high-volume dedicated cardiology center. In particular, this score may not apply to other nationalities, especially if socioeconomic characteristics and health plan reimbursement practices are significantly different from those described in the present report. Because most patients included in this study were treated with bare metal stents, applying this risk score in patients treated with drug-eluting stents should be done with caution because its accuracy may be different in this population. Only 3% of patients were treated with drug-eluting stents, and none of them had discontinued thienopyridine at 1-month follow-up compared with a discontinuation rate of 17% in patients treated with bare metal stents. This difference in rates of thienopyridine discontinuations between the
Coronary Artery Disease/A Risk Score for Thienopyridine Adherence
2 types of stents was not statistically significant. The fact that patients treated with drug-eluting stents had private health insurance and higher mean salaries, which were the strongest predictors of thienopyridine adherence at 30-day follow-up, definitely played a role in excellent medication compliance at 1-month in this group. Unfortunately, we do not have data available on longer-term compliance because we focused this analysis on 30-day outcomes. We believe that adherence rates in this time frame are particularly important because clopidogrel discontinuation in the first month after stenting is the most powerful predictor of stent thrombosis.1 Because we had low rates of cardiac events at 30 days, we could not make any meaningful association between clopidogrel discontinuation and outcomes. However, previous studies have demonstrated that premature discontinuation of thienopyridine is the main predictor of stent thrombosis, which carries a 70% risk of myocardial infarction and a 40% risk of death.1,3 1. Airoldi F, Colombo A, Morici N, Latib A, Cosgrave J, Buellesfeld L, Bonizzoni E, Carlino M, Gerckens U, Godino C, Melzi G, Micher I, Montorfano M, Sangiorgi GM, Qasim A, Chieffo A, Briguori C, Grube E. Incidence and predictors of drug-eluting stent thrombosis during and after discontinuation of thienopyridine treatment. Circulation 2007;116:745–754. 2. Cutlip DE, Baim DS, Ho KK, Popma JJ, Lansky AJ, Cohen DJ, Carrozza JP Jr, Chauhan MS, Rodriguez O, Kuntz RE. Stent thrombosis in the modern era: a pooled analysis of multicenter coronary stent clinical trials. Circulation 2001;103:1967–1971. 3. Iakovou I, Schmidt T, Bonizzoni E, Ge L, Sangiorgi GM, Stankovic G, Airoldi F, Chieffo A, Montorfano M, Carlino M, Michev I, Corvaja N, Briguori C, Gerckens U, Grube E, Colombo A. Incidence, predictors, and outcome of thrombosis after successful implantation of drugeluting stents. JAMA 2005;293:2126 –2130. 4. Pfisterer M, Brunner-La Rocca HP, Buser PT, Rickenbacher P, Hunziker P, Mueller C, Jeger R, Bader F, Osswald S, Kaiser C. Late clinical events after clopidogrel discontinuation may limit the benefit of drug-eluting stents: an observational study of drug-eluting versus bare-metal stents. J Am Coll Cardiol 2006;48:2584 –2591. 5. Grines CL, Bonow RO, Casey DE Jr, Gardner TJ, Lockhart PB, Moliterno DJ, O’Gara P, Whitlow P. Prevention of premature discontinuation of dual antiplatelet therapy in patients with coronary artery stents: a science advisory from the American Heart Association, American College of Cardiology, Society for Cardiovascular Angiography and Interventions, American College of Surgeons, and American Dental Association, with representation from the American College of Physicians. Circulation 2007;115:813– 818. 6. Spertus JA, Kettelkamp R, Vance C, Decker C, Jones PG, Rumsfeld JS, Messenger JC, Khanal S, Peterson ED, Bach RG, Krumholz HM, Cohen DJ. Prevalence, predictors, and outcomes of premature discontinuation of thienopyridine therapy after drug-eluting stent placement: results from the PREMIER registry. Circulation 2006;113:2803–2809. 7. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med 2005;353:487– 497.
689
8. Shemesh E, Yehuda R, Milo O, Dinur I, Rudnick A, Vered Z, Cotter G. Posttraumatic stress, nonadherence, and adverse outcome in survivors of a myocardial infarction. Psychosom Med 2004;66:521–526. 9. 2007 Writing Group to Review New Evidence and Update the ACC/ AHA/SCAI 2005 Guideline Update for Percutaneous Coronary Intervention, Writing on Behalf of the 2005 Writing Committee. 2007 Focused update of the ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2008;117:261–295. 10. Morisky DE, Green LW, Levine DM. Concurrent and predictive validity of a self-reported measure of medication adherence. Med Care 1986;24:67–74. 11. Goldberg AI, Cohen G, Rubin AH. Physician assessments of patient compliance with medical treatment. Soc Sci Med 1998;47:1873–1876. 12. Hosmer DW, Lemeshow S. Assessing the fit of the model. In: Hosmer DW, Lemeshow S, eds. Applied Logistic Regression, 1st Ed. New York: John Wiley & Sons, 1989:135–175. 13. Harrell FE Jr. Package ‘HMISC’. Updated June 4, 2010. Available at: cran.it.r-project.org/web/packages/Hmisc/Hmisc.pdf. Accessed June 17, 2010. 14. Heisler M, Hogan MM, Hofer TP, Schmittdiel JA, Pladevall M, Kerr EA. When more is not better: treatment intensification among hypertensive patients with poor medication adherence. Circulation 2008; 117:2884 –2892. 15. Kulkarni SP, Alexander KP, Lytle B, Heiss G, Peterson ED. Longterm adherence with cardiovascular drug regimens. Am Heart J 2006; 151:185–191. 16. Ho PM, Magid DJ, Shetterly SM, Olson KL, Maddox TM, Peterson PN, Masoudi FA, Rumsfeld JS. Medication nonadherence is associated with a broad range of adverse outcomes in patients with coronary artery disease. Am Heart J 2008;155:772–779. 17. Piette JD, Weinberger M, Kraemer FB, McPhee SJ. Impact of automated calls with nurse follow-up on diabetes treatment outcomes in a Department of Veterans Affairs Health Care System: a randomized controlled trial. Diabetes Care 2001;24:202–208. 18. Haynes RB, McDonald HP, Garg AX. Helping patients follow prescribed treatment: clinical applications. JAMA 2002;288:2880 –2883. 19. Simpson RJ Jr. Challenges for improving medication adherence. JAMA 2006;296:2614 –2616. 20. McDonald HP, Garg AX, Haynes RB. Interventions to enhance patient adherence to medication prescriptions: scientific review. JAMA 2002; 288:2868 –2879. 21. Lee JK, Grace KA, Taylor AJ. Effect of a pharmacy care program on medication adherence and persistence, blood pressure, and low-density lipoprotein cholesterol: a randomized controlled trial. JAMA 2006;296: 2563–2571. 22. Gitt AK, Bueno H, Danchin N, Fox K, Hochadel M, Kearney P, Maggioni AP, Opolski G, Seabra-Gomes R, Weidinger F. The role of cardiac registries in evidence-based medicine. Eur Heart J 2010;31: 525–529. 23. Hordijk-Trion M, Lenzen M, Wijns W, de Jaegere P, Simoons ML, Scholteop Reimer WJ, Bertrand ME, Mercado N, Boersma E. EHS-CR Investigators. Patients enrolled in coronary intervention trials are not representative of patients in clinical practice: results from the Euro Heart Survey on Coronary Revascularization. Eur Heart J 2006;27: 671– 678.
Therapeutic Benefit of Preventive Telehealth Counseling in the Community Outreach Heart Health and Risk Reduction Trial Robert P. Nolan, PhDa,b,*, Ross E.G. Upshur, MD, MScc, Hazel Lynn, MD, MHScd, Thomas Crichton, MDe, Ellen Rukholm, PhDf, Donna E. Stewart, MDa,b, David A. Alter, MD, PhDg, Caroline Chessex, MDa,b, Paula J. Harvey, MD, PhDa,b, Sherry L. Grace, PhDa,i, Louise Picard, MSc(A), RNh, Isabelle Michel, MA, RNh, Jan Angus, PhDb, Kim Corace, MAj, Susan M. Barry-Bianchi, PhDa, and Maggie H. Chen, MSca We evaluated whether telehealth counseling augments lifestyle change and risk factor decrease in subjects at high risk for primary or secondary cardiovascular events compared to a recommended guideline for brief preventive counseling. Subjects at high risk or with coronary heart disease (35 to 74 years of age, n ⴝ 680) were randomized to active control (risk factor feedback, brief advice, handouts) or telehealth lifestyle counseling (active control plus 6 weekly 1-hour teleconferenced sessions to groups of 4 to 8 subjects). Primary outcome was questionnaire assessment of adherence to daily exercise/physical activity and diet (daily vegetable and fruit intake and restriction of fat and salt) after treatment and at 6-month follow-up. Secondary outcomes were systolic and diastolic blood pressures, ratio of total to high-density lipoprotein cholesterol, and 10-year absolute risk for coronary disease. After treatment and at 6-month follow-up, adherence increased for telehealth versus control in exercise (29.3% and 18.4% vs 2.5% and 9.3%, respectively, odds ratio 1.60, 95% confidence interval 1.2 to 2.1) and diet (37.1% and 38.1% vs 16.7% and 33.3%, respectively, odds ratio 1.41, 95% confidence interval 1.1 to 1.9). Telehealth versus control had greater 6-month decreases in blood pressure (mean ⴞ SE, systolic ⴚ4.8 ⴞ 0.8 vs ⴚ2.8 ⴞ 0.9 mm Hg, p ⴝ 0.04; diastolic ⴚ2.7 ⴞ 0.5 vs ⴚ1.5 ⴞ 0.6 mm Hg, p ⴝ 0.04). Decreases in cholesterol ratio and 10-year absolute risk were significant for the 2 groups. In conclusion, telehealth counseling augments therapeutic lifestyle change in subjects at high risk for cardiovascular events compared to a recommended guideline for brief preventive counseling. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:690 – 696) The primary aim of this clinical trial was to evaluate whether a telehealth protocol that used motivational interviewing1,2 added therapeutic benefit for change in exercise, diet, or smoking in subjects at high risk for or with established cardiovascular disease. We used an active control intervention that represented a recommended guideline for brief preventive counseling.3 The extent to which telehealth augmented a decrease in cardiovascular risk factors after adjusting for medications was our secondary objective.
a University Health Network, Toronto, Ontario, Canada; bUniversity of Toronto, Toronto, Ontario, Canada; cSunnybrook Health Sciences Centre, Toronto, Ontario, Canada; dGrey Bruce Health Unit, Owen, Sound, Ontario, Canada; eNorthern Ontario School of Medicine, Sudbury, Ontario, Canada; fLaurentian University, Sudbury, Ontario, Canada; gInstitute for Clinical Evaluative Sciences and the Li Ka, Shing Knowledge Institute of St. Michael’s Hospital, Toronto Rehabilitation Institute and University of Toronto, Toronto, Ontario, Canada; hSudbury and District Health Unit, Sudbury, Ontario, Canada; iYork University, Toronto, Ontario, Canada; j Ottawa Hospital, Ottawa, Ontario, Canada. Manuscript received June 17, 2010; revised manuscript received and accepted October 13, 2010. This research was supported by Grant NA4146 from the Heart and Stroke Foundation of Canada, Ottawa, Ontario, Canada. *Corresponding author: Tel: 416-340-4800, ext 6008; fax: 416-3403162. E-mail address:
[email protected] (R.P. Nolan).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.050
Methods Subjects were at high risk for a cardiovascular event; they were 35 to 74 years of age and had a diagnosis of coronary heart disease or diabetes, or Framingham 10-year absolute risk for coronary heart disease ⱖ20,4 or ⱖ2 risk factors that included hypertension, dyslipidemia, smoking, or obesity (body mass index ⱖ30 kg/m2 or waist circumference ⬎88 cm in women or ⬎102 cm in men). Exclusion was based on psychiatric illness, alcohol or drug dependence in the previous year, or residence in an institutional setting. Medications were not altered by design. The Community Outreach Heart Health and Risk Reduction Trial (COHRT) was a 2 parallel-group, single-blind randomized controlled trial: active control versus telehealth counseling assessed at baseline, 2 weeks after treatment, and at 6-month follow-up. Subjects and their physicians were blinded to the research design. Research assistants providing the telehealth intervention had no involvement in outcome evaluations. Randomization was conducted by a computer program that stratified for gender, cardiovascular disease, diabetes, and depression (Beck Depression Inventory II ⱖ14).5 Randomization was blocked within our northern, rural, and urban recruitment sites in Ontario, Canada. A 2:3 sampling ratio was used for telehealth versus control groups to permit analysis of psychosocial determinants of www.ajconline.org
Preventive Cardiology/Preventive Telehealth Counseling
691
Table 1 Lifestyle assessment at baseline, after treatment, and at six-month follow-up Category
Item*
Content
Exercise
Planned exercise
I do a planned exercise for ⱖ20 minutes, 3–5 times/week (e.g., brisk walking, aerobics, jogging, swimming, skiing). I keep active in my daily habits 5–6 days/week (e.g., walking and climbing stairs at home and work, cutting grass, shoveling snow, washing floors). I eat 3–5 servings of vegetables each day (1/2 cup of raw or cooked vegetables is 1 serving). I eat 2–4 servings of fruit each day (1 piece of fruit, such as an apple, is 1 serving). At each meal, I eat ⬍30% of calories from fat by eating lean meat without the skin and avoiding fried foods (e.g., French fries) and high-fat comfort foods such as potato chips. At each meal, I avoid adding extra salt to my food, and I avoid eating salty foods such as chips, soy sauce, fast foods such as hamburgers, and prepared food mixes. I have a smoke-free lifestyle everyday, which does not include even 1 puff of a cigarette.
Daily activity Diet
Vegetables Fruit Fat intake Sodium intake
Smoking cessation
* Each questionnaire item was scaled by the transtheoretical model12: Precontemplation (not adhering to the target behavior and not ready to change within 6 months), contemplation (not adhering to the target behavior but ready to change within 6 months), preparation (not adhering to the target behavior but ready to change within 1 month), action (adhering to the target behavior for ⬍6 months), and maintenance (adhering to the target behavior for ⱖ6 months).
lifestyle change (not reported in the present study). Randomization codes were concealed in opaque sealed envelopes that were sorted by stratification features. Approval was obtained from research ethics boards of each institution. Recruitment initiatives included presentations to family medicine departments and patient groups, random-digit dialing within targeted area codes, and newspaper advertisements. Recruits were screened by telephone interview. Eligible subjects were mailed an information package and a patient profile form, which requested family physicians to confirm diagnoses and cardiovascular risk factors. We provided a prepaid requisition for assays of 12-hour fasting lipoprotein cholesterol if these were not assessed within the previous 6 months. Eligible subjects were scheduled for assessment at a local COHRT clinic after receipt of fasting blood tests and physician confirmation of diagnoses and cardiovascular risk factors. Randomization to telehealth versus control followed informed consent during the initial COHRT clinic visit. COHRT clinics were held in family medicine outpatient clinics and a behavioral cardiology research unit in 2 tertiary care hospitals (urban site) and collaborating family medicine practices and offices of 2 public health units (rural and northern sites) in Ontario, Canada. COHRT clinics were scheduled between 8:00 A.M. and 12:00 P.M. Subjects were instructed to refrain from smoking and strenuous exercise for ⱖ4 hours before their appointment. Baseline measurements were taken by trained research assistants for height, weight, body mass index, waist circumference, and blood pressure (2 measurements, 30 minutes apart). Blood assays of 12-hour fasting cholesterol (total, high-density lipoprotein, total/high-density lipoprotein ratio, and low-density lipoprotein) were obtained for the COHRT clinic visit. Procedures for assaying blood samples were standardized across sites using a common laboratory network. Lowdensity lipoprotein cholesterol was calculated using the Friedewald formula when triglycerides were ⬍4.52 mmol/L (400 mg/dl). The Framingham 10-year absolute risk index4 was estimated for subjects without established cardiovascular disease. Baseline exercise, diet, and smoking were measured by questionnaire6 (Table 1). At 2 weeks after treatment, this questionnaire was administered by telephone. However, the 6-month follow-up assessment was conducted
during a COHRT clinic visit that replicated the baseline protocol: anthropometric measurements, cardiovascular risk factors (with fasting blood assays obtained within 2 weeks of the visit), and questionnaires. Antihypertensive medications (angiotensin-converting enzyme inhibitors, calcium channel blockers, diuretics,  blockers, ␣-adrenergic blockers, or angiotensin II receptor blockers) or lipid-lowering agents were recorded at baseline and 6-month follow-up. Change in dosage or type of medications was coded for the baseline to 6-month follow-up interval. During the initial COHRT clinic visit, controls received a 10-minute intervention that included a review and written summary of their cardiovascular risk factor profiles. Subjects without established cardiovascular disease were given their Framingham 10-year absolute risk score. Brief advice for therapeutic change in exercise, diet, and smoking was provided with accompanying educational handouts7–9 and a list of community programs for lifestyle change. The profile of each subject’s cardiovascular risk factors, 10-year absolute risk,4 and severity of depressive symptoms5 was mailed to their family physician. The telehealth group received this intervention plus 6 weekly 1-hour sessions of lifestyle counseling by teleconference to small groups (n ⫽ 4 to 8). Subjects were matched according to a fixed schedule of weekly sessions. Access to each telehealth session was made by a toll-free number and a private access code. The protocol for telehealth sessions was standardized with a COHRT treatment manual2 that was a group-based application of motivational interviewing.1 Subjects identified their priority for lifestyle change (diet, exercise, or smoking) and they were taught to selfassess their stage of readiness for change: precontemplation, contemplation, preparation, action, and maintenance.6 Across sessions, stage-matched strategies to support lifestyle change were introduced and supported with take-home projects. Motivational interviewing guidelines were used to focus group discussion on (1) salient lifestyle goals identified by subjects, (2) progress in resolving ambivalence about lifestyle change, and (3) experiences of increased efficacy in initiating or maintaining change. Group facilitators included 2 public health nurses, 6 allied health professionals, and 8 PhD students in clinical psychology who
692
The American Journal of Cardiology (www.ajconline.org)
Completed telephone screening interview n = 2566
Excluded: n = 1739 1704 failed screening criteria 35 refused to participate
Eligible after initial screening: n = 827 302 (36.5%) Rural sites 125 (15.1%) Northern sites 500 (60.5%) Urban sites
Excluded: n = 144 139 CAD, Diabetes or high CAD risk was not confirmed at COHRT clinic visit 5 Withdrew
Randomized: n = 683 Randomized n = 683
Randomized to Active Control: n = 268
Randomized to Lifestyle Counseling: n = 415
< 50% sessions completed: n = 71 (17.2%)
Withdrew (due to time commitments, lack of availability, or moved): n =30 (11.2%)
Withdrew (due to time commitments, lack of availability, or moved): n = 32 (7.7%)
Deaths (unrelated to trial): n = 1 Subjects included in final analysis: n = 267
Deaths (unrelated to trial): n = 2 Subjects included in final analysis: n = 413
Figure 1. Subject enrollment and randomization in the Community Outreach Heart Health and Risk Reduction Trial. CAD ⫽ coronary artery disease.
completed a 3- to 4-day training program (with R.P.N.).2 Quality control of this intervention was maintained by weekly supervision by teleconference (with R.P.N. or K.C.). Primary outcomes were defined as adherence to Health Canada guidelines7,8 for exercise (planned weekly exercise or daily activity), diet (daily intake of vegetables and fruit, and daily restriction of fat and salt), and smoke-free living (Table 1). As in previous trials,10 adherence to each of these behaviors was defined by self-reported criteria for the action or maintenance stages of readiness for change.6 Questionnaire items for exercise and diet were previously validated according to grade of energy expenditure,11 food frequency questionnaires,12,13 and by body mass index and waist circumference.12,14 Secondary outcomes at 6-month follow-up included risk factors (systolic and diastolic blood pressures, total/high-density lipoprotein cholesterol) and the Framingham index of 10-year absolute risk for coronary heart disease.4 Criterion validity of self-reported adherence versus nonadherence to exercise and diet at baseline and 6-month follow-up was measured by body mass index, weight decrease, and Health Canada guidelines for active living15 using analysis of variance and Pearson chi-square test. Baseline characteristics of telehealth subjects versus controls were evaluated with analysis of variance and Pearson chi-square test. Generalized estimation equations assessed
whether a larger proportion of telehealth versus control subjects adhered to exercise and diet after treatment and at 6-month follow-up, controlling for baseline adherence to exercise and diet, age, gender, body mass index, and interval (after treatment and 6-month follow-up). Multivariable linear regression analyses evaluated whether telehealth versus control subjects had greater decreases at 6-month follow-up in systolic and diastolic blood pressures and total/ high-density lipoprotein cholesterol adjusted for baseline values of each variable, age, gender, body mass index, antihypertensive or lipid-lowering medications at baseline, and change in antihypertensive or lipid-lowering medications up to 6-month follow-up. This regression model also assessed group differences in decrease of the Framingham index of 10-year absolute risk in subjects without cardiovascular disease. COHRT was powered to detect increased adherence at 4-month follow-up after lifestyle counseling10 to the action/ maintenance stage of readiness for change6 in dietary fat decrease (odds ratio 2.15, 95% confidence interval 1.3 to 3.6) and physical activity (odds ratio 1.89, 95% confidence interval 1.1 to 3.4) with 80% power, 5% type 1 error rate, and 25% attrition. This estimate used the method of Wang et al16 to account for a 2:3 sampling ratio for telehealth versus control. Analyses were performed using the intention-to-treat approach. Missing data were managed by mul-
Preventive Cardiology/Preventive Telehealth Counseling
693
Table 2 Baseline characteristics Variables
Active Control (n ⫽ 267)
Telehealth (n ⫽ 413)
p Value
Age (years) Women Income groups (⬍$20,000 ⫽ 1, ⱖ$70,000 ⫽ 7) Body mass index (kg/m2) Waist circumference (⬎88 cm in women, ⬎102 cm in men) Hypertension Dyslipidemia Diabetes mellitus Smoker Coronary heart disease risk status ⱖ2 cardiovascular risk factors* Absolute 10-year coronary heart disease risk10 ⱖ20 or diabetes Coronary heart disease Medications  Blocker Calcium channel blocker Angiotensin-converting enzyme inhibitor Angiotensin II receptor blocker ␣-Receptor blocker Diuretic Lipid-lowering agent ⱖ1 antihypertensive drug† ⱖ1 antihypertensive or lipid-lowering drug Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Total/high-density lipoprotein cholesterol (mmol/L)‡ Total lipoprotein cholesterol (mmol/L)‡ High-density lipoprotein cholesterol (mmol/L)‡ Low-density lipoprotein cholesterol (mmol/L)‡
58.61 ⫾ 0.53 132 (49.4%) 4.91 ⫾ 0.13 30.58 ⫾ 0.38 188 (70.4%) 182 (68.2%) 181 (67.8%) 138 (51.7%) 40 (15.0%)
59.27 ⫾ 0.43 202 (48.9%) 4.50 ⫾ 0.10 31.60 ⫾ 0.33 321 (77.7%) 300 (72.6%) 286 (69.2%) 216 (52.3%) 53 (12.8%)
0.34 0.89 0.02 0.047 0.03 0.21 0.69 0.88 0.43 0.38
69 (25.8%) 132 (49.45%) 66 (24.75%)
97 (23.5%) 194 (47.0%) 122 (29.55%)
73 (27.3%) 56 (21.0%) 94 (35.2%) 32 (12.0%) 1 (0.4%) 75 (28.1%) 137 (51.3%) 187 (70.0%) 212 (79.4%) 128.39 ⫾ 0.97 77.02 ⫾ 0.56 4.42 ⫾ 0.09 5.21 ⫾ 0.09 1.26 ⫾ 0.03 3.05 ⫾ 0.07
114 (27.6%) 94 (22.8%) 156 (37.8%) 50 (12.1%) 4 (0.6%) 111 (26.9%) 217 (52.5%) 298 (72.2%) 332 (80.4%) 131.30 ⫾ 0.78 77.48 ⫾ 0.45 4.35 ⫾ 0.07 5.30 ⫾ 0.07 1.29 ⫾ 0.02 3.03 ⫾ 0.05
0.94 0.58 0.50 0.96 0.38 0.73 0.75 0.55 0.75 0.02 0.52 0.58 0.45 0.48 0.78
Data are presented as mean ⫾ SE or number of subjects (percentage). * Select cardiovascular risk factors include hypertension, dyslipidemia, smoking, or obesity. † Antihypertensive medications include  blockers, calcium channel blockers, angiotensin-converting enzyme inhibitors, ␣-adrenergic blockers, angiotensin II receptor blockers, or diuretics. ‡ To convert to milligrams per deciliter, multiply by 38.67. Table 3 Unadjusted adherence to exercise and diet for active control and telehealth
Analyses were conducted with SAS 9.1 (SAS Institute, Cary, North Carolina).
Variables
Results
Adherence to exercise* Active control (baseline, n ⫽ 161) Telehealth (baseline, n ⫽ 239) Adherence to diet† Active control (baseline, n ⫽ 132) Telehealth (baseline, n ⫽ 197)
After Treatment
6-Month Follow-Up
165 (2.5%) 309 (29.3%)
176 (9.3%) 283 (18.4%)
154 (16.7%) 270 (37.1%)
176 (33.3%) 272 (38.1%)
Data are presented as number of subjects (percent change). * Defined as planned exercise ⱖ20 minutes, 3 to 5 times per week, or active daily habits 5 to 6 days per week. † Defined as daily intake of vegetables (3 to 5 servings per day) and fruit (2 to 4 servings per day) and restriction of fat (⬍30% of daily calories) and salt (no extra salt at meals and avoidance of prepared foods and salty snacks).
tiple imputations using the Markov chain Monte Carlo method. Outcomes were similar across 5 imputations and each was consistent with the raw data. Statistical significance was defined by 2-tailed tests with a p value ⬍0.05.
The sample included 267 controls and 413 telehealth subjects (Figure 1). Withdrawals did not differ significantly between groups (controls, n ⫽ 30, 11.2%; telehealth subjects, n ⫽ 32, 7.7%, p ⫽ 0.12). Table 2 presents baseline characteristics. Antihypertensive or lipid-lowering drugs were prescribed to 80% of subjects. There was a high prevalence of change in antihypertensive or lipid-lowering medications from baseline to 6-month follow-up (controls, n ⫽ 113, 42.3%; telehealth subjects, n ⫽ 171, 41.4%, p ⫽ 0.74; Appendix 1, available on-line). Criterion validity for questionnaires assessing adherence to diet and exercise was supported by corroborating physiologic data. Subjects who reported baseline adherence versus nonadherence to diet had a lower body mass index (30.1 ⫾ 0.39 vs 31.8 ⫾ 0.32 kg/m2, respectively, p ⬍0.001). Body mass index was also lower in subjects who reported adherence versus nonadherence to exercise (30.1 ⫾ 0.29 vs 32.7 ⫾ 0.43 kg/m2, respectively, p ⬍0.001). Weight loss from baseline to 6-month follow-up (assessed in COHRT
694
The American Journal of Cardiology (www.ajconline.org)
Table 4 Generalized estimating equation models of subjects in adherence to exercise and diet after treatment and at six-month follow-up Variables Adherence to exercise* Baseline exercise: nonadherence versus adherence Age Gender† Body mass index Interval: after treatment versus 6-month follow-up‡ Intervention: active control versus telehealth Adherence to diet* Baseline diet: nonadherence versus adherence Age Gender† Body mass index Interval: after treatment versus 6-month follow-up‡ Intervention: active control versus telehealth
Odds Ratio
95% Confidence Interval
p Value
4.77
3.59–6.32
⬍0.0001
1.01 1.17 0.97 1.13
0.99–1.03 0.88–1.56 0.95–0.99 0.92–1.40
0.29 0.28 0.001 0.28
1.60
1.20–2.12
0.0009
5.12
3.82–6.86
⬍0.0001
1.04 1.46 0.99 0.56
1.02–1.06 1.11–1.92 0.97–1.02 0.46–0.68
⬍0.0001 0.007 0.65 ⬍0.0001
1.31
1.09–1.48
0.012
* Nonadherence ⫽ 0 and adherence ⫽ 1 for adherence to exercise or diet at baseline, after treatment, and at 6-month follow-up. † Women ⫽ 0, men ⫽ 1. ‡ After treatment ⫽ 1, 6-month follow-up ⫽ 2.
clinics) was greater in subjects who reported adherence versus nonadherence to exercise (⫺l.56 ⫾ 0.61 vs 0.96 ⫾ 0.88 lb, respectively, p ⫽ 0.02) or to diet (⫺1.96 ⫾ 0.70 vs 0.52 ⫾ 0.71 lb, respectively, p ⫽ 0.01). Subjects reporting adherence to exercise at baseline were more likely to meet conventional criteria for active living as assessed by screening items obtained from the Canadian Community Health Survey15 (Appendix 2, available online). The most prevalent self-selected goals for lifestyle change were exercise (n ⫽ 284, 41.8%) and diet (n ⫽ 280, 41.2%) or a combination of exercise and diet (n ⫽ 59, 8.7%). Smoking cessation was the sole primary goal for change for only 1 subject. At 6-month follow-up, smoking cessation was observed for 3 of 40 controls (7.5%) and 5 of 53 telehealth subjects (9.4%). Hence, additional analyses are not presented. A larger proportion of telehealth subjects versus controls reported adherence to exercise and diet after treatment and at 6-month follow-up as observed for unadjusted outcomes (Table 3) and after adjustment for covariates (Table 4). All risk factors decreased significantly for telehealth subjects and controls at 6-month follow-up (Table 5). Telehealth subjects demonstrated greater decreases in systolic and diastolic blood pressures but not total/high-density lipoprotein cholesterol or 10-year absolute risk of coronary heart disease (Table 6). Discussion We evaluated the efficacy of a telehealth intervention that used motivational interviewing in a group-based pro-
tocol that was standardized for session content and contact time with subjects.1,2 The control intervention was also standardized for contact time with subjects and in its application of a recommended guideline for brief preventive counseling.3 The major finding of COHRT is that a larger proportion of telehealth subjects versus controls reported adherence to exercise and diet behaviors after treatment and at 6-month follow-up. This benefit was achieved after 6 weekly telehealth sessions. An important goal for future telehealth trials is to determine whether long-term adherence to preventive lifestyle behaviors can be sustained with intermittent booster sessions or with Internet-based support.17 Systolic and diastolic blood pressures at 6-month follow-up were significantly improved with telehealth counseling. In contrast, total/high-density lipoprotein cholesterol and the Framingham 10-year absolute risk index were reduced for the telehealth and active control groups and the magnitude of improvement in these outcomes was consistent with previous trials.18,19 It is unclear whether similar improvement in lipoprotein cholesterol and the Framingham absolute risk index would have been achieved with usual care combined with a behavioral placebo that provided nonspecific support. Meta-analysis19 has shown that compared to usual care, telehealth counseling significantly improved lipid profiles at 6- to 48-month follow-up in subjects with cardiovascular disease. Therefore, the failure to observe a similar therapeutic benefit for telehealth in COHRT may be due to use of (1) a shorter (6-month) follow-up assessment or (2) an active control intervention that provided risk factor information to subjects and their family physicians, which is known to facilitate a decrease in cardiovascular risk factors.20,21 A large proportion of COHRT subjects (42%) reported a change in antihypertensive or lipid-lowering medications up to 6-month follow-up. We did not anticipate this outcome because subjects and their physicians were blinded to our research design. However, therapeutic adjustment in medications can be facilitated by providing patients with their cardiovascular risk profiles and by forwarding this information to their family physicians.20,21 In addition, change in medications during COHRT was associated with a decrease in all cardiovascular risk factors (Table 6), but a similar association was not observed for baseline assessment of medications. Trials of preventive lifestyle counseling may overestimate the effect of behavior change on risk factor decrease when only baseline medications are factored into the outcome analysis. Several studies have noted that lifestyle counseling is provided to patients at a suboptimal level due to barriers such as constrained time and resources in patient care settings.22 The telehealth protocol in COHRT was designed for administration to small groups (4 to 8 subjects) and it was provided by student trainees and allied health professionals. A group-based telehealth strategy merits consideration as a potential cost-efficient method to deliver preventive lifestyle counseling in settings where personnel resources are constrained. The findings of COHRT are limited to subjects who may require long-term access to an economical resource that can support therapeutic lifestyle change. Most subjects were at increased risk for an initial cardiovascular event. In contrast
Preventive Cardiology/Preventive Telehealth Counseling
695
Table 5 Unadjusted change in cardiovascular risk indexes for active control and telehealth groups Variables Systolic blood pressure (mm Hg) Active control Telehealth Diastolic blood pressure (mm Hg) Active control Telehealth Total/high-density lipoprotein cholesterol Active control Telehealth Framingham index of 10-year absolute risk* Active control Telehealth
Baseline
6-Month Follow-Up
Change
p Value for Change
128.39 ⫾ 0.93 131.30 ⫾ 0.80
125.55 ⫾ 0.93 126.47 ⫾ 0.75
⫺2.84 ⫾ 0.87 ⫺4.83 ⫾ 0.75
0.001 ⬍0.0001
77.02 ⫾ 0.53 77.48 ⫾ 0.47
75.53 ⫾ 0.58 74.75 ⫾ 0.49
⫺1.49 ⫾ 0.60 ⫺2.73 ⫾ 0.46
0.01 ⬍0.0001
4.42 ⫾ 0.09 4.39 ⫾ 0.03
3.95 ⫾ 0.08 4.07 ⫾ 0.07
⫺0.47 ⫾ 0.08 ⫺0.32 ⫾ 0.07
⬍0.0001 ⬍0.0001
12.83 ⫾ 0.28 13.73 ⫾ 0.19
11.06 ⫾ 0.45 12.61 ⫾ 0.34
⫺1.77 ⫾ 0.45 ⫺1.12 ⫾ 0.39
⬍0.0001 0.005
Data are presented as mean ⫾ SE. * Absolute 10-year risk in subjects without coronary heart disease. Table 6 Multivariable linear regression models of change in cardiovascular risk factors at 6-month follow-up
Systolic blood pressure (mm Hg) Baseline systolic blood pressure Age Sex Body mass index Antihypertensive medications at baseline* Change in antihypertensive medications† Interventions: Control vs. Telehealth Diastolic blood pressure (mm Hg) Baseline diastolic blood pressure Age Sex Body mass index Antihypertensive medications at baseline* Change in antihypertensive medications† Interventions: Control vs. Telehealth Total/High-Density Cholesterol Baseline total/high-density cholesterol Age Sex Body mass index Lipid lowering medications at baseline* Change in lipid lowering medications† Interventions: Control vs. Telehealth Framingham Index of 10-Year Absolute Risk‡ Baseline 10-year absolute risk Age Sex Body mass index Antihypertensive medications at baseline* Change in antihypertensive medications† Lipid lowering medications at baseline* Change in lipid lowering medications† Interventions: Control vs. Telehealth

95% CI
p Value
⫺14.70 ⫺0.05 0.20 0.24 0.19 ⫺3.62 ⫺2.09
⫺17.27, ⫺12.13 ⫺0.16, 0.07 ⫺1.94, 2.34 0.06, 0.42 ⫺2.01, 2.39 ⫺5.89, ⫺1.35 ⫺4.09, ⫺0.09
⬍.0001 0.40 0.85 0.006 0.87 0.002 0.04
⫺9.36 ⫺0.10 ⫺1.39 0.15 0.66 ⫺1.94 ⫺1.39
⫺11.87, ⫺6.85 ⫺0.18, ⫺0.02 ⫺2.78, 0.002 0.03, 0.27 ⫺0.81, 2.13 ⫺3.41, ⫺0.47 ⫺2.72, ⫺0.06
⬍.0001 0.03 0.05 0.01 0.38 0.009 0.04
⫺0.79 0.02 ⫺0.04 0.01 ⫺0.08 ⫺0.35 0.13
⫺0.99, ⫺0.59 0.0004, 0.04 ⫺0.24, 0.16 ⫺0.01, 0.03 ⫺0.28, 0.12 ⫺0.59, ⫺0.11 ⫺0.07, 0.33
⬍.0001 0.02 0.65 0.16 0.44 0.004 0.18
⫺5.74 0.05 ⫺0.79 0.06 ⫺0.08 ⫺0.68 ⫺0.13 ⫺0.94 0.65
⫺7.15, ⫺4.33 ⫺0.009, 0.11 ⫺2.02, 0.44 ⫺0.06, 0.18 ⫺1.47, 1.31 ⫺2.09, 0.73 ⫺1.31, 1.05 ⫺2.31, 0.43 ⫺0.53, 1.83
⬍.0001 0.16 0.21 0.30 0.91 0.35 0.82 0.18 0.28
Data for risk factor change ⫽ 6-month follow-up - baseline. * Antihypertensive medications as in Table 2. † Change in antihypertensive or lipid lowering medications: 0 ⫽ no change or not prescribed, 1 ⫽ change. ‡ Absolute 10-year risk among subjects without coronary heart disease.
to many secondary prevention trials, our sample was balanced according to gender. The recruitment procedures may have attracted subjects with greater motivation to change
their lifestyle behaviors. Nevertheless, telehealth was associated with improved lifestyle change compared to active controls. Behavioral outcomes were assessed by question-
696
The American Journal of Cardiology (www.ajconline.org)
naire. Future telehealth trials can improve upon this standard by using objective measurements of behavior change with accelerometry, food diaries, or 24-hour urine analysis of sodium excretion. Nevertheless, telehealth counseling may help extend the reach and efficacy of a recommended standard for brief preventive counseling3 in patients at high risk for primary or secondary cardiovascular events. Acknowledgment: The COHRT investigators extend sincere appreciation to the participants, collaborating organizations, students, and family physicians who made this trial possible. Supplementary Data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.amjcard.2010. 10.050. 1. Miller WR, Rollnick S. Motivational Interviewing: Preparing People for Change. New York: Guilford Press; 2002. 2. Nolan RP. The Community Outreach Heart Health and Risk Reduction Trial: Facilitator’s Guide. Toronto, Ontario, Canada: University Health Network, 2002. 3. Balady GJ, Williams MA, Ades PA, Bittner V, Comoss P, Foody JA, Franklin B, Sanderson B, Southard D. Core components of cardiac rehabilitation/secondary prevention programs: 2007 update: a scientific statement from the American Heart Association Exercise, Cardiac Rehabilitation, and Prevention Committee, the Council on Clinical Cardiology; the Councils on Cardiovascular Nursing, Epidemiology and Prevention, and Nutrition, Physical Activity, and Metabolism, and the American Association of Cardiovascular and Pulmonary Rehabilitation. J Cardiopulm Rehabil Prev 2007;27:121–129. 4. Grundy SM, Pasternak R, Greenland P, Smith S Jr, Fuster V. Assessment of cardiovascular risk by use of multiple-risk-factor assessment equations: a statement for healthcare professionals from the American Heart Association and the American College of Cardiology. Circulation 1999;100:1481–1492. 5. Beck AT, Steer RA. Internal consistencies of the original and revised Beck Depression Inventory. J Clin Psychol 1984;40:1365–1367. 6. Prochaska JO, Velicer WF, Rossi JS, Goldstein MG, Marcus BH, Rakowski W, Fiore C, Harlow LL, Redding CA, Rosenbloom D. Stages of change and decisional balance for 12 problem behaviors. Health Psychol 1994;13:39 – 46.
7. Health Canada. Canada Food Guide to Healthy Eating. Ottawa, Ontario, Canada: Health Canada, 1997. 8. Health Canada. Canada’s Physical Activity Guide to Healthy Active Living. Ottawa, Ontario, Canada: Health Canada, 1998. 9. Canadian Council on Smoking and Health. Your Guide to a Smoke Free Future. Ottawa, Ontario, Canada: Canadian Council on Smoking and Health, 1996. 10. Steptoe A, Kerry S, Rink E, Hilton S. The impact of behavioral counseling on stage of change in fat intake, physical activity, and cigarette smoking in adults at increased risk of coronary heart disease. Am J Public Health 2001;91:265–269. 11. Plotnikoff RC, Hotz SB, Birkett NJ, Courneya KS. Exercise and the transtheoretical model: a longitudinal test of a population sample. Prev Med 2001;33:441– 452. 12. Glanz K, Patterson RE, Kristal AR, DiClemente CC, Heimendinger J, Linnan L, McLerran DF. Stages of change in adopting healthy diets: fat, fiber, and correlates of nutrient intake. Health Educ Q 1994;21:499 –519. 13. Kristal AR, Glanz K, Curry SJ, Patterson RE. How can stages of change be best used in dietary interventions? J Am Diet Assoc 1999; 99:679 – 684. 14. Logue E, Sutton K, Jarjoura D, Smucker W. Obesity management in primary care: assessment of readiness to change among 284 family practice patients. J Am Board Fam Pract 2000;13:164 –171. 15. Beland Y. Canadian community health survey—methodological overview. Health Rep 2002;13:9 –14. 16. Wang H, Chow SC, Li G. On sample size calculation based on odds ratio in clinical trials. J Biopharm Stat 2002;12:471– 483. 17. Wister A, Loewen N, Kennedy-Symonds H, McGowan B, McCoy B, Singer J. One-year follow-up of a therapeutic lifestyle intervention targeting cardiovascular disease risk. CMAJ 2007;177:859 – 865. 18. Garcia-Lizana F, Sarria-Santamera A. New technologies for chronic disease management and control: a systematic review. J Telemed Telecare 2007;13:62– 68. 19. Neubeck L, Redfern J, Fernandez R, Briffa T, Bauman A, Freedman SB. Telehealth interventions for the secondary prevention of coronary heart disease: a systematic review. Eur J Cardiovasc Prev Rehabil 2009;16:281–289. 20. Grover SA, Lowensteyn I, Joseph L, Kaouache M, Marchand S, Coupal L, Boudreau G. Patient knowledge of coronary risk profile improves the effectiveness of dyslipidemia therapy: the CHECK-UP study: a randomized controlled trial. Arch Intern Med 2007;167:2296 –2303. 21. Grover SA, Lowensteyn I, Joseph L, Kaouache M, Marchand S, Coupal L, Boudreau G. Discussing coronary risk with patients to improve blood pressure treatment: secondary results from the CHECK-UP study. J Gen Intern Med 2009;24:33–39. 22. Castaldo J, Nester J, Wasser T, Masiado T, Rossi M, Young M, Napolitano JJ, Schwartz JS. Physician attitudes regarding cardiovascular risk reduction: the gaps between clinical importance, knowledge, and effectiveness. Dis Manag 2005;8:93–105.
Patterns of Ventricular Tachyarrhythmias Associated With Training, Deconditioning and Retraining in Elite Athletes Without Cardiovascular Abnormalities Alessandro Biffi, MDa,*, Barry J. Maron, MDb, Franco Culasso, MDc, Luisa Verdile, MDa, Fredrick Fernando, MDd, Barbara Di Giacinto, MDa, Fernando M. Di Paolo, MDa, Antonio Spataro, MDa, Pietro Delise, MDe, and Antonio Pelliccia, MDa Ventricular tachyarrhythmias commonly occur in trained athletes during ambulatory Holter electrocardiography and are usually associated with a benign course. Such arrhythmias have been demonstrated to be sensitive to short periods of athletic deconditioning; however, their response to retraining is not known. Twenty-four hour Holter electrocardiographic monitoring was performed at peak training and after 3 to 6 months of deconditioning and was repeated in the present study after 2, 6, and 12 months of retraining in 37 athletes with frequent and complex ventricular tachyarrhythmias and without cardiovascular abnormalities. These subjects showed partial (101 to 500 ventricular premature complexes [VPCs]/24 hours) or marked (<100 VPCs) reversibility of arrhythmias after deconditioning. Retraining initially resulted in a significant increase in arrhythmia frequency compared with deconditioning (from 280 ⴞ 475 to 1,542 ⴞ 2,186 VPCs; p ⴝ 0.005), couplets (0.14 ⴞ 0.42 to 4.4 ⴞ 8.2; p ⴝ 0.005), and nonsustained ventricular tachycardia (from 0 to 0.8 ⴞ 1.8; p ⴝ 0.02). Subsequently, a progressive reduction was seen in the frequency of all ventricular arrhythmias during the 1 year of training to well below that at the peak training levels (VPCs 917 ⴞ 1,630, couplets 1.8 ⴞ 4.2, and nonsustained ventricular tachycardia 0.4 ⴞ 1.2). Such annual arrhythmia reduction was significantly greater statistically in those athletes with marked reversibility after deconditioning than in the athletes with partial reversibility (69 ⴞ 139 vs 1,496 ⴞ 1,917 VPCs/24 hours, respectively; p ⴝ 0.007). No cardiac events or symptoms occurred during 1 year of follow-up. In conclusion, in elite athletes without cardiovascular disease, a resumption in intense training after deconditioning was associated with variable, but prolonged, suppression of ventricular ectopy. The absence of adverse clinical events or symptoms associated with the resumption of training supports the continued eligibility in competitive sports for such athletes and is also consistent with the benign nature of physiologic athlete’s heart syndrome. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:697–703) Ventricular tachyarrhythmias are not uncommon findings on ambulatory Holter electrocardiogram (ECG) in trained athletes1–3 and have usually been associated with benign outcomes in the absence of cardiovascular abnormalities.4,5 Such arrhythmias have been demonstrated to be sensitive to short periods of athletic deconditioning6 and largely independent of training-related physiologic left ventricular (LV) remodeling.7,8 However, the course of the ventricular tachyarrhythmias after the resumption of physical training and competition (occurring after a period of complete deconditioning) is unknown. This becomes a rela
Institute of Sports Medicine and Science, Italian National Olympic Committee, Rome, Italy, bHypertrophic Cardiomyopathy Center, Minneapolis Heart Institute Foundation, Minneapolis, Minnesota; cUniversity of Rome “La Sapienza,” Rome, Italy; dSant’Andrea Hospital, Rome, Italy; and eDivision of Cardiology, Hospital of Conegliano, Conegliano, Italy. Manuscript received September 1, 2010; manuscript received and accepted October 26, 2010. *Corresponding author: Tel: (⫹39) 06-3685-9185; fax: (⫹39) 063685-9256. E-mail address:
[email protected] (A. Biffi). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.049
evant issue in clinical practice, because athletes with frequent and complex ventricular arrhythmias in the absence of structural heart disease often represent clinical management dilemmas, including the decisions regarding sports disqualification versus eligibility. The aim of the present study was to assess, in a unique subset of elite athletes, the response and clinical significance of ventricular tachyarrhythmias on resumption of exercise and competition after a forced period of deconditioning. Methods The case records of the Institute of Sports Medicine and Science were reviewed, and 355 athletes who had met the following criteria were selected with 24-hour ambulatory (Holter) ECG: (1) ⱖ3 ventricular premature complexes (VPCs) on the at rest 12-lead ECG (n ⫽ 337); and/or (2) a history of palpitations (n ⫽ 18). Of the 355 athletes, 284 with ⬍2,000 VPCs/24 hours, who were allowed to continue in competition and training, were excluded from the present analysis.4 The remaining 71 athletes with frequent and/or complex ventricular arrhythmias (arbitrarily defined as ⱖ2,000 www.ajconline.org
Data are presented as mean ⫾ SD (ranges). * p ⬍0.005 deconditioning versus peak conditioning; † p ⬍0.005 retraining at 2 and 6 months versus deconditioning; ‡ p ⬍0.02 retraining at 2, 6, and 12 months versus deconditioning.
917 ⫾ 1,630‡ (0–8,029) 1.8 ⫾ 4.2‡ (0–20) 0.4 ⫾ 1.2 (0–7) 1,087 ⫾ 1,830 (0–7,780) 3.0 ⫾ 5.6† (0–22) 0.6 ⫾ 1.6 (0–8) 1,542 ⫾ 2,186 (0–9,069) 4.4 ⫾ 8.2† (0–34) 0.8 ⫾ 1.8‡ (0–10) 280 ⫾ 475* (0–2,342) 0.14 ⫾ 0.42* (0–2) 0.08 ⫾ 0.5* (0–3) Ventricular premature complexes Couplets Nonsustained ventricular tachycardia
10,405 ⫾ 9,605 (2,140–43,221) 35.2 ⫾ 73 (1–321) 7.7 ⫾ 29.5 (1–179)
†
†
6 2
Retraining (months) Deconditioning Peak Conditioning
VPCs and/or ⱖ1 burst of nonsustained ventricular tachycardia [NSVT]) were considered for inclusion in the present study. One of these athletes died suddenly from arrhythmogenic right ventricular cardiomyopathy (ARVC) and was excluded because he had not undergone detraining. Of the remaining 70 athletes, 50 were without detectable cardiovascular abnormalities, and the other 20 had structural cardiac diseases, including ARVC (n ⫽ 6), mitral valve prolapse (n ⫽ 6), myocarditis (n ⫽ 4), and dilated cardiomyopathy (n ⫽ 4). These latter athletes with cardiac disease were excluded from sports activity (and from the present study) in accordance with the current Italian guidelines.9 Of the 50 athletes with no cardiovascular abnormalities, 37 (75%) had shown a partial (101 to 500 VPCs/24 hours; n ⫽ 21) or marked (⬍100 VPCs; n ⫽ 16) reversibility of the ventricular arrhythmias within the 3- to 6-month period of deconditioning. These 37 athletes resumed training and competition (after deconditioning) for a 1-year period (mean 12 ⫾ 0.6 months) and represent the present study group. The inclusion and exclusion criteria, which have taken into account the criteria of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement,10 are presented in Figure 1. The study athletes were engaged in a variety of sports disciplines, most commonly soccer (n ⫽ 10 [27%]), basketball (n ⫽ 6 [16%]), and volleyball (n ⫽ 4 [10%]). They also presented a broad spectrum of athletic achievement, with 13
Variable
Figure 1. Flow chart showing inclusion (Left) and exclusion (Right) criteria used in this study design.
12
The American Journal of Cardiology (www.ajconline.org)
Table 1 Ventricular tachyarrhythmias on 24-hour ambulatory (Holter) electrocardiogram (ECG) at peak conditioning, deconditioning, and retraining in 37 elite athletes without cardiovascular abnormalities
698
Arrhythmias and Conduction Disturbances/Training-Related Ventricular Tachyarrhythmias
699
45000 40000
VPCs/24 hours
35000 30000 25000 20000 15000 10000 5000
M
O
12
6 ET R R
ET R R
AI NI NG
AI NI NG
2 AI NI NG R
ET R
DI T O N EC D
M
O M
G IO N
IN
AI NI NG TR PE AK
O
0
Figure 2. Scatter plot of number of VPCs/24 hours of each athlete at peak training, deconditioning, and retraining at 2, 6, and 12 months (connected by continued lines) showing gradual and progressive reduction of ventricular ectopy (with the exception of an initial increase after 2 months) during 1 year of retraining.
(35%) participating at an elite level, including 9 who had competed in the Olympic Games or World Championships. The mean age of the athletes was 24 ⫾ 10 years (range 18 to 33), and 28 (75%) were men. All athletes were asymptomatic. At the subsequent Holter electrocardiographic monitoring, no athlete was taking antiarrhythmic or other cardioactive medications. The operators who analyzed the Holter ECGs were unaware of the phase of deconditioning or retraining. The 24-hour ambulatory (Holter) ECGs were initially recorded during the periods of peak training, including conditioning sessions (average 1 hour in duration), similar to that generally performed by the athlete; the remaining time was occupied by the usual daily activities, which could have involved recreational physical activity. The data related to 24-hour Holter ECGs during peak training in the 37 athletes constituting the study group have been previously reported.4 The athletes underwent a complete deconditioning period of ⱖ3 consecutive months (mean 19 ⫾ 6 weeks, range 12 to 24). This period was selected, because it has been previously shown to be sufficient to reverse the cardiovascular adaptations induced by physical training, including LV hypertrophy.11 After deconditioning, each athlete underwent a second cardiovascular assessment that also included a 24-hour Holter ECG performed under the same conditions as at peak training. The data relative to this deconditioning period have been previously reported.6 After the period of deconditioning, each athlete resumed competitive sports without restriction, according to the usual program and intensity of their athletic training. Each athlete underwent additional cardiovascular assessments, including 24-hour ambulatory [Holter] ECGs performed under the same conditions as previously, at 2, 6, and 12 months of retraining. These latter 24-hour Holter ECGs included conditioning sessions similar to that performed at the initial (peak training) and second (deconditioning) 24-hour Holter ECGs. The data obtained from the athletes at 2, 6, and 12
months of retraining were compared to those obtained at peak conditioning and at deconditioning. In the previous study of deconditioning,6 we assembled as a control group, 148 athletes without structural heart disease, of similar age to the study subjects (26 ⫾ 10 years), with less frequent ventricular arrhythmias (⬍2,000 VPCs/ 24 hours, mean 1,211 ⫾ 850, and without NSVT). The 148 controls underwent a second Holter ECG 3 to 6 months after peak training, maintaining the same level of training, and without previous deconditioning. The period between these 2 Holter recordings obtained during training (19 ⫾ 4 weeks, range 12 to 24) was the same as between the active training and deconditioned phases in the 37 athletes of our study group. The echocardiographic studies were performed at peak training, after deconditioning, and 2.6 and 12 months of retraining using commercially available instruments (Sonos 5500, Philips, Cleveland, Ohio). Images of the heart were obtained in multiple cross-sectional planes using standard transducer positions. The LV cavity dimensions, anterior, ventricular, septal, and posterior free wall thicknesses and left atrial dimensions were obtained from the M-mode echocardiograms, in accordance with previous recommendations.12 The LV mass was calculated using the formula of Devereux et al13 and normalized to the body surface area. The mean, median, and SDs for VPCs, couplets, and NSVT were computed at different points (peak conditioning, deconditioning, and retraining at 2, 6, and 12 months). Comparisons between the mean values were performed using the paired t test or Mann-Whitney U test, as appropriate. The same analyses were performed in the 2 subgroups of athletes with either partial (ⱖ100 VPCs/24 hours) or marked (⬍100 VPCs/24 hours) reversibility of arrhythmias after deconditioning. The data were then log transformed (to achieve approximately normal distributions), and the differences in VPCs at retraining (2, 6, and 12 months) were compared with those at deconditioning. Finally, we applied a linear regression model (to avoid the problem of multiple
700
The American Journal of Cardiology (www.ajconline.org)
Table 2 Different response in ventricular ectopy during 1 year of retraining in elite athletes with partial or marked reversibility of arrhythmias after deconditioning Variable
Peak training Deconditioning Retraining (months) 2 6 12
VPCs
p Value
Athletes With Marked Reversibility (⬍100 VPCs/24 hours; n ⫽ 16)
Athletes With Partial Reversibility (ⱖ100 VPCs/24 hours; n ⫽ 21)
12,470 ⫾ 12,193 (2,140–43,221) 14.6 ⫾ 24 (0–63)
8,996.4 ⫾ 7,343 (2,235–25,580) 281.4 ⫾ 109 (120–498)
0.28 0.0001
2,054.6 ⫾ 2,446 (65–9,069) 1,734.4 ⫾ 2,152 (4–7) 1,496.0 ⫾ 1,917 (11–8,029)
0.084 0.007 0.007
790.2 ⫾ 1,514 (0–5,032) 139.4 ⫾ 216 (0–785) 69.3 ⫾ 139 (0–552)
Data are presented as mean ⫾ SD (range).
testing) to assess the relation of VPC frequency with retraining at 2, 6, and 12 months and at deconditioning. Results At the initial assessment, during the peak training period, the frequency of VPCs/24 hours was 2,140 to 43,221 (mean 10,405 ⫾ 9,605; Table 1 and Figure 2). Of the 37 athletes, 23 (62%) had ⱖ1 couplet (mean 35.2 ⫾ 73, range 1 to 321); 21 (57%) also had 1 to 179 bursts of NSVT (mean 7.7 ⫾ 29.5), consisting of 3 to 28 consecutive beats, at a heart rate of 130 to 270 beats/min. After the deconditioning period, the study group showed a significant reduction in VPCs (98%), couplets (98%), and NSVT (100%; p ⬍0.005 with respect to peak training). The individual subject analysis showed that after deconditioning, all 37 athletes had either partial (ⱖ100 to 500 VPCs/24 hours; n ⫽ 21) or marked (⬍100 VPCs; n ⫽ 16) reversibility of ventricular tachyarrhythmias (Figure 2). Only 4 of 23 athletes showed the persistence of ventricular couplets after deconditioning (3 athletes had 1 couplet/24 hours and 1 athlete had 2 couplets/24 hours). None of the 21 athletes with NSVT at peak training had NSVT after deconditioning. After deconditioning, the athletes resumed their training programs and participated in competitive sports events without restriction. After 2 months of retraining, the number of VPCs significantly increased compared to that at deconditioning to 1,542 ⫾ 2,186 (p ⬍0.005; 80% change). The number of couplets increased to 4.4 ⫾ 8.2 (p ⬍0.005; 95% change) and NSVT to 0.8 ⫾ 1.8 (p ⬍0.02; 90% change). Nevertheless, the VPC, couplet, and NSVT frequency remained at significantly less than that peak training levels (p ⬍0.01; Table 1). Individual subject analysis showed a significant increase in ventricular tachyarrhythmias within 2 months of retraining in 18 athletes (49% change; to 500 to 2,000 VPCs/24 hours; p ⬍0.001 compared with deconditioning), including 9 with a marked increase in this arrhythmia (to ⬎2,000 VPCs/24 hours). The remaining 19 athletes showed nonsignificant change in VPCs with respect to the deconditioning period (to 0 to 500 VPCs/24 hours; p ⫽ 0.78; Figure 2). Couplets significantly reappeared compared to the deconditioning levels after 2 months of retraining in 18 athletes (48% change; from 1 to 34 couplets/24 hours; p ⫽ 0.002) and NSVT in 13 athletes (35% change; from 1 to 10 episodes of NSVT/24 hours; p ⫽ 0.0001). Increased ar-
rhythmia frequency at 2 months of retraining was greater in the athletes with previous partial reversibility of VPC frequency than in those with marked reversibility (2,054.6 ⫾ 2,446 vs 790.2 ⫾ 1,514 VBCs/24 hours; p ⫽ 0.084; Table 2). After 6 months of retraining, the VPCs showed a decrease to 1,087 ⫾ 1,830 (i.e., 30% change with respect to that at 2 months of retraining; Figure 2). A similar trend was evident for complex arrhythmias, with couplets decreasing to 3.0 ⫾ 5.6 (32% change compared with at 2 months of retraining) and NSVT to 0.6 ⫾ 1.6 (25% change). After 6 months of retraining, the VPC, couplet, and NSVT frequency remained at significantly less than peak training levels (p ⬍0.01). Individual subject analysis showed no significant changes from 2 to 6 months in the number of athletes with 500 to 2,000 VPCs/24 hours (from 19 to 15 athletes; p ⫽ 0.53). No difference was also found in those athletes who showed a marked increase of arrhythmia at 2 months of retraining (to ⬎2,000 VPCs/24 hours) (from 9 athletes at 2 months to 8 athletes at 6 months of retraining; p ⫽ 0.94). Athletes with less-frequent ventricular arrhythmias (0 to 500 VPCs/24 hours) remained substantially unchanged from 2 to 6 months of retraining (18 vs 22 athletes, respectively; p ⫽ 0.52; Figure 2). The number of athletes with couplets and NSVT showed no significant variation from 2 to 6 months of retraining (from 18 to 16 athletes, p ⫽ 0.91; and from 13 to 11 athletes, p ⫽ 0.83, respectively). Furthermore, individual subject analysis showed that the 16 athletes with marked reversibility of ventricular tachyarrhythmias after deconditioning had a significantly greater decrease in arrhythmias at 6 months of retraining (to 139.4 ⫾ 216 VPCs/24 hours) compared to that of the 21 athletes with only partial reversibility, for whom this arrhythmia remained consistent (1,734.4 ⫾ 2,152 VPCs; p ⫽ 0.007; Table 2). After 1 year of retraining, VPCs continued to decrease to 917 ⫾ 1,630 (i.e., 16% change with respect to 6 months of retraining; Figure 2). Couplets and NSVT also showed an additional decrease to 1.8 ⫾ 4.2 (40% change compared to that at 6 months of retraining) and to 0.4 ⫾ 1.2 (35% change), respectively. VPCs, couplets, and NSVT continued to remain at significantly less than the peak training levels (p ⬍0.01; Table 1). Individual subject analysis showed no significant changes from 6 to 12 months in athletes with 500 to 2,000 VPCs/24 hours (from 15 to 14 athletes; p ⫽ 0.97). A not significant reduction was found in the athletes who showed a marked increase of arrhythmia at 2 months of
Arrhythmias and Conduction Disturbances/Training-Related Ventricular Tachyarrhythmias
retraining (to ⬎2,000 VPCs/24 hours; from 8 athletes at 6 months to 4 athletes at 12 months of retraining; p ⫽ 0.21). The athletes with less-frequent ventricular arrhythmias (0 to 500 VPCs/24 hours) remained substantially unvaried from 6 to 12 months of retraining (22 vs 23 athletes, respectively; p ⫽ 0.97; Figure 2). Also, the athletes with couplets and NSVT did not show a significant variation from 6 to 12 months (from 16 to 11 athletes, p ⫽ 0.36; and from 11 to 7 athletes, p ⫽ 0.40, respectively). Furthermore, individual subject analysis showed that athletes with marked reversibility of arrhythmias with deconditioning continued to have a more prolonged suppression of VPCs at 1 year of retraining compared to athletes with partial reversibility (69 ⫾ 139 vs 1,496 ⫾ 1,917 VPCs; p ⫽ 0.007; Table 2). Linear regression analysis applied to the differences in VPC frequency between deconditioning and the 3 phases of retraining showed a statistically significant association between retraining at 6 and 12 months compared to that at deconditioning. Of the 37 athletes, 30 (81%) showed a pattern of VPC morphology consistent with a right ventricular outflow tract origin and 7 a fascicular origin (19%). Of the 30 athletes with a right ventricular origin of VPCs, 28 had an inferior axis and 2 showed an indeterminate axis. Ventricular arrhythmias with marked reversibility showed a right ventricular outflow tract origin, with an inferior axis in all 16 athletes. At peak training, the LV mass index was 114 ⫾ 23 g/m2. After deconditioning, it had decreased to 94 ⫾ 19 g/m2 (p ⬍0.001). During retraining, the LV mass index had increased to 96 ⫾ 18 g/m2 at 2 months (p ⫽ 0.61), 101 ⫾ 21 g/m2 at 6 months (p ⫽ 0.27), and 110 ⫾ 19 g/m2 at 12 months (p ⬍0.05). The changes in LV mass index with training did not differ between the 18 athletes who experienced a reappearance of ventricular tachyarrhythmias with retraining and the 19 athletes without a reappearance of arrhythmias (95 ⫾ 11 vs 96 ⫾ 10 g/m2 at 2 months of retraining, p ⫽ 0.55). No significant structural or functional abnormality of the right ventricle was identified in any athlete. No cardiac events or symptoms occurred in the athletes with or without a reappearance of ventricular tachyarrhythmias during the 1-year follow-up period, during which training and competition had been resumed. The cardiovascular evaluations with at rest and exercise ECGs and 2-dimensional echocardiography at 2, 6, and 12 months of retraining in all athletes and selectively with cardiac magnetic resonance (n ⫽ 30), signal-averaged electrocardiography (n ⫽ 37), programmed ventricular stimulation (n ⫽ 13), and myocardial biopsy (n ⫽ 2) did not detect previously unrecognized cardiovascular abnormalities in any athlete. Only 2 of the 37 athletes incurred ventricular tachyarrhythmias during exercise testing; the remaining 35 athletes had a complete disappearance of arrhythmia during exercise. No athlete experienced cardiac symptoms during exercise testing. In the 13 athletes who underwent an electrophysiologic study, either no arrhythmia or only a nonsustained ventricular response (3 to 5 consecutive ectopic beats) was induced by programmed ventricular stimulation. None of the athletes underwent radiofrequency catheter ablation. In the 148 control athletes, who had not modified their physical conditioning, no significant variability in ventric-
701
ular arrhythmias was found between the 2 ambulatory Holter ECGs obtained during training (19 ⫾ 4 weeks apart during the training period). The mean number of VPCs on the first Holter ECG was 1,211 ⫾ 850 and was 1,050 ⫾ 648 on the second (p ⫽ 0.28). Also, no significant difference was found with regard to couplets and NSVT (0 vs 0). Discussion We had previously observed that intense athletic conditioning was not uncommonly associated with the occurrence of frequent and/or complex ventricular tachyarrhythmias on ambulatory Holter ECGs.4 Despite the presence of such tachyarrhythmias, the risk of sudden death has proved to be exceedingly low during an 8-year follow-up period (annual mortality rate 0.17%). In a subsequent study,6 we demonstrated that these arrhythmias are sensitive to shorts periods of deconditioning, independent of whether structural cardiovascular abnormalities are present. The present study has extended these observations by assessing a unique subset of 37 elite athletes without cardiovascular abnormalities who had demonstrated partial or marked reversibility of frequent and/or complex ventricular tachyarrhythmias with deconditioning, at which point they resumed training and competition. The decision to resume training in athletes with frequent ventricular tachyarrhythmias reversed by deconditioning was justified by the absence of structural cardiovascular abnormalities and by our previous experience with 284 athletes with less frequent arrhythmias (⬍2,000 VPCs/24 hours), who did not develop unfavorable consequences during an 8-year follow-up period during which training and competition continued.4 The major finding of the present study was that, although retraining was initially associated with an increase in ventricular tachyarrhythmias (after 2 months), a gradual and progressive regression of all forms of arrhythmia were subsequently observed during the remainder of the 1-year follow-up period (ultimately to a level significantly below that of peak training). Such regression was most evident in those athletes with marked reversibility of arrhythmias after deconditioning in contrast to that of athletes with only partial reversibility. Therefore, a single period of physical deconditioning appeared to convey a prolonged suppressive effect on ventricular ectopy, even after a 1-year resumption of intense exercise training. Furthermore, the resumed physical retraining was not associated with cardiac symptoms, clinical events, or evolving expression of cardiac disease14 in all athletes with frequent ventricular tachyarrhythmias sensitive to deconditioning. Therefore, we believe that the deconditioning strategy could be regarded as a useful clinical tool in managing competitive athletes with frequent and/or complex ventricular tachyarrhythmias in the absence of heart disease, supporting the recommendation stated in the recent international guidelines on sports eligibility.9,15,16 The observed changes in ventricular arrhythmias during the 1-year period could have resulted theoretically from the resolution of a previously unrecognized myocardial process (e.g., myocarditis) or might represent the preclinical expression of disease states such as ARVC or hypertrophic cardiomyopathy.14 However, the latter consideration seems highly unlikely to explain the findings in all or most athletes
702
The American Journal of Cardiology (www.ajconline.org)
in the present analysis. Also, a normal QT interval on a 12-lead ECG largely excluded ion channelopathies, which could have been responsible for the arrhythmogenicity in athletes with structurally normal hearts.2,17–19 Currently, we do not have data regarding the retraining in the other 13 athletes who were excluded from the present study because of frequent and/or complex ventricular tachyarrhythmias (ⱖ2,000 VPCs/24 hours) and were not allowed to resume training because of a lack of ventricular arrhythmia reversibility with deconditioning (Figure 1). Our prudent decision was determined by the consideration that, although these 13 athletes had a good prognosis at the 1-year follow-up examination, we could not exclude the possibility that such a favorable outcome was also influenced by the withdrawal from regular training. The persistence (or increased frequency) of complex ventricular tachyarrhythmias over time (and despite suspension of training) might represent the first expression of a clinically silent arrhythmogenic cardiomyopathy, as suggested by the athlete who continued competitive sports against medical advice and died suddenly from ARVC.4,18 The mechanisms that explain the changes in ventricular tachyarrhythmias with deconditioning and retraining are unresolved but appear to be independent of conditioningrelated physiologic cardiac remodeling. Our findings have demonstrated that changes in the LV mass with physical training did not differ between the athletes with and without a reappearance of ventricular tachyarrhythmias during retraining, and are consistent with our previous observations in which no correlation was evident between the frequency or complexity of ventricular ectopy and the magnitude of training-induced LV hypertrophy.7,8 Furthermore, the gradual increase in LV mass observed from 2 months to 1 year of retraining was associated with a parallel regression, and not an increase, in all forms of ventricular tachyarrhythmias during the same period. Although less intensive conditioning regimens could theoretically be responsible for a reduction in ventricular ectopy in some athletes during the retraining phase, this was an unlikely explanation, because retraining was permitted without restriction for all athletes (and was confirmed by a progressive increase in LV mass during the same period). Other factors and mechanisms can potentially promote ventricular arrhythmogenicity in conditioned athletes. Autonomic nervous system adaptations induced by a varying intensity of physical training have been reported to have a role in the genesis of ventricular tachyarrhythmias in susceptible subjects.20 –23 It is possible that in our unique cohort of elite athletes, a shift in cardiovascular autonomic modulation from parasympathetic to sympathetic predominance because of an intensive and systematic training regimen24 might have predisposed our athletes to electrical instability, triggering ventricular tachyarrhythmias. The resolution of arrhythmias with deconditioning and their reappearance in the early stage of retraining could have been consistent with such a neural hypothesis. However, it cannot explain the gradual reduction of ventricular tachyarrhythmias during continued training period during a 1-year follow-up period. Furthermore, the possible influence of the hormonal changes (in particular of the cortisol and adrenaline levels) induced by prolonged and intensive training on
myocardial irritability should be taken into account.25,26 Thus, we propose a prolonged “antiarrhythmic” effect of deconditioning in sensitive athletes, in which ventricular ectopic foci might become less prone to autonomic influences over time. However, we wish to underscore that our data need to be confirmed by other studies in larger athletic populations. Also, a limitation of the present study was the absence of a control group of athletes (with ⬎2,000 VPCs/24 hours and NSVT) who had not undergone periods of deconditioning and retraining and their relation to ventricular arrhythmias. The acquisition of these data could better exclude the influence of day-to-day variability of an ectopic burden as a confounder in the present study and of regression toward the mean as a potential explanation for reduction in ectopy over time. Finally, in none of our athletes were antiarrhythmic or other cardioactive medications administered. Each athlete underwent multiple comprehensive screenings for performance-enhancing drugs in accord with recommendations of the World AntiDoping Agency and International Olympic Committee,27 thereby excluding the presence of steroids, amphetamines, cocaine, growth hormones, and erythropoietin. 1. Zehender M, Meinertz T, Keul J, Just H. ECG variants and cardiac arrhythmias in athletes: clinical relevance and prognostic importance. Am Heart J 1990;119:1379 –1391. 2. Palatini P, Maraglino G, Sperti G, Calzavara EC, Libardoni M, Pessina AC, Dal Palù C. Prevalence and possible mechanisms of ventricular arrhythmias in athletes. Am Heart J 1985;110:560 –565. 3. Pantano JA, Oriel RJ. Prevalence and nature of cardiac arrhythmias in apparently normal well-trained runners. Am Heart J 1982;104:762– 768. 4. Biffi A, Pelliccia A, Verdile L, Fernando F, Spataro A, Caselli S, Santini M, Maron BJ. Long-term clinical significance of frequent and complex ventricular tachyarrhythmias in trained athletes. J Am Coll Cardiol 2002;40:446 – 452. 5. Bjornstad HH, Bjornstad TH, Urheim S, Hoff PI, Smith G, Maron BJ. Long-term assessment of electrocardiographic and echocardiographic findings in Norwegian elite endurance athletes. Cardiology 2009;112: 234 –241. 6. Biffi A, Maron BJ, Verdile L, Fernando F, Spataro A, Marcello G, Ciardo R, Ammirati F, Colivicchi F, Pelliccia A. Impact of physical deconditioning on ventricular tachyarrhythmias in trained athletes. J Am Coll Cardiol 2004;44:1053–1058. 7. Biffi A, Maron BJ, Di Giacinto B, Porcacchia P, Verdile L, Fernando F, Spataro A, Culasso F, Casasco M, Pelliccia A. Relation between training-induced left ventricular hypertrophy and risk for ventricular tachyarrhythmias in elite athletes. Am J Cardiol 2008;101:1792–1795. 8. Biffi A, Ansalone G, Verdile L, Fernando F, Caselli G, Ammirati F, Pelliccia A, Santini M. Ventricular arrhythmias and athlete’s heart: role of signal-averaged electrocardiography. Eur Heart J 1996;17: 557–563. 9. Cardiovascular guidelines for eligibility in competitive sports (COCIS– 4th edition). Med Sport 2010;63:5–136. 10. Von Elm E, Altman DG, Egger M. The strengthening the Reporting of Observational Studied in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008;61: 344 –349. 11. Maron BJ, Pelliccia A, Spataro A, Granata M. Reduction in left ventricular wall thickness after deconditioning in highly trained Olympic athletes. Br Heart J 1993;69:125–128. 12. Sahn DJ, DeMaria A, Kisslo J, Weyman A. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978;58:1072–1083. 13. Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, Reichek N. Echocardiographic assessment of left ventricular hypertrophy: comparison with necropsy findings. Am J Cardiol 1986;57:450 – 458.
Arrhythmias and Conduction Disturbances/Training-Related Ventricular Tachyarrhythmias 14. Pelliccia A, Di Paolo FM, Quattrini FM, Basso C, Culasso F, Popoli G, De Luca R, Spataro A, Biffi A, Thiene G, Maron BJ. Outcomes in athletes with marked ECG repolarization abnormalities. N Engl J Med 2008;358:152–161. 15. Pelliccia A, Fagard R, Bjørnstad HH, Anastassakis A, Arbustini E, Assanelli D, Biffi A, Borjesson M, Carrè F, Corrado D, Delise P, Dorwarth U, Hirth A, Heidbuchel H, Hoffmann E, Mellwig KP, Panhuyzen-Goedkoop N, Pisani A, Solberg EE, van-Buuren F, Vanhees L, Blomstrom-Lundqvist C, Deligiannis A, Dugmore D, Glikson M, Hoff PI, Hoffmann A, Hoffmann E, Horstkotte D, Nordrehaug JE, Oudhof J, McKenna WJ, Penco M, Priori S, Reybrouck T, Senden J, Spataro A, Thiene G; Study Group of Sports Cardiology of the Working Group of Cardiac Rehabilitation and Exercise Physiology; Working Group of Myocardial and Pericardial Diseases of the European Society of Cardiology. Recommendations for competitive sports participation in athletes with cardiovascular disease: a consensus document from the Study Group of Sports Cardiology of the Working Group of Cardiac Rehabilitation and Exercise Physiology and the Working group of Myocardial and Pericardial Diseases of the European Society of Cardiology. Eur Heart J 2005;26:1422–1445. 16. Maron BJ, Douglas PS, Graham TP, Nishimura RA, Thompson PD, Bethesda Conference 36th Task Force 1: preparticipation screening and diagnosis of cardiovascular disease in athletes. J Am Coll Cardiol 2005;45:1322–1326. 17. Heidbüchel H, Corrado D, Biffi A, Hoffmann E, Panhuyzen-Goedkoop N, Hoogsteen J, Delise P, Hoff PI, Pelliccia A; Study Group on Sports Cardiology of the European Association for Cardiovascular Prevention and Rehabilitation. Recommendations for participation in leisure-time physical activity and competitive sports of patients with arrhythmias and potentially arrhythmogenic conditions. Part II: Ventricular arrhythmias, channelopathies and implantable defibrillators. Eur J Cardiovasc Prev Rehabil 2006;13:676 – 686.
703
18. Corrado D, Basso C, Schiavon M, Thiene G. Does sports activity enhance the risk of sudden cardiac death? J Cardiovasc Med 2006;7: 228 –233. 19. Rajappan K, O’Connel C, Sheridan DJ. Changes in QT interval with exercise in elite male rowers and controls. Int J Cardiol 2003;87:217– 222. 20. Fazio G, Novo G, Sutera L, Di Gesaro G, Fazio M, D’Angelo L, Visconti C. Sympathetic tone and ventricular tachycardia. J Cardiovasc Med 2008;9:963–966. 21. Ueno LM, Moritani T. Effects of long-term exercise training on cardiac autonomic nervous activities and baroreflex sensitivity. Eur J Appl Physiol 2003;89:109 –114. 22. O’Sullivan SE, Bell C. The effects of exercise and training on human cardiovascular reflex control. J Auton Nerv Syst 2000;81:16 –24. 23. Chen LS, Zhou S, Fishbein MC, Chen PS. New perspectives on the role of autonomic nervous system in the genesis of arrhythmias. J Cardiovasc Electrophysiol 2007;18:123–127. 24. Iellamo F, Legramante JM, Pigozzi F, Spataro A, Norbiato G, Lucini D, Pagani M. Conversion from vagal to sympathetic predominance with strenuous training in high-performance world class athletes. Circulation 2002;105:2719 –2724. 25. Bonifazi M, Sardella F, Lupo C. Preparatory versus main competitions: differences in performances, lactate responses and pre-competition plasma cortisol concentrations in elite male swimmers. Eur J Appl Physiol 2000;82:368 –373. 26. Takimoto K, Levitan ES. Glucocorticoid induction of Kv1.5 K⫹ channel gene expression in ventricle of rat heart. Circ Res 1994;75: 1006 –1013. 27. Striegel H, Rossner D, Simon P, Niess AM. The World Anti-Doping Code 2003: consequences for physicians associated with elite athletes. Int J Sports Med 2005;29:1–14.
Safety of Lower Activated Clotting Times During Atrial Fibrillation Ablation Using Open Irrigated Tip Catheters and a Single Transseptal Puncture Roger A. Winkle, MD*, R. Hardwin Mead, MD, Gregory Engel, MD, and Rob A. Patrawala, MD Guidelines largely based on closed-tip catheters recommend activated clotting times (ACTs) >300 to 350 seconds during atrial fibrillation (AF) ablation to prevent thrombus and char formation. Open irrigated tip catheters (OITC) may decrease complications and permit lower ACTs. This study evaluated factors contributing to vascular and hemorrhagic complications during AF ablation with emphasis on catheter type, anticoagulation level, procedural and clinical variables, and gender. In 1,122 AF ablations we examined catheter used, ACT level, gender, and complications. Target ACTs initially were >300 seconds and were decreased to 225 seconds for the OITC. Average ACT ranges were created: <250, 250 to 299, 300 to 350, and >350 seconds. Average ACT was <250 seconds in 557 ablations (complication rate 1.62%). Cochran–Armitage analysis showed that complications increased linearly as ACT increased and peaked at 5.55% for ablations with ACTs >350 seconds (p ⴝ 0.038). Women were older (66 ⴞ 10 vs 60 ⴞ 10 years, p <0.001) and had more paroxysmal AF (43% vs 28%, p ⴝ 0.007) and more hypertension (50% vs 40%, p ⴝ 0.013). Women received less heparin but were over-represented in higher ACT ranges (p <0.0001) consistent with a pharmacokinetic gender difference. There was no difference in vascular or hemorrhagic complications between men and women (2.3% vs 2.9%, p ⴝ 0.668). Multivariate logistic regression showed that only use of the OITC was associated with lower complication rates (p ⴝ 0.024). In conclusion, AF ablation with the OITC is safe with a target ACT of 225 seconds. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011; 107:704 –708) Consensus guidelines on catheter ablation recommend an intraprocedural activated clotting time (ACT) of 300 to 350 seconds and a longer time of 350 to 400 seconds in patients with spontaneous echocardiographic contrast noted on transesophageal or intracardiac echocardiogram.1 These recommendations are largely based on intracardiac echocardiographic observation of apparent thrombus and not actual clinical events.2– 4 These studies were performed before the introduction of the open irrigated tip catheter (OITC). Animal studies have shown that OITCs have lower tissue interface temperature to achieve deeper tissue heating than closed-tip catheters (CTCs) and closed irrigated tip catheters, which results in less or no local thrombus formation.5–7 Because of decreased thrombus formation with the OITC in animal models, after the introduction of OITCs we decreased our target ACT to 225 seconds during atrial fibrillation (AF) radiofrequency ablation. This study reports our experience with the OITC for AF ablation at lower target ACTs and examines factors associated with vascular and hemorrhagic complications.
Cardiovascular Medicine and Cardiac Arrhythmias, East Palo Alto, California, and Sequoia Hospital, Redwood City, California. Manuscript received August 11, 2010; revised manuscript received and accepted October 13, 2010. *Corresponding author: Tel: 650-617-8100; fax: 650-327-2947. E-mail address:
[email protected] (R.A. Winkle). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.048
Methods Subjects were consecutive patients undergoing AF ablation at Sequoia Hospital, Redwood City, California from October 10, 2003 through December 31, 2009. All patients had symptomatic AF and signed written informed consent for ablation. Data analysis was retrospective and approved by the Sequoia Hospital institutional review board. AF type was categorized as paroxysmal (type 1, lasting ⬍1 week), persistent (type 2, lasting ⬎1 week and ⬍1 year or requiring pharmacologic or electrical cardioversion in ⬍1 week), and longstanding persistent (type 3, lasting ⬎1 year).1 General anesthesia was used in 96% of ablations. Venous access was from the groin and most patients had all sheaths placed in the right groin. A 63-cm 8Fr transseptal sheath (St. Jude [St. Paul, Minnesota] Fast Cath or Biosense Webster [Diamond Bar, California] Preface SL1), a 23-cm 9Fr sheath (St. Jude Ultimum), and a 12-cm 8Fr sheath (St. Jude Fast Cath) were used. A 7Fr duodeca catheter (St. Jude Livewire) was placed through the 8Fr sheath and around the tricuspid annulus with the distal poles in the coronary sinus. A 9Fr Boston Scientific (Natick, Massachusetts) Ultra Ice intracardiac ultrasound catheter was inserted through the 9Fr sheath for transseptal puncture, which was done using a 71-cm St. Jude BRK or a Baylis (Montreal, Quebec, Canada) NRG needle. All patients had a femoral or a radial arterial line placed. The St. Jude NavX system was used in all cases. Before January 2006, all ablations were done with a CTC (Boston Scientific Blazer II, n ⫽ 8, or Webster Celsius 8 www.ajconline.org
Arrhythmias and Conduction Disturbances/Less Intense Anticoagulation During AF Ablation
705
Table 1 Data and complications by activated clotting time ranges ACT Ranges (seconds)
Number of ablations Heparin bolus (units) Heparin bolus (U/kg) Average activated clotting time/ablation (seconds) Nonirrigated tip catheters Number of groins Women Severe protamine reaction Tamponade/pericardiocentesis Arteriovenous fistula Groin pseudoaneurysm Hematoma (transfusion or surgery) Stroke Transient ischemic attack
⬍250
250–299
300–350
⬎350
557 8,240 ⫾ 1,539 90 ⫾ 12 224 ⫾ 16 0.2% 1.09 ⫾ 0.28 23.5% 0 3 (0.5%) 2 (0.4%) 3 (0.5%) 0 1 (0.2%) 0
331 8,952 ⫾ 1,336 99 ⫾ 16 272 ⫾ 15 4.8% 1.23 ⫾ 0.42 27.5% 1 (0.30%) 3 (0.9%) 1 (0.3%) 2 (0.6%) 2 (0.6%) 1 (0.30%) 0
196 9,843 ⫾ 1,493 115 ⫾ 22 319 ⫾ 14 42% 1.26 ⫾ 0.44 33.1% 0 3 (1.5%) 1 (0.5%) 1 (0.5%) 0 1 (0.5%) 1 (0.5%)
36 10,333 ⫾ 1,394 135 ⫾ 30 389 ⫾ 29 50% 1.25 ⫾ 0.44 55.6% 0 1 (2.8%) 0 0 1 (2.8%) 0 0
mm, n ⫽ 110) until January 2006 after which all ablations were done with an OITC (Webster Thermocool 3.5 mm or St. Jude Cool Path 4.0 mm, n ⫽ 1,004). From January until August 2006 the OITC power was gradually increased from 35 to 50 W after which all received 50-W ablations. No special effort was made to keep the transseptal sheath in the right atrium. All patients underwent circumferential atrial ablation around all pulmonary veins and a left atrial roof line ablation. The 50-W ablations were done with a technique of “perpetual motion” where the catheter was moved back and forth across a small area and was not left at a single site for extended periods. The ultrasound catheter was removed and the ablation catheter placed into the left atrium with a circumferential mapping catheter (7Fr Webster Lasso or St. Jude Reflexion Spiral) across the single transseptal puncture site to confirm pulmonary vein isolation. Touchup was done as needed to isolate all pulmonary veins, which was the primary end point of the study. Patients with atrial flutter also underwent ablation of a mitral or caval-tricuspid isthmus line, and many patients had ablation at sites of complex fractionated electrograms in the atria or coronary sinus, ablation of a low posterior left atrial line, and isolation of the superior vena cava. Patients with more persistent AF received more aggressive ablation protocols. Ablation was avoided near the esophagus, which was marked by a thermistor catheter. Patients receiving warfarin continued it until 5 days before the procedure. Three days before the ablation patients started enoxaparin 1 mg/kg every 12 hours with the last dose 24 hours before ablation. Patients not already receiving warfarin did not receive anticoagulation drugs before the procedure. Patients with persistent or frequent paroxysmal AF underwent transesophageal echocardiography. When the transseptal sheath was in the left atrium an intravenous heparin bolus was given followed by a 1,000U/hour infusion through the transseptal sheath (protected by an air filter) at a concentration of 10 U/ml. Mapping and ablation were begun immediately and the first ACT was obtained 15 minutes later. ACTs were drawn every 20 to 30 minutes during the procedure. Additional heparin was given and the drip rate adjusted depending on the ACT. The OITC
was infused with normal saline containing a heparin concentration of 2 U/ml at 2 ml/min at baseline increased to 30 ml/min while ablating. Our initial target ACT was 300 to ⱖ350 seconds. After the OITC became available in January 2006 we gradually decreased our target ACT to 225 seconds. Enoxaparin 0.5 mg/kg every 12 hours was started when the sheaths were removed and discontinued when the international normalized ratio was 2.0. Warfarin was started the evening of the procedure and continued for ⱖ3 months. Vascular or hemorrhagic complications were defined as pericardial tamponade (or effusion requiring pericardiocentesis), groin or other bleeding requiring transfusion or surgical intervention, groin pseudoaneurysms and arteriovenous fistulas, systemic thromboembolism, and stroke or transient ischemic attack. All strokes, tamponades, and groin complications requiring surgery were considered major and transient ischemic attacks or complications requiring transfusion only or treated noninvasively or with a thrombin injection were considered minor. A single major protamine reaction was also included because it was related to anticoagulation. Most patients were discharged the morning after the procedure but all were contacted by telephone every few days for the first month after the procedure. For all ablations we recorded start (lidocaine injection), procedure end (sheath suture or removal), and left atrial times, heparin bolus, patient weight, and all ACT values. For each ablation mean ACT was calculated from individual ACTs. For data analysis a series of average ACT ranges was defined: ⬍250, 250 to 299, 300 to 350, and ⬎350 seconds. Percentage of patients with a complication in each ACT range was calculated. Heparin bolus, heparin bolus normalized for body weight, use of OITCs, and number of groins/procedure were calculated for each ACT range. Procedure time, left atrial time, heparin bolus, heparin bolus normalized for body weight, age, type of AF, number of antiarrhythmic drugs failed, and presence of hypertension were analyzed separately for men and women. Statistical analysis was done using XLSTAT 2010 (Paris, France). Continuous data were described as mean ⫾ SD and counts and percentages if categorical. Student’s t test, chisquare test, and Fisher’s exact test were used to compare
706
The American Journal of Cardiology (www.ajconline.org)
. .
Figure 1. As intensity of heparin anticoagulation increased, reflected by a longer average activated clotting time, vascular and hemorrhagic complications of atrial fibrillation ablation showed a linear and statistically significant increase
Table 2 Procedural gender differences for atrial fibrillation ablation
Procedure time (minutes) Left atrial time (minutes) Heparin bolus (units) Heparin bolus (U/kg) Number of groins used Complications
Men
Women
p Value
136 ⫾ 48 104 ⫾ 40 9,137 ⫾ 1,375 96 ⫾ 17 1.16 ⫾ 0.37 2.3%
139 ⫾ 48 99 ⫾ 40 7,902 ⫾ 1,818 103 ⫾ 25 1.17 ⫾ 0.38 2.9%
0.257 0.027 ⬍0.001 ⬍0.001 0.62 0.668
differences between men and women. The Cochran–Armitage trend test was used to evaluate complication rates and representation of women across ACT ranges. Logistic regression analysis was done to examine factors associated with procedural complications. Factors examined were age, left atrial size, average ACT, gender, irrigated versus CTC, duration of procedure, and paroxysmal versus nonparoxysmal AF. All tests were 2-sided and a p value ⬍0.05 was considered statistically significant. Results There were 1,122 ablations performed in 843 patients including 258 second ablations and 21 third ablations. Women accounted for 235 of patients (28%) and 307 (27%) of ablations. Average age at first ablation was 62 ⫾ 10 years and AF types were type 1 in 270 patients (32%), type 2 in 423 patients (50%), and type 3 in 150 patients (18%). Average procedure time was 138 ⫾ 48 minutes. For all 1,122 ablation procedures mean heparin dose was 8,799 ⫾ 1,606 U and mean heparin dose per kilogram of body weight was 98 ⫾ 20. Table 1 lists mean heparin bolus and dose per kilogram, which increased sequentially across ACT ranges. Among the 557 procedures with an average ACT ⬍250 seconds, 41 had an average ACT ⬍200 seconds. Table 1 presents vascular or hemorrhagic complications, which occurred in 28 of 1,122 procedures (2.5%). Seventeen of the 28 complications were major and 11 were minor. Cochran–Armitage test showed a significant increase in
Figure 2. Logistic regression data for factors contributing to vascular and hemorrhagic complications showed that only use of the open irrigated tip catheter compared to the closed-tip catheter was associated with a significant decrease in complications. Type of atrial fibrillation, gender, left atrial size, procedure time, mean activated clotting time, and age did not influence the complications rate. F ⫽ female; LA ⫽ left atrium; M ⫽ male
total complication rate (Figure 1) from 1.62% for procedures with ACTs ⬍250 seconds to 5.55% among those procedures with average ACTs ⬎350 seconds (p ⫽ 0.038). There were no deaths, atrial-esophageal fistulas, or pulmonary vein stenoses requiring intervention. Char was never seen on any irrigated tip catheter. At the time of first ablation the women were older (66 ⫾ 9 vs 60 ⫾ 10 years, p ⬍0.001) and had more hypertension (50% vs 43%, p ⫽ 0.013), more AF type 1 (43% vs 28%, p ⫽ 0.007), and less AF type 3 (11% vs 20%, p ⫽ 0.002) than men. Procedural data by gender are presented in Table 2. Procedure times were similar for women and men; however, left atrial times were slightly shorter for women. Number of groins used was similar for men and women. Despite a lower initial heparin loading bolus, women ended up with a higher dose than men when adjusted for body weight. Women were under-represented in the group with lower ACTs and over-represented in the groups with the longer ACTs (Table 1). Only 23.5% of procedures with ACTs ⬍250 seconds were in women. However, women represented 33.1% of procedures in the ACT range of 300 to 350 and 55.6% of procedures with ACTs ⬎350 (p ⬍0.0001). Of the 28 complications 9 occurred in women (2.9%) and 19 occurred in men (2.3%, p ⫽ 0.668). These univariate statistical analyses suggest that a higher ACT is associated with an increase in complications and that there is no gender differences in complications. When all clinical variables were evaluated without considering catheter type logistic analysis showed that only a higher mean ACT was predictive of complications (95% confidence interval 0.056 to 0.445, p ⫽ 0.012). When catheter type (irrigated vs closed tip) was added to the analysis there were still no gender differences in complications; however, catheter type (fewer complications with the OITC, 95% confidence intervals ⫺0.514 to ⫺0.036, p ⫽ 0.024) was the only significant contributor to complications (Figure 2). Although there was a trend for older age and higher ACTs
Arrhythmias and Conduction Disturbances/Less Intense Anticoagulation During AF Ablation
to increase complications and shorter procedure times to decrease complications, none of these variables achieved statistical significance. Given the very close linkage between use of the OITC and the lower ACTs it may be the combination of the 2 that improves procedural safety. We attempted to evaluate the influence of ACT on complications for ablations only using the OITC; however, there were too few complications with the OITC to draw any conclusions about the relative contribution of ACT level to complications. All ablations were done by the 4 authors and there was no physician difference in complications when operators were added to the analysis. Discussion AF ablation guidelines recommend much longer ACTs than we used in our procedures because of fear of systemic thromboembolism.1 These recommendations are based on echocardiographic observations of left atrial thrombi associated with the transseptal sheath and occurrence of char on CTCs.2– 4 Ren et al2 studied 511 patients undergoing AF ablation using 2 transseptal sheaths for left atrial catheterization. Left atrial thrombus occurred in 11.2% of patients when the ACT was 250 to 300 seconds and in 2.8% with ACTs from 300 to 410. In patients with spontaneous echocardiographic contrast in the left atrium, the lower ACT group had thrombus in 45% and the higher ACT group had thrombus in 4.6%. All these left atrial thrombi were aspirated and there were no embolic events in any patient. In this study “all transseptal sheaths were continually flushed with heparinized solution at 200 m/hour” but the heparin concentration was not specified. Maleki et al3 evaluated thrombus formation on the transseptal sheath detected by intracardiac echocardiography using a 2–transseptal sheath technique. Ninety patients had the sheath preflushed with a 2-U/ml heparin concentration and another 90 had the sheath preflushed with a 1,000-U/ml heparin concentration before obtaining vascular access. The target ACT was 250 to 300 seconds in the 2 groups and the 2 sheaths were perfused with heparin at 100 ml/hour at a 2-U/ml concentration. Thrombus was observed in 9% of sheaths preflushed with the lower heparin concentration and in only 1% preflushed with a higher concentration. All clots were removed with aspiration and no patients had a neurologic event. Wazni et al4 examined thromboembolic events and char formation in 785 patients undergoing AF ablation. They divided patients by ACT times (250 to 300, 300 to 350, and 350 to 400 seconds). Char formation was detected in 36% of patients in the lowest ACT group, 2.8% in the intermediate group, and 1.9% in the highest group. An embolic event occurred in 3.6% of the lowest ACT group, 1.7% in the intermediate group, and 0.5% in the highest ACT group. In the present study, we did not monitor patients for thrombus formation by ultrasound; however, no OITC was ever observed to have char formation regardless of ACT level. Ablation requires tissue heating to ⬎50°C to form scar through coagulation necrosis. The CTC requires significant heating of the catheter–tissue interface with only a modest difference of 3°C to 7°C between tissue and catheter tip temperatures.5 For the OITC there is only a few-degree increase in tip temperature associated with the 20°C to 30°C
707
increase in tissue temperature.6 In a dog thigh muscle preparation the OITC never reached 80°C, the catheter– tissue interface temperature associated with thrombus formation.7 The safety of our ablations using the OITC and a target ACT of 225 seconds are consistent with these experimental data. Several technical points may also account for our low incidence of hemorrhage and thromboembolism despite keeping our ACT levels low when using the OITC. We used a single transseptal puncture, which permits heparinization immediately after entering the left atrium. The 2-sheath technique forces the choice between giving heparin before the second puncture, which might increase the risk of pericardial tamponade, or wait until the 2 sheaths are placed, leaving the patient unprotected from thromboembolism between sheath placements. We had no difficulty manipulating the circumferential catheter and ablation catheter when these were across a single puncture site. Although we did not preflush our sheaths with a high heparin concentration, we did infuse our maintenance heparin through the transseptal sheath. This is in contrast to many centers that infuse the transseptal sheath with a standard 2-U/ml concentration of heparin. By infusing our maintenance heparin through the transseptal sheath, we may have eliminated thrombi associated with the transseptal sheaths. Because we frequently position the transseptal sheath in the left atrium to facilitate catheter movement and improve tissue contact, during these times we deliver a high heparin concentration directly into the left atrium. It is important to emphasize that our results cannot be extrapolated to the use of 2 transseptal sheaths, infusion of the maintenance heparin through a peripheral intravenous sheath or venous sheaths other than the transseptal sheath, or to use of closed irrigated tip catheters, which create more thrombus than the OITC in animal models.7 Our overall 2.5% vascular and hemorrhagic complication rate is below those reported for other series using more intense anticoagulation and our lowest rate of complications at 1.62% was in patients with the lowest ACTs. An updated worldwide survey on catheter ablation for AF reported on complications in 16,301 patients.8 Death, tamponade, hemothorax, femoral pseudoaneurysms and arteriovenous fistulas, strokes, and transient ischemic attacks occurred in 3.9% of patients. A recent meta-analysis of outcomes of AF ablation reported these complications in 2.7% of patients.9 Medicare beneficiaries in the United States have rates of 0.4% for mortality, 0.6% for strokes/transient ischemic attack, 3.1% for pericardial tamponade, and 4.8% for vascular complications, for a total complication rate of 8.7%.10 Similar to other reports, we found clinical gender differences. Patel et al11 noted that women accounted for only 16% of their ablations and that women were older, had less AF type 1, had more failed antiarrhythmic drugs, and had AF for a longer duration before ablation. Their complication rate for women was 5.0% versus 2.4% for men (p ⬍0.001). Hematomas occurred in 2.1% of women and 0.9% of men (p ⫽ 0.026) and pseudoaneurysms in 0.6% versus 0.1% (p ⫽ 0.031). Forleo et al12 reported fewer ablations in women who had more hypertension and larger left atria but no difference in type of AF, antiarrhythmic drug history, or complication rates (5.6% vs 4.7%, p ⫽ NS). In our series we
708
The American Journal of Cardiology (www.ajconline.org)
also found that most ablations were performed in men. Although the women were older and had more hypertension than the men, we did not observe more vascular or hemorrhagic complications in women. In our series, women were over-represented in the higher ACT groups consistent with different pharmacokinetics for heparin in women compared to men. Campbell et al13 reported gender differences during heparin administration in patients treated for proximal leg deep venous thrombosis. Men and women were given a standard heparin bolus and infusion and women had higher heparin levels and a higher activated partial thromboplastin time than men. After achieving a therapeutic activated partial thromboplastin time, women received lower heparin doses but still had higher heparin levels. After adjustments for body weight and age, older women had higher heparin levels for any dose of heparin. These investigators concluded that women showed alterations in the pharmacokinetics of heparin, which could explain a predisposition to bleeding complications. Because women undergoing AF ablation are older than men, gender and age differences could account for the predominance of women in the higher ACT groups. Had our target ACT level been ⬎300 or 350 seconds in most patients, we may have seen the increase in complications in women noted in some other series.11 Our data suggest that great care should be given in choosing the heparin bolus and infusion rates in women undergoing AF ablation. Our study has some limitations. It was a retrospective nonrandomized study. Our findings are applicable only to the single transseptal puncture technique with the infusion of maintenance heparin through the transseptal sheath. Because most OITC ablations were done with lower target ACT values, we could not determine if the safety was due to the catheter itself or the lower ACT level. Our findings should stimulate interest in re-evaluating current recommendations for anticoagulation during AF ablation using the OITC. Although a randomized trial would be more definitive than our retrospective analysis, the very low incidence of complications we observed with the OITC using a target ACT of 225 seconds suggests that such a study would require a very large number of patients and that such a study might only prove that higher ACTs with the OITC were equivalent to or worse than low-intensity anticoagulation. Acknowledgment: We acknowledge the assistance of Patricia Barberini, RN, Cynthia Lebsack, PharmD, and Alfie Pierantoni in data management and Glenda Rhodes for manuscript preparation.
1. Calkins H, Brugada J, Packer DL, Capppato R, Chen S, Crijns H, Damiano RJ, Davies DW, Haines DE, Haissaguerre M, Iesaka Y, Jackman W, Jais P, Kottkamp H, Kuck KH, Lindsay BD, Marchlinski FE, McCarthy PM, Mont JL, Morady F, Nademanee K, Natale A, Pappone C, Prystowsdy E, Raviele A, Ruskin JN, Shemin RJ. HRS/ EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for personnel, policy, procedures and follow-up. Heart Rhythm 2007;4:1– 46. 2. Ren JF, Marchlinski FE, Callans DJ, Gerstenfeld EP, Dixit S, Lin D, Nayak HM, Hsia HH. Increased intensity of anticoagulation may reduce risk of thrombus during atrial fibrillation ablation procedures in patients with spontaneous echo contrast. J Cardiovasc Electrophysiol 2005;16:474 – 477. 3. Maleki K, Mohammadi R, Hart D, Cotiga D, Farhat N, Steinberg JS. Intracardiac ultrasound detection of thrombus on transseptal sheath: incidence, treatment, and prevention. J Cardiovasc Electrophysiol 2005;16:561–565. 4. Wazni OM, Rossillo A, Marrouche NF, Saad EB, Martin DO, Bhargava M, Bash D, Beheiry S, Wexman M, Potenza D, Pisano E, Fanelli R, Bonso A, Themistoclakis S, Erciyes D, Saliba WI, Schweikert RA, Brachmann J, Raviele A, Natale A. Embolic events and char formation during pulmonary vein isolation in patients with atrial fibrillation: impact of different anticoagulation regimens and importance of intracardiac echo imaging. J Cardiovasc Electrophysiol 2005;16:576 –581. 5. Bunch TJ, Bruce GK, Johnson SB, Sarabanda A, Milton MA, Packer DL. Analysis of catheter-tip (8-mm) and actual tissue temperatures achieved during radiofrequency ablation at the orifice of the pulmonary vein. Circulation 2004;110:2988 –2995. 6. Bruce GK, Bunch TJ, Milton MA, Sarabanda A, Johnson SB, Packer DL. Discrepancies between catheter tip and tissue temperature in cooled-tip ablation. Relevance to guiding left atrial ablation. Circulation 2005;112:954 –960. 7. Yokoyama K, Nakagawa H, Wittkampf FHM, Pitha JV, Lazzara R, Jackman WM. Comparison of electrode cooling between internal and open irrigation in radiofrequency ablation lesion depth and incidence of thrombus and steam pop. Circulation 2006;113:11–19. 8. Cappato R, Calkins H, Chen SA, Chen S, Davies W, Iesaka Y, Kalman J, Kim Y, Klein G, Natale A, Packer D, Skanes A, Ambrogi F, Biganzoli E. Updated worldwide survey on the methods, efficacy and safety of catheter ablation for human atrial fibrillation. Circ Arrhythmia Electrophysiol 2010;3:32–38. 9. Calkins H, Reynolds MR, Spector P, Sondhi M, Xu Y, Martin A, Williams CJ, Sledge I. Treatment of atrial fibrillation with antiarrhythmic drugs or radiofrequency ablation: two systematic literature reviews and meta-analyses. Circ Arrhythmia Electrophysiol 2009;2: 349 –361. 10. Ellis ER, Culler SD, Simon AW, Reynolds MR. Trends in utilization and complications of catheter ablation for atrial fibrillation in Medicare beneficiaries. Heart Rhythm 2009;6:1267–1273. 11. Patel D, Mohanty P, Di Biase L, Sanchez JE, Shaheen MH, Burkhardt JD, Bassouni M, Cummings J, Wang Y, Lewis WR, Diaz A, Horton RP, Beheiry S, Hongo R, Gallinghouse GJ, Zagrodzky JD, Bailey SM, Al-Ahmad A, Wang P, Schweikert RA, Natale A. Outcomes and complications of catheter ablation for atrial fibrillation in females. Heart Rhythm 2010;7:167–172. 12. Forleo GB, Tondo C, De Luca L, Dello Russo A, Casella M, De Sanctis V, Clementi F, Fagundes RL, Leo R, Romeo F, Mantica M. Gender-related differences in catheter ablation of atrial fibrillation. Europace 2007;9:613– 620. 13. Campbell NRC, Hull RD, Brant R, Hogan DB, Pineo GF, Raskob GE. Different effects of heparin in males and females. Clin Invest Med 1998;21:71–79.
Sleep-Disordered Breathing in Patients With the Brugada Syndrome Paula G. Macedo, MDa, Josep Brugada, MDb, Pavel Leinveber, MSca,d, Begoña Benito, MDb, Irma Molina, MDb, Fatima Sert-Kuniyoshi, PhDa, Taro Adachi, MD, PhDa, Jan Bukartyk, MSca, Christelle van der Walt, RPSGTa, Tomas Konecny, MDa, Shantal Maharaja, Tomas Kara, MD, PhDa,d, Josep Montserrat, MDc, and Virend Somers, MD, PhDa,* We investigated breathing patterns and the occurrence of arrhythmias and ST-segment changes during sleep in patients with Brugada syndrome. Patients with Brugada syndrome are more likely to die from ventricular arrhythmias during sleep. ST-segment changes have been correlated with risk of sudden cardiac death. Whether sleep disturbances may contribute to arrhythmogenesis is unknown. Patients with Brugada syndrome underwent overnight polysomnography with simultaneous 12-lead electrocardiographic recording. A control group matched by age, gender, and body mass index (BMI) also underwent polysomnography. Twenty patients were included (50 ⴞ 15 years old, 75% men). Despite their normal BMI (24.7 ⴞ 2.7 kg/m2), 45% had sleep-disordered breathing (SDB), with a mean apnea-hypopnea index of 17.2 ⴞ 14 events/hour. In patients with a high risk of arrhythmias, 5 (63%) had SDB. In the control group, 27% had SDB. Atrial or ventricular arrhythmias were not observed. Spontaneous ST-segment changes occurred in 2 patients over 45 different time points. Most ST-segment changes were observed during rapid eye movement sleep (31%) or within 1 minute of arousals (44%). Regarding respiratory events, 25 (56%) of ST-segment changes were related to occurrence of apnea or hypopnea. In conclusion, patients with Brugada syndrome have a high prevalence of SDB even in the setting of normal BMI. The higher incidence of nocturnal death in patients with Brugada syndrome may be conceivably related to co-morbid SDB. Moreover, autonomic instability encountered in rapid eye movement sleep and arousals could potentiate the risk of arrhythmias. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:709 –713) Brugada syndrome was described in 1992 in patients presenting with recurrent aborted sudden cardiac death (SCD) and typical electrocardiographic sign of ST-segment elevation in leads V1 to V3.1 It was later shown that patients die from ventricular tachycardia or fibrillation often during sleep.2,3 The arrhythmogenic role of sleep is being increasingly recognized. Nocturnal occurrence of apnea and hypopneas has been related to SCD and atrial fibrillation.4 Apneic events elicit significant autonomic responses, which have been linked to ST-segment elevation in patients with Brugada syndrome.5,6 In addition to respiratory events, changes from nonrapid eye movement (non-REM) sleep to REM sleep are associated with autonomic instability, which a Division of Cardiovascular Diseases, Department of Internal Medicine, Mayo Clinic and Foundation, Rochester, Minnesota; bCardiology Department, cPneumology Department, The Thorax Institute, Hospital Clinic, University of Barcelona, Barcelona, Spain, and dICRC-Department of Cardiovascular Diseases, St. Anne’s University Hospital Brno, Brno, Czech Republic. Manuscript received July 29, 2010; revised manuscript received and accepted October 19, 2010. Dr. Somers is supported by Grants HL65176 and 1 UL1 RR 024150 from the National Institutes of Health, Bethesda, Maryland. Mr. Leinveber, Mr. Bukartyk, Dr. Konecny, and Dr. Kara were partially supported from Grants NS10099-3/20008, NS10098-3/2008, and NT11401-5/2011 from the Czech Ministry of Health. *Corresponding author: Tel: 507-255-1144; fax: 507-255-7070. E-mail address:
[email protected] (V.K. Somers).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.046
can play a role in arrhythmogenesis.7,8 Therefore, this study evaluated breathing patterns during sleep in patients with Brugada syndrome and investigated the occurrence of arrhythmias and ST-segment changes during different sleep stages. Methods Consecutive patients being evaluated at the Hospital Clinic, University of Barcelona, were invited to participate in the present study. Patients were recruited from sudden death and implantable cardioverter– defibrillator outpatient clinics, and about 10% of those invited declined participation. Inclusion criteria were age ⬎21 years and a definite diagnosis of Brugada syndrome based on spontaneous or drug-induced type 1 Brugada electrocardiogram (ECG). Patients underwent overnight polysomnography with simultaneous recording of 12-lead ECG throughout the night. The control group (recruited in Rochester, Minnesota) consisted of healthy subjects of similar age, gender, and body mass index (BMI) as patients. This study was approved by the institutional review boards of the University of Barcelona and the Mayo Clinic (Rochester, Minnesota) and all patients signed an informed consent. Patients were classified as high risk for fatal arrhythmias according to the following criteria: (1) previous aborted SCD associated with type 1 ECG (spontaneous or induced) or (2) unexplained syncope with spontaneous type 1 ECG. www.ajconline.org
710
The American Journal of Cardiology (www.ajconline.org)
Table 1 Clinical and polysomnographic characteristics in patients with Brugada syndrome according to presence or absence of sleep-disordered breathing (listed by ascending apnea– hypopnea index) Patient Number
Patients without sleep-disordered breathing 1 2 3 4 5 6 7 8 9 10 11 Patients with sleepdisordered breathing 12 13 14 15 16 17 18 19 20
Age (years)/ Sex
BMI (kg/m2)
Familial History of SCD
VF/SCD
Syncope
High Arrhythmic Risk
HTN
DM
Sleep Efficiency (%)
REM (%)
Lowest SpO2 (%)
AHI (events/ hour)
32/F 66/F 31/M 31/M 60/M 29/M 22/M 50/M 56/M 39/M 43/M
24 21 26 23 27 24 25 23 28 24 24
⫹ 0 0 ⫹ 0 0 0 0 ⫹ 0 0
0 0 ⫹ 0 ⫹ 0 0 0 0 0 0
0 ⫹ ⫹ 0 ⫹ 0 ⫹ 0 0 0 0
0 0 ⫹ 0 ⫹ 0 ⫹ 0 0 0 0
0 ⫹ 0 0 0 0 0 0 ⫹ 0 0
0 0 0 0 0 0 0 0 0 0 0
93 89 90 75 78 80 87 86 69 81 75
19 15 16 12 28 24 19 28 9 21 18
97 94 86 93 90 93 94 87 89 91 86
0 0 0.1 0.4 0.5 0.6 0.8 1.5 1.5 1.7 4.4
65/F 71/F 56/M 62/F 64/M 52/M 51/M 58/M 61/M
26 22 28 21 26 22 24 23 28
0 0 ⫹ ⫹ 0 ⫹ 0 0 ⫹
⫹ 0 0 0 0 0 0 0 0
⫹ ⫹ 0 0 ⫹ 0 ⫹ ⫹ 0
⫹ ⫹ 0 0 ⫹ 0 ⫹ ⫹ 0
⫹ ⫹ 0 0 0 0 0 0 ⫹
0 0 0 0 0 0 0 0 ⫹
72 57 94 91 61 79 69 62 86
10 7 8 14 11 11 13 4 29
85 87 84 79 89 78 84 78 72
5.3 6.5 10.4 10.7 10.7 14.5 16.5 32.4 48.1
0 ⫽ absent; ⫹ ⫽ present; AHI ⫽ apnea– hypopnea index; DM ⫽ diabetes mellitus; HTN ⫽ arterial hypertension; SpO2 ⫽ arterial oxygen saturation; VF ⫽ ventricular fibrillation.
All subjects underwent full-night polysomnography using a Compumedics Siesta802 wireless amplifier/recorder (Compumedics, Abbotsford, Victoria, Australia). Airflow was monitored by an oronasal thermal airflow sensor and respiratory effort was monitored by calibrated respiratory impedance plethysmography. During all polysomnographic procedures, electroencephalogram, electro-oculogram, and submental electromyogram were recorded according to American Academy of Sleep Medicine standards. Oxyhemoglobin saturation was recorded by finger pulse oximetry. Polysomnograms were scored by an experienced polysomnographic technologist who was blind to subjects’ status and arrhythmic risk. Apneas were defined as a ⱖ90% decrease in peak sensor excursion from baseline for ⱖ10 seconds. Hypopneas were defined by a ⱖ50% decrease in sensor excursion for ⱖ10 seconds accompanied by an oxyhemoglobin desaturation of ⱖ4%. Disordered breathing events were quantified by the apnea– hypopnea index. An apnea– hypopnea index ⱖ5 events/hour established the diagnosis of sleep-disordered breathing (SDB). A 12-lead ECG was simultaneously recorded in LabVIEW (National Instruments Corporation, Austin, Texas) and processed using ScopeWin software (Institute for Scientific Instruments, Brno, Czech Republic). The ECG was analyzed by a cardiac electrophysiologist blinded to patients’ arrhythmic risk and to sleep stages. A spontaneous change in the ST segment was considered if there was modification in the type of Bru-
gada ECG or a visible change in the T wave in the right precordial leads according to a previously published method.9 Time points where these changes began were correlated to sleep stages and respiratory events on polysomnographic tracings. Data are summarized as frequencies for categorical variables and means ⫾ SDs for continuous variables. Group differences were evaluated by Wilcoxon rank-sum test. Differences in proportions were tested by Fisher’s exact test. Analyses were performed with JMP 7 (SAS Institute, Cary, North Carolina). For all comparisons a p value ⬍0.05 for a 2-tailed test was considered statistically significant. Results Twenty patients were included in the study. Mean age was 50 ⫾ 15 years, BMI was 24.7 ⫾ 2.7 kg/m2, and most patients were men (75%). Six patients (30%) had unexplained syncope and 3 (15%) had aborted SCD. A family history of SCD was reported by 7 patients (35%; Table 1). Twelve patients (60%) had an implantable cardioverter– defibrillator, 2 of whom were also on medication, 1 on quinidine and the other on a  blocker. The control group consisted of 11 subjects with mean age 43 ⫾ 16 years and mean BMI 25 ⫾ 3 kg/m2 and 82% were men. Nine patients (45%) had SDB on polysomnogram with a mean apnea– hypopnea index of 17.2 ⫾ 14 events/hour. Six
Arrhythmias and Conduction Disturbances/Sleep-Disordered Breathing in Brugada Syndrome
711
Figure 1. Twelve-lead electrocardiograms show progressive augmentation of ST-segment elevation in leads V1 and V2 and changes in ST-segment pattern in lead V3 on type 3 Brugada electrocardiogram (A) to type 2 (B) and type 1 (C). These tracings were recorded during the same rapid eye movement period within a 13-second time frame in 1 patient with Brugada syndrome.
patients had a diagnosis of obstructive sleep apnea, 2 with central sleep apnea and 1 patient had mixed sleep apnea. In the control group, 3 (27%) had SDB and all were diagnosed as having obstructive sleep apnea. Regarding clinical characteristics, groups with or without SDB were very similar except that patients with SDB were older (mean age 60 ⫾ 6 vs 42 ⫾ 14 years, p ⫽ 0.008). There was no difference in mean BMI between the SDB and non-SDB groups (25 vs 24.4 kg/m2, p ⫽ 0.85). Clinically defined high-risk patients were more frequently observed in the SDB group (56%) than in the non-SDB group (27%). During sleep, patients with SDB spent less time in REM (12% vs 19%, p ⫽ 0.02), had a lower arterial oxygen saturation nadir (82% vs 91%, p ⬍0.001) and a higher apnea– hypopnea index (17.2 vs 1.0 events/hour, p ⬍0.001; Table 1). Eight patients (40%, 75% men, mean age 53 ⫾ 17 years) were classified as having a high risk for fatal arrhythmias as described earlier. Despite their relatively normal BMI (25 ⫾ 1.6 kg/m2), 5 patients (63%) had sleep apnea in this subgroup with a mean apnea– hypopnea index of 14.3 ⫾ 11 events/hour. There was no significant difference in mean age or BMI between patients with or without SDB (62 vs 38 years old, p ⫽ 0.10; 24.2 vs 25.9 kg/m2, p ⫽ 0.29, respectively).
Atrial or ventricular arrhythmias were not observed in this study population. Analysis of bradyarrhythmias was limited because 60% of patients had an implantable cardioverter– defibrillator with backup pace setting. In 2 patients without an implantable cardioverter– defibrillator, 6 episodes of significant bradycardia were observed: 5 sinus pauses of a maximal duration of 3.1 seconds and 1 episode of Mobitz I second-degree atrioventricular block (pause 2.5 seconds). None of them had SDB or high arrhythmic risk. Spontaneous ST-segment changes were found in 2 patients (10%). These 2 patients were previously asymptomatic, had low arrhythmic risk, and were not on medications. However, these 2 patients were in their third decade of life and had a history of SCD in a first-degree relative. One had an apnea– hypopnea index of 14.5 events/hour and the other an index 1.5 events/hour. Changes observed in the first patient were spontaneous augmentation of ST-segment elevation in leads V1 and V2 and a shift in lead V3 pattern from type 3 to type 2 or 1 (Figure 1). In the second patient, there was a visible change in T-wave structure and amplitude in lead V1. ST-segment changes occurred over 45 different time points. Most ST-segment changes were observed dur-
712
The American Journal of Cardiology (www.ajconline.org)
Figure 2. Distribution of ST-segment changes in 2 patients according to changes in sleep stages. 1 ⫽ stage 1; 2 ⫽ stage 2; 3/4 ⫽ stage 3 and 4; ARL ⫽ arousals; R ⫽ REM; W ⫽ wakefulness.
ing REM sleep (31%) or within 1 minute of arousals (44%). There were no changes in ST segments during sleep stage 3 or 4 (Figure 2). With regard to respiratory events, 25 (56%) ST-segment changes occurred during or within 1 minute after an episode of apnea or hypopnea. Discussion In the present study we found that SDB is relatively common in patients with Brugada syndrome, especially in patients at high risk for fatal arrhythmic events. Our data also suggest that spontaneous changes in the ST segment, a risk factor for SCD, may be predominantly present during REM sleep or after arousals. Recently, a very large prospective study analyzed predictive risk factors for SCD in patients with Brugada syndrome, but presence of SDB was not included in the analysis.10 To the best of our knowledge, this is the first study to analyze polysomnographic and electrocardiographic characteristics in this patient population. The reported prevalence of SDB in the general population is around 25%.11,12 Similarly, in the present study, 27% of control subjects had a diagnosis of SDB. However, we found that SDB is present in 45% of patients with Brugada syndrome and in 63% of patients with Brugada syndrome who fulfill clinical criteria of high arrhythmic risk. These values are surprisingly high, especially considering that these patients had a normal BMI. Peppard et al13 found a mean apnea– hypopnea index of 1.2 events/hour in a lean population, whereas we observed a mean apnea– hypopnea index of 8 events/hour. Brugada syndrome is a rare disorder and is associated with SCD in seemingly healthy patients. Approximately 10% of patients die annually, usually in their fourth decade of life.14 Matsuo et al3 showed that patients with Brugada syndrome are more likely to die while sleeping. They observed that 87% of ventricular fibrillation episodes occur between midnight and 6 A.M. Another condition previously called “sudden unexplained nocturnal death syndrome,” more common in southeastern Asian countries, is currently recognized as probably the same disorder as Brugada syndrome.15 Sleep apnea is also associated with SCD.7,16 Patients with SDB are also more likely to die during sleep, and it is
believed that oscillations in autonomic nervous activation and cardiopulmonary dynamics may play a key role. Obstructive apneas are associated with surges in arterial blood pressure, vascular sympathetic nerve activation, and increased cardiac vagal drive.17,18 Furthermore, in patients with obstructive sleep apnea, hypoxemia is associated with an increased incidence of complex ventricular ectopy, tachycardia, and surges in blood pressure.4,19 In contrast, bradyarrhythmias with long pauses may occur due to vagal activation triggered by the diving reflex.18,20,21 Low oxygen and increased myocardial demand may facilitate the occurrence of myocardial ischemia and related arrhythmias. Indeed, nocturnal myocardial infarction occurs more frequently in patients with obstructive sleep apnea.22 Presence of spontaneous coved-type ST-segment elevation (type 1 electrocardiographic pattern) has been associated with a higher incidence of appropriate implantable cardioverter– defibrillator shocks.23 Moreover, repeated ECGs in patients with Brugada syndrome show very frequent spontaneous fluctuations between diagnostic and nondiagnostic electrocardiographic patterns, which have been linked to ventricular arrhythmias.9,23,24 A prospective study including 124 patients found that spontaneous changes in the ST segment were associated with the highest risk for ventricular tachyarrhythmias and SCD with a relative hazard ratio of 9.2.9 It has been suggested that acute modifications in transmembrane ionic currents participating in early phases of repolarization may underlie ST-segment changes in patients with Brugada syndrome.25 Oscillations in the autonomic system, through direct effects on ionic currents, seem to play an important role in ST-segment fluctuations.5,25 Therefore, we analyzed ST-segment characteristics during polysomnographic monitoring of sleep stages, when significant changes in autonomic status are known to occur.26 We observed changes in the ST segment in 2 patients, most frequently during REM sleep and after arousals, when 34 (75.5%) of the 45 recorded events occurred. Although less significant, those changes were also more common during apneic/hypopneic events (56%). No changes were observed during sleep stage 3 or 4, and only 5 (11%) were detected in stages 1 and 2. In normal subjects, REM has been associated with surges in sympathetic activity and abrupt vagal discharges, explaining the higher incidence of tachycardias and bradyarrhythmias in this period.8,26,27 Arousals are also accompanied by transient bursts of sympathetic discharges.28,29 It has been hypothesized that the vivid and emotional dreams experienced during REM sleep may also contribute to nocturnal arrhythmogenesis.7 Therefore, autonomic instability during REM sleep and after arousals could play an important role in triggering arrhythmias. This may be especially true in patients with Brugada syndrome in whom an intrinsic autonomic impairment has been described.30 –32 Studies using iodine-123 MIBG single-photon emission computed tomography have shown decreased tracer uptake in the left ventricle of patients with Brugada syndrome, which is compatible with a dysfunction in myocardial sympathetic innervation. Those patients may also have cardiac autonomic neuropathy based on abnormalities evident in tests of cardiovascular sympathetic function.30 Accordingly, dynamic changes in the ST segment were more frequent during REM sleep and after arousals.
Arrhythmias and Conduction Disturbances/Sleep-Disordered Breathing in Brugada Syndrome
Spontaneous ST-segment changes may provide a marker of increased arrhythmic risk for nocturnal SCD in these patients. Strengths of our study were the use of overnight polysomnography in all patients, which is a gold standard for diagnosis of SDB, use of 12-lead electrocardiographic monitoring throughout polysomnography, and blinded analysis of the data. However, this was a cross-sectional study with a relatively small sample, and therefore the high frequency of SDB encountered in these patients with Brugada syndrome should be interpreted with caution. Nevertheless, it is notable that SDB was evident in a large proportion of patients despite having a normal BMI. Although these data do not support a causal relation between SDB and STsegment changes, they provide a compelling basis for further investigation. Prospective controlled studies are needed to address the role of SDB in triggering ventricular arrhythmias in patients with Brugada syndrome. 1. Brugada P, Brugada J. Right bundle branch block, persistent ST segment elevation and sudden cardiac death: a distinct clinical and electrocardiographic syndrome. A multicenter report. J Am Coll Cardiol 1992;20:1391–1396. 2. Brugada J, Brugada P, Brugada R. The syndrome of right bundle branch block ST segment elevation in V1 to V3 and sudden death—the Brugada syndrome. Europace 1999;1:156 –166. 3. Matsuo K, Kurita T, Inagaki M, Kakishita M, Aihara N, Shimizu W, Taguchi A, Suyama K, Kamakura S, Shimomura K. The circadian pattern of the development of ventricular fibrillation in patients with Brugada syndrome. Eur Heart J 1999;20:465– 470. 4. Gami AS, Somers VK. Implications of obstructive sleep apnea for atrial fibrillation and sudden cardiac death. J Cardiovasc Electrophysiol 2008;19:997–1003. 5. Mizumaki K, Fujiki A, Tsuneda T, Sakabe M, Nishida K, Sugao M, Inoue H. Vagal activity modulates spontaneous augmentation of ST elevation in the daily life of patients with Brugada syndrome. J Cardiovasc Electrophysiol 2004;15:667– 673. 6. Noda T, Shimizu W, Taguchi A, Satomi K, Suyama K, Kurita T, Aihara N, Kamakura S. ST-segment elevation and ventricular fibrillation without coronary spasm by intracoronary injection of acetylcholine and/or ergonovine maleate in patients with Brugada syndrome. J Am Coll Cardiol 2002;40:1841–1847. 7. Verrier RL, Josephson ME. Impact of sleep on arrhythmogenesis. Circ Arrhythmia Electrophysiol 2009;2:450 – 459. 8. Guilleminault C, Pool P, Motta J, Gillis AM. Sinus arrest during REM sleep in young adults. N Engl J Med 1984;311:1006 –1010. 9. Ikeda T, Takami M, Sugi K, Mizusawa Y, Sakurada H, Yoshino H. Noninvasive risk stratification of subjects with a Brugada-type electrocardiogram and no history of cardiac arrest. Ann Noninvasive Electrocardiol 2005;10:396 – 403. 10. Probst V, Veltmann C, Eckardt L, Meregalli PG, Gaita F, Tan HL, Babuty D, Sacher F, Giustetto C, Schulze-Bahr E, Borggrefe M, Haissaguerre M, Mabo P, Le Marec H, Wolpert C, Wilde AA. Longterm prognosis of patients diagnosed with Brugada syndrome: results from the FINGER Brugada Syndrome Registry. Circulation 2010;121: 635– 643. 11. Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med 1993;328:1230 –1235. 12. Duran J, Esnaola S, Rubio R, Iztueta A. Obstructive sleep apneahypopnea and related clinical features in a population-based sample of subjects aged 30 to 70 yr. Am J Respir Crit Care Med 2001;163:685– 689.
713
13. Peppard PE, Young T, Palta M, Skatrud J. Prospective study of the association between sleep-disordered breathing and hypertension. N Engl J Med 2000;342:1378 –1384. 14. Naccarelli GV, Antzelevitch C. The Brugada syndrome: clinical, genetic, cellular, and molecular abnormalities. Am J Med 2001;110:573– 581. 15. Fowler SJ, Priori SG. Clinical spectrum of patients with a Brugada ECG. Curr Opin Cardiol 2009;24:74 – 81. 16. Gami AS, Howard DE, Olson EJ, Somers VK. Day-night pattern of sudden death in obstructive sleep apnea. N Engl J Med 2005;352: 1206 –1214. 17. Somers VK, Dyken ME, Clary MP, Abboud FM. Sympathetic neural mechanisms in obstructive sleep apnea. J Clin Invest 1995;96:1897– 1904. 18. Somers VK, Dyken ME, Mark AL, Abboud FM. Parasympathetic hyperresponsiveness and bradyarrhythmias during apnoea in hypertension. Clin Auton Res 1992;2:171–176. 19. Mehra R, Stone KL, Varosy PD, Hoffman AR, Marcus GM, Blackwell T, Ibrahim OA, Salem R, Redline S. Nocturnal arrhythmias across a spectrum of obstructive and central sleep-disordered breathing in older men: outcomes of sleep disorders in older men (MrOS sleep) study. Arch Intern Med 2009;169:1147–1155. 20. Grimm W, Hoffmann J, Menz V, Kohler U, Heitmann J, Peter JH, Maisch B. Electrophysiologic evaluation of sinus node function and atrioventricular conduction in patients with prolonged ventricular asystole during obstructive sleep apnea. Am J Cardiol 1996;77:1310 – 1314. 21. Caples SM, Gami AS, Somers VK. Obstructive sleep apnea. Ann Intern Med 2005;142:187–197. 22. Kuniyoshi FH, Garcia-Touchard A, Gami AS, Romero-Corral A, van der Walt C, Pusalavidyasagar S, Kara T, Caples SM, Pressman GS, Vasquez EC, Lopez-Jimenez F, Somers VK. Day-night variation of acute myocardial infarction in obstructive sleep apnea. J Am Coll Cardiol 2008;52:343–346. 23. Richter S, Sarkozy A, Veltmann C, Chierchia GB, Boussy T, Wolpert C, Schimpf R, Brugada J, Brugada R, Borggrefe M, Brugada P. Variability of the diagnostic ECG pattern in an ICD patient population with Brugada syndrome. J Cardiovasc Electrophysiol 2009;20:69 –75. 24. Veltmann C, Schimpf R, Echternach C, Eckardt L, Kuschyk J, Streitner F, Spehl S, Borggrefe M, Wolpert C. A prospective study on spontaneous fluctuations between diagnostic and non-diagnostic ECGs in Brugada syndrome: implications for correct phenotyping and risk stratification. Eur Heart J 2006;27:2544 –2552. 25. Antzelevitch C. Brugada syndrome. Pacing Clin Electrophysiol 2006; 29:1130 –1159. 26. Somers VK, Dyken ME, Mark AL, Abboud FM. Sympathetic-nerve activity during sleep in normal subjects. N Engl J Med 1993;328:303– 307. 27. Verrier RL, Lau TR, Wallooppillai U, Quattrochi J, Nearing BD, Moreno R, Hobson JA. Primary vagally mediated decelerations in heart rate during tonic rapid eye movement sleep in cats. Am J Physiol Regul Integr Comp Physiol 1998;274:R1136 –R1141. 28. Morgan BJ, Crabtree DC, Puleo DS, Badr MS, Toiber F, Skatrud JB. Neurocirculatory consequences of abrupt change in sleep state in humans. J Appl Physiol 1996;80:1627–1636. 29. Blasi A, Jo J, Valladares E, Morgan BJ, Skatrud JB, Khoo MC. Cardiovascular variability after arousal from sleep: time-varying spectral analysis. J Appl Physiol 2003;95:1394 –1404. 30. Bigi MA, Aslani A, Aslani A. Significance of cardiac autonomic neuropathy in risk stratification of Brugada syndrome. Europace 2008; 10:821– 824. 31. Nomura M, Nada T, Endo J, Kondo Y, Yukinaka M, Saito K, Ito S, Mori H, Nakaya Y, Shinomiya H. Brugada syndrome associated with an autonomic disorder. Heart 1998;80:194 –196. 32. Wichter T, Matheja P, Eckardt L, Kies P, Schafers K, Schulze-Bahr E, Haverkamp W, Borggrefe M, Schober O, Breithardt G, Schafers M. Cardiac autonomic dysfunction in Brugada syndrome. Circulation 2002;105:702–706.
Frequency of Cardiac Events at Four Years Among Initially Asymptomatic Filipinos with the Brugada Type 1 Electrocardiographic Pattern Giselle Gervacio Domingo, MD*, Gabriel Jocson, MD, and Antonio Dans, MD Brugada type 1 electrocardiographic (ECG) pattern occurs in 0.2% of Filipinos. A knowledge gap exists on the natural course of asymptomatic patients with Brugada type 1 ECG pattern. Most studies that reported cohort event rates were taken from hospitals or referral centers. This is the first cohort from an entire country where the subjects were selected randomly. The objective of this study was to describe the frequency of cardiac events at 4 and 6 years of 7 patients with Brugada type 1 ECG pattern of 3,907 patients previously screened from the general population of the Philippines during the National Nutrition and Health Survey. Personal interviews at year 4 using a structured questionnaire were conducted by 1 of the investigators. Occurrences of major (syncope, seizure, unexplained accidents, sudden death) and minor events in subjects and their first- and second-degree relatives were elicited. Six-year follow-up by text messaging was conducted to ascertain vital status and occurrence of cardiac events. All 7 patients with Brugada type 1 ECG pattern were men. Three of the 7 initially asymptomatic subjects (43%, 95 confidence interval 6 to 80) developed a major cardiac event by the fourth year. Those with events were younger than those without events. All 7 were alive by the sixth year. No additional events were noted between the fourth and sixth years. In conclusion, cardiac events are considerable in initially asymptomatic Filipinos with Brugada type 1 ECG pattern. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:714 –716) Brugada type 1 electrocardiographic (ECG) pattern occurs frequently in Filipinos. During the 2003 Philippine National Nutrition and Health Survey (NNHES), 7 of 3,907 subjects (0.2%) from the general population selected in a stratified randomized sampling based on age and geographic region were found to have Brugada type 1 ECG pattern. All subjects were asymptomatic on initial diagnosis.1 The prognostic significance of this Brugada type 1 ECG sign in initially asymptomatic patients is unknown.2 In symptomatic patients (syncope, seizure, or sudden cardiac death) this ECG marker identifies a high risk of symptom recurrence or sudden cardiac death, but its significance in those who are asymptomatic is uncertain.3,4 Conflicting data exist on event rates in asymptomatic patients with Brugada ECG pattern. Priori et al5 reported no events in this group, whereas Brugada et al6 reported a 27% 2-year event rate. This study describes the frequency of cardiac events at 4 and 6 years in the 7 initially asymptomatic patients identified during the 2003 NNHES. Major (sudden death, syncope, seizure, presyncope) and minor (e.g., moaning during sleep but patients easily aroused) events are described.
University of the Philippines, Philippine General Hospital, Manila, Philippines. Manuscript received September 26, 2010; revised manuscript received and accepted October 26, 2010. This study was funded in full by the Philippine Heart Association, Pasig City, Philippines. *Corresponding author: Tel: 632-723-2092; fax: 632-723-2092. E-mail address:
[email protected] (G.G. Domingo). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.047
Methods The 7 patients with Brugada type 1 ECG pattern (2-mm J-point elevation and ⱖ1-mm convex ST-segment elevation in leads V1 and V2) randomly identified from the 3,907 screened from the general population during the 2003 NNHES were initially contacted by a third party to obtain preliminary consent. This was done to protect patients’ privacy. After receipt of this preliminary consent by the third party, formal informed consent was obtained by the investigators. A visit was done 4 years (2007) after the initial ECG identification. During the visit a validated questionnaire focusing on identifying major and minor cardiac events, identifying family history of similar events, and excluding alternative causes for sudden death was administered to the individual by 1 of the investigators (G.J.). Additional questions not included in the questionnaire were allowed for clarification of points. A 12-lead ECG recording was then obtained. A follow-up on the sixth year was performed by telephone interview and/or text messaging to ascertain whether any more cardiac events had occurred. No statistical analysis was performed. Event rate was reported as mean with its 95% confidence interval (CI). Results There were 7 patients with Brugada type 1 ECG pattern of 3,907 chosen randomly from the general population in the Philippines during the 2003 NNHES. All were men with mean age of 50 years (37 to 62). Three subjects were ⬍50 www.ajconline.org
Arrhythmias and Conduction Disturbances/Cardiac Events in Asymptomatic Brugada
715
Table 1 Summary of events and clinical characteristics of seven identified subjects with Brugada syndrome Age (years)*/Sex 49/M 32/M 37/M 54/M 55/M 61/M 62/M
Place of Residence
Major Event†
Minor Event‡
First-Degree Relative With Sudden Cardiac Death
Second-Degree Relative With Sudden Cardiac Death
Sarangani Samal, Davao La Trinidad, Benguet Binangonan, Rizal Bulacan Isabela Bolinao, Pangasinan
2 4 2 0 0 0 0
0 0 1 0 0 0 0
0 0 Father at 71 years of age None Brother at 38 years of age Brother at 51 years of age 0
0 0 0 0 2 nephews (30, 42 years of age) 2 nephews (32, 40 years of age) 0
* Age at time of diagnosis. Unexplained death, successful resuscitation, near death requiring hospitalization, unexplained accidents requiring confinement, syncope or near syncope that can be attributed to nonsustained ventricular tachycardia, or witnessed nocturnal agonal respiration. ‡ Being awakened from sleep spontaneously or by vigorous stimulation, with symptoms of severe distress or impending doom. †
years of age. All were asymptomatic at the time of ECG identification, reporting no previous episodes of syncope, presyncope, seizure, unexplained accidents, palpitations, chest pains or shortness of breath before ECG identification in 2003. Five of these 7 were from Luzon, the largest island in the Philippines, and 2 were from Mindanao, the southern tip of the archipelago. All 7 were alive at follow-up visits 4 years (2007) and 6 years (2009) after initial ECG diagnosis. Repeat 12-lead electrocardiograms in the fourth year still showed Brugada type 1 ECG sign in all subjects. Of the 7 subjects, 3 had a major cardiac event (syncope, seizure, prolonged nonarousability), representing possible spontaneously terminating ventricular arrhythmia. This represents an event rate of 43% (95% CI 6 to 80). The 3 who reported major cardiac events were in their third or fourth decade of life. This is in contrast to the other 4 patients with no reported events who were in their fifth and sixth decades of life. The 3 who reported events had excellent functional capacity, denied angina, and had no known cardiac condition. Only 1 had a known cardiac risk factor (hypertension). These 3 subjects had brief spontaneously terminating major events (syncope and near syncope) during daytime activity (farming, sweeping, 1 while seated); 2 of these 3 also had nocturnal events described as labored breathing, profuse sweating, moaning, and prolonged nonarousability during sleep. The 2 were eventually aroused; however, 1 was brought to the emergency room due to difficulty in arousing him. Upon reaching the emergency room he was fully conscious, although sweating profusely (Table 1). Only 1 of the 3 subjects with events had a first-degree relative (father) who had sudden cardiac death. However, the father was 71 years of at time of death. The remaining 4 of the 7 subjects did not report any major or minor event at 4 and 6 years. These 4 subjects were in the fifth and sixth decades of life. Only 1 had hypertension. However, 2 had first- and second-degree relatives who died suddenly. Most of these relatives’ deaths occurred in young men in their 30s or 40s. Six-year follow-up by telephone interview and text messaging showed that none of the 7 patients had died and none had occurrence or recurrence of cardiac events. There were no new events reported for the first- and second-degree relatives of the 7 subjects.
Discussion Sudden unexplained death syndrome occurs in 0.043% of the general population in the Philippines.7 Brugada type 1 ECG sign occurs in 0.2% of the Philippine general population, although these patients were asymptomatic at initial ECG identification.1 Symptomatic patients clearly benefit from an implantable cardioverter– defibrillator but a knowledge gap exists in the prognosis and treatment of asymptomatic patients with Brugada type 1 ECG pattern.8 There are limited data sources and conflicting evidence.5 Priori et al5 reported no cardiac events at 2 years, whereas Brugada et al6 reported a 27% event rate in asymptomatic patients. This figure decreased to 8% and then to 5%, a phenomenon that Viskin and Rogowski9 called the “founder’s effect.” The apparent discrepancy was explained by the observation that many subjects in the study by Brugada et al were relatives of patients with Brugada syndrome and may have had a higher baseline risk. Our data are the first derived from the general population on a nationwide scale, where subjects were selected in random fashion and recruitment was not driven by symptoms or family history. Hence, selection bias was not a factor. Data from previous studies were culled from city-based or hospital-based health examinations in Japan. In a study by Miyasaka et al10 covering annual physical examinations of Osaka residents, the prevalence of asymptomatic Brugada syndrome was 0.12% and the event rate was 1% in 2.6 years. A study by Matsuo et al11 covered atomic bomb survivors from Nagasaki with a prevalence of 0.014% and an event rate of 15.6% in 40 years. It is unknown if the atomic bomb survivors’ exposure to radiation may have influenced their risk. Furuhashi et al2 recruited patients from the Asahikawa Red Cross and reported that 0.14% of their patients had asymptomatic Brugada ECG findings and no events at 2 years were found. In a European cohort with periodic health examinations followed for 10 years, prevalence was 0.19% and event rate was 0.3% per year.12 Length of follow-up of these subjects with Brugada ECG pattern may be another important factor in event rate. Follow-up periods ⬍3 years (in studies by Miyasaka et al10 and Furuhashi et al2) are likely to result in low event
716
The American Journal of Cardiology (www.ajconline.org)
rates; studies that followed patients longer picked up more events. We followed our subjects for 6 years and found a 4-year event rate of 43% (95% CI 6 to 80). Even with the small sample the best-case scenario showed a significantly high cardiac event rate of 6% in 4 years. This is a novel finding. All 7 subjects were men and were healthy and asymptomatic before initial diagnosis in 2003. The predominance of this condition in men is compatible with many reports in the literature.13 Two variant findings include occurrence of events during activity and sleep. Three subjects had events during activity, whereas only 2 of these 3 had events while asleep. This suggests that the autonomic system modulates the occurrence of arrhythmia but does not prevent its occurrence. The second divergent finding is that family history did not appear to differ between those with and those without events. Of the 3 subjects with major cardiac events, 2 did not have symptomatic first- or second-degree relatives. Paradoxically, 2 of the 4 subjects without events reported a positive family history of sudden death; this would seem to contradict the theory by Priori et al5 that the high cardiac event rate described by Brugada et al6 could be attributed to the recruitment of asymptomatic relatives of index patients with Brugada syndrome. Our data would seem to confirm a high crude event rate even in asymptomatic patients with the ECG sign without a family history. Although our sample is too small to serve as the basis for recommending treatment direction, 2 points must be considered. First, Brugada syndrome is not a common condition; hence, it would be difficult to gather a large enough sample directly from the community. Second, with a mean frequency for major events of 43% (95% CI 6 to 80), even the best scenario shows a risk of 6% at 4 years, which appears to be considerable, given that these patients were asymptomatic at time of ECG identification. 1. Gervacio-Domingo G, Isidro J, Tirona J, Gabreil E, David G, Amarillo ML, Morales D, Dans A. The Brugada type 1 electrocardiographic
2.
3.
4. 5.
6.
7.
8.
9. 10. 11.
12.
13.
pattern is common among Filipinos. J Clin Epidemiol 2008;61: 1067–1072. Furuhashi M, Uno K, Tsuchihashi K, Naghara D, Hyakukoku M. Prevalence of asymptomatic ST segment elevation in right precordial leads with right bundle branch block (Brugada-type ST shift) among the general Japanese population. Heart 2001;86:161–166. Brugada J, Brugada R, Brugada P. Determinants of sudden cardiac death in individuals with the electrocardiographic pattern of Brugada syndrome and no previous cardiac arrest. Circulation 2003;108:3092– 3096. Otto A. Brugada sign: a normal variant or a bad omen? Insights for risk stratification and prognostication. Eur Heart J 2004;25:810 – 811. Priori SG, Napolitano C, Gasparini M, Pappone C, Della Bella P, Brignole M, Giordano U, Giovannini T, Menozzi C, Bloise R, Crotti L, Terreni L, Schwartz PJ. Clinical and genetic heterogeneity of right bundle branch block and ST-segment elevation syndrome: a prospective evaluation of 52 families. Circulation 2000;102:2509 –2515. Brugada P, Brugada R, Rivero M, Brugada J. Natural history of Brugada syndrome: the prognostic value of programmed electrical stimulation of the heart. J Cardiovasc Electrophysiol 2003;14:455– 457. Gervacio-Domingo G, Punzalan FE, Amarillo ML, Dans A. Sudden unexplained death occurred commonly in the general population in the Philippines: a substudy of the National Nutrition and Health Survey. J Clin Epidemiol 2007;60:567–571. Eckardt L, Probst V, Smits JPP, Schulze Bahr E, Wolpert C, Schimpf R, Wichter T, Boisseau P, Heinecke A, Breithardt G, Borggrefe M, LeMarec H, Böcker D, Wilde AAM. Long-term prognosis of individuals with right precordial ST-segment-elevation Brugada syndrome. Circulation 2005;111:257–263. Viskin S, Rogowski O. Asymptomatic Brugada syndrome: a cardiac ticking time bomb. Europace 2007;9:707–710. Miyasaka Y, Tsuji H, Yamada K, Tokunaga S, Saito D, Imuro Y, Matsumoto N. Prevalence and mortality of the Brugada-type electrocardiogram in one city in Japan. J Am Coll Cardiol 2001;38:771–774. Matsuo K, Akahoshi M, Nakashima E, Suyama A, Seto S, Hayano M, Yano K. The prevalence, incidence and prognostic value of the Brugada-type electrocardiogram: a population-based study of four decades. J Am Coll Cardiol 2001;38:765–770. Gallagher MM, Forleo GB, Behr ER, Magliano G, De Luca L, Morgia V, De Liberato F, Romeo F. Prevalence and significance of Brugadatype ECG in 12,012 apparently healthy European subjects. Int J Cardiol 2008;130:44 – 48. Atarashi H, Ogawa S, Harumi K, Hayakawa H, Sugimoto T, Inoue H, Murayama M, Toyama J. Characteristics of patients with right bundle branch block and ST-segment elevation in right precordial leads. Am J Cardiol 1996;78:581–583.
Differential Effect of Elevated Blood Pressure on Left Ventricular Geometry Types in Black and White Young Adults in a Community (from the Bogalusa Heart Study) Jian Wang, MD, PhDa,b, Wei Chen, MD, PhDa, Litao Ruan, MD, PhDa,c, Ahmet Toprak, MDa, Sathanur R. Srinivasan, PhDa, and Gerald S. Berenson, MDa,* Hypertension and left ventricular (LV) hypertrophy are both more common in blacks than in whites. The aim of the present study was to test the hypothesis that blood pressure (BP) has a differential effect on the LV geometry types in black versus white asymptomatic young adults. As a part of the Bogalusa Heart Study, echocardiography and cardiovascular risk factor measurements were performed in 780 white and 343 black subjects (aged 24 to 47 years). Four LV geometry types were identified as normal, concentric remodeling, eccentric, and concentric hypertrophy. Compared to the white subjects, the black subjects had a greater prevalence of eccentric (15.7% vs 9.1%, p <0.001) and concentric (9.3% vs 4.1%, p <0.001) hypertrophy. On multivariate logistic regression analyses, adjusting for age, gender, body mass index, lipids, and glucose, the black subjects showed a significantly stronger association of LV concentric hypertrophy with BP (systolic BP, odds ratio [OR] 3.74, p <0.001; diastolic BP, OR 2.86, p <0.001) than whites (systolic BP, OR 1.50, p ⴝ 0.037; and diastolic BP, OR 1.35, p ⴝ 0.167), with p values for the race difference of 0.007 for systolic BP and 0.026 for diastolic BP. LV eccentric hypertrophy showed similar trends for the race difference in the ORs; however, the association between eccentric hypertrophy and BP was not significant in the white subjects. With respect to LV concentric remodeling, its association with BP was not significant in either blacks or whites. In conclusion, elevated BP levels have a greater detrimental effect on LV hypertrophy patterns in the black versus white young adults. These findings suggest that blacks might be more susceptible than whites to BP-related adverse cardiac remodeling. © 2011 Published by Elsevier Inc. (Am J Cardiol 2011;107:717–722)
Hypertension is an important risk factor for left ventricular (LV) hypertrophy,1,2 a subclinical cardiovascular disease that represents one of the strongest independent predictors of cardiovascular morbidity and mortality in the general population.3–5 Although it is well known that hypertension and LV hypertrophy are both more common in blacks than in their white counterparts,6 – 8 data on whether high blood pressure (BP) levels are more closely associated with an increased LV mass in blacks than in whites have not been consistent.8 –13 Also, information has been limited on how BP levels influence LV geometry and remodeling in black and white young adults. The objective of the present study was to examine the racial differences in the effect of BP on LV geometric patterns in black and white young adults enrolled in the Bogalusa Heart Study. a
Center for Cardiovascular Health, Tulane University, New Orleans, Louisiana; bFirst Affiliated Hospital, Shanxi Medical University, Taiyuan, China; and cFirst Affiliated Hospital, School of Medicine, Xi’an Jiaotong University, Xi’an, China. Manuscript received July 28, 2010; manuscript received and accepted October 11, 2010. This study was supported by grants HD-061437 and HD-062783 from the National Institute of Child Health and Human Development, grant 0855082E from the American Heart Association, and grant AG-16592 from the National Institute on Aging, Bethesda, Maryland. *Corresponding author: Tel: (504) 988-7197; fax: (504) 988-7194. E-mail address:
[email protected] (G.S. Berenson). 0002-9149/11/$ – see front matter © 2011 Published by Elsevier Inc. doi:10.1016/j.amjcard.2010.10.053
Methods As a part of the Bogalusa Heart Study, a biracial (black– white) community-based investigation of the early nature history of cardiovascular disease, a total of 1,123 subjects (780 whites and 343 blacks, 42.2% men, age 24 to 47 years), who were residing in the community of Bogalusa, Louisiana, were examined for LV dimensions and cardiovascular risk factor variables in 2004 to 2008. All subjects in the present study gave informed consent for examination. The institutional review board of the Tulane University Medical Center approved the study protocols. The BP levels were measured from the right arm of the subjects, with the subjects in a sitting position by 2 trained observers (3 replicates each). The systolic BP (SBP) and diastolic BP (DBP) were recorded at the first and fifth Korotkoff phases, respectively, using a mercury sphygmomanometer. The average of the 6 BP readings was used for analysis. For the 147 subjects who were taking medications for hypertension at the examination, the recorded BP values were adjusted by adding 10 mm Hg to the SBP and 5 mm Hg to the DBP, as determined by the average treatment effects.14,15 We tried to include these subjects because they represented a subgroup with the greatest BP. The LV dimensions were assessed using 2-dimensional guided M-mode echocardiography, with 2.25- and 3.5-MHz transducers according to the American Society of Echocarwww.ajconline.org
718
The American Journal of Cardiology (www.ajconline.org)
Table 1 Characteristics of study variables by race and patterns of left ventricular (LV) remodeling Variable
Blacks Normal (n ⫽ 230)
Age (years) Gender (n) Men Women Body mass index (kg/m2) Triglycerides (mg/dl) High-density lipoprotein cholesterol (mg/dl) Triglyceride/high-density lipoprotein cholesterol ratio Low-density lipoprotein cholesterol (mg/dl) Glucose (mg/dl) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Left ventricular mass index (g/m2.7) Relative wall thickness
Whites
CR (n ⫽ 27)
EH (n ⫽ 54)
CH (n ⫽ 32)
38 ⫾ 5
39 ⫾ 5
40 ⫾ 4*
39 ⫾ 5
87 143 29 ⫾ 7 104 ⫾ 67 53 ⫾ 15
13 14 29 ⫾ 6 86 ⫾ 40 54 ⫾ 13
23 31 37 ⫾ 9* 123 ⫾ 111 49 ⫾ 12†
2.3 ⫾ 2.2
1.7 ⫾ 0.9
121 ⫾ 35 94 ⫾ 37 119 ⫾ 15 74 ⫾ 11 33 ⫾ 7
Normal (n ⫽ 597)
EH (n ⫽ 71)
CH (n ⫽ 32)
CR (n ⫽ 80)
39 ⫾ 4
39 ⫾ 5
39 ⫾ 5
40 ⫾ 3
8 24† 38 ⫾ 10* 115 ⫾ 58 50 ⫾ 13
241 356 28 ⫾ 6 139 ⫾ 105 49 ⫾ 13
49 31* 29 ⫾ 5 157 ⫾ 129 44 ⫾ 11
44 27 36 ⫾ 7* 165 ⫾ 122 42 ⫾ 12*
9 23† 34 ⫾ 6* 169 ⫾ 95 44 ⫾ 10*
2.6 ⫾ 2.1
2.6 ⫾ 1.9
3.3 ⫾ 3.2
4.2 ⫾ 5.2
4.4 ⫾ 3.8†
4.0 ⫾ 2.4
109 ⫾ 31
116 ⫾ 46
122 ⫾ 36
127 ⫾ 33
127 ⫾ 32
127 ⫾ 31
135 ⫾ 32
85 ⫾ 13 121 ⫾ 14 74 ⫾ 10 36 ⫾ 7†
89 ⫾ 20 134 ⫾ 19* 83 ⫾ 12* 57 ⫾ 8*
107 ⫾ 39* 139 ⫾ 19* 86 ⫾ 13* 59 ⫾ 13*
87 ⫾ 18 113 ⫾ 12 71 ⫾ 9 32 ⫾ 8
90 ⫾ 26 113 ⫾ 11 71 ⫾ 8 36 ⫾ 7*
101 ⫾ 31* 121 ⫾ 13* 76 ⫾ 8* 55 ⫾ 8*
100 ⫾ 48* 121 ⫾ 15* 76 ⫾ 9* 55 ⫾ 10*
0.32 ⫾ 0.05 0.46 ⫾ 0.03* 0.35 ⫾ 0.04* 0.47 ⫾ 0.05* 0.32 ⫾ 0.05 0.47 ⫾ 0.05* 0.36 ⫾ 0.05* 0.48 ⫾ 0.06*
Data are presented as mean ⫾ SD for continuous variables. Compared to normal, adjusting for gender and age, * p ⬍0.01; † p ⬍0.05. CH ⫽ concentric hypertrophy; CR ⫽ concentric remodeling; EH ⫽ eccentric hypertrophy.
diography recommendations.16 The parasternal long- and short-axis views were used to measure the LV end-diastolic and end-systolic measurements in duplicate, which were then averaged. The LV mass was calculated from a necropsy-validated formula according to thick wall prolate ellipsoidal geometry.17 To take the body size into account, the LV mass was indexed for body height (m2.7). The LV relative wall thickness was calculated as the septal wall thickness plus the posterior wall thickness divided by the LV end-diastolic diameter.18 The presence of LV hypertrophy was defined by a gender-specific cutoff of the LV mass index of ⬎46.7g/m2.7 for women and ⬎49.2g/m2.7 for men. The LV geometry was considered concentric when the relative wall thickness was ⬎0.42.19 Four different patterns of LV geometry were defined: normal LV geometry, normal relative wall thickness with no LV hypertrophy; concentric remodeling, increased relative wall thickness but no LV hypertrophy; eccentric hypertrophy, normal relative wall thickness with LV hypertrophy; and concentric hypertrophy, increased relative wall thickness with LV hypertrophy.1,4,9,18 –21 The cholesterol and triglyceride serum levels were assayed using enzymatic procedures on a Hitachi 902 automatic analyzer (Roche Diagnostics, Indianapolis, Indiana). The glucose levels were measured as a part of a multiple chemistry profile (SMA20) using enzymatic procedures with a multichannel Olympus, Au-5000 analyzer (Olympus, Lake Success, New York). The laboratory was monitored for precision and accuracy of lipid measurements by the Lipid Standardization and Surveillance Program of the Centers for Disease Control and Prevention (Atlanta, Georgia). All the statistical analyses were performed using SAS, version 9.1 (SAS Institute, Cary, North Carolina). Analyses of covariance were performed using general linear models
Figure 1. Prevalence of LV hypertrophy and geometric remodeling in blacks and whites: the Bogalusa Heart Study.
to test the differences in the study variables between the normal group and LV geometric remodeling groups. Multivariate logistic regression analyses were performed to examine the associations of BP with concentric remodeling and different patterns of LV hypertrophy in separate models using normal geometry as a control. The race differences in odds ratios (ORs) for LV geometric patterns were tested by including the respective race–BP and race– covariate interaction terms in separate models. The SBP, DBP and covariates, except for gender, were standardized into Z scores with a mean of 0 and SD of 1 before logistic regression analyses to make the ORs comparable. The triglycerides and triglyceride/high-density lipoprotein cholesterol ratio were log-transformed to improve the normality of the distribution in the general linear models and logistic regression models;
Systemic Hypertension/Blood Pressure and Left Ventricular Geometry
719
Table 2 Odds ratio (OR) for different left ventricular (LV) geometry patterns by race Variable
CR Versus Normal
Model 1 Age Women Body mass index Triglyceride/high-density lipoprotein cholesterol ratio Low-density lipoprotein cholesterol Glucose Systolic blood pressure Model 2 Age Women Body mass index Triglyceride/high-density lipoprotein cholesterol ratio Low-density lipoprotein cholesterol Glucose Diastolic blood pressure
EH Versus Normal
CH Versus Normal
Whites
Blacks
p Value*
Whites
Blacks
p Value*
Whites
Blacks
p Value*
1.05 0.43** 1.11 1.10
1.28 0.58 1.28 0.63
0.501 0.537 0.611 0.077
1.25 0.73 2.82‡ 1.04
1.63† 1.24 3.67‡ 1.13
0.398 0.382 0.391 0.838
1.59† 3.06† 1.92‡ 1.11
1.13 7.07‡ 2.63‡ 0.93
0.356 0.724 0.203 0.795
0.95 1.12 0.83
0.63 0.77 1.02
0.088 0.214 0.700
0.75† 1.21 0.95
0.58‡ 0.48 1.82‡
0.403 0.077 0.016
0.93 1.31† 1.50†
0.82 1.20 3.74‡
0.817 0.670 0.007
1.04 0.42‡ 1.08 1.10
1.32 0.57 1.35 0.65
0.484 0.526 0.607 0.076
1.27 0.73 2.95‡ 1.05
1.66† 0.81 3.36‡ 1.20
0.402 0.404 0.447 0.889
1.65† 3.09† 2.06‡ 1.10
1.16 5.91‡ 2.49‡ 0.96
0.267 0.713 0.353 0.755
0.95 1.11 0.90
0.62 0.77 0.83
0.091 0.211 0.601
0.76 1.22 0.84
0.59‡ 0.47 1.62*
0.344 0.065 0.017
0.94 1.33† 1.35
0.76 1.16 2.86‡
0.627 0.492 0.026
Continuous variables were standardized into Z scores, with mean ⫽ 0 and SD ⫽ 1. * p Values for race difference in ORs; † p ⬍0.05, ‡ p ⬍0.01 for significance of OR. Abbreviations as in Table 1. Table 3 Odds ratios (ORs) for different left ventricular (LV) geometry patterns by race Variable
Model 1 Systolic blood pressure Diastolic blood pressure Model 2 Hypertension
CR Versus Normal
EH Versus Normal
CH Versus Normal
Whites
Blacks
p Value*
Whites
Blacks
p Value*
Whites
Blacks
p Value*
0.81 0.87
1.06 0.79
0.685 0.542
0.70 0.74
1.70† 1.75†
0.008 0.004
1.23 1.03
4.69‡ 4.45‡
0.003 0.004
1.01
1.28
0.864
1.68
2.60†
0.127
2.14*
6.21‡
0.014
Systolic and diastolic blood pressure were standardized into Z-scores, with mean ⫽ 0 and SD ⫽ 1. Same covariates as those listed in Table 2 were included in models 1 and 2. Hypertension was coded as 0 ⫽ no or 1 ⫽ yes. * p Values for race difference in ORs; † p ⬍0.05, ‡ p ⬍0.01 for significance of OR. Model 1, subjects receiving treatment (n ⫽ 174) were excluded; model 2, all subjects were included. Abbreviations as in Table 1.
however, the mean values in the original scales are listed in Table 1 for descriptive purposes. Results Figure 1 shows the prevalence of the LV geometric patterns by race. The black patients had a greater prevalence of eccentric hypertrophy (15.7% vs 9.1%, p ⬍0.001) and concentric hypertrophy (9.3% vs 4.1%, p ⬍0.001) than the white patients. Furthermore, the blacks had a greater prevalence of hypertension than the whites (33.2% vs 15.1%, p ⬍0.001). With respect to the overall average BP level by race groups, significant race differences in the BP mean levels were noted among those who were normotensive (SBP 111.1 mm Hg in whites and 115.1 mm Hg in blacks, p ⬍0.001; DBP 75.1 mm Hg in whites and 76.6 mm Hg in blacks, p ⬍0.001). Among those with hypertension, a significant race difference was noted in the SBP (135.7 mm Hg
in whites and 143.8 mm Hg in blacks, p ⬍0.001) but not in DBP (94.3 mm Hg in whites and 95.5 mm Hg in blacks, p ⫽ 0.355). Table 1 lists the characteristics of the study variables by race and LV geometric patterns. Compared to the normal geometry group, those with eccentric and concentric hypertrophy showed significantly greater levels of body mass index (BMI), SBP, and DBP, adjusting for age and gender, in both blacks and whites. The mean BMI levels, lipid variables, glucose levels, and BP did not differ significantly between the normal geometry and concentric remodeling groups. The association parameters (ORs) of BP with different LV geometric patterns derived from logistic regression models using the normal geometry as a control are listed in Table 2. The SBP and DBP levels (Z scores) were significantly associated with eccentric hypertrophy and concentric
720
The American Journal of Cardiology (www.ajconline.org)
hypertrophy in blacks. In contrast, only SBP was associated with concentric hypertrophy among the whites. The BP levels were not associated with concentric remodeling for both blacks and whites. The BMI showed significant and consistent associations with eccentric hypertrophy and concentric hypertrophy in blacks and whites, but not with concentric remodeling. Female gender was associated with a lower risk of concentric remodeling; however, women were more likely to have concentric hypertrophy, especially in black women (OR 5.91 to 7.07, p ⬍0.01). With respect to race difference, the associations of SBP and DBP with eccentric and concentric hypertrophy were significantly stronger in the blacks than in the whites; the other study variables did not show race differences in the ORs. The ORs with BP levels and hypertension status for LV geometric patterns, including the same covariates as those in Table 2, but using a different cohort (147 subjects receiving treatment excluded) in model 1 or a different predictor (hypertension status) in model 2 are listed in Table 3. In model 1, compared to the corresponding results listed in Table 2, the race difference in ORs with SBP and DBP levels (Z scores) did not change much for eccentric hypertrophy; however, the magnitude of the race difference in ORs was considerably increased for concentric hypertrophy. Using hypertension (yes vs no) as a predictor variable in model 2, a similar pattern in the race difference in the ORs for concentric hypertrophy was seen. Discussion The evidence that blacks had a greater prevalence of hypertension and LV hypertrophy than whites is almost indisputable; however, the data on racial differences in the magnitude of the BP–LV mass association have been inconsistent.8 –13 In the present biracial community-based study of asymptomatic young adults, the main finding was that the BP levels were more strongly associated with LV eccentric and concentric hypertrophy in the blacks than in the whites. In addition, the well-documented black–white difference in BP levels and the prevalence of hypertension was also noted in the present study cohort. These observations suggest that in addition to the black–white difference in the prevalence of hypertension, black patients with hypertension are more susceptible to LV hypertrophy than their white counterparts. Comparing the ORs for LV hypertrophy between the blacks and whites derived from the logistic regression interaction models in the present study provided appropriate and straightforward parameters to measure the racial difference in the effect of BP on LV hypertrophy. The Coronary Artery Risk Development in Young Adults (CARDIA) study,9 which had a similar age range to our cohort, reported Pearson correlation coefficients between the BP and LV mass in a longitudinal analysis and a BP trend with tertile groups of LV mass in a cross-sectional analysis. The black subjects had greater association parameters than did the whites, but the significance of the racial difference was not tested in the Coronary Artery Risk Development in Young Adults study. Another approach has been to use the adjustment of BP levels to examine its effect on the racial differences in the LV mass. Studies using the adjustment method
found that after multivariate adjustment for BP, anthropometric and lipid variables, and other covariates, the LV mass still remained greater in the blacks than in the whites.8,10 These findings suggest that the greater prevalence of LV hypertrophy cannot be fully accounted for by the greater prevalence of hypertension in the black population. In contrast, other studies have reported inconsistent or conflicting results regarding the BP–LV mass association in blacks and whites.12,13,22 Among a number of influencing factors, the sample size and selection, age range, and the definition of LV hypertrophy are possibly responsible for the discrepancies of the results in this regard. Furthermore, on the univariate analyses (Table 1), in the present study, those with eccentric and concentric hypertrophy showed significantly greater levels of BMI, SBP, and DBP, adjusting for age and gender, in both blacks and whites compared to the normal geometry group. However, the BP– eccentric hypertrophy associations became nonsignificant in whites on multivariate analyses (Table 2), suggesting the complex interactions between BP and other cardiovascular risk factors, in particular, the interaction between BP and BMI. Echocardiography has allowed identification of different forms of LV geometric remodeling, including eccentric or concentric hypertrophy and disproportionate septal thickness. Although the significance of the different forms, however, has not yet been well defined, concentric hypertrophy has been considered to carry the greatest risk of cardiovascular events4,5 and has been the predominant form in middle-age and elderly patients with hypertension.23 In the young adult cohort from the general population in the present study, the prevalence of concentric hypertrophy was 9.3% in blacks and 4.1% in whites. In the Framingham Offspring study, which consisted largely of whites with a mean age of 44 years, the prevalence of concentric hypertrophy was 3.4% in men and 1.8% in women.23 In a study of patients with hypertension and a mean age of 46 years in blacks and 47 years in whites, the blacks had twice the prevalence of LV hypertrophy (including both eccentric and concentric hypertrophy) in the whites (41% vs 19%).24 The black–white difference in the prevalence of concentric hypertrophy was noted even in children with a mean age of 14.7 years (13.1% in black children and 3.8% in white children).25 Obviously, different age groups, BP levels, and cutoff points of relative wall thickness and LV mass have been largely responsible for the differences in the prevalence rates reported by different studies. Obesity is an established risk factor of LV hypertrophy.26 –28 Obesity and hypertension occurring together place a dual burden on the left ventricle and are associated with metabolic abnormalities.27 The association between obesity and LV hypertrophy is reported to originate in childhood.29 In our previous longitudinal study, childhood obesity was found to predict the adulthood LV mass in a cohort from the same population.30 In some early reports, obesity appears to be a more consistent and stronger determinant of LV hypertrophy than BP levels or hypertension.8 –10,27 In the present study, the BMI showed a more consistent association with LV hypertrophy and stronger association with eccentric hypertrophy than BP in both blacks and whites by comparing the ORs derived using standardized variables. However, the effect of BMI on LV hypertrophy did not
Systemic Hypertension/Blood Pressure and Left Ventricular Geometry
differ significantly between the blacks and whites, suggesting that despite the greater effect of BMI, the BP might contribute more than obesity to the black–white difference in the prevalence of LV hypertrophy. Importantly, as mentioned, the adjustment for BMI and other cardiovascular risk variables resulted in a nonsignificant association between eccentric hypertrophy and BP in whites, suggesting a potential interaction between the BP and BMI. Gender is a strong determinant of both body size and LV mass. Therefore, we used gender-specific cutoffs to identify the LV geometric patterns in the present study. In the multivariate analyses (Table 2), female gender was associated with a lower risk of concentric remodeling; however, women were more likely to have concentric hypertrophy, especially for black women. Among the white subjects in the Framingham study, the prevalence of concentric hypertrophy was 10.6% in the men and 15.9% in the women in the original cohort (mean age 68 years) and 3.4% in men and 1.8% in women for the offspring-spouse cohort (mean age 44 years).23 Data were not available for comparison in blacks regarding the prevalence of concentric hypertrophy by gender. The observation of the greater risk of concentric hypertrophy in women in the present study, particularly in black women, needs to be studied further. In the present community-based epidemiologic study, a major limitation was the adjustment of adding 10 mm Hg and 5 mm Hg to the SBP and DBP value, respectively, for subjects taking antihypertensive medications. Although this approach has been commonly used according to the average treatment effects,14,15 it is an approximate estimation and might have led to a bias in the data analyses to some extent, because information on the drugs taken for the treatment was not available. Therefore, we also analyzed the data by excluding these subjects. The results indicated that the bias might lead to an underestimation of the race difference in the association. The present study have provided evidence supporting the hypothesis that the greater prevalence of LV hypertrophy in blacks than in whites has resulted from both a greater prevalence of hypertension and a stronger influence of BP on LV hypertrophy in blacks. In contrast, the influence of obesity on the BP–LV hypertrophy relation was more evident in the whites than in the blacks. These findings reflect the black–white divergence in the pathogenesis of LV hypertrophy and the potential role in the development of cardiovascular disease. A better understanding of the disparities of the pathophysiologic basis of LV hypertrophy by race and ethnicity might help clinicians and public health professionals develop culturally sensitive interventions, prevention programs, and services specifically targeted toward risk burdens in different populations.
1. Frohlich ED, Apstein C, Chobanian AV, Devereux RB, Dustan HP, Dzau V, Fauad-Tarazi F, Horan MJ, Marcus M, Massie B, Pfeffer MA, Re RN, Roccella EJ, Savage D, Shub C. The heart in hypertension. N Engl J Med 1992;327:998 –1008. 2. Diamond JA, Phillips RA. Hypertensive heart disease. Hypertens Res 2005;28:191–202. 3. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular
4.
5.
6.
7.
8.
9.
10.
11.
12. 13. 14. 15.
16. 17.
18. 19.
20.
21.
22.
721
mass in the Framingham Heart Study. N Engl J Med 1990;322: 1561–1566. Verdecchia P, Angeli F, Achilli P, Castellani C, Broccatelli A, Gattobigio R, Cavallini C. Echocardiographic left ventricular hypertrophy in hypertension: marker for future events or mediator of events? Curr Opin Cardiol 2007;22:329 –334. Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass and geometry to morbidity and mortality in uncomplicated essential hypertension. Ann Intern Med 1991;114:345– 352. Burt VL, Whelton P, Roccella EJ, Brown C, Cutler JA, Higgins M, Horan MJ, Labarthe D. Prevalence of hypertension in the US adult population: results from the Third National Health and Nutrition Examination Survey, 1988 –1991. Hypertension 1995;25:305–313. Drazner MH, Dries DL, Peshock RM, Cooper RS, Klassen C, Kazi F, Willett D, Victor RG. Left ventricular hypertrophy is more prevalent in blacks than whites in the general population: the Dallas Heart Study. Hypertension 2005;46:124 –129. Kizer JR, Arnett DK, Bella JN, Paranicas M, Rao DC, Province MA, Oberman A, Kitzman DW, Hopkins PN, Liu JE, Devereux RB. Differences in left ventricular structure between black and white hypertensive adults: the Hypertension Genetic Epidemiology Network study. Hypertension 2004;43:1182–1188. Lorber R, Gidding SS, Daviglus ML, Colangelo LA, Liu K, Gardin JM. Influence of systolic blood pressure and body mass index on left ventricular structure in healthy African-American and white young adults: the CARDIA study. J Am Coll Cardiol 2003;41:955–960. Gardin JM, Wagenknecht LE, Anton-Culver H, Flack J, Gidding S, Kurosaki T, Wong ND, Manolio TA. Relationship of cardiovascular risk factors to echocardiographic left ventricular mass in healthy young black and white adult men and women: the CARDIA study. Circulation 1995;92:380 –387. Dunn FG, Oigman W, Sugaard-Riise MFH, Ventura H, Reisin E, Frohlich ED. Racial differences in cardiac adaptation to essential hypertension determined by echocardiographic indexes. J Am Coll Cardiol 1983;5:1348 –1351. Mayet J, Chapman N, Li CKC, Shahi M, Poulter NR, Sever PS, Foale RA, Thom SA. Ethnic differences in the hypertensive heart and 24hour blood pressure profile. Hypertension 1998;31:1190 –1194. Hammond IW, Alderman MH, Devereux RB, Lutas EM, Laragh JH. Contrast in cardiac anatomy and function between black and white patients with hypertension. J Natl Med Assoc 1984;76:247–255. Cui J, Hopper JL, Harrap SB. Genes and family environment explain correlations between blood pressure and body mass index. Hypertension 2002;40:7–12. Neaton JD, Grimm RH Jr, Prineas RJ, Stamler J, Grandits GA, Elmer PJ, Cutler JA, Flack JM, Schoenberger JA, McDonald R. Treatment of mild hypertension study: final results—Treatment of Mild Hypertension Study Research group. JAMA 1993;270:713–724. Sahn DJ, DeMaria A, Kisslo J, Weyman A. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978;58:1072–1083. Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, Reichek N. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol 1986;57:450 – 458. Foppa M, Duncan BB, Rohde L. Echocardiography-based left ventricular mass estimation: how should we define hypertrophy? Cardiovasc Ultrasound 2005;3:17–29. de Simone G, Kitzman DW, Chinali M, Oberman A, Hopkins PN, Rao DC, Arnett DK, Devereux RB. Left ventricular concentric geometry is associated with impaired relaxation in hypertension: the HyperGEN study. Eur Heart J 2005;26:1039 –1045. Ganau A, Devereux RB, Roman MJ, de Simone G, Pickering TG, Saba PS, Vargiu P, Simongini I, Laragh JH. Patterns of left ventricular hypertrophy and geometric remodeling in essential hypertension. J Am Coll Cardiol 1992;19:1550 –1558. Toprak A, Reddy J, Chen W, Srinivasan S, Berenson G. Relation of pulse pressure and arterial stiffness to concentric left ventricular hypertrophy in young men (from the Bogalusa Heart Study). Am J Cardiol 2009;103:978 –984. Skelton TN, Andrew ME, Arnett DK, Burchfiel CM, Garrison RJ, Samdarshi TE, Taylor HA, Hutchinson RG. Echocardiographic left ventricular mass in African-Americans: the Jackson cohort of the
722
23.
24. 25. 26.
The American Journal of Cardiology (www.ajconline.org) Atherosclerosis Risk in Communities Study. Echocardiography 2003; 20:111–120. Savage DD, Garrison RJ, Kannel WB, Levy D, Anderson SJ, Stokes J 3rd, Feinleib M, Castelli WP. The spectrum of left ventricular hypertrophy in a general population sample: the Framingham Study. Circulation 1987;75:I-26-I–I-2633. Koren MJ, Mensah GA, Blake J, Laragh JH, Devereux RB. Comparison of left ventricular mass and geometry in black and white patients with essential hypertension. Am J Hypertens 1993;6:815– 823. Daniels SR, Loggie JMH, Khoury P, Kimball TR. Left ventricular geometry and severe left ventricular hypertrophy in children and adolescents with essential hypertension. Circulation 1998;97:1907–1911. Chirinos JA, Segers P, De Buyzere ML, Kronmal RA, Raja MW, De Bacquer D, Claessens T, Gillebert TC, St. John-Sutton M, Rietzschel
27. 28. 29. 30.
ER. Left ventricular mass: allometric scaling, normative values, effect of obesity, and prognostic performance. Hypertension 2010;56:91–98. Lavie CJ, Ventura HO, Messerli FH. Left ventricular hypertrophy: its relationship to obesity and hypertension. Postgrad Med 1992;91:135– 138. Kuch B, Hense HW, Gneiting B, Döring A, Muscholl M, Bröckel U, Schunkert H. Body composition and prevalence of left ventricular hypertrophy. Circulation 2000;102:405– 410. Dhuper S, Abdullah RA, Weichbrod L, Mahdi E, Cohen HW. Association of obesity and hypertension with left ventricular geometry and function in children and adolescents. Obesity Epub 2010 Jun 17. Li X, Li S, Ulusoy E, Chen W, Srinivasan SR, Berenson GS. Childhood adiposity as a predictor of cardiac mass in adulthood: the Bogalusa Heart Study. Circulation 2004;110:3488 –3492.
Predictive Value of Depressive Symptoms and B-Type Natriuretic Peptide for New-Onset Heart Failure and Mortality Krista C. van den Broek, PhDa,b, Christopher R. deFilippi, MDb, Robert H. Christenson, PhDc, Stephen L. Seliger, MDb, John S. Gottdiener, MDb, and Willem J. Kop, PhDa,b,* Depression and natriuretic peptides predict heart failure (HF) progression, but the unique contributions of depression and biomarkers associated with HF outcomes are not known. The present study determined the additive predictive value of depression and aminoterminal pro-B-type natriuretic peptide (NT-proBNP) for new-onset HF in HF-free subjects and mortality in patients with HF. The participants in the Cardiovascular Health Study were assessed for depressive symptoms using the Center for Epidemiologic Studies Depression Scale and NT-proBNP using an electrochemiluminescence immunoassay. The validated cutoff values for depression (Center for Epidemiologic Studies Depression Scale >8) and NT-proBNP (>190 pg/ml) were used. The risks of incident HF and mortality (cardiovascular disease-related and all-cause) were examined during a median follow-up of 11 years, adjusting for demographics, clinical factors, and health behaviors. In patients with HF (n ⴝ 208), depression was associated with an elevated risk of cardiovascular disease mortality (hazard ratios [HR] 2.07, 95% confidence interval [CI] 1.31 to 3.27) and all-cause mortality (HR 1.49, 95% CI 1.05 to 2.11), independent of the NT-proBNP level and covariates. The combined presence of depression and elevated NT-proBNP was associated with substantially elevated covariate-adjusted risks of cardiovascular disease mortality (HR 5.42, 95% CI 2.38 to 12.36) and all-cause mortality (HR 3.72, 95% CI 2.20 to 6.37). In the 4,114 HF-free subjects, new-onset HF was independently predicted by an elevated NT-proBNP level (HR 2.27, 95% CI 1.97 to 2.62) but not depression (HR 1.08, 95% CI 0.92 to 1.26) in covariateadjusted analysis. In conclusion, depression and NT-proBNP displayed additive predictive value for mortality in patients with HF. These associations can be explained by complementary pathophysiologic mechanisms. The presence of both elevated depression and NT-proBNP levels might improve the identification of patients with HF with a high risk of mortality. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:723–729)
a Center of Research on Psychology in Somatic diseases (CoRPS), Department of Medical Psychology and Neuropsychology, Tilburg University, Tilburg, The Netherlands; Departments of bMedicine and cPathology, University of Maryland School of Medicine, Baltimore, Maryland. Manuscript received September 21, 2010; manuscript received and accepted October 11, 2010. This research was supported by the National Heart, Lung, and Blood Institute (Bethesda, Maryland) grants N01-HC-85079 through N01-HC85086, N01-HC-35129, N01-HC-15103, N01-HC-55222, N01-HC-75150, N01-HC-45133, U01 HL080295, and R0-1 HL62181 and HL079376 (to W.J. Kop), with additional contribution from the National Institute of Neurological Disorders and Stroke, Bethesda, Maryland and Roche Diagnostics, Indianapolis, Indiana, and a travel grant, 2009R001, from the Dutch Heart Foundation, The Hague, The Netherlands (to K.C. van den Broek). A full list of the principal Cardiovascular Health Study investigators and institutions can be found at www.chs-nhlbi.org/pi.htm. Dr. deFilippi received research grant support (⬎$10,000) and honorarium/consulting fees (⬎$10,000) from Siemens, Glasgow, Delaware, Roche Diagnostics, Indianapolis, Indiana, BG Medicine, Waltham, Massachusetts, and Critical Diagnostics, San Diego, California. Dr. Christenson reports that funding from Roche Diagnostics, Indianapolis, Indiana was supplied for NT-proBNP testing. Siemens Healthcare Diagnostics, Glasgow, Delaware, and Response Biomedical, Vancouver, Canada have supported research efforts and market NT-proBNP assays. Dr. Seliger received a research grant from Roche, Inc, Indianapolis, Indiana. *Corresponding author: Tel: ⫹⫹31-13-466-8738; fax: ⫹⫹31-13-466-2067. E-mail address:
[email protected] (W.J. Kop).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.055
Depression has been associated with poor clinical outcomes in patients with heart failure (HF); however, the pathophysiologic mechanisms and potential synergy between depression and biomarkers in relation to an adverse HF prognosis are insufficiently understood.1,2 Research findings regarding the relation between depressive symptoms and B-type natriuretic peptide or aminoterminal proB-type natriuretic peptide (NT-proBNP) levels have been inconsistent.3– 8 The objectives of the present study were to determine whether depression was associated with elevated NT-proBNP levels in subjects with and without HF; to investigate whether depression and NT-proBNP are additive predictors of increased risk of new-onset HF in those free of HF; and to determine the additive predictive value of depression and NT-proBNP for cardiovascular and all-cause mortality in patients with HF. Methods The design of the Cardiovascular Health Study (CHS) has been previously described.9 In brief, CHS was a prospective, community-based, observational study of subjects aged ⱖ65 years, with the main objective to determine the factors related to the onset and course of coronary heart disease (CHD) and stroke.9 The participants were enrolled in 4 geographically distinct communities across the United www.ajconline.org
724
The American Journal of Cardiology (www.ajconline.org)
Table 1 Baseline characteristics of patients at risk of new-onset heart failure (HF) stratified by depression status Variable
Age (years) Men Black race Systolic blood pressure (mm Hg) Diabetes mellitus Total cholesterol (mg/dl) Body mass index (kg/m2) Current smoker Reduced physical activity Coronary heart disease Reduced left ventricular ejection fraction (⬍55%) Left ventricular hypertrophy  blocker Angiotensin-converting enzyme inhibitor Diuretics Antidepressants
Total (n ⫽ 4,114)
72.9 ⫾ 5.5 1,677 (40.8%) 584 (14.2%) 136.4 ⫾ 21.4 729 (17.7%) 212.0 ⫾ 39.0 26.6 ⫾ 4.6 449 (10.9%) 2,020 (49.1%) 728 (17.7%) 315 (7.7%) 181 (4.4%) 550 (13.4%) 264 (6.4%) 1,020 (24.8%) 145 (3.5%)
Depression
p Value
No (n ⫽ 3,274; 79.6%)
Yes (n ⫽ 840; 20.4%)
72.8 ⫾ 5.4 1,427 (43.6%) 407 (12.4%) 136.3 ⫾ 21.1 553 (16.9%) 211.5 ⫾ 38.9 26.6 ⫾ 4.4 332 (10.1%) 1,520 (46.4%) 532 (16.2%) 248 (7.6%) 129 (3.9%) 430 (13.2%) 197 (6.0%) 777 (23.8%) 83 (2.5%)
73.1 ⫾ 5.6 250 (29.8%) 177 (21.1%) 136.7 ⫾ 22.4 176 (21.0%) 214.2 ⫾ 39.5 26.9 ⫾ 5.0 117 (13.9%) 500 (59.5%) 196 (23.3%) 67 (8.0%) 52 (6.2%) 120 (14.3%) 67 (8.0%) 243 (28.9%) 62 (7.4%)
0.12 ⬍0.001* ⬍0.001* 0.64 0.006* 0.068 0.071 0.002* ⬍0.001* ⬍0.001* 0.70 0.005* 0.39 0.039* 0.002* ⬍0.001*
Data are presented as mean ⫾ SD, unless otherwise stated. * Statistically significant. Table 2 Baseline characteristics of participants with heart failure (HF) stratified by depression status Variable
Age (years) Men Black race Systolic blood pressure (mm Hg) Diabetes mellitus Total cholesterol (mg/dl) Body mass index (kg/m2) Current smoker Reduced physical activity† Coronary heart disease Reduced left ventricular ejection fraction Left ventricular hypertrophy  Blocker Angiotensin-converting enzyme inhibitor Diuretics Antidepressants
Total (n ⫽ 208)
75.2 ⫾ 6.1 102 (49.0%) 45 (21.6%) 136.7 ⫾ 24.9 77 (36.1%) 195.2 ⫾ 37.8 27.5 ⫾ 6.0 19 (9.1%) 140 (67.3%) 130 (62.5%) 76 (36.5%) 30 (14.4%) 31 (14.9%) 66 (31.7%) 139 (66.8%) 8 (3.8%)
Depression
p Value
No (n ⫽ 133; 63.9%)
Yes (n ⫽ 75; 36.1%)
75.3 ⫾ 5.9 77 (57.9%) 23 (17.3%) 136.1 ⫾ 24.9 43 (32.3%) 195.3 ⫾ 34.5 27.0 ⫾ 5.1 8 (6.0%) 83 (62.4%) 84 (63.2%) 49 (36.8%) 21 (15.8%) 19 (14.3%) 35 (26.3%) 80 (60.2%) 3 (2.3%)
74.9 ⫾ 6.5 25 (33.3%) 22 (29.3%) 137.9 ⫾ 25.0 32 (42.7%) 195.2 ⫾ 43.3 28.4 ⫾ 7.1 11 (14.7%) 57 (76.0%) 46 (61.3%) 27 (36.0%) 9 (12.0%) 12 (16.0%) 31 (41.3%) 59 (78.7%) 5 (6.7%)
0.63 0.001* 0.043* 0.61 0.14 0.98 0.13 0.038* 0.045* 0.79 0.90 0.46 0.74 0.025* 0.006* 0.12
Data are presented as mean ⫾ SD, unless otherwise stated. * Statistically significant. † ⬍1,072 kcal/wk.
States. The institutional review boards of the participating hospitals approved the CHS, and all patients gave written informed consent. The institutional review board of the University of Maryland (Baltimore, Maryland) approved the present analysis. The initial main CHS cohort included 5,201 participants recruited from 1989 to 1990 and an additional minority cohort of 687 participants recruited from 1992 to 1993. To enable simultaneous assessment of the echocardiographic data (obtained in 1994 to 1995 for the added minority cohort), the assessment visit obtained in 1994 to 1995 of the
minority cohort was considered their baseline evaluation for depression, NT-proBNP level, and covariates. The present study stratified participants by HF status, because depression and NT-proBNP could have differential roles for new-onset HF versus mortality in patients with pre-existing HF. Of the 5,888 participants in the CHS, 5,565 did not have prevalent HF at baseline. Of these 5,565 patients, 4,114 (74%) had complete data for depression, NT-proBNP level, and covariates. Of the 295 participants with prevalent HF at baseline, 208 (71%) with complete data were included in the analyses. The main reason for
Heart Failure/Depression, NT-proBNP, and Heart Failure
725
Table 3 Hierarchical Cox model for new-onset heart failure in participants free of heart failure at study entry (n ⫽ 4,114) Variable
Depression Elevated NT-proBNP
Adjusted Demographics*
Demographics* and CVD Risk Factors†
Demographics*, CVD Risk Factors†, Health Behaviors‡
Demographics*, CVD Risk Factors†, Health Behaviors‡, CHD Indexes§
1.18 (1.01–1.37)¶ 2.66 (2.33–3.04)¶
1.15 (0.99–1.35) 2.63 (2.30–3.02)¶
1.12 (0.96–1.31) 2.62 (2.29–3.01)¶
1.08 (0.92–1.26) 2.27 (1.97–2.62)¶
* Including age, gender, and race. † Including systolic blood pressure, cholesterol, and diabetes mellitus status. ‡ Including body mass index, smoking, and reduced physical activity. § Including CHD at baseline, reduced left ventricular ejection fraction, and left ventricular hypertrophy. ¶ HR (95% CI), statistically significant.
excluding the participants from the analyses was missing data for NT-proBNP resulting from the unavailability of stored samples in the blood repository (1,332 participants [24%] free of HF and 76 participants [26%] with HF at screening). As described previously, the differences between those with and without NT-proBNP data were minimal.10 The 10-item Center for Epidemiologic Studies Depression Scale (CES-D) was used to measure the symptoms of depression.11–13 The CES-D items cover the previous week using a 4-point Likert scale ranging from 0, rarely or none of the time (⬍1 day), to 3, all of the time (5 to 7 days). The CES-D has good psychometric properties.11 The total score can range from 0 to 30, with a score ⱖ8 indicating clinically relevant depression,11,12 similar to the ⱖ16 cutoff for the 20-item version of the CES-D. NT-proBNP was measured in serum using an electrochemiluminescence immunoassay (ECIA, Elecsys 2010 System, Roche Diagnostics, Indianapolis, Indiana), as described previously.10 The coefficient of variation for the NT-proBNP assay was 2% to 5%, and the analytic measurement range for NT-proBNP was 5 to 35,000 pg/ml. All samples were stored at ⫺70° to ⫺80°C before testing (maximum of 3 freeze-thaw cycles). The measurements of NTproBNP using this assay did not change after 5 freeze-thaw cycles.14 The cutoff NT-proBNP level of ⬎190 pg/ml was applied to indicate a high NT-proBNP level, because we have previously shown that this cutoff corresponds to the inflection point for an increased risk of new-onset HF in the CHS.10 The demographic and clinical data were obtained by trained interviewers and physical examination. The demographic variables included age, gender, and race (white vs black and other). The clinical risk factors for cardiovascular disease (CVD), including HF and CHD, were systolic blood pressure, diabetes mellitus status, and cholesterol level. The measures related to health behaviors included current smoking status, body mass index, and physical activity. Physical activity was assessed using self-reported activity levels, which were converted to kcal/wk.15 Because of the nonnormal distribution of this measure, the median level was used (⬍1,072 kcal/wk) to indicate those with low physical activity levels.16 The indexes of CHD status at study entry were a history of CHD at baseline (e.g., history of myocardial infarction, coronary revascularization by angioplasty or bypass surgery, or angina pectoris), echocardiographically defined ejection fraction (normal vs borderline and abnor-
Figure 1. Predictive value of depression and NT-proBNP for incident HF. Fully adjusted HR 2.33 (95% CI 1.85 to 2.95) for depressed/high NTproBNP group; HR 2.36 (95% CI 2.02 to 2.76) for nondepressed/high NT-proBNP group; and HR 1.18 (95% CI 0.95 to 1.46) for depressed/low NT-proBNP group using participants with low depression and low NTproBNP levels as reference group.
mal), and electrocardiographically determined left ventricular hypertrophy. Two-dimensional echocardiograms were obtained using a standardized protocol and interpreted at a centralized core laboratory by 2 trained and independent readers, who were unaware of the clinical information. The assessment protocol for cardiovascular outcomes has been previously described in detail.17 In brief, the participants were asked whether they had experienced cardiovascular outcomes during the semiannual interviews and annual examinations. The hospital discharge summaries, outpatient physician notes, and results of diagnostic tests were obtained for the participants who reported these outcomes. The Events Subcommittee adjudicated the outcomes by review of the medical records, including interview, physical examination, and diagnostic study data.17 New-onset HF was confirmed if a physician diagnosis of HF had been made and the subjects had received medical therapy for HF.17–19 The documentation of HF signs and symptoms and supporting diagnostic data were considered sufficient, but not necessary, to validate a diagnosis of HF.18 CVD mortality and all-cause mortality were examined by the Events Subcommittee and adjudicated for cardiovascu-
726
The American Journal of Cardiology (www.ajconline.org)
Table 4 Multivariate Cox model for interval to cardiovascular-related mortality in patients with heart failure (HF) at study entry (n ⫽ 208) Variable
Depression Elevated NT-proBNP
Adjusted Demographics*
Demographics*, CVD Risk Factors†
Demographics*, CVD Risk Factors†, Health Behaviors‡
Demographics*, CVD Risk Factors†, Health Behaviors‡, CHD Indexes§
2.04 (1.33–3.14) 2.83 (1.64–4.90)
2.19 (1.40–3.42) 3.03 (1.75–5.25)
2.08 (1.33–3.27) 3.17 (1.80–5.58)
2.07 (1.31–3.27) 2.70 (1.47–4.95)
All data statistically significant, HR (95% CI). * Including age, gender, and race. † Including systolic blood pressure, cholesterol, and diabetes mellitus status. ‡ Including body mass index, smoking, and reduced physical activity. § Including CHD at baseline, reduced left ventricular ejection fraction, and left ventricular hypertrophy.
lar cause by review of the medical records, death certificate, or autopsy reports and the Medicare database.17 The data are presented as the mean ⫾ SD or median with the interquartile range for variables with a non-normal skewed distribution. Chi-square tests and t tests for independent samples were applied to examine the differences between the depressed and nondepressed participants in baseline characteristics. The distribution of CES-D scores was positively skewed, and a square root transformation was applied before the parametric analyses, as described previously.12 The NT-proBNP levels were positively skewed, and logarithmic transformations were applied before the analyses.10 Transformed scores of depression and NT-proBNP were used in the analyses of continuous depression and NT-proBNP measures. The association between continuous depression and NT-proBNP levels was examined with Pearson’s correlation, and hierarchical linear regression analyses were used to determine whether the relation between depression and NT-proBNP was independent of the covariates. To test the additive independent value of depression and NT-proBNP for adverse clinical outcomes (new-onset HF in HF-free subjects and mortality in participants with HF at screening), hierarchical Cox regression analyses were performed. Four hierarchical sets of covariates were consecutively used in the multivariate models. These variables were selected, because they have been associated with depression, HF, and/or CHD progression. The first set of covariates included depression, NT-proBNP level, and demographics (i.e., age, gender, and race). The second set included CVD risk factors (i.e., systolic blood pressure, diabetes mellitus, and cholesterol). The third set contained factors related to health behaviors (i.e., smoking status, body mass index, and physical activity levels). The final set included indexes of cardiac disease status (CHD at baseline, left ventricular ejection fraction, and left ventricular hypertrophy). The role of cardiac medications was explored in separate models. The medications included  blockers, angiotensin-converting enzyme inhibitors, and diuretics. Using Cox proportional hazards analyses, the combined effect of depression and elevated NT-proBNP level on the outcomes was investigated by forming 4 groups according to the depression (CES-D ⱖ8) and NT-proBNP (ⱖ190 pg/ml) status. For all multivariate Cox proportional hazards models, graphic and formal methods were used to test the assumption of proportionality. We also examined whether the analyses using continuous CES-D depression scores and
Figure 2. Predictive value of depression and NT-proBNP for cardiovascular-related mortality in patients with HF. Fully adjusted HR 5.42 (95% CI 2.38 to 12.36) for depressed/high NT-proBNP group; HR 2.58 (95% CI 1.14 to 5.84) for nondepressed/high NT-proBNP group; and HR 1.91 (95% CI 0.65 to 5.64) for depressed/low NT-proBNP group using participants with low depression/low NT-proBNP levels as reference group.
NT-proBNP levels revealed the same results as those obtained when evaluating the dichotomized measures. To investigate whether multiplicative effects between depression and NT-proBNP were observed, the interaction between these 2 factors was examined. The analyses were performed using the Statistical Package for Social Sciences, version 17.0 (SPSS, Chicago, Illinois). A 2-sided p value of ⬍0.05 was used to indicate statistical significance. Results The baseline data for participants free of HF at study entry with complete data on depression, NT-proBNP level, and covariates (n ⫽ 4,114) are listed in Table 1. A total of 840 (20%) participants without HF reported clinically relevant levels of depressive symptoms (CES-D ⱖ8). Depression in those without HF was associated with female gender, black race, diabetes mellitus, current smoking status, reduced physical activity levels, CHD at baseline, left ventricular hypertrophy, and the use of angiotensin-converting enzyme inhibitors, diuretics, and antidepressive medications compared to nondepressed participants.
Heart Failure/Depression, NT-proBNP, and Heart Failure
727
Table 5 Multivariate Cox model for interval to all-cause mortality in patients with heart failure (HF) at study entry (n ⫽ 208) Variable
Depression Elevated NT-proBNP
Adjusted Demographics*
Demographics*, CVD Risk Factors†
Demographics*, CVD Risk Factors†, Health Behaviors‡
Demographics*, CVD Risk Factors†, Health Behaviors‡, CHD Indexes§
1.53 (1.10–2.13) 2.40 (1.61–3.57)
1.53 (1.08–2.15) 2.49 (1.67–3.72)
1.50 (1.07–2.12) 2.42 (1.59–3.66)
1.49 (1.05–2.11) 2.19 (1.40–3.43)
All data statistically significant, HR (95% CI). * Including age, gender, and race. † Including systolic blood pressure, cholesterol, and diabetes mellitus status. ‡ Including body mass index, smoking, and reduced physical activity. § Including CHD at baseline, reduced left ventricular ejection fraction, and left ventricular hypertrophy.
The baseline data of those with HF (n ⫽ 208) are listed in Table 2. Of these subjects, 36% were classified as having clinically relevant levels of depression. The patients with HF and depression were more often women, black, and current smokers, had lower physical activity levels, and were more often taking angiotensin-converting enzyme inhibitors and diuretics than did the patients with HF but without depression. The median NT-proBNP value in those without HF was 112 pg/ml (interquartile range 57 to 222 pg/ml). The participants without HF but with depression tended to have a greater median NT-proBNP level than those without HF and without depression (118 pg/ml, interquartile range 62 to 242, vs 110 pg/ml, interquartile range 56 to 218; p ⫽ 0.069). Greater continuous depression scores correlated weakly with greater NT-proBNP levels (r ⫽ 0.03, p ⫽ 0.043) in those without HF. However, this association lost significance when adjusting for demographics ( ⫽ 0.03, p ⫽ 0.070) and remained nonsignificant when additionally adjusting for CVD risk factors, health behaviors, and CHD indexes ( ⫽ 0.01, p ⫽ 0.61). In patients with HF at baseline, the mean NT-proBNP levels did not differ between those with and without depression (median 496 pg/ ml, interquartile range 159 to 1,632 vs median 520 pg/ml, interquartile range 148 to 1,716, p ⫽ 0.85). Similarly, no associations were found between the continuous depression scores and NT-proBNP levels (r ⫽ 0.001, p ⫽ 0.99), which remained nonsignificant in multivariate analyses. New-onset HF developed in 970 participants (24%) who were free of HF at screening (median follow-up 10.7 years, range 0.01 to 14). Depression (CES-D ⱖ8) was associated with an increased odds of incident HF when adjusting for demographics (hazard ratio [HR] 1.21, 95% confidence interval [CI] 1.04 to 1.42). The results remained similar when additionally adjusting for NT-proBNP (Table 3). However, this elevated depression-related risk was attenuated and nonsignificant after additional adjustment for CVD risk factors and remained nonsignificant after additional adjustment for health behaviors and indexes of CHD (Table 3) and cardiac medications (HR 1.13, 95% CI 0.96 to 1.32). The analyses of the continuous depression scores revealed the same pattern of results (demographic- and NT-proBNP– adjusted HR 1.09, 95% CI 1.02 to 1.17 per square root CES-D unit). No interaction was found between depression status and NT-proBNP level in the prediction of new-onset HF (p ⫽ 0.11, adjusting for demographics).
When comparing 4 groups according to the presence or absence of depression status and elevated NT-proBNP levels at baseline, the results of the demographic-adjusted models showed an elevated risk of new-onset HF in the depressed/high NT-proBNP group (HR 2.91, 95% CI 2.32 to 3.65, n ⫽ 282), nondepressed/high NT-proBNP group (HR 2.81, 95% CI 2.42 to 3.27, n ⫽ 948), and depressed/low NT-proBNP group (HR 1.33, 95% CI 1.07 to 1.64, n ⫽ 558), using the nondepressed/low NT-proBNP group as the reference (n ⫽ 2,326). Depression did not add to the effect of high NT-proBNP levels (nondepressed/high NT-proBNP group vs depressed/high NT-proBNP group, p ⫽ 0.76). Figure 1 shows that similar results were found when adjusting for all covariates. Of the 208 participants with HF at screening, 168 (81%) died during follow-up, with 97 patients dying from CVDrelated causes. The median duration until death was 5.8 years (range 0.06 to 14). The demographic-adjusted predictive value of depression for CVD-related mortality was significantly elevated (HR 2.17, 95% CI 1.41 to 3.32) and remained significant when adjusting for NT-proBNP, CVD risk factors, health behaviors, and cardiac disease indexes (Table 4). When adding cardiac medications, the predictive value of depression also remained significant (HR 1.76, 95% CI 1.08 to 2.88). The results were comparable in the model with continuous depression scores (HR 1.26, 95% CI 1.01 to 1.56 per square root CES-D unit, adjusting for demographics and NT-proBNP). No interaction was found between depression status and elevated NT-proBNP for increased CVD mortality risk (p ⫽ 0.78 in the demographicadjusted model). The combined risks of depression and NT-proBNP status for CVD-related mortality, adjusting for demographics, were as follows. The depressed/high NT-proBNP group showed the greatest risk of CVD death (HR 6.02, 95% CI 2.86 to 12.67, n ⫽ 56), followed by the nondepressed/high NT-proBNP group (HR 3.03, 95% CI 1.46 to 6.26, n ⫽ 95) and the depressed/low NT-proBNP group (HR 2.32, 95% CI 0.85 to 6.31, n ⫽ 19) compared to the nondepressed/low NT-proBNP group (n ⫽ 38). The risk of participants with both depression and high NT-proBNP levels was significantly greater than the risk of participants with high NTproBNP but no depression (p ⫽ 0.004). When adjusting for all covariates, the results remained basically similar (Figure 2). Among the patients with HF, depression was also associated with a significantly increased risk of all-cause mor-
728
The American Journal of Cardiology (www.ajconline.org)
Figure 3. Predictive value of depression and NT-proBNP for all-cause mortality in patients with HF. Fully adjusted HR 3.15 (95% CI 1.75 to 5.69) for depressed/high NT-proBNP group; HR 2.05 (95% CI 1.16 to 3.65) for nondepressed/high NT-proBNP group; and HR 1.32 (95% CI 0.60 to 2.88) for depressed/low NT-proBNP group using participants with low depression and low NT-proBNP levels as reference group.
tality (HR 1.59, 95% CI 1.14 to 2.22), after adjusting for demographics. Depression remained associated with a significant all-cause mortality risk (Table 5), after additional adjustment for NT-proBNP, CVD risk factors, health behaviors, and CHD indexes. The addition of cardiac medications decreased the predictive value of depression (HR 1.43, 95% CI 0.99 to 2.07). Continuous depression scores were also associated with an increased all-cause mortality risk (HR 1.21, 95% CI 1.02 to 1.43 per square root CES-D unit), adjusting for NT-proBNP and demographics. The interaction between an elevated depression status and increased NT-proBNP levels was not significant (p ⫽ 0.79 in the demographic-adjusted model). Investigating the combined effect of depression status and NT-proBNP showed that the depressed/high NTproBNP group had a 3.7-fold increased risk (95% CI 2.20 to 6.37, n ⫽ 56) for all-cause mortality, followed by the nondepressed/high NT-proBNP (HR 2.50, 95% CI 1.51 to 4.12, n ⫽ 95) compared to the reference category of nondepressed/low NT-proBNP levels (n ⫽ 38), adjusting for demographics. The depressed/low NT-proBNP group (n ⫽ 19) was not at a significantly elevated risk of all-cause mortality (HR 1.67, 95% CI 0.81 to 3.45). Of the participants with high NT-proBNP levels, depression significantly added to the risk of all-cause mortality (p ⫽ 0.031). The risks in the model with all covariates were similar (Figure 3). Discussion The results of the present investigation have shown that depression and NT-proBNP are independent and additive predictors of cardiovascular mortality and all-cause mortality in patients with HF during a median follow-up of 10 years. In contrast to the predictive value of mortality in patients with HF, depression did not independently predict new-onset HF in the covariate-adjusted models.
Investigation of the association between depressive symptoms and NT-proBNP in a large sample of the general population aged ⱖ65 years who were free of HF is a unique component of the present study. Previous studies have largely focused on patients with HF in clinical settings and found positive results, including those by Parissis et al3 and Laederach-Hofmann et al,4 as well as an absence of marked relationships.4 – 8 The psychological factors that partially overlap with depression, such as anxiety, negative affectivity, social inhibition, and type D personality (i.e., the joint presence of negative affectivity and social inhibition), have also not revealed significant associations with natriuretic peptide levels.8,20 One study found that pro-atrial natriuretic peptide correlated negatively with anxiety in patients with HF.21 Thus, the evidence regarding psychological factors and BNP and NT-proBNP has been mixed, but most studies have suggested that these factors are minimally related. Depression and NT-proBNP might therefore adversely affect HF progression by way of independent pathophysiologic pathways. Depression at baseline was an important predictor of mortality among patients with HF, consistent with the results from previous studies1 (for a meta-analysis, see the study by Nicholson et al22). The CHS has also confirmed a positive relation between depression and all-cause mortality in adults aged ⱖ65 years from the general population.13 The elevated risk associated with depression in the present study occurred over and above the risk of the simultaneously obtained NT-proBNP levels. Exploratory analyses revealed that depression and NT-proBNP were independently predictive of mortality in patients with HF with preserved left ventricular function. In contrast, the predictive value of depression was nonsignificant in patients with systolic HF (data not shown). The results of the present study have extended the current published data by demonstrating that depression adds to the risk of mortality in patients with HF over and above the effects of NT-proBNP. Depression was not independently associated with newonset HF, corroborating the results of previous studies that indicated that the role of depression for new-onset HF might be less strong in community-based samples23 than in samples with hypertension24 or coronary artery disease.25 It is possible that the association between depression and newonset HF is primarily mediated by cardiovascular co-morbidities that predispose patients to developing HF, including myocardial infarction and hypertension. The results of the present study should be interpreted in light of the following limitations. It is possible that depression reflects underlying clinical or subclinical disease processes. The present study adjusted for multiple possible “common factors,” such as CHD status and diabetes, which attenuated the risks of depression for incident HF. However, the common indexes of cardiac disease severity were not related to depression in those with HF (Table 2). In addition, the associations between depression and mortality in those with HF remained significant when taking these co-morbidities into account. Additional studies are needed to examine changes in depression and NT-proBNP (either in response to treatment or in observational studies) to elucidate the time trajectories and causal pathways linking depression and plausible biologic pathways of HF progression to ad-
Heart Failure/Depression, NT-proBNP, and Heart Failure
verse clinical outcomes. Selective survival might have resulted in an underestimation of the observed associations between depression and NT-proBNP, particularly in patients with HF, because high levels of both factors might have been associated with increased mortality risk before study enrollment. Because all participants with available data were included at study entry, selective survival was not likely to have been a primary factor in the observed relations between depression and NT-proBNP. Finally, depression and NT-proBNP were evaluated at one point (i.e., baseline) only, and no information was available on the long-term psychological treatment of depression. In conclusion, depression and NT-proBNP are independent risk factors for CVD-related and all-cause mortality in elderly individuals with HF. Recent evidence suggests that the detrimental effect of depression may be attributable to behavioral factors, such as physical inactivity and lack of compliance, as well as biological factors.26 Data from this investigation suggest that depression may add to the risk stratification of HF patients over and above HF-related biomarkers. 1. Rutledge T, Reis VA, Linke SE, Greenberg BH, Mills PJ. Depression in heart failure a meta-analytic review of prevalence, intervention effects, and associations with clinical outcomes. J Am Coll Cardiol 2006;48:1527–1537. 2. Kop WJ, Synowski SJ, Gottlieb SS. Depression in heart failure: biobehavioral mechanisms. Heart Fail Clin 2010;7:23–28. 3. Parissis JT, Farmakis D, Nikolaou M, Birmpa D, Bistola V, Paraskevaidis I, Ikonomidis I, Gaitani S, Venetsanou K, Filippatos G, Kremastinos DT. Plasma B-type natriuretic peptide and anti-inflammatory cytokine interleukin-10 levels predict adverse clinical outcome in chronic heart failure patients with depressive symptoms: a 1-year follow-up study. Eur J Heart Fail 2009;11:967–972. 4. Laederach-Hofmann K, Roher-Gubeli R, Messerli N, Meyer K. Comprehensive rehabilitation in chronic heart failure— better psycho-emotional status related to quality of life, brain natriuretic peptide concentrations, and clinical severity of disease. Clin Invest Med 2007;30: E54 –E62. 5. Gottlieb SS, Kop WJ, Ellis SJ, Binkley P, Howlett J, O’Connor C, Blumenthal JA, Fletcher G, Swank AM, Cooper L. Relation of depression to severity of illness in heart failure (from Heart Failure and a Controlled Trial Investigating Outcomes of Exercise Training [HFACTION]). Am J Cardiol 2009;103:1285–1289. 6. Redwine LS, Mills PJ, Hong S, Rutledge T, Reis V, Maisel A, Irwin MR. Cardiac-related hospitalization and/or death associated with immune dysregulation and symptoms of depression in heart failure patients. Psychosom Med 2007;69:23–29. 7. Sherwood A, Blumenthal JA, Trivedi R, Johnson KS, O’Connor CM, Adams KF, Jr., Dupree CS, Waugh RA, Bensimhon DR, Gaulden L, Christenson RH, Koch GG, Hinderliter AL. Relationship of depression to death or hospitalization in patients with heart failure. Arch Intern Med 2007;167:367–373. 8. Feola M, Rosso GL, Peano M, Agostini M, Aspromonte N, Carena G, Salvatico L, Valle R. Correlation between cognitive impairment and prognostic parameters in patients with congestive heart failure. Arch Med Res 2007;38:234 –239. 9. Fried LP, Borhani NO, Enright P, Furberg CD, Gardin JM, Kronmal RA, Kuller LH, Manolio TA, Mittelmark MB, Newman A. The Cardiovascular Health Study: design and rationale. Ann Epidemiol 1991; 1:263–276.
729
10. DeFilippi CR, Christenson RH, Gottdiener JS, Kop WJ, Seliger SL. Dynamic cardiovascular risk assessment in the elderly: the role of amino terminal pro-B-type natriuretic peptide testing. J Am Coll Cardiol 2010;55:441– 450. 11. Andresen EM, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). Am J Prev Med 1994;10:77– 84. 12. Kop WJ, Gottdiener JS, Tangen CM, Fried LP, McBurnie MA, Walston J, Newman A, Hirsch C, Tracy RP. Inflammation and coagulation factors in persons ⬎65 years of age with symptoms of depression but without evidence of myocardial ischemia. Am J Cardiol 2002;89:419 – 424. 13. Schulz R, Beach SR, Ives DG, Martire LM, Ariyo AA, Kop WJ. Association between depression and mortality in older adults: the Cardiovascular Health Study. Arch Intern Med 2000;160:1761–1768. 14. Ordonez-Llanos J, Collinson PO, Christenson RH. Amino-terminal pro-B-type natriuretic peptide: analytic considerations. Am J Cardiol 2008;101:9 –15. 15. Hirsch CH, Fried LP, Harris T, Fitzpatrick A, Enright P, Schulz R. Correlates of performance-based measures of muscle function in the elderly: the Cardiovascular Health Study. J Gerontol A Biol Sci Med Sci 1997;52:M192–M200. 16. Mozaffarian D, Furberg CD, Psaty BM, Siscovick D. Physical activity and incidence of atrial fibrillation in older adults: the cardiovascular health study. Circulation 2008;118:800 – 807. 17. Ives DG, Fitzpatrick AL, Bild DE, Psaty BM, Kuller LH, Crowley PM, Cruise RG, Theroux S. Surveillance and ascertainment of cardiovascular events: the Cardiovascular Health Study. Ann Epidemiol 1995; 5:278 –285. 18. Gottdiener JS, McClelland RL, Marshall R, Shemanski L, Furberg CD, Kitzman DW, Cushman M, Polak J, Gardin JM, Gersh BJ, Aurigemma GP, Manolio TA. Outcome of congestive heart failure in elderly persons: influence of left ventricular systolic function. The Cardiovascular Health Study. Ann Intern Med 2002;137:631– 639. 19. Parashar S, Katz R, Smith NL, Arnold AM, Vaccarino V, Wenger NK, Gottdiener JS. Race, gender, and mortality in adults ⬎ or ⫽65 years of age with incident heart failure (from the Cardiovascular Health Study). Am J Cardiol 2009;103:1120 –1127. 20. Pelle AJ, van den Broek KC, Szabo B, Kupper N. The relationship between type D personality and chronic heart failure is not confounded by disease severity as assessed by BNP. Int J Cardiol 2010;145: 82– 83. 21. Herrmann-Lingen C, Binder L, Klinge M, Sander J, Schenker W, Beyermann B, von Lewinski D, Pieske B. High plasma levels of N-terminal pro-atrial natriuretic peptide associated with low anxiety in severe heart failure. Psychosom Med 2003;65:517–522. 22. Nicholson A, Kuper H, Hemingway H. Depression as an aetiologic and prognostic factor in coronary heart disease: a meta-analysis of 6362 events among 146,538 participants in 54 observational studies. Eur Heart J 2006;27:2763–2774. 23. Williams SA, Kasl SV, Heiat A, Abramson JL, Krumholz HM, Vaccarino V. Depression and risk of heart failure among the elderly: a prospective community-based study. Psychosom Med 2002;64:6 –12. 24. Abramson J, Berger A, Krumholz HM, Vaccarino V. Depression and risk of heart failure among older persons with isolated systolic hypertension. Arch Intern Med 2001;161:1725–1730. 25. May HT, Horne BD, Carlquist JF, Sheng X, Joy E, Catinella AP. Depression after coronary artery disease is associated with heart failure. J Am Coll Cardiol 2009;53:1440 –1447. 26. Whooley MA, de Jonge P, Vittinghoff E, Otte C, Moos R, Carney RM, Ali S, Dowray S, Na B, Feldman MD, Schiller NB, Browner WS. Depressive symptoms, health behaviors, and risk of cardiovascular events in patients with coronary heart disease. JAMA 2008;300:2379 – 2388.
Effect and Clinical Prediction of Worsening Renal Function in Acute Decompensated Heart Failure Tobias Breidthardt, MD*, Thenral Socrates, MD, Markus Noveanu, MD, Theresia Klima, MD, Corinna Heinisch, MD, Tobias Reichlin, MD, Mihael Potocki, MD, Albina Nowak, MD, Christopher Tschung, BSc, Nisha Arenja, MD, Roland Bingisser, MD, and Christian Mueller, MD We aimed to establish the prevalence and effect of worsening renal function (WRF) on survival among patients with acute decompensated heart failure. Furthermore, we sought to establish a risk score for the prediction of WRF and externally validate the previously established Forman risk score. A total of 657 consecutive patients with acute decompensated heart failure presenting to the emergency department and undergoing serial creatinine measurements were enrolled. The potential of the clinical parameters at admission to predict WRF was assessed as the primary end point. The secondary end point was all-cause mortality at 360 days. Of the 657 patients, 136 (21%) developed WRF, and 220 patients had died during the first year. WRF was more common in the nonsurvivors (30% vs 41%, p ⴝ 0.03). Multivariate regression analysis found WRF to independently predict mortality (hazard ratio 1.92, p <0.01). In a single parameter model, previously diagnosed chronic kidney disease was the only independent predictor of WRF and achieved an area under the receiver operating characteristic curve of 0.60. After the inclusion of the blood gas analysis parameters into the model history of chronic kidney disease (hazard ratio 2.13, p ⴝ 0.03), outpatient diuretics (hazard ratio 5.75, p <0.01), and bicarbonate (hazard ratio 0.91, p <0.01) were all predictive of WRF. A risk score was developed using these predictors. On receiver operating characteristic curve analysis, the Forman and Basel prediction rules achieved an area under the curve of 0.65 and 0.71, respectively. In conclusion, WRF was common in patients with acute decompensated heart failure and was linked to significantly worse outcomes. However, the clinical parameters failed to adequately predict its occurrence, making a tailored therapy approach impossible. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:730 –735) Worsening renal function (WRF) is a common complication during the treatment of acute decompensated heart failure (ADHF) and has even earned the term “cardiorenal syndrome type I.”1,2 About 20% to 40% of patients hospitalized for ADHF concurrently develop WRF.3 The increased short-term morbidity, treatment costs, and in-hospital mortality of patients with heart failure who develop WRF have been well established. Recent studies have also showed WRF to possess a graded association with longterm mortality.4 Although WRF has been reported to occur more frequently in the presence of co-morbid conditions, such as impaired baseline renal function, lowered left ventricular function, age, and arterial hypertension,5–7 and a clinical risk score has been advocated,8 its prediction remains a daily clinical challenge. We, therefore, aimed to determine the clinical predictors of WRF in a large cohort of patients with ADHF presenting to the emergency department. Furthermore, we sought to establish a novel risk score
Department of Internal Medicine, University Hospital, Basel, Switzerland. Manuscript received July 27, 2010; manuscript received and accepted October 13, 2010. *Corresponding author: Tel: (0041) 61-265-2525; fax: 0041-61-2655353. E-mail address:
[email protected] (T. Breidthardt). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.056
for the prediction of WRF and externally validate the previously established Forman risk score. Methods During the enrollment periods (from May 2001 to April 2002 and April 2006 to April 2010), we prospectively screened all patients presenting to the participating emergency departments (University Hospital, Basel, Kantonsspital Aarau, Kantonsspital, Lucerne) with a chief complaint of acute dyspnea. A total of 767 patients were diagnosed with ADHF according to the current guidelines.9 The 657 patients with ADHF hospitalized for ⬎48 hours and undergoing serial creatinine measurements constituted the study population for the present analysis. The 1-year follow-up data were complete for 575 patients. The patients ⬍18 years of age, those undergoing hemodialysis, and trauma cases were excluded. The study was performed according to the principles of the Declaration of Helsinki. The local ethical committee approved the present study. All participants provided written informed consent. At enrollment, all patients underwent an initial clinical assessment, including clinical history, physical examination, pulse oximetry, blood tests, and chest radiography. Echocardiography, blood gas analyses, and pulmonary function tests were perwww.ajconline.org
Heart Failure/Clinical Prediction of WRF in ADHF
731
Table 1 Demographic and clinical characteristics Variable
Age (years) Women Arterial hypertension Heart failure Coronary artery disease Diabetes mellitus Chronic kidney disease Neoplastic disease Diuretics at admission  Blockers at admission Nitrates at admission Renin angiotensin system blockers at admission Digitalis at admission Aspirin at admission Calcium channel blockers at admission Anticoagulation at admission Blood pressure (mm Hg) Systolic Diastolic Heart rate (beats/min) Oxygen saturation (%) Left ventricular ejection fraction (%) Urea (mmol/L) Creatinine (mol/L) Glomerular filtration rate (ml/min) B-type natriuretic peptide (pg/ml) Troponin T pH* Bicarbonate (mmol/L)* Percutaneous coronary intervention during hospitalization Contrast-enhanced computed tomography during hospitalization Hospital stay (days) In-hospital mortality 360-Day mortality*
All Patients (n ⫽ 657)
79 [71–85] 293 (45%) 460 (70%) 356 (55%) 353 (54%) 196 (31%) 268 (42%) 137 (21%) 456 (70%) 359 (56%) 157 (24%) 390 (60%) 63 (10%) 305 (47%) 146 (21%) 253 (39%) 137 [120–160] 83 [70–96] 91 [75–111] 96 [92–98] 40 [30–60] 11.0 [6.9–18.5] 107 [80–148] 49 [32–71] 981 (534–1493) 0.04 [0.01–0.4] 7.42 [7.35–7.45] 23.2 [20.6–26.9] 70 (11%)
WRF No (n ⫽ 521) 79 [70–85] 240 (46%) 364 (70%) 274 (54%) 281 (54%) 148 (29%) 190 (37%) 111 (21%) 349 (67%) 275 (54%) 127 (24%) 298 (58%) 53 (10%) 248 (47%) 100 (20%) 198 (38%) 136 83 96 95 45 10.0 104 50 925 0.03 7.42 24.0
[120–158] [70–96] [77–111] [90–98] [30–60] [6.6–18.0] [78–146] [33–72] [520–1441] [0.01–0.355] [7.36–7.46] [21.0–24.5] 55 (11%)
p Value Yes (n ⫽ 136) 79 [72–85] 53 (39%) 96 (71%) 82 (61%) 72 (53%) 48 (36%) 78 (58%) 26 (19%) 107 (79%) 84 (63%) 30 (22%) 82 (61%) 10 (11%) 57 (42%) 33 (24%) 55 (41%) 144 [125–164] 84 [74–97] 89 [71–120] 93 [90–98] 40 [30–55] 12.6 [8.6–22.1] 115 [86–156] 49 [21–68] 1,071 (709–1714) 0.03 [0.01–0.300] 7.41 [7.35–7.44] 21.7 [18.2–25.4] 15 (11%)
0.36 0.08 0.55 0.59 0.49 0.08 ⬍0.01 0.89 0.01 0.04 0.95 0.40 0.33 0.76 0.50 0.34 0.02 0.54 0.73 0.86 0.69 ⬍0.01 0.02 0.02 0.02 0.14 0.06 0.84 0.87
103 (16%)
80 (15%)
23 (17%)
0.98
13 [7–19] 56 (9%) 220 (38%)
14 [9–20] 33 (6%) 171 (30%)
15 [8–23] 23 (17%) 49 (41%)
0.42 ⬍0.01 0.03
Data are presented as n (%) or median [interquartile range]. * pH and bicarbonate levels available for 256 patients and 360-day mortality for 575 patients.
formed according to the treating physicians’ recommendations. In accordance with the clinical practice guidelines for chronic kidney disease (CKD) from the American National Kidney Foundation10 and for comparison with previous studies,11,12 renal function was divided into 3 groups according to the glomerular filtration rate (GFR; ⬍30 ml/min/ 1.73 m2, 30 to 60 ml/min/1.73 m2, and ⬎60 ml/min/1.73 m2) on presentation to the emergency room. The GFR was calculated using the abbreviated Modification of Diet in Renal Disease Study (MDRD) equation.13 This method is currently considered the best method clinically available for this purpose14 and has been validated in patients with advanced HF.15 WRF was defined as an in-hospital increase in serum creatinine ⬎0.3 mg/dl (26.5 mol/L) from admission.8,11,12 We defined CKD as a stably reduced estimated GFR of ⬍60 ml/min/1.73 m2 that had persisted for ⬎3 months before hospitalization.
The establishment of a tool to stratify patients with ADHF at admission according to their risk of developing WRF during the hospitalization, as well as the potential of this score and the Forman risk score to predict WRF, was the primary end point of the present study. All-cause mortality at 360 days of follow-up was assessed as the secondary end point. Mortality was prospectively assessed during follow-up. Patients were interviewed by telephone at 6 and 12 months after their initial presentation. In addition, referring physicians and the administrative databases of the respective hometowns were interviewed/accessed in the case of uncertainty regarding the health status or additional hospitalizations. The ADHF diagnosis was adjudicated by 2 independent cardiologists not involved in the emergency department care using all available medical records pertaining to the individual patient, including the response to therapy and the autopsy data for those patients who had died in hospital.
732
The American Journal of Cardiology (www.ajconline.org)
Table 2 Prediction of in-hospital worsening renal function (WRF) on univariate and multivariate logistic regression analysis Predictor Univariate analysis Glomerular filtration rate class History of diabetes History of chronic kidney disease Outpatient diuretics Outpatient  blockers Bicarbonate Multivariate analysis History of chronic kidney disease Multivariate analysis including blood gas analyses (n ⫽ 253) History of chronic kidney disease Outpatient diuretics Bicarbonate
Hazard Ratio (95% CI)
p Value
1.35 (1.04–1.73) 1.36 (0.91–2.04) 2.31 (1.57–3.40) 1.89 (1.15–2.82) 1.44 (0.97–2.13) 0.92 (0.87–0.97)
0.02 0.13 ⬍0.01 0.01 0.06 0.03
2.07 (1.29–3.30)
⬍0.01
2.13 (1.08–4.19) 5.75 (2.10–15.73) 0.91 (0.86–0.97)
0.03 ⬍0.01 ⬍0.01
Age, gender, vital signs, all other co-morbidities, outpatient medications, and laboratory parameters, including left ventricular ejection fraction, blood gas analysis parameters, cardiac troponin T, and B-type natriuretic peptide, failed to predict WRF. Table 3 Basel risk score Class
0 1 2 3
WRF No (n ⫽ 204)
Yes (n ⫽ 51)
49 (24) 78 (38) 64 (31) 13 (6)
1 (1) 18 (35) 14 (27) 18 (35)
Data are presented as n (%). Jonckheere-Terpstra test for occurrence of WRF, p ⫽ 0.07.
The statistical analyses were performed using Statistical Package for Social Sciences for Windows, version 15.0 (SPSS, Chicago, Illinois). A statistical significance level of p ⬍0.05 was used. The comparison between the 2 groups was done using the chi-square test and Fisher’s exact test for categorical variables and the t test for continuous variables if normally distributed or the Mann-Whitney U test if not normally distributed. Cumulative survival was calculated using Kaplan-Meier analysis and differences between the curves were evaluated using log-rank statistics. The independent predictors of WRF were identified using multivariate regression analysis. The analyzed variables included age, gender, vital signs, co-morbidities, outpatient medications, and laboratory parameters. The laboratory parameters included the left ventricular ejection fraction, blood gas analysis parameters, and cardiac troponin T and B-type natriuretic peptide levels. Overall, 48 parameters were evaluated. The variables were entered at an entry level of significance p ⬍0.2 and kept in the model at an existence level of p ⬍0.05. A clinically applicable risk score was calculated as the sum of point values assigned to each independent predictor variable. The number of points assigned to each predictor depended on the regression coefficient . Because the regression coefficient  values of all
Table 4 External validation of Forman risk score Forman Class
0 1 2 3 4 5 6
WRF No (n ⫽ 496)
Yes (n ⫽ 129)
115 (23) 153 (30) 80 (16) 48 (10) 65 (10) 30 (13) 5 (1)
19 (10) 31 (24) 24 (12) 19 (10) 18 (9) 17 (9) 1 (1)
Data are presented as n (%). Jonckheere-Terpstra test for occurrence of WRF, p ⫽ 0.29. Table 5 Prediction of in-hospital mortality on univariate and multivariate logistic regression analysis Predictor Univariate analysis Systolic blood pressure Glomerular filtration rate class In-hospital worsening renal function B-type natriuretic peptide (for 100pg/ml increase) Multivariate analysis Systolic blood pressure Glomerular filtration rate class In-hospital worsening renal function
Hazard Ratio (95% CI)
p Value
0.98 (0.97–0.99) 1.49 (1.02–2.12) 5.01 (2.04–7.03) 1.01 (0.99–1.03)
⬍0.01 0.04 ⬍0.01 0.23
0.98 (0.96–0.99) 1.17 (0.74–1.84) 5.53 (2.81–10.88)
⬍0.01 0.49 ⬍0.01
parameters in the final model were similar, no weighing of the point values (i.e., more points for certain parameters) was necessary. For continuous variables, a point value was assigned for values above (1) or below (0) the best calculated cut-off value using Youden’s J index. In the complex, computer-based risk score model, continuous values and actual regression coefficient values were used. In a first step, the baseline characteristics, vital signs, and standard laboratory parameters available for all 522 patients were assessed. In a second step, the blood gas analysis parameters were added to the analysis; these parameters were available for 223 patients. Furthermore, the previously described Forman prediction rule was calculated. The relation among the Basel risk score, Forman risk score, and development of WRF was assessed using the Jonckheere-Terpstra test. The prognostic accuracy of the different models was evaluated using receiver operating characteristic (ROC) curve analysis. The areas under the ROC curves were compared using MedCalc software, version 9.2 (MedCalc Software, Mariakerke, Belgium). Results The detailed baseline characteristics of the study population are listed in Table 1. Overall, 136 patients (21%) experienced WRF during the hospitalization. No differences were seen in the co-morbidities, vital signs, outpatient treatment, admission laboratory parameters, or the incidence of
Heart Failure/Clinical Prediction of WRF in ADHF Table 6 Prediction of 360-day mortality on univariate and multivariate logistic regression analysis Predictor Univariate analysis Age Systolic blood pressure Glomerular filtration rate class Glomerular filtration rate Glomerular filtration rate class at discharge Glomerular filtration rate at discharge Urea (mmol/L) Uric acid (mmol/L) In-hospital worsening renal function B-type natriuretic peptide (for 100-pg/ml increase) Multivariate analysis Systolic blood pressure In-hospital worsening renal function B-type natriuretic peptide (for 100-pg/ml increase) Glomerular filtration rate class
Hazard Ratio (95% CI)
p Value
1.04 (1.03–1.06) 0.99 (0.98–0.99) 1.54 (1.28–1.87) 0.99 (0.99–0.99) 1.49 (0.94–2.09)
⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 0.08
0.99 (0.98–0.99)
⬍0.01
1.01 (1.00–1.01) 1.00 (1.00–1.00) 1.67 (1.21–2.38) 1.03 (1.02–1.03)
0.03 ⬍0.01 0.02 ⬍0.01
0.99 (0.98–0.99) 1.92 (1.29–2.59) 1.01 (1.00–1.03)
⬍0.01 ⬍0.01 ⬍0.01
1.34 (1.03–1.75)
0.03
1.0 P<0.01
Survival rate
0.8
0.6 0.4 GFR >60, no WRF GFR 30-60, no WRF GFR <30, no WRF
0.2
GFR >60, WRF GFR 30-60, WRF GFR <30, WRF
0.0 0
90
180
270
360
450
540
Follow-up period [days]
Figure 1. Survival stratified according to GFR classes at admission and occurrence of WRF.
WRF between the patients undergoing blood gas analyses and those without blood gas analyses. The risk factors for the development of WRF are listed in Table 2. Previously diagnosed CKD was the only independent predictor of WRF. The creatinine values (p ⫽ 0.38), urea values (p ⫽ 0.87), GFR rate (p ⫽ 0.74), or GFR classes (p ⫽ 0.98) did not independently predict WRF. In a ROC curve analysis for the development of WRF, previously diagnosed CKD achieved an area under the curve of 0.60 (95% confidence interval [CI] 0.55 to 0.66). However, only 28% of patients with previously diagnosed CKD developed WRF compared to 15% of patients without known CKD. Furthermore, 42% of all WRF cases developed in patients without known CKD.
733
Including the blood gas analysis parameters into the calculation, previously diagnosed CKD, outpatient treatment with diuretics, and lowered bicarbonate levels were independently associated with the development of WRF. Again, the creatinine levels (p ⫽ 0.60), urea values (p ⫽ 0.99), GFR rate (p ⫽ 0.52), and GFR classes (p ⫽ 0.32) failed to predict WRF. A risk score was developed using the predictors. One point each was assigned for outpatient diuretic treatment, known CKD, and bicarbonate level ⬍21 mmol/L. The relation between the calculated Basel risk score and the development of WRF is listed in Table 3. Of the 20% of patients in risk group 0, 2% developed WRF, and of the 43% of patients in risk groups 2 and 3, 29% developed WRF. However, ⬎1/3 (37%) of all WRF cases developed in the low-risk (0) and intermediate-risk (1) groups. In a ROC curve analysis for the development of WRF, the Basel risk score achieved an area under the curve of 0.71 (95% CI 0.63 to 0.79), significantly surpassing the diagnostic accuracy of the single-parameter model (p ⫽ 0.02). A computer-based, complex, exponential Basel risk assessment model achieved an area under the curve of 0.75 (95% CI 0.67 to 0.82). The relation between the Forman risk score and the development of WRF is listed in Table 4. Greater Forman risk scores were not associated with a greater incidence of WRF (0.09). Hence, 14% of patients in risk group 0 developed WRF compared to 27% of patients in risk groups 4 or greater. Additionally, ⬎50% of WRF cases developed in the low-risk groups, risk groups 2 or less. In a ROC curve analysis for the prediction of WRF, the Forman risk score achieved an area under the curve of 0.65 (95% CI 0.57 to 0.71). The prognostic accuracy of the novel Basel score and Forman risk score was similar (p ⫽ 0.24). Overall, 56 patients died in hospital and 220 patients died during the 360 days after admission. Independent of the time of death, WRF was more common in patients who died during follow-up (in-hospital, p ⬍0.01; 360 days, p ⫽ 0.03). Thus, multivariate Cox regression analysis found WRF to independently predict short-term (Table 5) and long-term (Table 6) mortality. Spot measurements of renal function described by the creatinine values, urea values, or continuous GFR values at admission failed to predict in-hospital mortality. Nevertheless, morbidity was greater in the patients with a lower GFR class, as reflected by the longer duration of the initial hospitalization (GFR ⬍30 ml/min, 16 days, range 9 to 24; vs GFR 30 to 60 ml/min, 12 days, range 6 to 19; vs GFR ⬎60 ml/min, 11 days, range 7 to 17; p ⬍0.01). No relevant differences were seen in the discharge medication prescribed for HF among the patients with and without renal dysfunction. However, the 90- and 360-day mortality rates were significantly greater in patients with a lower admission GFR class (90-day, 28% vs 15% vs 12%, p ⬍0.01; 360-day, 48% vs 33% vs 24%, p ⬍0.01). As shown in Figure 1, the risk of death clearly increased with decreasing GFR class. However, the occurrence of WRF further increased the risk of death independently of the baseline GFR class. Discussion In the present investigation, we examined the effect of renal dysfunction, as defined by a decreased GFR at pre-
734
The American Journal of Cardiology (www.ajconline.org)
sentation and in-hospital WRF, on the outcome of patients with ADHF. In agreement with previous studies, we found impaired renal function was associated with increased morbidity and short- and long-term mortality. Additionally, we determined that WRF is common and exceeds the prognostic potential of admission renal function when determining the mortality risk. Furthermore, we had 4 major findings. First, a history of CKD, with a hazard ratio of 2.07, was the single most powerful predictor for the development of WRF. The presence of CKD alone achieved an area under the ROC curve for the prediction of WRF of 0.6. However, 42% of all WRF cases developed in patients without known CKD. Second, after the inclusion of the blood gas analysis parameters to the analysis, a history of CKD, outpatient diuretics, and bicarbonate were all predictive of in-hospital WRF. A novel risk score using these clinical predictors achieved an area under the ROC curve of 0.71. Third, a Basel risk score of 0 was associated with a very low risk WRF (rule out). Finally, the prognostic accuracy of the novel Basel score was comparable to that of the previously described Forman risk score. Our results are in full agreement with the findings of a recent meta-analysis.4 In the present large cohort, WRF developed in 25% of all patients and was associated with a greater risk of mortality and hospitalization. Additional studies have linked the development of WRF to greater short- and long-term all-cause and cardiovascular mortality, prolonged hospitalization, increased readmission rates, accelerated progression to CKD stage 4 and 5, and greater healthcare costs.3 Therefore, the assessment of WRF is of utmost importance in patients with ADHF. Additionally, WRF has gained widespread clinical attention in noncardiac settings. A recent international consensus introduced the term “acute kidney injury” for acute creatinine changes ⬎0.3 mg/dl from the baseline value or any 1.5-fold creatinine increase.16 The cutoff value of 0.3 mg/dl was chosen, in analogy to the definition of WRF in the HF setting. In a meta-analysis of ⬎47,000 patients from 48 studies, Coca et al17 found in-hospital acute kidney injury to be independently associated with mortality and myocardial infarction. The alternative classification with the acronym RIFLE (risk, injury, and failure, sustained loss, and endstage kidney disease)18 defined acute kidney injury as a 1.5-fold increase in creatinine value or a 25% decrease in the estimated GFR. Although this classification has been evaluated extensively in the intensive care unit setting,19 it has not been evaluated in patients with HF. Additionally, this classification of acute kidney injury will miss the small creatinine changes frequently observed during ADHF treatment, which have repeatedly been associated with a worse patient outcome.20,21 The pathophysiologic mechanisms underlying WRF in ADHF appear to be multifactorial and have not been well defined. Overall, an imbalance among the failing heart, the neurohumoral system, and the inflammatory responses has been associated with the occurrence of WRF.22 Additionally, hemodynamic factors, including the adequacy of renal perfusion and the degree of venous congestion, as well as drug nephrotoxicity, appear to contribute to the development of WRF. Hence, the multiple pathologic features underlying kidney injury in ADHF complicate the prediction,
tailored therapy, and avoidance of WRF. This has been clearly demonstrated by our results, with areas under the ROC curve of about 0.7 for the 2 available risk prediction scores. Both risk scores failed to adequately identify patients developing WRF. Using the Basel risk score, 35% of WRF cases developed in the intermediate-risk group, and approximately 50% of WRF cases developed in the Forman low-risk groups. Furthermore, in both risk scores, ⬍40% of high-risk patients eventually developed WRF. Similarly, even invasive hemodynamic measurements were recently shown to not predict the occurrence of WRF in patients with advanced HF.23 However, the novel Basel risk score was able to identify a patient subgroup at a minimal risk of developing WRF (rule out) and, consequently, a more favorable outcome. Only 2% of patients without CKD, who had not received diuretic treatment, and who had had normal bicarbonate levels developed WRF. Our external validation of the Forman risk score found the incidence of WRF in the lower risk groups 0 to 2 to equal the results described by Forman et al8 (incidence of WRF in group 0, 14% vs 13%; incidence of WRF in groups 1 to 2, 19% vs 16%). However, our analysis could not validate the high incidence of WRF cases in the high-risk groups of 4 and greater. We believe that substantial differences in the baseline characteristics of our patient cohort (older patient age, greater prevalence of previously known CKD [42% vs 23%], lower prevalence of diabetes [31% vs 40%], and previously known HF [55% vs 63%]), as well as the more extensive baseline antihypertensive therapy (systolic blood pressure ⬎160 mm Hg in 25% vs 32%), accounted for the limited prognostic accuracy of the Forman risk score. However, these baseline differences were in line with the slightly varying co-morbidities observed in different large-scale HF registries.24,25 Novel renal biomarkers have recently received growing attention26 and might in the future simplify the prediction of WRF. Neutrophil gelatinase-associated lipocalin appears to be the most promising candidate biomarker. Various studies have demonstrated that neutrophil gelatinase-associated lipocalin represents a novel specific biomarker for the early identification of acute kidney injury after pediatric cardiac surgery27 and contrast agent administration.28,29 However, the potential of neutrophil gelatinase-associated lipocalin in ADHF has not yet been evaluated and might be limited by the multiple underlying pathologic findings and the potentially serial insults.22 The potential limitations of the present study merit consideration. First, only Swiss centers participated in the present study. However, because the baseline characteristics, treatment, and in-hospital mortality rates were similar to those observed in large registry studies, we consider our results representative. We believe that instead of replicating our data in a larger study, future research efforts should be directed toward finding suitable biomarkers for the prediction of WRF in patients presenting to the emergency department. Second, the timing of creatinine sampling during the index hospitalization was left at the discretion of the treating physicians. Thus, we were unable to comment on the time of occurrence of WRF. Previous studies, however,
Heart Failure/Clinical Prediction of WRF in ADHF
have continuously described WRF to predominantly occur during the first week of ADHF hospitalization.8,30 1. Ronco C, Haapio M, House AA, Anavekar N, Bellomo R. Cardiorenal syndrome. J Am Coll Cardiol 2008;52:1527–1539. 2. Breidthardt T, Mebazaa A, Mueller CE. Predicting progression in nondiabetic kidney disease: the importance of cardiorenal interactions. Kidney Int 2009;75:253–255. 3. Ronco C, McCullough P, Anker SD, Anand I, Aspromonte N, Bagshaw SM, Bellomo R, Berl T, Bobek I, Cruz DN, Daliento L, Davenport A, Haapio M, Hillege H, House AA, Katz N, Maisel A, Mankad S, Zanco P, Mebazaa A, Palazzuoli A, Ronco F, Shaw A, Sheinfeld G, Soni S, Vescovo G, Zamperetti N, Ponikowski P. Cardio-renal syndromes: report from the consensus conference of the acute dialysis quality initiative. Eur Heart J. 2010;31:703–711. 4. Damman K, Navis G, Voors AA, Asselbergs FW, Smilde TD, Cleland JG, van Veldhuisen DJ, Hillege HL. Worsening renal function and prognosis in heart failure: systematic review and meta-analysis. J Card Fail 2007;13:599 – 608. 5. Khan NA, Ma I, Thompson CR, Humphries K, Salem DN, Sarnak MJ, Levin A. Kidney function and mortality among patients with left ventricular systolic dysfunction. J Am Soc Nephrol 2006;17:244 –253. 6. Krumholz HM, Chen YT, Vaccarino V, Wang Y, Radford MJ, Bradford WD, Horwitz RI. Correlates and impact on outcomes of worsening renal function in patients ⬎ or ⫽ 65 years of age with heart failure. Am J Cardiol 2000;85:1110 –1113. 7. de Silva R, Nikitin NP, Witte KK, Rigby AS, Goode K, Bhandari S, Clark AL, Cleland JG. Incidence of renal dysfunction over 6 months in patients with chronic heart failure due to left ventricular systolic dysfunction: contributing factors and relationship to prognosis. Eur Heart J 2006;27:569 –581. 8. Forman DE, Butler J, Wang Y, Abraham WT, O’Connor CM, Gottlieb SS, Loh E, Massie BM, Rich MW, Stevenson LW, Young JB, Krumholz HM. Incidence, predictors at admission, and impact of worsening renal function among patients hospitalized with heart failure. J Am Coll Cardiol 2004;43:61– 67. 9. Dickstein K. ESC guidelines for the diagnosis and treatment of acute and chronic heart failure 2008: application of natriuretic peptides: reply. Eur Heart J Epub 2008 Dec 24. 10. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002;39:S1–266. 11. van Kimmenade RR, Januzzi JL Jr, Baggish AL, Lainchbury JG, Bayes-Genis A, Richards AM, Pinto YM. Amino-terminal pro-brain natriuretic peptide, renal function, and outcomes in acute heart failure: redefining the cardiorenal interaction? J Am Coll Cardiol 2006;48: 1621–1627. 12. Smith GL, Lichtman JH, Bracken MB, Shlipak MG, Phillips CO, DiCapua P, Krumholz HM. Renal impairment and outcomes in heart failure: systematic review and meta-analysis. J Am Coll Cardiol 2006; 47:1987–1996. 13. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D; Modification of Diet in Renal Disease Study Group. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;130:461– 470. 14. Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney function—measured and estimated glomerular filtration rate. N Engl J Med 2006;354:2473–2483.
735
15. O’Meara E, Chong KS, Gardner RS, Jardine AG, Neilly JB, McDonagh TA. The modification of diet in renal disease (MDRD) equations provide valid estimations of glomerular filtration rates in patients with advanced heart failure. Eur J Heart Fail 2006;8:63– 67. 16. Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, Levin A. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 2007;11:R31. 17. Coca SG, Yusuf B, Shlipak MG, Garg AX, Parikh CR. Long-term risk of mortality and other adverse outcomes after acute kidney injury: a systematic review and meta-analysis. Am J Kidney Dis 2009;53:961– 973. 18. Kellum JA, Bellomo R, Ronco C. The concept of acute kidney injury and the RIFLE criteria. Contrib Nephrol 2007;156:10 –16. 19. Ricci Z, Cruz D, Ronco C. The RIFLE criteria and mortality in acute kidney injury: A systematic review. Kidney Int 2008;73:538 –546. 20. Coca SG, Peixoto AJ, Garg AX, Krumholz HM, Parikh CR. The prognostic importance of a small acute decrement in kidney function in hospitalized patients: a systematic review and meta-analysis. Am J Kidney Dis 2007;50:712–720. 21. Gottlieb SS, Abraham W, Butler J, Forman DE, Loh E, Massie BM, O’Connor CM, Rich MW, Stevenson LW, Young J, Krumholz HM. The prognostic importance of different definitions of worsening renal function in congestive heart failure. J Card Fail 2002;8:136 –141. 22. Tang WH, Mullens W. Cardiorenal syndrome in decompensated heart failure. Heart. 2010;96:255–260. 23. Weinfeld MS, Chertow GM, Stevenson LW. Aggravated renal dysfunction during intensive therapy for advanced chronic heart failure. Am Heart J 1999;138:285–290. 24. De Luca L, Fonarow GC, Adams KF Jr, Mebazaa A, Tavazzi L, Swedberg K, Gheorghiade M. Acute heart failure syndromes: clinical scenarios and pathophysiologic targets for therapy. Heart Fail Rev 2007;12:97–104. 25. Blair JE, Zannad F, Konstam MA, Cook T, Traver B, Burnett JC Jr, Grinfeld L, Krasa H, Maggioni AP, Orlandi C, Swedberg K, Udelson JE, Zimmer C, Gheorghiade M. Continental differences in clinical characteristics, management, and outcomes in patients hospitalized with worsening heart failure: results from the EVEREST (Efficacy of Vasopressin Antagonism in Heart Failure: Outcome Study with Tolvaptan) program. J Am Coll Cardiol 2008;52:1640 –1648. 26. Coca SG, Yalavarthy R, Concato J, Parikh CR. Biomarkers for the diagnosis and risk stratification of acute kidney injury: a systematic review. Kidney Int 2008;73:1008 –1016. 27. Mishra J, Dent C, Tarabishi R, Mitsnefes MM, Ma Q, Kelly C, Ruff SM, Zahedi K, Shao M, Bean J, Mori K, Barasch J, Devarajan P. Neutrophil gelatinase-associated lipocalin (NGAL) as a biomarker for acute renal injury after cardiac surgery. Lancet 2005;365:1231–1238. 28. Bachorzewska-Gajewska H, Malyszko J, Sitniewska E, Malyszko JS, Dobrzycki S. Neutrophil-gelatinase-associated lipocalin and renal function after percutaneous coronary interventions. Am J Nephrol 2006;26:287–292. 29. Hirsch R, Dent C, Pfriem H, Allen J, Beekman RH III, Ma Q, Dastrala S, Bennett M, Mitsnefes M, Devarajan P. NGAL is an early predictive biomarker of contrast-induced nephropathy in children. Pediatr Nephrol 2007;22:2089 –2095. 30. Aronson D, Burger AJ. The relationship between transient and persistent worsening renal function and mortality in patients with acute decompensated heart failure. J Card Fail. 2010;16:541–547.
Value of the Surface Electrocardiogram in Detecting Right Ventricular Dilatation in the Presence of Left Bundle Branch Block Rutger J. Van Bommel, MDa, Nina Ajmone Marsan, MDa, Victoria Delgado, MDa, Eva P.M. van Rijnsoever, MSca, Martin J. Schalij, MD, PhDa, Jeroen J. Bax, MD, PhDa,*, and Hein J. Wellens, MD, PhDb Approximately 20% of patients with heart failure have left bundle branch block (LBBB) on surface electrocardiogram (ECG). In this group of patients, detection of right ventricular (RV) dilatation on standard ECG can be of clinical relevance because RV enlargement is an important prognostic marker. Consequently, the aim of this study was to evaluate diagnostic accuracy for several electrocardiographic criteria in determining significant RV dilatation in these patients. Standard 12-lead ECGs were obtained in 173 patients with heart failure and known LBBB. From the ECG, 3 criteria for RV dilatation were defined: presence of terminal positivity in lead aVR (late R wave in lead aVR), low voltage (<0.6 mV) in all extremity leads, and an R/S ratio <1 in lead V5. In addition, all patients underwent comprehensive echocardiographic evaluation including assessment of RV dimensions. Measurements were performed blinded to electrocardiographic results. Significant RV dilatation was defined as an RV base-to-apex length >86 mm or an RV diastolic area >33 cm2. Eighty-six patients (50%) had a late R wave in lead aVR, 36 patients (21%) had low voltage in extremity leads, and 67 patients (39%) had an R/S ratio <1 in lead V5. An RV base-to-apex length >86 mm was present in 67 patients (39%), and 62 patients (36%) had an RV diastolic area >33 cm2. Any combination of 2 to 3 positive criteria could predict an RV base-to-apex length >86 mm with a positive predictive value of 89% and a negative predictive value of 88%. Similarly, an RV diastolic area >33 cm2 was predicted with a positive predictive value of 80% and a negative predictive value of 88%. In conclusion, combining 2 to 3 distinct electrocardiographic criteria allows for accurate detection of RV dilatation in patients with heart failure and LBBB. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:736 –740) Approximately 20% of patients with heart failure have left bundle branch block (LBBB) on surface electrocardiogram (ECG).1,2 In patients with LBBB, ventricular activation occurs abnormally, affecting the application of conventional electrocardiographic rules for diagnosing abnormalities in blood supply, scar location, and ventricular hypertrophy because these are based on normal conduction over the bundle branch system. Attempts have been made to recognize those abnormalities in the presence of LBBB. More specifically, the possibility of detecting right ventricular (RV) dilatation using a standard ECG can be of clinical relevance because RV enlargement is a prognostic marker in these patients.3 Consequently, the aim of this study was to evaluate the diagnostic accuracy for several electrocardiographic criteria in determining RV dilatation in patients with heart failure and LBBB.
a Department of Cardiology, Leiden University Medical Center, Leiden, The Netherlands; bCardiovascular Research Institute, Maastricht, The Netherlands. Manuscript received September 17, 2010; revised manuscript received and accepted October 26, 2010. *Corresponding author: Tel: 31-71-526-2020; fax: 31-71-526-6809. E-mail address:
[email protected] (J.J. Bax).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.051
Methods In total 173 patients with heart failure were selected according to the following criteria: New York Heart Association functional classes III and IV and an LBBB configuration on surface ECG. Cause of heart failure was considered ischemic in the presence of significant coronary artery disease (ⱖ50% stenosis in ⱖ1 major coronary artery) and/or a history of myocardial infarction or previous revascularization. In all patients a 12-lead ECG was obtained and extensive echocardiographic evaluation including assessment of RV size and function was performed. All echocardiographic measurements were performed blinded to electrocardiographic results. From the 12-lead ECG, 3 criteria for RV dilatation were defined. Diagnostic accuracy of these electrocardiographic criteria for detecting RV dilatation was evaluated. A standard supine artifact free 12-lead ECG (filter range 0.15 to 100 Hz, AC filter 60 Hz, 25 mm/s, 10 mm/mV) showing a supraventricular rhythm (sinus or atrial fibrillation) with an LBBB pattern was available for analysis. The interval between ECG and echocardiogram was ⬍10 days in all patients. The ECG was analyzed manually and the following electrocardiographic parameters were obtained: frontal plane axis, ventricular rate, QRS width, frontal plane axis of the full and second half of the QRS complex, voltage of www.ajconline.org
Heart Failure/ECG to Predict RV Dilatation in Presence of LBBB
737
chotomous data are presented as numbers and percentages. Comparison of data between patient groups was performed using independent-samples t test for continuous data. Fisher’s exact tests or chi-square tests were used as appropriate to compare dichotomous data. Comparisons between ⬎2 patient groups were performed using 1-way analysis of variance with Bonferroni post hoc testing. All analyses were performed with SPSS 16.0 for Windows (SPSS, Inc., Chicago, Illinois). All statistical tests were 2-sided. A p value ⬍0.05 was considered statistically significant. Results
Figure 1. Twelve-lead electrocardiogram showing 3 electrocardiographic criteria evaluated to detect right ventricular dilatation.
extremity and precordial leads, T-wave axis in frontal plane, QRS-T angle in frontal plane, presence of terminal positivity in lead aVR, presence of an R/S ratio ⬍1 in lead V5, and presence of low QRS voltage (⬍0.6 mV) in all 6 extremity leads with a normal or increased voltage in precordial leads. For reasons outlined in the Discussion, 3 electrocardiographic criteria were selected as indicators for RV dilatation: presence of terminal positivity in lead aVR, low QRS voltage (⬍0.6 mV) in all extremity leads, and an R/S ratio ⬍1 in lead V5 (Figure 1). All patients underwent echocardiography in the left lateral decubitus position. Imaging was performed using a commercially available echocardiographic system (VIVID 7, General Electric Vingmed Ultrasound, Milwaukee, Wisconsin). Images were obtained using a 3.5-MHz transducer at a depth of 16 cm in the parasternal and apical (2- and 4-chamber) views. Standard 2-dimensional and color Doppler data triggered to the QRS complex were saved in cineloop format and analyzed off-line with commercial software (EchoPac 108.1.5, General Electric Vingmed Ultrasound). Left ventricular (LV) end-diastolic and LV end-systolic volumes were determined from conventional apical 2- and 4-chamber views and LV ejection fraction was calculated using the biplane Simpson technique.4 Severity of tricuspid regurgitation and mitral regurgitation was assessed according to current guidelines.5,6 For quantification of RV size, the apical 4-chamber view was used. Special care was taken to obtain a true nonforeshortened apical 4-chamber view oriented to obtain maximum RV dimension. The RV base-to-apex length and RV diastolic area (Figure 2) were assessed as described previously.7 Significant RV dilatation was defined as an RV base-to-apex length ⱖ86 mm or an RV diastolic area ⱖ33 cm2 (moderate to severe RV dilatation).7,8 These specific measurements were selected because these were expected to best identify RV dilatation defined by the chosen electrocardiographic criteria. For assessment of RV function, the tricuspid annular plane systolic excursion was measured.9 Pulmonary artery systolic pressure was derived from the RV to right atrial pressure gradient or tricuspid regurgitation jet gradient and calculated with the modified Bernoulli equation.10 Continuous data are presented as mean ⫾ SD and di-
Baseline characteristics of the patient population are presented in Table 1. Most patients were men (64%) and cause of heart failure was ischemic cardiomyopathy in 83 patients (48%). No differences were observed between patients with ischemic heart failure and those with nonischemic heart failure, except that ischemic patients were older (68 ⫾ 9 vs 63 ⫾ 10 years, p ⬍0.001) and more frequently men (76% vs 53%, p ⫽ 0.002). Mean heart rate during electrocardiographic registration was 71 ⫾ 13 beats/min and mean QRS duration was 178 ⫾ 16 ms. Other electrocardiographic parameters are listed in Table 1. When assessing the predefined electrocardiographic criteria, the following observations were made: (1) 86 patients (50%) patients had a late R wave in lead aVR, (2) 36 patients (21%) had low voltage in extremity leads, and (3) 67 patients (39%) had an R/S ratio ⬍1 in lead V5. There were 51 patients (29%) without any positive criterion for RV dilatation, 61 patients (35%) had 1 positive criterion, 55 patients (32%) had 2 positive criteria, and 6 patients (3%) had all 3 positive criteria. Mean values for all echocardiographic findings are listed in Table 1. Significant RV dilatation defined as an RV base-to-apex length ⱖ86 mm was present in 67 patients (39%), and an RV diastolic area ⱖ33 cm2 was noted in 62 patients (36%). Of note, a concordance between the 2 definitions was found in 160 patients (92%). Mean values of RV size and function measurements in patients with 0 or 1 criterion 2 or 3 positive electrocardiographic criteria are listed in Table 2. There were significant differences between groups in all tested RV size and RV function measurements. In particular, after post hoc testing, patients with 2 or 3 positive electrocardiographic criteria had a longer RV baseto-apex length, larger RV diastolic area, higher tricuspid regurgitation grade, and higher pulmonary artery systolic pressure (p ⬍0.05 for all tests) compared to patients with 0 positive electrocardiographic criterion (Table 2). Subsequently, the diagnostic accuracy of the 3 electrocardiographic criteria for detecting RV dilatation was tested. An RV base-to-apex length ⱖ86 mm was observed in 62% of patients with a late R wave in lead aVR, and an RV diastolic area ⱖ33 cm2 was observed in 57% of these patients (Table 3). Conversely, an RV base-to-apex length ⱖ86 mm was observed in 61% of patients with low voltage in all extremity leads, and an RV diastolic area ⱖ33 cm2 was observed in 56% of patients with low voltage in all extremity leads (Table 3). An RV base-to-apex length ⱖ86 mm was observed in 69% of patients with an R/S ratio ⱕ1 in lead V5, and an RV diastolic area ⱖ33 cm2 was observed in 63% of these patients (Table 3). Because no single elec-
738
The American Journal of Cardiology (www.ajconline.org)
Figure 2. Examples of echocardiographic measurements to evaluate right ventricular dilatation according to (A) right ventricular base-to-apex length (87 mm in this example) and (B) right ventricular diastolic area (34.2 cm2 in this example). LA ⫽ left atrium; LV ⫽ left ventricle; RA ⫽ right atrium; RV ⫽ right ventricle. Table 1 Patient characteristics (n ⫽ 173) Clinical Parameters Age (years) Men/women New York Heart Association functional class III IV Cause of heart failure Ischemic Nonischemic Medications Angiotensin-converting enzyme inhibitors Diuretics  Blockers Spironolactone Digoxin Electrocardiographic parameters Heart rate (beats/min) QRS duration (ms) QRS axis (°) QRS axis second half (°) QRS-T angle (°) Echocardiographic parameters Left ventricular end-diastolic volume (ml) Left ventricular end-systolic volume (ml) Left ventricular ejection fraction (%) Mitral regurgitation (grade) Right ventricular base-to-apex length (mm) Right ventricular diastolic area (cm2) Tricuspid annular plane systolic excursion (mm) Pulmonary artery systolic pressure (mm Hg) Tricuspid regurgitation (grade)
65 ⫾ 10 111/62 162 (94%) 11 (6%) 83 (48%) 90 (52%) 151 (87%) 158 (91%) 123 (71%) 83 (48%) 25 (14%) 71 ⫾ 13 178 ⫾ 16 ⫺22 ⫾ 39 ⫺50 ⫾ 43 136 ⫾ 43 234 ⫾ 88 180 ⫾ 77 24 ⫾ 8 1.7 ⫾ 1.0 82 ⫾ 8 28 ⫾ 7 18 ⫾ 3 34 ⫾ 9 1.5 ⫾ 1.0
trocardiographic criterion could predict significant RV dilatation defined as an RV base-to-apex length ⱖ86 mm or an RV diastolic area ⱖ33 cm2 with a positive predictive value ⬎69% and patients with 2 or 3 positive criteria had a longer RV base-to-apex length and larger RV diastolic area than patients with 0 or 1 positive criterion (Table 2), separate electrocardiographic criteria were combined and 2 subgroups were constituted: patients with 0 to 1 positive elec-
trocardiographic criterion and patients with 2 to 3 positive criteria. An RV base-to-apex length ⱖ86 mm was observed in only 13 of 112 patients (12%) with 0 to 1 positive criterion, and an RV diastolic area ⱖ33 cm2 was present in 13 of 112 patients (12%) with 0 to 1 positive criterion. Consequently, any combination of 2 to 3 positive criteria was able to predict an RV base-to-apex length ⱖ86 mm with a positive predictive value of 89% and a negative predictive value of 88% (Table 3). Similarly, an RV diastolic area ⱖ33 cm2 could be predicted with a positive predictive value of 80% and a negative predictive value of 88% (Table 3). Discussion Current experience with cardiac resynchronization therapy (CRT) in patients with heart failure indicates that the best results are obtained in patients with LBBB with a QRS width ⬎140 ms.11 This electrocardiographic pattern is found in approximately 20% of patients with heart failure.1,2 Nevertheless, about 30% to 40% of those patients do not respond favorably to CRT. Several patient and device-related factors play a role in nonresponders.12 Lack of improvement can be related to patient and/or pacing characteristics. Examples of patient-related causes are a too large left ventricle or too much scar, class IV heart failure, renal dysfunction, and (supra-) ventricular arrhythmias. Pacingrelated causes include suboptimal LV pacing site, pacing in a scarred area of the left ventricle, suboptimal RV pacing site, no optimal LV–RV activation interval, and no optimal atrioventricular activation interval. In patients with heart failure and LBBB, the value of the ECG has been studied in relation to outcome of CRT. Baseline QRS duration was found to be helpful in several, but not all, reported studies.13–16 Recently, Sweeney et al,17 when analyzing LV activation on baseline 12-lead ECG, described 2 findings able to predict CRT outcome. A favorable outcome was related to LV activation time and an unfavorable outcome when LV scar was present as determined by the Selvester LV scar score.18 It is well known that in a patient with heart
Heart Failure/ECG to Predict RV Dilatation in Presence of LBBB
739
Table 2 Differences in right ventricular dimensions and function in patients with 0, 1, 2, or 3 positive electrocardiographic criteria 0 Positive (n ⫽ 51)
1 Positive (n ⫽ 61)
2 Positive (n ⫽ 55)
3 Positive (n ⫽ 6)
p Value
78 ⫾ 6 25 ⫾ 6 18 ⫾ 3 30 ⫾ 8 1.1 ⫾ 0.8
78 ⫾ 6 25 ⫾ 7 19 ⫾ 2 30 ⫾ 9 1.4 ⫾ 0.9
90 ⫾ 5* 34 ⫾ 5* 18 ⫾ 4 38 ⫾ 7* 2.0 ⫾ 0.9*
92 ⫾ 6* 36 ⫾ 3* 16 ⫾ 4 46 ⫾ 2* 2.2 ⫾ 1.3*
⬍0.001 ⬍0.001 0.034 ⬍0.001 0.002
Right ventricular base-to-apex length (mm) Right ventricular diastolic area (cm2) Tricuspid annular plane systolic excursion (mm) Pulmonary artery systolic pressure (mm Hg) Tricuspid regurgitation (grade)
Provided p values are for trend (least squares regression) between subgroups. * p ⬍0.05 versus 0 positive electrocardiographic criterion. Table 3 Diagnostic accuracy of presence (or combination) of electrocardiographic criteria to detect right ventricular base-to-apex length ⱖ86 mm and right ventricular diastolic area ⱖ33 cm2 Positive Predictive Value Right ventricular base-to-apex length ⱖ86 mm Late R wave in lead aVR Low voltage in all extremity leads R/S ratio ⬍1 in lead V5 2–3 positive criteria Right ventricular diastolic area ⱖ33 cm2 Late R wave in lead aVR Low voltage in all extremity leads R/S ratio ⬍1 in lead V5 2–3 positive criteria
Negative Predictive Value
Sensitivity
Specificity
62%
84%
79%
69%
61%
67%
33%
87%
69%
80%
69%
80%
89%
88%
81%
93%
57%
85%
79%
67%
56%
69%
33%
86%
63%
81%
68%
78%
80%
88%
79%
89%
failure and LV disease additional RV dysfunction worsens prognosis.3 The aim of our study therefore was to investigate the possibility of using standard 12-lead ECG to detect RV dilation in patients with heart failure, LV systolic dysfunction, and complete LBBB. Three electrocardiographic findings were considered of possible value: (1) terminal positivity of QRS voltage in lead aVR, (2) an R/S ratio ⬍1 in lead V5, and (3) low QRS voltage in extremity leads with normal or increased QRS voltage in precordial leads. Positivity in the terminal portion of QRS voltage in lead aVR suggests that the last part of ventricular activation occurs in the direction of the right shoulder, suggesting a longer delay in activation of the basal part of the right than of the left ventricle. Clockwise rotation in precordial leads as manifested by an R/S ⬍1 in lead V5 suggests that the right ventricle extends farther to the left in the horizontal plane. As pointed out by Goldberger et al19 many years ago, voltage in bipolar extremity leads decreases when intracardiac fluid volume increases. This is not the case with unipolar precordial leads. When values of the 3 criteria were examined separately, none reached a high predictive value
for RV enlargement. However, presence of any combination of 2 to 3 positive criteria could predict an RV base-to-apex length ⱖ86 mm and an RV diastolic area ⱖ33 cm2 with satisfactory predictive values. 1. Baldasseroni S, Opasich C, Gorini M, Lucci D, Marchionni N, Marini M, Campana C, Perini G, Deorsola A, Masotti G, Tavazzi L, Maggioni AP. Left bundle-branch block is associated with increased 1-year sudden and total mortality rate in 5517 outpatients with congestive heart failure: a report from the Italian network on congestive heart failure. Am Heart J 2002;143:398 – 405. 2. Sandhu R, Bahler RC. Prevalence of QRS prolongation in a community hospital cohort of patients with heart failure and its relation to left ventricular systolic dysfunction. Am J Cardiol 2004;93:244 –246. 3. Oakley C. Importance of right ventricular function in congestive heart failure. Am J Cardiol 1988;62(suppl):14A–19A. 4. Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H, Gutgesell H, Reichek N, Sahn D, Schnittger I. Recommendations for quantitation of the left ventricle by twodimensional echocardiography. American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr 1989; 2:358 –367. 5. Zoghbi WA, Enriquez-Sarano M, Foster E, Grayburn PA, Kraft CD, Levine RA, Nihoyannopoulos P, Otto CM, Quinones MA, Rakowski H, Stewart WJ, Waggoner A, Weissman NJ. Recommendations for evaluation of the severity of native valvular regurgitation with twodimensional and Doppler echocardiography. J Am Soc Echocardiogr 2003;16:777– 802. 6. Bonow RO, Carabello BA, Chatterjee K, De LA Jr, Faxon DP, Freed MD, Gaasch WH, Lytle BW, Nishimura RA, O’Gara PT, O’Rourke RA, Otto CM, Shah PM, Shanewise JS. 2008 Focused update incorporated into the ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the 1998 guidelines for the management of patients with valvular heart disease). Endorsed by the Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol 2008;52(suppl):e1– e142. 7. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, Picard MH, Roman MJ, Seward J, Shanewise J, Solomon S, Spencer KT, St John SM, Stewart W. Recommendations for chamber quantification. Eur J Echocardiogr 2006;7:79 –108. 8. Foale R, Nihoyannopoulos P, McKenna W, Kleinebenne A, Nadazdin A, Rowland E, Smith G. Echocardiographic measurement of the normal adult right ventricle. Br Heart J 1986;56:33– 44. 9. Kaul S, Tei C, Hopkins JM, Shah PM. Assessment of right ventricular function using two-dimensional echocardiography. Am Heart J 1984; 107:526 –531. 10. Yock PG, Popp RL. Noninvasive estimation of right ventricular systolic pressure by Doppler ultrasound in patients with tricuspid regurgitation. Circulation 1984;70:657– 662. 11. Hawkins NM, Petrie MC, MacDonald MR, Hogg KJ, McMurray JJ. Selecting patients for cardiac resynchronization therapy: electrical or mechanical dyssynchrony? Eur Heart J 2006;27:1270 –1281.
740
The American Journal of Cardiology (www.ajconline.org)
12. Wellens HJ. Cardiac arrhythmias: the quest for a cure: a historical perspective. J Am Coll Cardiol 2004;44:1155–1163. 13. Lecoq G, Leclercq C, Leray E, Crocq C, Alonso C, De PC, Mabo P, Daubert C. Clinical and electrocardiographic predictors of a positive response to cardiac resynchronization therapy in advanced heart failure. Eur Heart J 2005;26:1094 –1100. 14. Mollema SA, Bleeker GB, van der Wall EE, Schalij MJ, Bax JJ. Usefulness of QRS duration to predict response to cardiac resynchronization therapy in patients with end-stage heart failure. Am J Cardiol 2007;100:1665–1670. 15. Adelstein EC, Saba S. Usefulness of baseline electrocardiographic QRS complex pattern to predict response to cardiac resynchronization. Am J Cardiol 2009;103:238 –242. 16. Gervais R, Leclercq C, Shankar A, Jacobs S, Eiskjaer H, Johannessen A, Freemantle N, Cleland JG, Tavazzi L, Daubert C. Surface electrocardio-
gram to predict outcome in candidates for cardiac resynchronization therapy: a sub-analysis of the CARE-HF trial. Eur J Heart Fail 2009; 11:699 –705. 17. Sweeney MO, van Bommel RJ, Schalij MJ, Borleffs CJ, Hellkamp AS, Bax JJ. Analysis of ventricular activation using surface electrocardiography to predict left ventricular reverse volumetric remodeling during cardiac resynchronization therapy. Circulation 2010;121: 626 – 634. 18. Strauss DG, Selvester RH. The QRS complex—a biomarker that “images” the heart: QRS scores to quantify myocardial scar in the presence of normal and abnormal ventricular conduction. J Electrocardiol 2009;42:85–96. 19. Goldberger AL, Dresselhaus T, Bhargava V. Dilated cardiomyopathy: utility of the transverse: frontal plane QRS voltage ratio. J Electrocardiol 1985;18:35– 40.
Relation of Aortic Valve Weight to Severity of Aortic Stenosis Renato Razzolini, MDa,*, Susy Longhi, MDa, Giuseppe Tarantini, MD, PhDa, Stefania Rizzo, MDb, Massimo Napodano, MDa, Elena Abate, MDa, Chiara Fraccaro, MDa, Gaetano Thiene, MDb, Sabino Iliceto, MDa, Gino Gerosa, MDa, and Cristina Basso, MD, PhDb The purpose of this study was to analyze the relation of aortic valve weight to transvalvular gradient and area, with special regard to valve anatomy, size of calcific deposits, gender, and body size. Two hundred forty-two surgically excised stenotic aortic valves of patients (139 men, mean age 72 ⴞ 9 years) who had undergone preoperative cardiac catheterization and echocardiography were weighed and examined with respect to number of cusps (tricuspid vs bicuspid), size of calcium deposits (microaggregates vs nodular macroaggregates), and presence of cholesterol clefts. The relation among valve weight, gradient, and area was studied. Transvalvular gradient was independent of gender or valve anatomy and was linearly correlated with valve weight absolutely (r ⴝ 0.33, p <0.01) or normalized by body surface area (r ⴝ 0.40, p <0.01). No correlation was evident between valve area and weight. Calcium macroaggregates were mainly present in men (51%) and in bicuspid valves (67%) and were seen to be strong determinants of valve weight (2.84 ⴞ 1.03 g with macroaggregates vs 1.63 ⴞ 0.56 g with microaggregates, p <0.001) but not of transvalvular gradient. Calcium microaggregates characterized tricuspid valves (62%), where transvalvular gradient was determined by valve weight (p ⴝ 0.0026). In conclusion, the heavier the valve, the less frequent were hypercholesterolemia, valve cholesterol clefts, hypertension, and diabetes mellitus. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107: 741–746) Aortic stenosis in Western countries is mostly owing to dystrophic calcification of bicuspid or tricuspid aortic valves, which are characterized by increased leaflet thickness and calcium deposition, without commissural fusion.1 Recently, it has been demonstrated that the weight of the aortic valve, excised during surgical replacement, is linearly related to transvalvular gradient and not to valve area.2– 8 The aim of our study was to ascertain whether this relation holds true irrespective of the anatomy of the valve (bileaflets vs trileaflets), gender, and body size. Moreover, we searched for a correlation between valve weight and hypercholesterolemia, valvular cholesterol, hypertension, and diabetes mellitus. Methods In the University of Padua (Padua, Italy) all aortic valves excised during valve replacement are sent to the pathology department. Two of the authors (S.R., C.B.) weighed and evaluated pathologically 269 stenotic aortic valves of patients operated because of aortic valve stenosis. Surgical pathologic evaluation aimed to determine number of cusps (trileaflets vs bileaflets), size of calcium deposits (diffuse microaggregates, ⱕ4 mm, vs nodular macroaggregates, ⬎4 mm), and cause of disease (dystrophic calcification vs rheu-
Departments of aCardiac, Thoracic, and Vascular Sciences and bMedical Diagnostic Sciences and Special Therapies, University of Padua, Padua, Italy. Manuscript received June 21, 2010; revised manuscript received and accepted October 11, 2010. *Corresponding author: Tel: 39-049-821-2324; fax: 39-049-876-1764. E-mail address:
[email protected] (R. Razzolini). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.052
Table 1 Patient characteristics (n ⫽ 242) Variable Men Age (years), mean ⫾ SD Bileaflet valves Trileaflet valves Aortic mean gradient (mm Hg), mean ⫾ SD Aortic valve area (cm2), mean ⫾ SD Valve weight (g), mean ⫾ SD Normalized valve weight (g/m2), mean ⫾ SD Microscopic calcium deposits Macroscopic calcium deposits Valvular cholesterol deposits Systolic/diastolic pressure (mm Hg), mean ⫾ SD Diabetes mellitus Atherosclerosis Coronary Carotid Peripheral
Values 139 (57%) 72 ⫾ 9 62 (26%) 180 (74%) 57 ⫾ 24 0.74 ⫾ 0.21 2.19 ⫾ 1.01 1.2 ⫾ 0.58 131 (54%) 111 (46%) 102 (42%) 146 ⫾ 27/88 ⫾ 18 56 (23%) 133 (55%) 61 (25%) 44 (18%)
Aortic gradient and aortic valve area measured by cardiac catheterization.
matic). Moreover, histologic study addressed the presence of cholesterol cleft deposits within leaflets. Our study focused on 242 valves with dystrophic calcification, among which 180 were trileaflets and 62 were bileaflets. All patients had undergone cardiac catheterization and echocardiography within 1 month before surgery. Catheter gradients were measured as peak-to-peak gradient during pull-back from the left ventricle. Aortic valve area was calculated with the Gorlin equation. Echocardiographic gradients were measured with continuous Doppler recordings and valve areas were computed with a continuity equation. www.ajconline.org
742
The American Journal of Cardiology (www.ajconline.org)
Table 2 Distribution of age, valve weight, gradient, and area with regard to gender, valve anatomy, and size of calcific deposits
Valves Men Women Trileaflets Bileaflets Macroaggregates Microaggregates
Number
Age (years), Mean ⫾ SD
Weight (g), Mean (range)
Gradient (mm Hg), Mean (range)
Peak Gradient (mm Hg), Mean (range)
Area (cm2), Mean ⫾ SD
242 139 103 180 62 111 131
72 ⫾ 9 70 ⫾ 10* 75 ⫾ 7* 75 ⫾ 6* 64 ⫾ 12* 70 ⫾ 10 74 ⫾ 7
2.19 (0.56–6.5) 2.5 (0.79–6.5)* 1.7 (0.56–3.6)* 1.98 (0.56–4.9)* 2.77 (1.1–6.5)* 2.8 (1–6.5) 1.6 (0.56–3.4)
57 (28–89) 56 (10–110) 58 (20–126) 56 (20–121) 60 (10–128) 65 (10–128) 50 (20–115)
77 (23–194) 75 (23–155)† 82 (32–194)† 79 (34–194) 75 (23–136) 85 (36–194) 71 (23–130)
0.74 ⫾ 0.21 0.77 ⫾ 0.20† 0.70 ⫾ 0.22† 0.75 ⫾ 0.20 0.70 ⫾ 0.25 0.69 ⫾ 0.2 0.7 ⫾ 0.2
* p ⬍0.001 for men versus women and trileaflets versus bileaflets. † p ⬍0.05 for men versus women.
Figure 1. Correlation between aortic valve gradient measured during cardiac catheterization and valve weight for (A) absolute weight (p ⫽ 0.021 for interaction between gender and valve weight) and (B) valve weight indexed by body surface area (p ⫽ 0.028 for interaction between gender and indexed valve weight). Correlations are shown separately for women (p ⬍0.001) (squares) and men (p ⬍0.001) (circles).
Figure 2. Correlation between aortic valve area and valve weight.
Valvular Heart Disease/Aortic Valve Weight
743
Figure 3. Aortic stenosis owing to dystrophic calcification of a tricuspid aortic valve with calcium macroaggregates. (A) Gross view of surgically resected aortic trileaflet; note coarse nodular calcium macroaggregates (⬎4 mm). (B) Histology of a leaflet showing massive intrinsic and extrinsic calcification.
Figure 4. Aortic stenosis owing to dystrophic calcification of a tricuspid aortic valve with calcium microaggregates. (A) Gross view of surgically resected aortic trileaflet; note smaller (⬍4 mm) calcium deposits. (B) Histology of a leaflet showing focal intrinsic calcification with fibrous thickening.
744
The American Journal of Cardiology (www.ajconline.org)
Figure 5. Correlation between aortic valve gradient and valve weight with regard to microscopic (p ⫽ 0.0026) (circles) versus macroscopic (p ⫽ 0.26) (squares) valvular calcific deposits.
We compared valve weight between groups and evaluated correlations between valve gradients and areas in all cases and within each group. Comparison between groups was carried out with analysis of variance, and correlations were computed with the least squares method. The statistical package STATISTICA 7.1 was used. Results General characteristics of patients are presented in Table 1. Of 242 patients, 139 were men and 103 were women. Women were significantly older (75 ⫾ 7 vs 70 ⫾ 10 years old, p ⬍0.001). There were 180 trileaflet and 62 bileaflet valves. Bileaflet valves weighed much more than trileaflet valves (2.77 ⫾ 1.3 vs 1.98 ⫾ 0.8 g, p ⬍0.001), and patients with bileaflet valves were younger than patients with trileaflet valves (64 ⫾ 12 vs 75 ⫾ 6 years, p ⬍0.001). Plasma cholesterol levels were 184 ⫾ 38 mg/dl in men and 201 ⫾ 34 mg/dl in women. Transaortic gradient was very similar irrespective of gender or number of leaflets, whereas absolute valve area was slightly larger in men (Table 2). The latter difference disappeared when valve area was normalized by body surface area (0.41 ⫾ 0.09 cm2/m2 in men vs 0.43 ⫾ 0.13 cm2/m2 in women, p ⫽ 0.41). A good linear correlation existed between aortic gradient measured during cardiac catheterization and valve weight by absolute weight or normalized by body surface area (Figure 1) in men and women. In contrast, no correlation was evident between valve weight and valve area (Figure 2). At gross examination macroscopic calcium deposits (macroaggregates) were present in 111 (46%) aortic valves and microscopic (or microaggregates) in 131 (54%; Figures 3 and 4). Nodular calcium macroaggregates were mainly present in bileaflet valves (67%), whereas in trileaflets microscopic aggregates prevailed (62%, chi-square 17.08, p ⫽ 0.000036). When present, calcium macroaggregates were important determinants of valve weight (2.84 ⫾ 1.03 g
Table 3 Relation between normalized valve weight and valve cholesterol, hypertension, and diabetes Variable Valvular cholesterol Yes No Hypertension* Yes No Diabetes mellitus Yes No
Normalized Valve Weight (g/cm2)
p Value 0.001
1.00 ⫾ 0.44 1.26 ⫾ 0.64 0.01 1.07 ⫾ 0.5 1.3 ⫾ 0.59 0.23 1.05 ⫾ 0.5 1.19 ⫾ 0.6
* Pressure ⬎140/90 mm Hg or treated hypertension.
when macroaggregates are present vs 1.63 ⫾ 0.56 g when microaggregates are present, p ⬍0.001). In contrast, macroaggregates were not determinants of valve gradient or area. In fact, when calcium was present as diffuse microaggregates, valve gradient was determined by valve weight (Figure 5), whereas no significant relation was present when calcium was macroaggregated. Distribution in calcium macroaggregates was slightly unbalanced in favor of men (51%) compared to women (39%, chi-square 3.78, p ⫽ 0.052). This may be the reason why the slope of the relation between aortic gradient and valve weight was much steeper in women than in men (Figure 1); therefore, the gradient increases faster in women than in men. We compared normalized valve weight in different subgroups of patients according to plasma cholesterol concentration or cholesterol deposits within valve leaflets, hypertension, and diabetes. The heavier the valve, the less frequent were valvular cholesterol deposits and hypertension (Table 3). Moreover, valve weight and area were independent of plasma cholesterol concentration (Figure 6), valvular cholesterol, hypertension, and diabetes.
Valvular Heart Disease/Aortic Valve Weight
745
Figure 6. Correlation between aortic valve (A) weight and (B) area and plasma cholesterol concentration. No significant correlation was found.
Discussion This study confirms in part the results of other investigators who found a relation between valve weight, owing almost exclusively to calcium deposits, and transvalvular gradient in aortic stenosis.2– 8 By far the heaviest valves are the bileaflets. This is understandable because these are larger and offer wide support for intrinsic and extrinsic nodular calcium deposits, which are larger (i.e., macroaggregates). In contrast, calcium in trileaflet aortic valves is diffusely distributed in microaggregates. Because valve weight is mainly determined by calcium macroaggregates and these are found especially in bileaflet valves, trileaflet valves are comparatively lighter. However, trileaflet valves are more severely stenotic and determine a higher transvalvular gradient. This study demonstrates that in aortic stenosis owing to dystrophic calcification, in bileaflet and trileaflet valves the gradient depends on valve weight only when calcium is arranged in microaggregates. Although calcium macroaggregates contribute greatly to valve weight, these add far less to valvular gradient. The pathophysiologic interpretation of this result could be that when calcium deposits are smaller (microaggregates) they may not encroach significantly on the movement of cusps. Conversely, when deposits coalesce, the cusps are stiffened and their movements are hindered, and a gradient is necessary to push these apart during ventricular ejection. Other investigators have observed a similar correlation between weight and gradient,2– 8 but no consideration has ever been given to the different effect of calcium deposit size, i.e., microaggregates versus macroaggregates. Although merely speculative, the distinction between calcium macro- and microaggregates could carry relevant implications for transcatheter aortic valve implantation because the native dystrophic valve is “left in place.”9 However, although computed tomography is currently used in aortic stenosis as a preoperative imaging technique,10 –12 it has not yet been demonstrated to differentiate calcific macro- from microaggregates. An intriguing observation is that the correlation between valve weight and gradient is closer for women than for men.
A partial explanation may be that in women calcium is mainly present as diffuse microaggregates that contribute more to transvalvular gradient. Absence of correlation between weight and area has been reported by others.2,3 We were able to confirm this observation in degenerative aortic stenosis. Absence of correlation between valve weight and area in aortic stenosis could reflect the fact that in these cases there is no commissural fusion, and therefore a definite “valve area” is hard to appreciate and to measure. In this setting, the disease is because of a difficulty in opening the stiffened leaflets, and transvalvular gradient seems an appropriate measurement of this difficulty. Senile aortic valve stenosis owing to dystrophic calcification is no longer considered a passive age-related phenomenon but rather an active disease process that shares many aspects with atherosclerosis.13–22 Surprisingly, valve weight and valve area were independent of plasma cholesterol concentration, cholesterol cleft deposits within leaflets, hypertension, and diabetes. In our opinion, this does not mean that these risk factors are not involved in the pathogenesis of aortic stenosis but rather may suggest that, once aortic stenosis is established, removing risk factors may have no influence on severity of aortic stenosis. 1. Freeman RV, Otto CM. Spectrum of calcific aortic valve disease: pathogenesis, disease progression, and treatment strategies. Circulation 2005;111:3316 –3326. 2. Roberts WC, Ko JM. Relation of weights of operatively excised stenotic aortic valves to preoperative transvalvular peak systolic pressure gradients and to calculated aortic valve areas. J Am Coll Cardiol 2004;44:1847–1855. 3. Roberts WC, Ko JM, Filardo G. Comparison of heavier versus lighter operatively excised stenotic aortic valves in adults with aortic stenosis and implications for percutaneous aortic valve implantation without replacement. Am J Cardiol 2009;104:393– 405. 4. Roberts WC, Ko JM. Weights of operatively-excised stenotic unicuspid, bicuspid, and tricuspid aortic valves and their relation to age, sex, body mass index, and presence or absence of concomitant coronary artery bypass grafting. Am J Cardiol 2003;92:1057–1065. 5. Roberts WC, Ko JM. Weights of individual cusps in operativelyexcised congenitally bicuspid stenotic aortic valves. Am J Cardiol 2004;94:678 – 681.
746
The American Journal of Cardiology (www.ajconline.org)
6. Roberts WC, Ko JM. Weights of individual cusps in operativelyexcised stenotic three-cuspid aortic valves. Am J Cardiol 2004;94: 681– 684. 7. Roberts WC, Ko JM, Hamilton C. Comparison of valve structure, valve weight, and severity of the valve obstruction in 1849 patients having isolated aortic valve replacement for aortic valve stenosis (with or without associated aortic regurgitation) studied at 3 different medical centers in 2 different time periods. Circulation 2005;112:3919 – 3929. 8. Sims JB, Roberts BJ, Roberts WC, Hebeler RF, Grayburn PA. The heaviest known operatively-excised aortic valve. Am J Cardiol 2006; 97:588 –589. 9. Masson JB, Kovac J, Schuler G, Ye J, Cheung A, Kapadia S, Tuzcu ME, Kodali S, Leon MB, Webb JG. Transcatheter aortic valve implantation: review of the nature, management, and avoidance of procedural complications. JACC Cardiovasc Interv 2009;2:811– 820. 10. Willmann JK, Weishaupt D, Lachat M, Kobza R, Roos JE, Seifert B, Lüscher TF, Marincek B, Hilfiker PR. Electrocardiographically gated multi-detector row CT for assessment of valvular morphology and calcification in aortic stenosis. Radiology 2002;225:120 –128. 11. Koos R, Mahnken AH, Sinha AM, Wildberger JE, Hoffmann R, Kühl HP. Aortic valve calcification as a marker for aortic stenosis severity: assessment on 16-MDCT. AJR Am J Roentgenol 2004;183:1813– 1818. 12. Liu F, Coursey CA, Grahame-Clarke C, Sciacca RR, Rozenshtein A, Homma S, Austin JH. Aortic valve calcification as an incidental finding at CT of the elderly: severity and location as predictors of aortic stenosis. AJR Am J Roentgenol 2006;186:342–349. 13. Otto CM, Kuusisto J, Reichenbach DD, Gown AM, O’Brien KD. Characterization of the early lesion of “degenerative” valvular aortic
14.
15.
16.
17.
18.
19. 20. 21. 22.
stenosis: histological and immunohistochemical studies. Circulation 1994;90:844 – 853. Pohle K, Mäffert R, Ropers D, Moshage W, Stilianakis N, Daniel WG, Achenbach S. Progression of aortic valve calcification: association with coronary atherosclerosis and cardiovascular risk factors. Circulation 2001;104:1927–1932. Poggianti E, Venneri L, Chubuchny V, Jambrik Z, Baroncini LA, Picano E. Aortic valve sclerosis is associated with systemic endothelial dysfunction. J Am Coll Cardiol 2003;41:136 –141. Cowell SJ, Newby DE, Burton J, White A, Northridge DB, Boon NA, Reid J. Aortic valve calcification on computed tomography predicts the severity of aortic stenosis. Clin Radiol 2003;58:712–716. Kuusisto J, Räsänen K, Särkioja T, Alarakkola E, Kosma VM. Atherosclerosis-like lesions of the aortic valve are common in adults of all ages: a necropsy study. Heart 2005;91:576 –582. Allison MA, Cheung P, Criqui MH, Langer RD, Wright CM. Mitral and aortic annular calcification are highly associated with systemic calcified arteriosclerosis. Circulation 2006;113:861– 866. Mohler ER III. Mechanisms of aortic valve calcification. Am J Cardiol 2004;94:1396 –1401. Otto CM, O’Brien KD. Why is there discordance between calcific aortic stenosis and coronary artery disease? Heart 2001;85:601– 602. Cowell SJ, Newby DE, Boon NA, Elder AT. Calcific aortic stenosis: same old story? Age Ageing 2004;33:538 –544. Caira FC, Stock SR, Gleason TG, McGee EC, Huang J, Bonow RO, Spelsberg TC, McCarthy PM, Rahimtoola SH, Rajamannan NM. Human degenerative valve disease is associated with up-regulation of low-density lipoprotein receptor-related protein 5 receptor-mediated bone formation. J Am Coll Cardiol 2006;47:1707–1712.
Incidence, Predictors, and Outcome of Conduction Disorders After Transcatheter Self-Expandable Aortic Valve Implantation Chiara Fraccaro, MDa,*, Gianfranco Buja, MDa, Giuseppe Tarantini, MD, PhDa, Valeria Gasparetto, MDa, Loira Leoni, MD, PhDa, Renato Razzolini, MDa, Domenico Corrado, MD, PhDa, Raffaele Bonato, MDb, Cristina Basso, MD, PhDc, Gaetano Thiene, MDc, Gino Gerosa, MDd, Giambattista Isabella, MDa, Sabino Iliceto, MDa, and Massimo Napodano, MDa The aims of the present study were to investigate the incidence and characteristics of conduction disorders (CDs) after transcatheter aortic valve implantation (TAVI), to analyze the predictors of permanent pacemaker (PPM) implantation, and to evaluate the outcomes of CDs over time. In particular, we sought to investigate whether the depth of deployment and other technical aspects of valve implantation might predict the need for PPM implantation after TAVI. TAVI has been reported to favor the onset or worsening of CDs often requiring PPM implantation. A total of 70 patients with aortic stenosis due to dystrophic calcification underwent TAVI with third-generation CoreValve Revalving System from May 2007 to April 2009. We collected electrocardiograms at baseline, during TAVI, during hospitalization and at the 1-, 3-, 6-, and 12-month follow-up visits thereafter. The clinical, anatomic, and procedural variables were tested to identify the predictors of PPM implantation. The PPM dependency at follow-up was analyzed. Six patients were excluded from the analysis because of a pre-existing PPM. Of the 64 patients, 32 (50%) had one or more atrioventricular-intraventricular CDs at baseline. TAVI induced a worsening in the CDs in 49 (77%) of the 64 patients, with 25 (39%) requiring in-hospital PPM implantation. On multivariate analysis, the independent predictors of PPM implantation were the depth of the prosthesis implantation (p ⴝ 0.039) and the pre-existing right bundle branch block (p ⴝ 0.046). A trend in the recovery of the CDs over time was recorded, although 2 patients required PPM implantation 1 month after discharge for late complete atrioventricular block. In conclusion, TAVI often induces or worsens CDs, requiring PPM in more than one third of patients, although a trend in the recovery of CDs during the midterm was recorded. The independent predictors of PPM implantation were the depth of prosthesis implantation and pre-existing right bundle branch block. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:747–754) The aim of the present study was to investigate the incidence and characteristics of conduction disorders (CDs) in patients undergoing transcatheter aortic valve implantation (TAVI) and the need for subsequent permanent pacemaker (PPM) implantation. In addition, to help identify the clinical, anatomic, and procedural predictors of postoperative PPM implantation and the outcome of CDs over time. In particular, we sought to investigate whether the depth of deployment and other technical aspects of valve implantation might predict the need for PPM implantation after TAVI.
a Division of Cardiology and dDivision of Cardiac Surgery, Department of Cardiac, Thoracic, and Vascular Sciences, bInstitute of Anesthesia, and c Department of Medical Diagnostic Sciences and Special Therapies, University of Padova, Padua, Italy. Manuscript received July 28, 2010; manuscript received and accepted October 19, 2010. Dr. Fraccaro is a Ph.D. fellow and was supported in part by a Grant of the Italian Society of Cardiology. *Corresponding author: Tel: (⫹39) 049-821-1844; fax: (⫹39) 049876-1764. E-mail address:
[email protected] (C. Fraccaro).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.054
Methods The data from 70 consecutive patients with aortic valve stenosis undergoing TAVI at our Department at Padova University from May 2007 to April 2009 were analyzed. A total of 6 patients were excluded from the analysis because they already had undergone PPM implantation before TAVI. All patients were a part of the multicenter, expanded evaluation registry after conformité européenne mark approval.1 The inclusion and exclusion criteria for TAVI have been previously reported.1–3 All procedures were performed using the third-generation self-expanding CoreValve Revalving System (Medtronic, Minneapolis, Minnesota), as previously described,1,2 using a transfemoral or transubclavian approach, according to the anatomy of the iliac and femoral arteries.3 Implantation success was defined as the correct positioning and performance of the prosthesis. Procedural success was defined as the success of implantation, with the patient leaving the catheterization laboratory alive. The good performance of the bioprosthesis was defined as a reduction in the mean transaortic gradient to ⬍20 mm Hg and aortic regurgitation www.ajconline.org
748
The American Journal of Cardiology (www.ajconline.org)
Table 1 Baseline characteristics Variable Age (years) Mean ⫾ SD Range Men Logistic EuroSCORE Mean ⫾ SD Range New York Heart Association I II III IV Canadian cardiovascular society 0 1 2 3 4 Calcium score ⱕ2 3 4 Coronary artery disease Congestive heart failure Cerebral vascular accident Chronic kidney disease Chronic obstructive pulmonary disease Peripheral vascular disease Previous cardiac surgery Neurologic dysfunction Liver cirrhosis Porcelain aorta* Hostile thorax† Aortic valve area (cm2) Ejection fraction (%)
Total (n ⫽ 64)
PPM After TAVI (n ⫽ 25)
No PPM After TAVI (n ⫽ 39)
80.97 ⫾ 6.55 55–91 29 (45%)
81.56 ⫾ 5.10
80.59 ⫾ 7.37
15 (60%)
14 (36%)
23.64 ⫾ 14.72 3–71
25.47 ⫾ 15.70
p Value 0.567
0.058 0.464
22.66 ⫾ 14.00 0.719
9 (14%) 19 (30%) 32 (50%) 4 (6%)
2 (8%) 8 (32%) 13 (52%) 2 (8%)
7 (18%) 11 (28%) 19 (49%) 2 (5%)
39 (61%) 2 (3%) 6 (9%) 10 (16%) 7 (11%)
17 (68%) 1 (4%) 3 (12%) 2 (8%) 2 (8%)
22 (56%) 1 (3%) 3 (8%) 8 (21%) 5 (13%)
0.629
0.595 17 (27%) 30 (47%) 17 (27%) 40 (63%) 30 (47%) 7 (11%) 35 (55%) 14 (22%) 22 (34.4%) 16 (25.0%) 10 (15.6%) 5 (7.8%) 15 (23.4%) 9 (14.1%) 0.78 ⫾ 0.21 52.32 ⫾ 13.24
8 (32%) 10 (40%) 17 (28%) 17 (71%) 13 (52%) 2 (8%) 17 (68%) 7 (29%) 8 (32.0%) 9 (36%) 3 (12%) 1 (4%) 3 (12%) 3 (12.0%) 0.80 ⫾ 0.20 51.68 ⫾ 12.95
9 (23%) 20 (51%) 10 (26%) 23 (61%) 17 (44%) 5 (13%) 18 (46%) 7 (18%) 14 (35.9%) 7 (17.9%) 7 (17.9%) 4 (10.3%) 12 (30.8%) 6 (15.4%) 0.77 ⫾ 0.22 52.74 ⫾ 13.58
0.586 0.511 0.547 0.087 0.298 0.749 0.104 0.523 0.363 0.084 0.704 0.544 0.759
Data are presented as mean ⫾ SD or n (%), unless otherwise noted. * Defined as an aorta with diffuse, circumferential, plate-like calcification involving the whole proximal ascending aorta, precluding cannulation or cross-clamping. † Included patients with a severe deformity of the thorax, severe connective tissue disease, and previous mediastinal/thorax radiotherapy.
to ⱕ2⫹/4, as evaluated on an aortic angiogram or echocardiogram.1 The degree of aortic valve calcium was scored according to the presence and extent of cusp calcification as it appeared on the aortic angiogram. The grading was as follows: grade 1, no calcification; grade 2, mild calcification appearing as a thin marginal rim in one or more cusps; grade 3, moderate calcification characterized by a thick rim occupying the entire surface of one or more cusps; and grade 4, severe calcification, defined as the presence of heavy calcification of all cusps or bulky calcification. The depth of bioprosthesis implantation was measured in the right anterior oblique projection as the distance (in millimeters) of the aortic prosthesis within the left ventricular outflow tract, from the lower edge of the noncoronary cusp (D1) and from the lower edge of the left coronary cusp (D2) to the ventricular end of the prosthesis frame using quantitative angiographic digital techniques (Allura, Philips Medical System, Best, The Netherlands).4 The difference
between D2 and D1 was calculated as the coaxial index. Prosthesis implantation was defined as coaxial when the coaxial index ranged from ⫺1.0 mm to ⫹1.0 mm and noncoaxial when the coaxial index was ⬎⫹1.0 mm or ⬍⫺1.0 mm. The ratio between the prosthesis nominal diameter and native annulus size was calculated as the prosthesis/annulus ratio. The ratio between the diameter of the deployed prosthesis measured at the level of the aortic annulus and the native aortic annulus was calculated as the prosthesis expansion index. All patients underwent standard 12-lead electrocardiography before the procedure. To assess intraoperative CDs, 3-lead continuous electrocardiographic monitoring was recorded and electronically stored throughout the procedure. After the procedure, continuous monitoring was routinely performed in all patients during the hospital stay. Postoperatively, 12-lead electrocardiography was performed daily during hospitalization and at 1-, 3-, 6-, and 12-month follow-up visits thereafter to detect any modifications in the
Valvular Heart Disease/Conduction Defects After Aortic Revalving Therapy
749
Table 2 Procedural data Variable
Total (n ⫽ 64)
PPM After TAVI (n ⫽ 25)
No PPM After TAVI (n ⫽ 39)
p Value
Procedural success General anesthesia Double-lumen intubation Transesophageal echocardiography Access Transfemoral Transubclavian Prosthesis size (mm) 26 29 Predilation Postdilation Valve-in-valve Prosthesis/annulus diameter ratio Depth of implantation (mm) D1 D2 Coaxial index (mm) Noncoaxial alignment Prosthesis expansion index Procedural duration (minutes) Hospital stay (days)
61 (95%) 15 (23%) 8 (13%) 13 (20%)
23 (92%) 8 (33%) 3 (12%) 6 (27%)
38 (97%) 7 (18%) 5 (13%) 7 (21%)
0.315 0.360 0.419 0.604
60 (94%) 4 (6%)
22 (88%) 3 (12%)
38 (97%) 1 (3%)
0.128
36 (56%) 28 (44%) 62 (97%) 19 (30%) 5 (8%) 1.21 ⫾ 8.89
12 (48%) 13 (52%) 24 (96%) 6 (24%) 3 (12%) 1.20 ⫾ 1.05
24 (62%) 15 (39%) 38 (97%) 13 (33%) 2 (5%) 1.21 ⫾ 7.80
0.287 0.747 0.425 0.318 0.561
10.25 ⫾ 3.39 11.41 ⫾ 3.27 1.15 ⫾ 1.59 44 (69%) 0.93 ⫾ 0.11 72.7 ⫾ 37.6 12.34 ⫾ 8.37
11.34 ⫾ 3.62 12.16 ⫾ 3.25 0.81 ⫾ 1.47 15 (58%) 0.93 ⫾ 0.11 72.52 ⫾ 28.33 15.27 ⫾ 11.06
9.50 ⫾ 3.04 10.89 ⫾ 3.23 1.39 ⫾ 1.64 29 (76%) 0.93 ⫾ 0.12 72.74 ⫾ 42.79 10.24 ⫾ 5.24
0.031 0.108 0.155 0.114 0.727 0.982 0.022
Data are presented as n (%) or mean ⫾ SD. Table 3 Change in conduction parameters over time Variable
Atrial fibrillation Left bundle branch block Right bundle branch block Anterior hemiblock Posterior hemiblock PR interval (ms) QRS width (ms) QT (ms)
Before TAVI
10/64 (16%) 9/64 (14%) 8/64 (13%) 11/64 (17%) 2/64 (3%) 182.9 ⫾ 5.7 103.6 ⫾ 4.1 406.2 ⫾ 45.7
After TAVI
11/64 (17%) 37/64 (58%) 3/64 (5%) 3/64 (5%) 2/64 (3%) 211.1 ⫾ 5.5 144.3 ⫾ 3.6 424.4 ⫾ 47.7
Last Follow-Up (6.0 ⫾ 4.2 mo)
10/57* (18%) 25/57* (44%) 2/57* (4%) 4/57* (7%) 1/57* (2%) 182.7 ⫾ 29.9 125.3 ⫾ 26.3 411.5 ⫾ 30.0
p Value Before Versus After TAVI
After TAVI Versus Last Follow-Up
Before TAVI Versus Last Follow-Up
0.705 ⬍0.0001 0.059 0.005 1.000 ⬍0.0001 ⬍0.0001 0.028
0.655 0.108 0.317 0.317 0.317 ⬍0.0001 ⬍0.0001 0.039
0.655 ⬍0.0001 0.046 0.034 1.000 0.761 ⬍0.0001 0.486
Data are presented as n (%) or mean ⫾ SD. * Three cases of periprocedural mortality and four with paced rhythm without adequate ventricular response during PPM inhibition excluded because analysis was not applicable.
atrioventricular (AV) and intraventricular conduction. The analyses of the records were performed by an experienced electrophysiologist. The presence of CDs at any time was defined by the presence of at least one of the following abnormalities: first-, second-, or third-degree AV block, left bundle branch block (BBB), right BBB, and/or left anterior or posterior hemiblock. The currently accepted criteria were used to code for each of these CD.5,6 The requirement for PPM implantation was determined by the attending cardiologist according to the standardized criteria from the American College of Cardiology/American Heart Association/Heart Rhythm Society 2008 Guidelines for DeviceBased Therapy of Cardiac Rhythm Abnormalities.7 All systems were implanted using a transvenous subclavian approach.
Follow-up data were collected at 1, 3, 6, and 12 months and yearly thereafter. Periprocedural death was defined as any death within 30 days after the procedure or any death before discharge. The clinical follow-up events included death from all causes, cardiac death (including all unexplained deaths), acute myocardial infarction, stroke, cardiac heart failure requiring rehospitalization, and PPM implantation. Moreover, at each temporal step, a 12-lead electrocardiogram was collected in all patients, to record modifications in AV and intraventricular conduction. In patients with a PPM, the percentage of ventricular pacing was detected by PPM interrogation. Moreover, to evaluate the PPM dependency in patients with a paced baseline electrocardiogram, the pacemaker was programmed to VVI at the lowest rate possible and the underlying rhythm was ob-
750
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Views of 85-year old woman who had developed complete AV block requiring PPM implantation after TAVI who died 13 days after the procedure. (A) Electrocardiogram showing right BBB and 60 mm Hg transaortic gradient at baseline. (B) Electrocardiogram showing complete AV block immediately after TAVI. (C) Computed tomography scan of heart explanted at autopsy. Note, deep positioning of CoreValve within left ventricular outflow tract, overlapping membranous septum (dotted circle) and crest of interventricular septum. (D) Gross anatomic view of left ventricular outflow seen from below. Note, expansion of prosthesis frames in subaortic region compressing ventricular septum and overlapping proximal branching of left bundle branch (dotted lines). Ao, aorta; LV, left ventricle; MS, membranous septum; MV, mitral valve; RV, right ventricle; VS, ventricular septum.
tained. The patients were considered pacemaker dependent if they continued to be paced or had complete AV block or atrial fibrillation with inadequate ventricular response. The patients were considered as nonpacemaker dependent if they had sinus rhythm or atrial fibrillation with an adequate ventricular response. The categorical data are expressed as numbers and percentages and compared by Fisher’s or chi-square exact test as appropriate. The continuous variables are expressed as the mean ⫾ SD and compared using Student’s t test. The preoperative clinical variables, anatomic characteristics, and procedural data thought likely to influence the conducting system were tested by univariate logistic regression analysis to determine the predictors of postoperative PPM implantation. This model included all the variables with a biologically relevant correlation to the onset of CD: age, gender, anatomic characteristics (e.g., aortic valve area, calcium score, aortic regurgitation, left ventricular mass index), effects of drugs (type of anesthesia), technical aspects that might mechanically effect the conduction system (e.g., val-
Figure 2. Incidence of PPM implantation according to baseline CD. Right BBB was the only baseline CD significantly related to PPM implantation.
vuloplasty balloon diameter, prosthesis size, prosthesis/annulus diameter ratio, need for postdilation, depth of implantation, prosthesis expansion index, and valve-in-valve), all pre-existing CDs (AV block I, left BBB, anterior hemi-
Valvular Heart Disease/Conduction Defects After Aortic Revalving Therapy
751
Table 4 Predictors of permanent pacemaker (PPM) implantation after transcatheter aortic valve implantation (TAVI) Variable
Age Male gender Aortic valve area Calcium score Aortic regurgitation Left ventricular mass index Anesthesia type Valvuloplasty balloon diameter Prosthesis size Prosthesis/annulus diameter ratio Postdilation Depth of implantation Prosthesis expansion index Valve-in-valve First-degree atrioventricular block at baseline Left bundle branch block at baseline Anterior hemiblock at baseline Posterior hemiblock at baseline Right bundle branch block at baseline Bifascicular block at baseline
Univariate Analysis
Multivariate Analysis
OR
95% CI
p Value
OR
95% CI
p Value
0.992 3.077 2.530 0.836 0.773 0.992 1.265 1.312 0.500 0.211 0.577 1.200 1.371 6.727 1.968 0.696 1.230 1.480 5.400 1.522
0.919–1.070 1.092–8.671 0.226–28.313 0.433–1.615 0.431–1.387 0.968–1.016 0.495–3.230 0.935–1.840 0.181–1.379 0.001–66.539 0.186–1.790 1.006–1.431 0.015–128.370 0.706–64.079 0.612–6.335 0.157–3.074 0.332–4.558 0.088–24.777 0.995–29.297 0.200–11.579
0.828 0.033 0.451 0.594 0.388 0.507 0.624 0.116 0.180 0.596 0.341 0.042 0.892 0.097 0.256 0.632 0.757 0.785 0.051 0.685
— 1.182 — — — — — — — — — 1.210 — 2.768 — — — — 6.132 —
— 0.982–1.423 — — — — — — — — — 1.010–1.449 — 0.226–33.843 — — — — 1.030–36.519 —
NS 0.127 NS NS NS NS NS NS NS NS NS 0.039 NS 0.425 NS NS NS NS 0.046 NS
CI, confidence interval; OR, odds ratio; NS, not significant.
block, posterior hemiblock, right BBB, bifascicular block). Multiple stepwise logistic regression analyses of those significant variables (p ⬍0.10) on univariate analysis were performed to identify independent predictors of PPM implantation. Univariate and multivariate analyses were also performed considering the same variables to identify predictors of worsening in CDs. The odds ratios and their corresponding 95% confidence intervals are provided. A p value of ⬍0.05 with a 2-tailed test was considered statistically significant. Statistical analysis was performed using the statistical software Statistical Package for Social Sciences, version 17.0, for Windows (SPSS, Chicago, Illinois). Results The baseline characteristics were similar between those patients who required PPM implantation and those who did not (Table 1). Implantation success was achieved in 62 (97%) of the 64 patients, with procedural success in 61 (95%) of 64. The procedural data were similar between those who required PPM implantation and those who did not, except for the depth of prosthesis implantation measured from the lower edge of the noncoronary cusp (D1), which was significantly deeper (i.e., more ventricular) in the patients who underwent PPM implantation than in those who did not. Also, the hospital stay was longer in patients who underwent PPM implantation (Table 2). Of the 64 patients, 32 (50%) had one or more degrees of AV-intraventricular CDs before TAVI, including first degree AV block (n ⫽ 15), complete left BBB (n ⫽ 9), right BBB (n ⫽ 8), anterior hemiblock (n ⫽ 11), and posterior hemiblock (n ⫽ 2; Table 3). After TAVI, worsening or new-onset CD appeared in most patients (77%). Left BBB was the most frequent new CD, appearing in 28 patients (44%), isolated or associated
to new first-degree AV block (Table 3). A complete AV block appeared in 19 patients (Figure 1). During the hospitalization, 25 patients (39%) underwent PPM implantation. The indications for PPM implantation were permanent or transient complete AV block in 16, second-degree AV block associated with left BBB in 6, sick sinus syndrome in 1, and trifascicular block in 1 patient. Indeed, 1 patient who did not have any CD before TAVI underwent PPM implantation because of the development of first-degree AV block associated with new left BBB after TAVI, with progressive prolongation of PR and QRS intervals during hospital stay. Of the 25 patients requiring PPM implantation, 9 had no AV or intraventricular CD before TAVI, and 16 patients showed at least one CD (Figure 2). On univariate analysis, male gender, right BBB at baseline, and the depth of prosthesis implantation resulted in a greater prevalence of PPM implantation after TAVI (Table 4). After adjustment by multivariate analysis for the baseline clinical, anatomic, and operative characteristics, right BBB at baseline, and the depth of prosthesis implantation remained the only independent predictors of PPM implantation (Table 4). After TAVI, 6 (75%) of 8 patients with right BBB at baseline required PPM implantation versus 19 (34%) of 56 patients, who had not had right BBB before TAVI (p ⫽ 0.026). Right BBB was the only baseline CD that significantly affected the occurrence of PPM implantation after TAVI (Figure 2 and Table 4). However, considering as a dependent variable the worsening in the CD, rather than the need for PPM implantation, the predictors of CD worsening remained the depth of implantation (odds ratio 1.30, 95% confidence interval 1.00 to 1.69, p ⫽ 0.05) and pre-existing left BBB (odds ratio 0.07, 95% confidence interval 0.01 to 0.49, p ⫽ 0.007). The 30-day mortality rate was 5% (3 of 64), with no
752
The American Journal of Cardiology (www.ajconline.org)
Figure 3. Impact of CoreValve positioning on electrical conduction system. Diagram illustrating anatomic relation between prosthesis across aortic annulus and conduction system. A deeper prosthesis implantation into left ventricular outflow tract (B) might effect electrical conduction system more than a higher one (A).
difference between those patients who required PPM implantation and those who did not (4% vs 5%; p ⫽ 0.83). The causes of death were left ventricular perforation, pneumonia, and cerebral edema owing to electrolyte disturbances. During hospitalization, the rate of stroke and myocardial infarction was 2% (1 of 64) and 2% (1 of 64), respectively, with no differences between the groups. At a mean follow-up of 6.0 ⫾ 4.2 months (range 30 days to 13.3 months), the mortality rate was 18% (11 of 61); 29% of patients who needed PPM versus 11% of patients who did not (p ⫽ 0.07). It was a cardiac death in only 1 case. Of 61 patients, myocardial infarction occurred in 1 (2%), stroke in 1 (2%), heart failure in 7 (12%), with no differences between the 2 groups (p ⫽ 0.39, p ⫽ 0.39, and p ⫽ 0.31, respectively). During the follow-up, the PR, QRS, and QT intervals decreased significantly within the first month after the procedure (Table 3). Two patients, discharged with first-degree AV block plus left BBB and with left BBB alone, respectively, underwent PPM implantation 1 month after TAVI for late onset of complete AV block. In these patients, the prosthesis/ annulus was similar to those of patients with early PPM implantation or no PPM implantation (1.10 vs 1.20 and 1.21, respectively). No sudden death occurred. Analyzing the patient’s pacemaker dependency at follow-up, of the 17 PPM patients who were alive, 12 presented with a spontaneous rhythm with a mean ventricular pacing percentage of 19% (range 0.2% to 62%; 7 patients ⬍20%). One patient underwent pacing at baseline and had an adequate ventricular response during pacemaker inhibition, with a ventricular pacing percentage of 4%. Finally, 4 patients underwent pacing at baseline electrocardiography and presented with an inadequate ventricular response at the lowest rate programmable (ventricular pacing percentage ⬎95% in all cases).
Discussion Transcatheter revalving therapy has been reported as valuable alternative strategy to aortic valve replacement in high-risk patients with severe symptomatic aortic valve stenosis.1,8,9 However, TAVI has often been associated with worsening or new-onset CD, particularly when the selfexpandable CoreValve device is used,1,4 as confirmed by our data showing PPM implantation in more than one third of cases. The geometry and design of this device, based on a 53- to 55-mm-long stent cage extending from ascending aorta to left ventricular outflow, might lead to a variable amount of prosthesis frames, pushing aside the interventricular septum and the underlying conduction tissue. In particular, the lower one third of the prosthesis stent frames, characterized by high radial forces for secure anchoring of the stent against the native annulus and outflow septum, might account for CDs due to compression on the left bundle branch, which runs superficially just below the endocardium in the uppermost part of the leftward ventricular septum (Figures 1 and 3). In this scenario, it is reasonable that the deeper the prosthesis has been implanted into the left ventricular outflow tract, the greater the risk of compression by the prosthesis against the left bundle branch and, consequently, of the development of severe CD requiring PPM implantation (Figures 1 and 3). These observations are consistent with the greater rate of PPM implantation reported with CoreValve device4,10,11 than with the Edwards Sapien prosthesis (Edwards Lifesciences, Irvine, California), which extends just a few millimeters below the annular plane.12,13 Piazza et al4 described a significant correlation between the depth of prosthesis implantation and new-onset left BBB after CoreValve implantation, suggesting that deploying the prosthesis in a more superior position within the left ventricular outflow tract might limit the risk of AV block and PPM implantation. In light of this hypoth-
Valvular Heart Disease/Conduction Defects After Aortic Revalving Therapy
esis, we performed a multivariate analysis to identify independent predictors of PPM implantation after TAVI. Our study has confirmed this preliminary hypothesis, identifying the depth of prosthesis implantation into the left ventricular outflow tract as an independent predictor of PPM implantation after CoreValve revalving therapy. In particular, the most relevant measurement predicting PPM implantation was the depth of prosthesis at level of noncoronary cusp, and neither the depth at the level of the left coronary cusp nor the coaxial implantation of prosthesis were related to PPM requirement. Moreover, it is important to emphasize that, as previously reported by Calvi et al,11 the most frequent CD recorded after TAVI in our series was new-onset left BBB. This CD was likely related to the close anatomic relation between the left bundle branch and the aortic valve apparatus and might favor the development of complete AV block when a right BBB is present before TAVI. Consistent with this assumption, the presence of right BBB before procedure appeared the most powerful predictor of PPM implantation in our analysis, confirming previous observations.4 A lower rate of PPM was associated with a low prevalence of right BBB at baseline in other series.14 A greater prevalence of PPM implantation was reported in our study compared to other series.11 We decided to perform early PPM implantation to minimize the risk related to transvenous temporary pacing support, such as infection and ventricular perforation, coupled with potential problems related to the immobilization and long hospital stay for an elderly patient. However, our results have confirmed that a recovery in intraventricular conduction (i.e., a decrease in the frequency of left BBB and PPM dependency) can occur over the time. The possible explanations for this include transient inflammation, edema, ischemia, and mechanical trauma with subsequent recovery of conduction. Postmortem specimens from patients who had developed new AV block after TAVI have demonstrated microscopic evidence of myocardial injury in the interventricular septum, as well as localized hematoma at the site of prosthesis expansion, which might account for mechanical compression of the conduction system coursing in the subendocardium close to the membranous septum.15 Balloon aortic valvuloplasty itself has been associated in the past with the occurrence of CD.16 Similarly, the balloon used to predilate the native aortic valve before TAVI could account for reversible mechanical or ischemic effects, explaining the possible conduction recovery with time after TAVI. In contrast, in our series, 2 patients who were discharged with first-degree AV block plus left BBB and with left BBB alone, respectively, experienced late complete AV block. Progressive degeneration of the conduction system, accelerated by mechanical injury and fibrosis in the upper interventricular septum, might have occurred in these patients. Some limitations of our study should be taken into account to place our findings in the proper perspective. First, because we decided to perform early PPM implantation to maximize the patient’s safety, the relatively greater rate of PPM resulting from this might have influenced our analyses. However, even though most CDs recovered over time, the PPM dependency was still present in most patients
753
during follow-up. In addition, to understand the relation between prosthesis size and aortic annulus, we assessed the prosthesis dimensions by angiography at the end of the implantation procedure. However, because the full expansion of the nitinol stent frames seems to be reached some days after deployment,17 this might have led to underestimating the maximal dimensions of the prosthesis and consequently of its effect on annulus stretching.
Acknowledgment: The authors are indebted to Mr. Claudio Bellini for his valuable support in graphic elaborations. 1. Piazza N, Grube E, Gerckens U, den Heijer P, Linke A, Luha O, Ramondo A, Ussia G, Laborde JC, de Jaegere P, Serruys PW. Procedural and 30-day outcomes following transcatheter aortic valve implantation using the third generation (18 Fr) CoreValve revalving system: results from the multicentre, expanded evaluation registry 1-year following CE mark approval. EuroIntervention 2008;4:242– 249. 2. Napodano M, Cutolo A, Fraccaro C, Tarantini G, Bonato R, Bianco R, Gerosa G, Iliceto S, Ramondo A. Totally percutaneous valve replacement for severe aortic regurgitation in a degenerating bioprosthesis. J Thorac Cardiovasc Surg 2009;138:1027–1028. 3. Fraccaro C, Napodano M, Tarantini G, Gasparetto V, Gerosa G, Bianco R, Bonato R, Pittarello D, Isabella G, Iliceto S, Ramondo A. Expanding the eligibility for transcatheter aortic valve implantation the trans-subclavian retrograde approach using: the III generation CoreValve revalving system. J Am Coll Cardiol Interv 2009;2:828 – 833. 4. Piazza N, Onuma Y, Jesserun E, Kint PP, Maugenest AM, Anderson RH, de Jaegere PP, Serruys PW. Early and persistent intraventricular conduction abnormalities and requirements for pacemaking after percutaneous replacement of the aortic valve. J Am Coll Cardiol Interv 2008;1:310 –316. 5. Willems JL, Robles de Medina EO, Bernard R, Coumel P, Fisch C, Krikler D, Mogensen L, Moret P, Mogensen L, Moret P. Criteria for intraventricular conduction disturbances and pre-excitation: World Health Organizational/International Society and Federation for Cardiology Task Force Ad Hoc. J Am Coll Cardiol 1985;5:1261–1275. 6. Elizari MV, Acunzo RS, Ferreiro M. Hemiblocks revisited. Circulation 2007;115:1154 –1163. 7. Writing Committee to Revise the ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices. ACC/AHA/HRS 2008 Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the American Association for Thoracic Surgery and Society of Thoracic Surgeons. Circulation 2008;117:e350 – e408. 8. Cribier A, Eltchaninoff H, Tron C, Bauer F, Agatiello C, Nercolini D, Tapiero S, Bessou JP, Babaliaros V. Treatment of calcific aortic stenosis with the percutaneous heart valve: mid-term follow-up from the initial feasibility studies: the French experience. J Am Coll Cardiol 2006;47:1214 –1223. 9. Grube E, Schuler G, Buellesfeld L, Gerckens U, Linke A, Wenaweser P, Walther T, Zickmann B, Felderhoff T, Cartier R, Bonan R. Percutaneous aortic valve replacement for severe aortic stenosis in high-risk patients using the second- and current third-generation self-expanding CoreValve prosthesis: device success and 30-day clinical outcome. J Am Coll Cardiol 2007;50:69 –76. 10. Jilaihawi H, Chin D, Vasa-Nicotera M, Jeilan M, Spyt T, Ng GA, Bence J, Logtens E, Kovac J. Predictors for permanent pacemaker requirement after transcatheter aortic valve implantation with the CoreValve bioprosthesis. Am Heart J 2009;157:860 – 866. 11. Calvi V, Puzzangara E, Pruiti GP, Conti S, Di Grazia A, Ussia GP, Capodanno D, Tamburino C. Early conduction disorders following
754
The American Journal of Cardiology (www.ajconline.org)
percutaneous aortic valve replacement. Pacing Clin Electrophysiol 2009;32:S126 –S130. 12. Sinhal A, Altwegg L, Pasupati S, Humphries KH, Allard M, Martin P, Kerr C, Lichtenstein SV, Webb JG. Atrioventricular block after transcatheter balloon expandable aortic valve implantation. J Am Coll Cardiol Interv 2008;1:305–309. 13. Godin M, Eltchaninoff H, Furuta A, Tron C, Anselme F, Bejar K, Sanchez-Giron C, Bauer F, Litzler PY, Bessou JP, Cribier A. Frequency of conduction disturbances after transcatheter implantation of an Edwards Sapien aortic valve prosthesis. Am J Cardiol 2010;106: 707–712. 14. Gutiérrez M, Rodés-Cabau J, Bagur R, Doyle D, DeLarochellière R, Bergeron S, Bertrand OF, Pibarot P, Dumont E. Electrocardiographic
changes and clinical outcomes after transapical aortic valve implantation. Am Heart J 2009;158:302–308. 15. Moreno R, Dobarro D, Lopez de Sa E, Prieto M, Morales C, Calvo Orbe L, Galeote G, Jiménez-Valero S, Lopez-Sendon JL. Cause of complete atrioventricular block after percutaneous aortic valve implantation: insights from a necropsy study. Circulation 2009;120:e29 – e30. 16. Serruys PW, Luijten HE, Beatt KJ, Di Mario C, de Feyter PJ, Essed CE, Roelandt JR, van den Brand M. Percutaneous balloon valvuloplasty for calcific aortic stenosis. A treatment “sine cure?” Eur Heart J 1988;9:782–794. 17. Napodano M, Tarantini G, Ramondo A. Is it reasonable to treat all calcified stenotic valves with a valve stent? Probably yes if we get a full stent expansion. J Am Coll Cardiol 2009;53:219.
Incidence, Epidemiology, and Prognosis of Residual Pulmonary Hypertension After Mitral Valve Repair for Degenerative Mitral Regurgitation Andrew B. Goldstone, BAa, Joanna Chikwe, MDa, Sean P. Pinney, MDb, Anelechi C. Anyanwu, MDa, Samuel A. Funt, MDa, Antonio Polanco, BAa, and David H. Adams, MDa,* Pulmonary hypertension (PH) is a common sequela of degenerative mitral valve disease, but the regression of PH after mitral surgery is often incomplete. We sought to identify the preoperative risk factors for residual PH after mitral valve repair and its effect on the clinical outcome. The outcomes in 71 patients with preoperative PH (mean pulmonary arterial pressure >25 mm Hg) were compared according to the presence or absence of residual PH 24 hours after mitral valve surgery. Of 71 patients, 33 (46%) had residual PH. The remainder experienced significant reductions in the mean pulmonary arterial pressure without changes in pulmonary vascular resistance. Patients with residual PH had significantly elevated postoperative pulmonary vascular resistance (despite a significant decrease from the preoperative baseline) compared to those without residual PH. Residual PH was an independent risk factor for postoperative morbidity, mortality, and a prolonged intensive care unit stay (odds ratio 4.0, 95% confidence interval 1.2 to 13.1, p ⴝ 0.02), independent of the preoperative mean pulmonary arterial pressure. A decreased left ventricular ejection fraction (odds ratio 0.9, 95% confidence interval 0.8 to 1.0, p ⴝ 0.007) and fibroelastic deficiency (odds ratio 3.6, 95% confidence interval 1.1 to 11.8, p ⴝ 0.03) were independent predictors of residual PH. In conclusion, residual PH is a clinically important entity common after mitral valve repair for degenerative disease and is associated with clinical variables that aid in the preoperative prediction of at-risk patients. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:755–760) The main objectives of the present study were to define preoperative risk factors that may be used to identify those patients more likely to experience residual elevation in pulmonary artery pressure after mitral repair for degenerative mitral valve disease, and to determine the prognostic effect of residual pulmonary hypertension (PH). Methods We retrospectively analyzed the data from patients with preoperative PH who had undergone mitral valve surgery for pure mitral regurgitation at the Mount Sinai Medical Center (New York, New York) from September 2005 to July 2008. Preoperative PH was defined as either a mean pulmonary arterial pressure (PAP) of ⱖ25 mm Hg by way of right heart catheterization or evidence of at least moder-
a
Department of Cardiothoracic Surgery and bDivision of Cardiology, Mount Sinai School of Medicine, New York, New York. Manuscript received July 26, 2010; manuscript received and accepted October 26, 2010. This work was supported by a grant from the Doris Duke Charitable Foundation, New York, New York to the Mount Sinai School of Medicine to fund Clinical Research Fellow A. B. Goldstone. *Corresponding author: Tel: (212) 659-6800; fax: (212) 659-6818. E-mail address:
[email protected] (D.H. Adams). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.057
ate PH on the preoperative echocardiogram. The exclusion criteria included concomitant aortic valve surgery, nondegenerative mitral valve disease, planned ventricular assist device implantation, death within 24 hours of surgery, and incomplete postoperative pulmonary artery catheterization data. Degenerative mitral valve disease was defined echocardiographically and confirmed intraoperatively as single or multisegment leaflet prolapse resulting from chordal elongation or rupture and caused by myxomatous valve disease. Patients with small valves and single-segment prolapse were classified as having fibroelastic deficiency, and patients with large valves and multisegment prolapse were classified as having Barlow’s disease.1 During the study period, 431 patients underwent mitral valve surgery with or without concomitant tricuspid repair and coronary artery bypass grafting. A total of 71 patients met the inclusion and exclusion criteria, with a mean age of 65 ⫾ 12 years. Of the 71 patients, 44% were women. The data from these patients were collected from the prospective clinical databases and chart review. The local institutional review board approved the protocol, and the study adhered to the Health Insurance Portability and Accountability Act regulations and ethical guidelines of the 1975 Declaration of Helsinki. All patients underwent preoperative echocardiography. The systolic PAP was derived by adding the right ventricwww.ajconline.org
756
The American Journal of Cardiology (www.ajconline.org)
Table 1 Preoperative clinical and operative data Variable
Preoperative data Age (years) Women Body mass index (kg/m2) Body mass index ⬎30 kg/m2 Left ventricular ejection fraction (%) Left ventricular ejection fraction ⱕ50% New York Heart Association class III-IV Previous myocardial infarction Previous stroke Systemic hypertension* Diabetes mellitus Peripheral vascular disease Chronic obstructive airways disease Asthma Sleep apnea Renal failure Atrial fibrillation Reoperation Barlow’s disease Logistic EuroSCORE Median Interquartile range Echocardiographic data Mitral regurgitation Moderate Severe Right ventricular systolic pressure (mm Hg) Left ventricular end-diastolic diameter (cm) Operative data Mitral repair Tricuspid repair Coronary artery bypass grafting Maze ablation Patent foramen ovale closure Cardiopulmonary bypass time (min) Cross-clamp time (min) Intra-aortic balloon pump placement
Residual PH
p Value
No (n ⫽ 38)
Yes (n ⫽ 33)
63 ⫾ 13 20 (53%) 26 ⫾ 10 3 (8%) 61 ⫾ 7 4 (12%)
68 ⫾ 11 11 (33%) 27 ⫾ 3 4 (12%) 56 ⫾ 9 8 (26%)
0.125 0.102 0.587 0.697 0.007 0.145
12 (32%)
12 (36%)
0.671
1 (3%) 1 (3%) 17 (45%) 1 (3%) 1 (3%) 2 (5%) 3 (8%) 1 (3%) 0 (0%) 7 (18%) 1 (3%) 16 (42%)
1 (3%) 2 (6%) 22 (68%) 2 (6%) 1 (3%) 0 (0%) 1 (3%) 0 (0%) 0 (0%) 14 (42%) 1 (3%) 8 (24%)
1.000 0.594 0.064 0.474 1.000 0.495 0.618 1.000 — 0.027 1.000 0.113 0.760
4 3–12
5 3–10
0 38 (100%) 45 ⫾ 10
1 (3%) 32 (97%) 53 ⫾ 13
0.023
5.7 ⫾ 0.7
5.5 ⫾ 0.8
0.352
38 (100%) 34 (90%) 7 (18%) 8 (21%) 2 (5%) 183 ⫾ 63 145 ⫾ 43 3 (8%)
33 (100%) 28 (85%) 8 (24%) 13 (39%) 1 (3%) 194 ⫾ 57 155 ⫾ 51 5 (15%)
— 0.724 0.549 0.091 1.000 0.442 0.381 0.459
0.549
* Defined by one of the following: documented history of hypertension diagnosed and treated with medication, diet, and/or exercise; previous documentation of blood pressure ⬎140 mm Hg systolic or 90 mm Hg diastolic for patients without diabetes or chronic kidney disease, or previous documentation of blood pressure ⬎130 mm Hg systolic or 80 mm Hg diastolic on ⱖ2 occasions for patients with diabetes or chronic kidney disease; and current pharmacological therapy to control hypertension.
ular systolic pressure to the estimated right atrial pressure. The right ventricular systolic pressure was estimated from the peak velocity of tricuspid regurgitation and the simplified Bernoulli equation (⌬P ⫽ 4V2). The right atrial pressure was estimated according to the hepatic vein flow, right atrial size, or the degree of inspiratory collapse of the inferior vena cava.
Preoperative right heart catheterization was performed in 44 patients (59%). Cardiac output was determined using the Fick equation, assuming an oxygen consumption at rest of 125 ml/min/m2 and converted to the cardiac index. Postoperative hemodynamic data were obtained for ⱕ24 hours after surgery. Aortocaval bypass was instituted using a median sternotomy, hemisternotomy, or right thoracotomy, and the mitral valve was approached through a left atriotomy in Sondegaard’s groove. Myocardial preservation was achieved using intermittent cold blood cardioplegia given anterogradely and retrogradely. The mitral valve was systematically evaluated using transesophageal echocardiography and then under direct vision. Mitral valve repair was performed according to Carpentier’s reconstructive principles.1 Concomitant tricuspid valve annuloplasty was performed for patients with moderate to severe PH with significant annular dilation or tricuspid regurgitation of at least moderate severity. The patients were categorized into 1 of 2 groups according to the absence or presence of residual PH (defined as a mean PAP of ⱖ25 mm Hg at rest 24 hours after surgery, measured using a pulmonary artery catheter). The secondary end points focused on the effect of residual PH on postoperative mortality, major complications (e.g., respiratory failure, renal failure, deep sternal wound infection, bleeding requiring reoperation, stroke, and gastrointestinal complications), length of the intensive care unit stay, duration of mechanical ventilation, and duration of inotropic support. Respiratory failure was defined as ventilator therapy for ⬎72 hours, the need for reintubation, or the need for tracheostomy. Renal failure was classified as serum creatinine ⬎2.5 mg/dl for ⱖ7 days postoperatively or a new dialysis requirement. Stroke was defined as a new, permanent neurologic deficit of cerebrovascular cause. Gastrointestinal complications included upper or lower gastrointestinal bleeding and the need for laparotomy. Survival data were obtained by cross-referencing the patient Social Security numbers with the Web-based Social Security death index. Normally distributed continuous variables are expressed as the mean ⫾ SD and non-normally distributed variables as the median with the interquartile range. Categorical variables are presented as proportions. Differences between groups were assessed using the chi-square test or Fischer’s exact test for categorical variables, the independent 2-tailed Student t test for normally distributed continuous variables, and the Mann-Whitney U test for non-normally distributed continuous variables. A univariate logistic regression analysis was performed, including all preoperative candidate variables in Tables 1 and 2 (mean PAP and pulmonary vascular resistance [PVR]), to evaluate the potential risk factors for residual PH. Variables with p ⬍0.15 were included in a stepwise multivariate model. The predicted probabilities for residual PH were calculated for patients in the study population, and receiver operating characteristic curves were generated. A c-statistic of ⬎0.7 was considered accurate. Midterm survival was evaluated using KaplanMeier survival analysis, and the groups were compared using log-rank testing. A Cox proportional hazard regression analysis was performed to determine the independent predictors of reduced survival. The results of the
Valvular Heart Disease/Residual Pulmonary Hypertension
757
Table 2 Pairwise comparison of preoperative and postoperative hemodynamics in patients with and without residual pulmonary hypertension Variable No residual pulmonary hypertension Mean arterial pressure (mm Hg) Mean pulmonary artery pressure (mm Hg) Systolic pulmonary artery pressure (mm Hg) Diastolic pulmonary artery pressure (mm Hg) Central venous pressure (mm Hg) Cardiac index (L/min/m2) Pulmonary vascular resistance (dyn ⫻ s ⫻ cm⫺5) Residual pulmonary hypertension Mean arterial pressure (mm Hg) Mean pulmonary artery pressure (mm Hg) Systolic pulmonary artery pressure (mm Hg) Diastolic pulmonary artery pressure (mm Hg) Central venous pressure (mm Hg) Cardiac index (L/min/m2) Pulmonary vascular resistance (dyn ⫻ s ⫻ cm⫺5)
Preoperative
Postoperative
p Value
84 ⫾ 11 31 ⫾ 4 45 ⫾ 5 21 ⫾ 5 8⫾3 2.4 ⫾ 0.6 117 ⫾ 43
77 ⫾ 11 21 ⫾ 2* 30 ⫾ 4* 14 ⫾ 3* 9 ⫾ 3* 2.7 ⫾ 0.6 114 ⫾ 46†
0.109 ⬍0.001 ⬍0.001 ⬍0.001 0.620 0.102 0.838
92 ⫾ 15 35 ⫾ 8 54 ⫾ 12 23 ⫾ 9 10 ⫾ 5 2.2 ⫾ 0.5 258 ⫾ 156
79 ⫾ 9 30 ⫾ 4* 42 ⫾ 6* 23 ⫾ 3* 15 ⫾ 3* 2.6 ⫾ 0.5 147 ⫾ 54†
0.003 0.015 ⬍0.001 0.691 ⬍0.001 0.030 0.024
* p ⬍0.001; † p ⬍0.05, comparison of corresponding postoperative hemodynamic parameters between patients with and without residual PH. Table 3 Early mortality and morbidity in patients with and without residual pulmonary hypertension Variable
Stroke Sternal wound infection Bleeding Sepsis Gastrointestinal complications Renal failure requiring dialysis Respiratory failure Any complication Hospital mortality Intensive care unit stay (days) Median Interquartile range Duration of inotropic support (days) Median Interquartile range Duration on ventilator (days) Median Interquartile range
Residual PH No (n ⫽ 38)
Yes (n ⫽ 33)
0 (0%) 1 (3%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 4 (11%) 4 (11%) 0 (0%)
1 (3%) 2 (6%) 1 (3%) 1 (3%) 0 (0%) 2 (6%) 7 (21%) 8 (24%) 2 (6%)
1 1–3
3 1–6
0.7 0.4–1.3
1.7 0.9–3.8
0.4 0.3–0.6
0.7 0.5–1.6
p Value
0.465 0.594 0.465 0.465 — 0.212 0.215 0.137 0.219 0.006
⬍0.001
0.004
regression analyses are presented as the odds ratios (ORs) or hazard ratios, with the corresponding 95% confidence intervals (CIs). All tests were 2-tailed, and p ⬍0.05 was considered statistically significant. The statistical analysis was performed using the Statistical Package for Social Sciences for Macintosh, version 18.0 (SPSS, Chicago, Illinois).
Results The preoperative patient characteristics and operative procedures are listed in Table 1. Of the 71 patients, 33 (incidence of 46%) were found to have residual PH.
The pre- and postoperative hemodynamic data are listed in Table 2. The mean, systolic, and diastolic PAP, PVR, and central venous pressure were significantly greater in those patients with residual PH than in those whose pulmonary pressures had normalized. In contrast, the mean arterial pressure and cardiac index did not differ between the 2 groups postoperatively. Of the patients without residual PH, the mean, systolic, and diastolic PAP had decreased significantly from the preoperative baseline values without any change in the PVR. In patients with residual PH, the PVR, systolic and mean PAP had significantly decreased without a corresponding decrease in the diastolic PAP. The postoperative cardiac index and central venous pressure had increased significantly from baseline. Most patients underwent postoperative transthoracic echocardiography before discharge (97%, n ⫽ 69). Of the 69 patients, 65 (94%) had trace or no mitral regurgitation postoperatively, and 4 had mild mitral regurgitation (3 without residual PH and 1 with residual PH). No significant difference was found in the incidence of postoperative right ventricular dysfunction in patients with or without residual PH (49% [n ⫽ 18] vs 47% [n ⫽ 15], respectively, p ⫽ 0.883). Although the mean gradient across the mitral valve did not differ between the patients with and without residual PH postoperatively, the patients with fibroelastic deficiency had greater mean gradients compared to that of patients with Barlow’s disease (5.6 ⫾ 3.7 mm Hg vs 3.9 ⫾ 1.2 mm Hg, respectively, p ⫽ 0.049). The early postoperative mortality and morbidity data are listed in Table 3. The operative mortality rate for the entire study population was 3% (n ⫽ 2), with all deaths occurring in the residual PH group. Both patients were octogenarians: 1 patient developed respiratory failure due to aspiration pneumonia several weeks post-operatively, and 1 patient who presented in congestive heart failure died from cardiogenic shock due to right ventricular dysfunction. The composite incidence of experiencing any major early postoperative adverse event was twofold greater for patients with residual PH (24% [n ⫽ 8] vs 11% [n ⫽ 4], respectively,
758
The American Journal of Cardiology (www.ajconline.org)
Multivariate analysis revealed the following independent predictors of residual PH: decreased left ventricular ejection fraction (OR 0.9, 95% CI 0.8 to 1.0, p ⫽ 0.007) and fibroelastic deficiency (compared with Barlow’s disease; OR 3.6, 95% CI 1.1 to 11.8, p ⫽ 0.033; Table 4). Although a significant determinant on univariate analysis, the preoperative mean PAP was not an independent predictor of residual PH on multivariate analysis. The area under the receiver operating characteristic curve for the multivariate model was 0.74. In a subgroup analysis of patients with complete pre- and postoperative right heart catheterization data, an elevated preoperative PVR was a significant predictor of residual PH.
Discussion
Figure 1. Kaplan-Meier actuarial survival curve for all patients with or without residual PH.
Table 4 Multivariate logistic regression analysis for independent determinants of residual pulmonary hypertension (PH)* Variable Fibroelastic deficiency Left ventricular ejection fraction Constant
Coefficient 
OR
95% CI
p Value
1.283 ⫺0.107
3.6 0.9
(1.1–11.8) (0.8–1.0)
0.033 0.007
⫺5.345
* c-Statistic ⫽ 0.74; Hosmer-Lemeshow p ⫽ 0.67.
p ⫽ 0.137). The patients with residual PH were also more likely to develop multiple postoperative complications (12% [n ⫽ 4] vs 0% [n ⫽ 0], p ⫽ 0.042), to require a longer time in the intensive care unit (p ⫽ 0.006), and to require longer durations of inotropic support (p ⬍0.001) and mechanical ventilation (p ⫽ 0.004). On multivariate analysis, residual PH (OR 4.0, 95% CI 1.2 to 13.1, p ⫽ 0.023) and age (OR 1.1, 95% CI 1.0 to 1.1, p ⫽ 0.009) were found to be independent predictors of postoperative morbidity, hospital mortality, or prolonged intensive care unit stay. Despite an increased operative mortality rate in patients with residual PH, the midterm survival was similar between the 2 groups. The 1- and 4-year survival rate was 93 ⫾ 5% and 93 ⫾ 3% compared to 100 ⫾ 1% and 94 ⫾ 6% for patients in whom the pulmonary pressures had normalized, respectively (p ⫽ 0.432; Figure 1). In a Cox proportional hazards multivariate analysis, residual PH was not an independent predictor of decreased midterm survival.
The results of our study suggest that PH does not resolve after mitral valve repair in nearly 1/2 of patients and that residual PH is associated with a poorer prognosis in the early postoperative period. Additionally, we have shown a moderate ability to predict which patients will have residual PH. Despite optimal mitral repair, 46% of the study population continued to have an elevated mean PAP on postoperative day 1. Two recent studies evaluated postoperative PH using Doppler echocardiography in patients with and without preoperative PH and reported persistent PH 1 to 2 years after mitral valve surgery in 54% and 64% of patients, respectively.2,3 A comparison of the incidence reported in the present study to that of the latter 2 studies supports the notion that the most significant changes in PAP are immediate and sustainable. Three principal mechanisms are implicated in the development of PH in the setting of mitral valve disease: (1) passive retrograde transmission of elevated left atrial pressure resulting from chronic pressure (mitral stenosis) or volume (mitral regurgitation) overload; (2) reactive pulmonary vasoconstriction; and (3) pulmonary vascular remodeling.4 –7 These mechanisms might partly explain how, in patients without residual PH, the mean, systolic and diastolic PAP normalized without a change in the vascular resistance across the pulmonary bed. A complete resolution of PH in these patients might have resulted from successful decompression of the left atrium alone, although it is impossible to state this with any certainty without measuring the changes in the left atrial pressure. In contrast to patients in whom PH resolved, the patients with residual PH had very different hemodynamic profiles. The incomplete resolution of PH in these cases might have resulted from reactive pulmonary vasoconstriction and structural remodeling of the pulmonary vasculature. Reactive pulmonary vasoconstriction should, in theory, resolve with left atrial decompression, but chronic morphologic changes within the pulmonary vasculature will not be promptly reversed. Thus, residual elevations in PVR and PAP suggest pulmonary vascular remodeling has occurred. Preoperative PH is a major risk factor for right ventricular failure and perioperative mortality in patients undergoing mitral valve surgery.8 –13 However, the effect of residual PH on the clinical course of patients in the immediate
Valvular Heart Disease/Residual Pulmonary Hypertension
postoperative period is not well defined. We have demonstrated that residual PH is a significant risk factor for postoperative morbidity and mortality, independent of the preoperative mean PAP. One explanation for this observation is that patients with residual PH are more likely to require prolonged mechanical ventilation, potentially contributing to the increased incidence of sepsis, multiorgan dysfunction, and perioperative mortality. A perhaps more likely reason is that residual PH is a marker for greater preoperative disease severity and adverse intraoperative events that contribute to the worse postoperative outcomes. Residual PH seemingly has a much greater influence on the clinical course of patients in the early postoperative period than in the long term. Several explanations are possible. First, the mean PAP might regress further over time, thereby decreasing its effect on survival, although longer term data have not supported this theory.2,3 Second, other risk factors for decreased survival, such as ischemic heart disease and congestive heart failure, might dominate the long-term survival trends. Finally, the present study might have been underpowered to detect any significant difference in survival between the 2 groups, particularly in a population consisting entirely of patients with degenerative mitral valve disease. In a multivariate analysis of all patients, decreased left ventricular ejection fraction and fibroelastic deficiency were found to be the only independent determinants of residual PH. In general, patients with degenerative mitral valve disease are much less likely to experience residual PH. This might reflect the timing of mitral valve repair, in that patients are likely referred for surgery earlier in the disease course and before the onset of significant left ventricular dysfunction. In a study of 179 patients undergoing mitral valve repair or replacement, patients undergoing repair for degenerative mitral valve disease tended to have a lower PAP than patients with nondegenerative disease.3 In the present study, a decreased ejection fraction was a significant risk factor for residual PH, possibly because it is a surrogate for disease severity and duration. Accordingly, patients referred for surgery early and before the onset of significant left ventricular dysfunction would be less likely to experience irreversible pulmonary vascular remodeling. It is less clear why fibroelastic deficiency was associated with a greater risk of residual PH. Although both have been classified within the umbrella of degenerative mitral valve disease, Barlow’s disease and fibroelastic deficiency are different entities, with unique differentiating characteristics on clinical and echocardiographic assessment.1 Compared to patients with Barlow’s disease, those with fibroelastic deficiency tend to be older at presentation, with much smaller valves that require smaller annuloplasty rings during repair. Possible explanations for the association between fibroelastic deficiency and residual PH might therefore include the greater cardiovascular disease burden in older patient populations and the greater transvalvular gradient observed postoperatively in these patients, which might be partly attributable to the use of smaller annuloplasty rings. In a subgroup analysis of patients with complete pre- and postoperative right heart catheterization data, an elevated preoperative PVR, and not the mean PAP, was a significant predictor of residual PH, supporting the theory that reactive
759
pulmonary vasoconstriction and morphologic changes in the pulmonary vasculature contribute to the persistent elevation in PAP after valve surgery. A better understanding of molecular modulators of secondary PH might inform new screening and management strategies for patients with residual PH. Given that residual PH is a significant risk factor for postoperative morbidity and mortality, it is important for physicians to be able to identify, preoperatively, those patients in whom PAP is less likely to normalize completely. Patients at risk might benefit from preoperative optimization with pharmacotherapy aimed at reversing the underlying mechanism of the PH. At our institution, patients with severe PH and impaired ventricular function receive nesiritide preoperatively, which was previously reported to significantly improve perioperative outcomes in this high-risk population.14 Pulmonary vasodilators, including bosentan, nitric oxide, sildenafil, and epoprostenol, effectively lower PAP and PVR and might be useful adjuncts in patients at risk of residual PH. Targeted preoperative optimization therapy should be studied further, because it could considerably reduce the incidence of residual PH and its adverse effect on the postoperative course. The limitations of the present study included its retrospective method, the inability to obtain preoperative values of right heart catheterization parameters for all patients, the lack of standardization of preoperative echocardiography and catheterization, and a limited sample size. The lack of longer term follow-up on the pulmonary hemodynamics, functional capacity, and cause of death data limited our conclusions regarding the natural history of residual PH and its effect on patients’ quality of life and long-term survival. Our data have suggested that residual PH in the immediate postoperative period is a clinically important entity and independently predicts postoperative outcomes. Prospective studies of larger cohorts using left atrial pressure monitoring should help to better define the mechanisms and prognostic factors for residual PH. 1. Anyanwu AC, Adams DH. Etiologic classification of degenerative mitral valve disease: Barlow’s disease and fibroelastic deficiency. Semin Thorac Cardiovasc Surg 2007;19:90 –96. 2. Li M, Dumesnil JG, Mathieu P, Pibarot P. Impact of valve prosthesispatient mismatch on pulmonary arterial pressure after mitral valve replacement. J Am Coll Cardiol 2005;45:1034 –1040. 3. Walls MC, Cimino N, Bolling SF, Bach DS. Persistent pulmonary hypertension after mitral valve surgery: does surgical procedure affect outcome? J Heart Valve Dis 2008;17:1–9. 4. Braunwald E, Braunwald NS, Ross J Jr, Morrow AG. Effects of mitral-valve replacement on the pulmonary vascular dynamics of patients with pulmonary hypertension. N Engl J Med 1965;273: 509 –514. 5. Dalen JE, Matloff JM, Evans GL, Hoppin FG Jr, Bhardwaj P, Harken DE, Dexter L. Early reduction of pulmonary vascular resistance after mitral-valve replacement. N Engl J Med 1967;277:387–394. 6. McIlduff JB, Daggett WM, Buckley MJ, Lappas DG. Systemic and pulmonary hemodynamic changes immediately following mitral valve replacement in man. J Cardiovasc Surg 1980;21:261–266. 7. Foltz BD, Hessel EA II, Ivey TD. The early course of pulmonary artery hypertension in patients undergoing mitral valve replacement with Cardioplegic arrest. J Thorac Cardiovasc Surg 1984;88:238 – 247.
760
The American Journal of Cardiology (www.ajconline.org)
8. Vincens JJ, Temizer D, Post JR, Edmunds LH Jr, Herrmann HC. Long-term outcome of cardiac surgery in patients with mitral stenosis and severe pulmonary hypertension. Circulation 1995;92:II137–II142. 9. Crawford MH, Souchek J, Oprian CA, Miller DC, Rahimtoola S, Giacomini JC, Sethi G, Hammermeister KE. Determinants of survival and left ventricular performance after mitral valve replacement: Department of Veterans Affairs Cooperative Study on Valvular Heart Disease. Circulation 1990;81:1173–1181. 10. Cevese PG, Gallucci V, Valfre C, Giacomin A, Mazzucco A, Casarotto D. Pulmonary hypertension in mitral valve surgery. J Cardiovasc Surg 1980;21:7–10. 11. Ward C, Hancock BW. Extreme pulmonary hypertension caused by mitral valve disease. Natural history and results of surgery. Br Heart J 1975;37:74 –78.
12. D’Ambra MN, LaRaia PJ, Philbin DM, Watkins WD, Hilgenberg AD, Buckley MJ. Prostaglandin E1. A new therapy for refractory right heart failure and pulmonary hypertension after mitral valve replacement. J Thorac Cardiovasc Surg 1985;89:567–572. 13. Roques F, Nashef SA, Michel P, Gauducheau E, de Vincentiis C, Baudet E, Cortina J, David M, Faichney A, Gabrielle F, Gams E, Harjula A, Jones MT, Pintor PP, Salamon R, Thulin L. Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19030 patients. Eur J Cardiothorac Surg 1999;15:816 – 822. 14. Salzberg SP, Filsoufi F, Anyanwu A, von Harbou K, Gass A, Pinney SP, Carpentier A, Adams DH. High-risk mitral valve surgery: perioperative hemodynamic optimization with nesiritide (BNP). Ann Thorac Surg 2005;80:502–506.
Spontaneous Rupture of Atrioventricular Valve Tensor Apparatus as Late Manifestation of Anti-Ro/SSA Antibody-Mediated Cardiac Disease Bettina F. Cuneo, MDa,*, Deborah Fruitman, MDb, D. Woodrow Benson, MD, PhDc, Bo-Yee Ngan, MDd, Michael R. Liske, MDe, Marie Wahren-Herlineus, MD, PhDf, S. Yen Ho, PhDg, and Edgar Jaeggi, MDh Atrioventricular (AV) block and endocardial fibroelastosis associated with dilated cardiomyopathy are the most common clinical manifestations of anti-Ro/SSA-mediated fetal cardiac disease. Valvar dysfunction has not been a prominent feature of this disease; however, recent anecdotal cases have suggested an association between rupture of the AV valve tensor apparatus and maternal anti-Ro/SSA antibodies. In the present study, we have described the clinical and laboratory findings and reviewed the published data for infants of anti-Ro/SSA-positive pregnancies with AV valve insufficiency due to chordal rupture from the papillary muscles. The histopathologic features of the papillary muscle and ventricular free wall and septum biopsy specimens were examined and compared to the sections of AV leaflets from 6 autopsied fetuses with anti-Ro/SSA-mediated complete AV block without chordal disruption. Specific epitopes to the p200 region of Ro52, and Ro60 antibodies were evaluated in cases with chordal rupture. Severe AV valve insufficiency was detected prenatally (as early as 34 weeks of gestation) or postnatally (as late as 182 days) after areas of patchy echogenicity were noted in the papillary muscle at 19 to 22 weeks of gestation. Postnatally, urgent valve surgery was performed in 5 of 6 patients; 1 of 6 patients died preoperatively. All patients tested positive for Ro52. Valve leaflet tissue from the autopsy specimens was normal. The ventricular free wall and septum biopsy specimens from a patient with chordal rupture showed normal tissue; however, the papillary muscle biopsy specimens demonstrated severe atrophy with near total replacement of myocytes by fibrosis and dystrophic calcifications, and negative immunochemistry findings. In conclusion, these findings have defined an underappreciated complication of fetal antibodymediated cardiac inflammation. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:761–766) The spectrum of maternal anti-Ro/SSA antibody-mediated fetal cardiac disease includes conduction system disease primarily involving, but not limited to, the atrioventricular (AV) node, and myocardial disease in the form of endocardial fibroelastosis and dilated cardiomyopathy.1– 4 Prenatal echocardiographic signs of fibrosis will manifest as areas of patchy echogenicity on the endocardial surfaces of the fetal heart. These often resolve but occasionally progress
a
Section of Perinatal Cardiology, The Heart Institute for Children, Hope Children’s Hospital, Oak Lawn, Illinois; bSection of Cardiology, Department of Pediatrics, Alberta Children’s Hospital, Calgery, Alberta, Canada; cHeart Institute, Department of Pediatrics, University of Cincinnati Hospital, Cincinnati, Ohio; dSection of Pathology, Department of Pediatric Laboratory Medicine and hSection of Cardiology, Department of Pediatrics, Hospital for Sick Children, Toronto, Ontario, Canada; eDivision of Cardiology, Department of Pediatrics, Monroe Carell Jr. Children’s Hospital, Vanderbilt University, Nashville, Tennessee; fRheumatology Unit, Department of Medicine, Karolinska Institute, Stockholm, Sweden; and gCardiac Morphology Unit, Royal Brompton Hospital, London, United Kingdom. Manuscript received August 28, 2010; manuscript received and accepted October 19, 2010. *Corresponding author: Tel: (708) 684-5580; fax: (708) 684-4068. E-mail address:
[email protected] (B.F. Cuneo). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.059
to endocardial fibroelastosis, ventricular dysfunction, and dilated cardiomyopathy.3,4 Although the papillary muscles and chordae of the AV valves are common sites of patchy echogenicity, severe AV-valvar insufficiency due to dysfunction of the tensor apparatus has rarely been reported. Recently, we cared for 2 infants from anti-Ro/SSA-positive pregnancies with fetal endocardial fibroelastosis and spontaneous disruption of the tensor apparatus of the AV valve (Figure 1), resulting in sudden-onset, severe AV valve insufficiency. Four similar cases have been previously reported.5– 8 The purpose of the present report was to summarize the clinical findings from all 6 cases and to provide new pathologic-immunologic insights into this unusual manifestation of immune-mediated cardiac disease. We speculated that the spontaneous disruption of the AV valve tensor apparatus might become more common as the survival of affected fetuses from anti-Ro/SSA-positive pregnancies continues to improve.9 –11 Case Reports The 4 previously reported patients have been designated patients 3 to 65– 8 and the new patients as patients 1 and 2. www.ajconline.org
762
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Diagram of ruptured chordae from distal tips of papillary muscle in SSA/Ro antibody-mediated cardiac disease according to the intraoperative findings.
birth), with no change in the patchy echogenicity of the papillary muscles. A 2.2 kg infant was delivered electively at term (38 weeks of gestation). The infant received 1 dose of intravenous immunoglobulin (2 g/kg) after delivery and oral prednisone for 2 months. The postnatal echocardiograms at birth and at 6 weeks showed mild residual patchy echogenicity of the tricuspid valve papillary muscles but no tricuspid valve insufficiency. At 6 months of age, the infant developed cyanosis and respiratory distress. The preoperative echocardiogram revealed severe tricuspid insufficiency due to a flailed anterior tricuspid valve leaflet. The intraoperative findings included disruption of the chordal attachments of the anterior leaflet from the head of the papillary muscle. The chordae of the posterior leaflet appeared brittle and fibrotic; the chordae of the septal leaflet appeared unaffected. The disrupted chordae were replaced by 5-0 Gore-Tex, and back-up chordae were placed to augment the posterior leaflet. Biopsy specimens from the ventricular septum and papillary muscles were obtained. The child did well and, at age 3 years, had normal biventricular function, mild tricuspid insufficiency, and an echogenic tricuspid tensor apparatus. Patient 2: A secundigravida anti-Ro/SSA-positive 25year-old woman presented at 19 weeks of gestation with fetal tachycardia. The fetal echocardiogram showed a structurally normal heart, a small pericardial effusion, echogenic tricuspid valve papillary muscles, and nonsinus tachycardia with 1:1 AV association and a rate of 205 beats/min. The biventricular function was subjectively normal. During the next few weeks, the tachycardia resolved. At 34 weeks of gestation, tricuspid valve insufficiency with dilation of the right atrium and right ventricle was noted. Despite treatment with transplacental dexamethasone 4 mg/day and one dose of intravenous immunoglobulin (1 g/kg), the insufficiency rapidly worsened, prompting delivery at 36 weeks of gestation. The infant weighed 2.8 kg. The postnatal echocardiogram showed normal biventricular function, a flailed anterior tricuspid valve leaflet with free tricuspid insufficiency, and mild pulmonary insufficiency. At 23 days, the anterior leaflet of the mitral valve became flail with severe mitral insufficiency. The infant developed pulmonary hemorrhage and renal failure and died at 32 days of age. A consent for autopsy was refused. Methods
Figure 2. Two-dimensional fetal echocardiogram of patient 1 at 21 weeks of gestation. Localized areas of increased echogenicity in posterior medial papillary muscle of tricuspid valve, both papillary muscles of mitral valve, and proximal portion of intra-atrial septum shown (arrows).
Patient 1: A 27-year-old anti-Ro/SSA-positive secundigravida mother was referred because of fetal bradycardia at 21 weeks of gestation. The fetal echocardiogram revealed a 3° AV block and focal areas of increased echogenicity in the papillary muscles of both AV valves and in the left atrium (Figure 2). The mother was given oral dexamethasone (8 mg/day for 2 weeks, 4 mg/day from 23 to 30 weeks, and 2 mg/day until delivery) and intravenous immunoglobulin (70 g every 2 weeks to
Intraoperative biopsies of the papillary muscles, ventricular septum, and anterior wall of the right ventricle were immediately fixed in buffered formalin for 24 hours. The tissues were embedded in paraffin, and 5-m tissue sections were prepared for microscopy and histochemical staining, including hematoxylin-eosin staining, periodic acid-Schiff staining for the demonstration of tissue glycogen, and Movat pentachrome staining for the demonstration of elastin fibers, tissue protein, and collagen. Immunostain procedures were performed using an automated immunostainer (Benchmark, Ventana/Roche Diagnostics, Tucson, Arizona). Antigen retrieval procedures were applied to the dewaxed and rehydrated tissue slides. Biotinylated antibodies to cleaved caspase 3 were used at a dilution of 1/200.
Congenital Heart Disease/SSA/Ro Antibody-Mediated AV Valve Disease
763
Table 1 Findings in 6 patients with SSA/Ro antibody-mediated disruption of atrioventricular valve tensor apparatus Pt. No.
1 2 37 48 59 610
GA (wk)
21 19 19 20 — 22
Maternal Antigen
Ro52 Ro60, p200 Ro52 Ro60 Ro52, Ro60, P200 Ro52, Ro60, p200, La Ro and La Ro
Location of Echogenicity
Rhythm 3° AVB
TV
MV
PM*
⫹ ⫹ ⫹ ⫹ — ⫹
⫹ 0 ⫹ ⫹ — ⫹
⫹ ⫹ ⫹ ⫹ — ⫹
⫹ 0 0 ⫹ — ⫹
Treatment
D D D D
⫹ IVIG ⫹ IVIG ⫹ IVIG ⫹T — —
Age of Insufficiency (days) 182 21 90 90 21 14
Affected Valve TV
MV
⫹ ⫹ ⫹ ⫹ 0 ⫹
0 ⫹ ⫹ 0 ⫹ ⫹
TV Repair
MV Repair
Alive
⫹ 0 0 ⫹ 0 ⫹
0 0 ⫹† 0 ⫹† ⫹
⫹ 0 ⫹ ⫹ ⫹ ⫹
AVB ⫽ atrioventricular block; D ⫽ dexamethasone; GA ⫽ gestational age; IVIG ⫽ intravenous immunoglobulin; MV ⫽ mitral valve; PM ⫽ papillary muscle; Pt. No. ⫽ patient number; T ⫽ terbutaline; TV ⫽ tricuspid valve. * All patients with PM involvement had chordal involvement. † Valve replaced.
Terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling (TUNEL) analysis was performed on the papillary muscle tissue. In situ hybridization was performed using an automated stainer (Discovery model, Ventana/Roche Diagnostics). The slides were incubated with probe and recombinant terminal deoxynucleotidyl enzyme (Gibco, Carlsbad, California), (Roche, Basel, Switzerland), and the in situ hybridization reaction performed under conditions preset by the Autostainer program. The tissue slides were counter stained with hematoxylin. From the archive at the Royal Brompton Hospital, we reviewed the histologic features of the mitral and tricuspid valve leaflets of 6 fetal hearts with maternal anti-Ro/SSA antibodies. These cases had complete AV block without disruption of the tensor apparatus of the AV valves. The tissue blocks had been serially cut at 10 m, and sections at 250-m intervals were stained with Masson’s trichrome. From the remainder, we retrieved representative sections of the AV valves and stained them with elastic-Van Gieson stain for examination of the leaflet architecture. These cases were examined for comparison with patient 1. We obtained maternal serum from patients 1 and 2 and the previously published patients 3 and 45,6 for Ro52, Ro60, and La antibody testing. Using the INNO-LIA ANA Update (Innogenetics, Ghent, Belgium) line immunoblot assay according to the manufacturer’s instructions, Ro52 antibodies and antibodies to a specific epitope encoded within amino acid 200-239 (p200) of Ro52 were quantified using enzymelinked immunosorbent assay with recombinant, bacterially expressed, and purified Ro52 antigen encoded in the pMAL vector system and a synthetized p200 peptide, as previously described.12 The sera were tested at a dilution of 1:500, and bound antibodies were detected by affinity-purified alkaline phosphatase-conjugated anti-IgG antibodies using p-nitrophenyl phosphate disodium salt as a substrate and registration of the optical density at 405 nm. Sera with high and low Ro52 and p200 antibody levels and mouse anti-human Ro52 monoclonal antibodies generated by us were used as internal controls.12 Results The results of the antibody testing, pre- and postnatal clinical findings, and treatment are summarized in Table 1.
Compared to the normal papillary muscle (Figure 3), the papillary muscles of patient 1 showed severe atrophy with near total replacement of myocytes by fibrosis and dystrophic calcifications (Figure 3). The residual myocytes had features of compensatory hypertrophy, including bizarre enlarged nuclei and increased sarcoplasm (Figure 3). TUNEL staining demonstrated limited apoptosis, confirmed by immunohistochemical staining, with an antibody to cleaved caspase 3, one of the enzymes that executes apoptosis (Figure 3). No acute inflammatory infiltrates were found in the vascular or perivascular spaces of the papillary muscles. The right ventricular septal and anterior muscular biopsy specimens showed normal-size myocytes and only insignificant interstitial fibrosis and mild subendocardial elastosis. The immunostaining results for cleaved caspase 3 and TUNEL staining were negative. In contrast to the findings from patient 1, the composition and structure of the AV valve leaflets from the 6 cases of fetuses with anti-Ro/SSA-mediated AV block were not different from those of the normal specimens (Figure 4). At 19 weeks, the basal part of the leaflets showed dense fibrous tissue with strands of elastic tissue, with the expanded distal part containing mainly pale-staining mesenchymal cells. At 37 weeks, the leaflet had matured into 4 recognizable layers of atrialis, spongiosa, fibrosa, and ventricularis. The chordae showed normal features of densely aligned fibrous tissue in the core and elastic fibers in the periphery. Discussion All 6 patients had in common the exposure to maternal anti-Ro/SSA antibodies and the findings of papillary muscle and chordal echogenicity followed by late prenatal or postnatal onset of severe AV valve insufficiency due to rupture of the AV valve chordae from the distal head of a papillary muscle. From the inspection at surgery and histopathologic assessment of the biopsy specimens, the pathologic findings in the papillary muscle and valve-supporting apparatus were markedly abnormal. The major histopathologic findings showed severe atrophy with near total replacement of the myocytes by fibrosis and dystrophic calcifications of the papillary muscle. In contrast, the valve leaflets from the 6 infants with anti-Ro/SSA-mediated conduction system disease were quite normal. This was not surprising, given the
764
The American Journal of Cardiology (www.ajconline.org)
Figure 3. Pathologic and immunohistochemistry studies of papillary muscle biopsy specimens from (A) normal subject and (B–D) patient 1. (A) Hematoxylin-eosin stain of normal papillary muscle. (B) Hematoxylin-eosin stain showing marked fibrosis and loss of muscle cells, with only few residual myocytes (small arrow). Calcified remains of myocytes seen as dystrophic calcifications within fibrous tissue (large arrow). (C) Movat pentachrome stain showing compensatory hypertrophy of residual myocytes (pink cells). Adjacent stroma was fibrotic with myxoid degeneration. Early deposition of elastin fibers noted (black staining). (D) Immunohistologic stain using antibody to cleaved caspase 3. Arrow indicates myocyte undergoing apoptosis.
different origins of the valve leaflets and chordae (endocardium) compared to the papillary muscle (myocardium).13,14 Together, these findings are consistent with those from previous studies suggesting that the immune-mediated insult to the fetal heart is directed toward cell lines derived from myocardium. That the tricuspid valve was usually involved might relate to the systemic right ventricular pressure and dominance of the right ventricle in fetal circulatory physiology. Differences among the 6 patients included the timing of the chordal disruption from the papillary muscle, in utero and postnatal treatments, and variations in the presence and extent of conduction system disease. From the cohort we examined, the occurrence of spontaneous AV valve disruption could not be predicted from maternal antibody specificity or other clinical findings, nor was the complication prevented by transplacental treatment. The only consistent risk factor was papillary muscle echogenicity. Other poten-
tial contributing mechanisms for chordal rupture that could not be absolutely excluded were steroid-induced necrosis, ischemia, and maternal intake of prostaglandin synthetase inhibitors. No history of the latter was reported, and steroids were not given in any cases described. A popular hypothesis to explain the initiation of anti-Ro/ SSA cardiac disease has been that the intracellular Ro/SSA, La/SSB proteins are translocated to the cell surface secondary to the normal developmental apoptosis of cardiac cells and thus become accessible to binding by maternal antibodies.15 This interaction of cell surface expression of the Ro/SSA particle with the maternal anti-Ro/SSB, anti-La/ SSB antibodies would then not only inhibit clearance of the apoptotic cells by fetal cardiocytes but allow macrophages to engulf the antigen–antibody complexes, leading to the secretion of pro-inflammatory and fibrosing cytokines and leading to cardiac damage.16 Thus, myocardial disease would predominantly be the result of a complex cascade
Congenital Heart Disease/SSA/Ro Antibody-Mediated AV Valve Disease
765
Figure 4. Leaflet tissues from ruptured valve apparatus unavailable for examination. Instead, histologic sections of 2 patients with congenital heart block and intact valve apparatus shown, alongside sections from 2 normal patients. Basal and middle parts of leaflets were compact with fibrous and elastic tissue at 19 weeks (dark coloration). As shown in normal specimens at 28 weeks with magnification of tricuspid valve, elastic and fibrous tissues became more organized and resembled layered arrangement of atrialis (A), spongiosa (S), fibrosa (F), and ventricularis (V) at 37 weeks. Tricuspid valve in patient with heart block at 37 weeks magnified to show layered appearance comparable to normal specimen at same age. Chordae (cord) appeared normal with dense and aligned arrangement of collagen fibers. Elastic van Geison stain.
of maternal antibody-triggered inflammatory and fibrotic events. The pathogenesis of chordal rupture from the papillary muscle might be similar to that proposed for AV block, namely that maternal anti-Ro/SSA antibodies bind to the specialized cells of the fetal electrical conduction system and cardiac myocytes and subsequently evoke inflammation and fibrosis.3,17 This has been supported by the findings of anti-Ro/SSA antibodies in all 6 affected mothers and the positive TUNEL staining results and caspase-3 activity. We speculated that antibody-mediated inflammation followed weeks later by extensive fibrosis of the chordae and/or papillary muscles might make them fragile and susceptible to rupture. At rupture, only scar tissue without signs of acute inflammation would be present. We have further speculated that the initial inflammatory insult might be suppressed, but not reversed, by steroids and intravenous immunoglobulin. Thus, the full expression of immune-mediated AV valve disease would not be prevented but would be postponed until late gestation or, more commonly, after birth. Although fibrosis is an end-stage result of myocyte loss that occurs in both dilated cardiomyopathy and rupture of the AV valve tensor apparatus, the pathologic findings differ between the 2. In dilated cardiomyopathy, widespread fibrosis and necrosis of the myocytes will be present, with diffuse deposition of IgG in the endocardium and in the
mitral valve papillary muscles.3,4 In contrast, the patient we reported showed fibrotic replacement of the myocytes limited to the tensor apparatus, with negative immunohistochemical findings, and the biopsy specimens taken from other right ventricular areas showed neither inflammation nor fibrosis. We speculate that that ongoing loss of the myocytes led to the ongoing occurrence of papillary muscle fibrosis and chordal rupture from the heads of the affected papillary muscle. Thus, it would appear that chordal rupture and dilated cardiomyopathy are 2 distinct manifestations of anti-Ro/SSA-mediated fetal cardiac disease. Acknowledgment: The authors thank Shyam Sathanandam, MD, for Figure 1 and Rita Allen, BA, for manuscript preparation. 1. Buyon JP, Hiebert R, Copel J, Craft J, Friedman D, Katholi M, Lee LA, Provost TT, Reichlin M, Rider L, Rupel A, Saleeb S, Weston WL, Skovron ML. Autoimmune-associated congenital heart block: demographics, mortality, morbidity and recurrence rates obtained from a national neonatal lupus registry. J Am Coll Cardiol 1998;31:1658 – 1666. 2. Hogg GR. Congenital, acute lupus erythematosus associated with subendocardial fibroelastosis. Am J Clin Pathol 1957;28:648 – 654. 3. Nield LE, Silverman ED, Taylor GP, Smallhorn JF, Mullen BM, Silverman NH, Finley JP, Law YM, Human DG, Seaward G, Hamilton RM,
766
4.
5.
6.
7.
8.
9.
10.
The American Journal of Cardiology (www.ajconline.org) Hornberger LK. Maternal anti-Ro and anti-La antibody-associated endocardial fibroelastosis. Circulation 2002;105:843– 848. Nield LE, Silverman ED, Smallhorn JF, Taylor GP, Mullen BM, Benson LN, Hornberger LK. Endocardial fibroelastosis associated with maternal anti-Ro and anti-La antibodies in the absence of atrioventricular block. J Am Coll Cardiol 2002;40:796 – 802. Cuneo BF, Strasburger JF, Niksch A, Ovadia M, Wakai RT. An expanded phenotype of maternal SSA/SSB antibody associated fetal cardiac disease. J Matern Fetal Neonat Med 2009;22:223–228. Fleming GA, Scholl FG, Kavanaugh-McHugh A, Liske MR. A case of an infant with flail tricuspid valve due to spontaneous papillary muscle rupture: was neonatal lupus the culprit? Pediatr Cardiol 2008;29:442– 445. Hamaoka A, Shiraishi I, Yamagishi M, Hamaoka K. A neonate with the rupture of mitral chordae tendineae associated with maternalderived anti-SSA antibody. Eur J Pediatr 2009;168:741–743. Weber HS, Myers JL. Maternal collagen vascular disease associated with fetal heart block and degenerative changes of the AV valves. Pediatr Cardiol 1994;15:204 –206. Jaeggi ET, Hamilton RM, Silverman ED, Zamora SA, Hornberger LK. Outcome of children with fetal, neonatal or childhood diagnosis of isolated congenital atrioventricular block. J Am Coll Cardiol 2002;39: 130 –137. Cuneo BF, Lee M, Roberson D, Niksch A, Ovadia M, Parilla BV, Benson DW. A management strategy for fetal immune-mediated atrioventricular block. J Matern Fetal Neonat Med 2010;23:1400 –1405.
11. Lopes LM, Tavares GMP, Damiano AP, Lopes MAB, Aiello VD, Schultz R, Zugaib M. Perinatal outcome of fetal atrioventricular block: one hundred and sixteen cases from a single institution. Circulation 2008;118:1268 –1275. 12. Strandberg L, Salomonsson S, Bremme K, Sonesson SE, WahrenHerlenius M. Ro52, Ro60 and La IgG autoantibody levels and Ro52 IgG subclass profiles longitudinally throughout pregnancy in congenital heart block risk pregnancies. Lupus 2006;15:346 –353. 13. Lincoln J, Alfieri CM, Yutzey KE. Development of heart valve leaflets and supporting apparatus in chick and mouse embryos. Dev Dyn 2004;230:239 –250. 14. de Lange FJ, Moorman AFM, Anderson RH, Manner J, Soufan AT, de Gier-de Vries C, Schneider MD, Webb S, van den Hoff MJB, Christoffels VM. Lineage and morphogenetic analysis of the cardiac valves. Circ Res 2004;95:645– 654. 15. Miranda ME, Tseng CE, Rashbaum W, Ochs RL, Casiano CA, Di Donato F, Chan EK, Buyon JP. Accessibility of SSA/Ro and SSB/La antigens to maternal autoantibodies in apoptotic human fetal cardiac myocytes. J Immunol 1998;161:5061–5069. 16. Clancy RM, Neufing PJ, Zheng P, O’Mahony M, Nimmerijahn F, Gordon TP, Buyon JP. Impaired clearance of apoptotic cardiocytes is linked to anti-SSA/Ro and -SSB/La antibodies in the pathogenesis of congenital heart block. J Clin Invest 2006;116:2413–2422. 17. Silverman E, Mamula M, Hardin JA, Laxer R. Importance of the immune response to the Ro/La particle in the development of congenital heart block and neonatal lupus erythematosus. J Rheumatol 1991; 18:120 –124.
Cardiac Magnetic Resonance Imaging and the Assessment of Ebstein Anomaly in Adults Sergey Yalonetsky, MDa, Daniel Tobler, MDa, Matthias Greutmann, MDa, Andrew M. Crean, MDa,b, Bernd J. Wintersperger, MDb,c, Elsie T. Nguyen, MDb, Erwin N. Oechslin, MDa, Candice K. Silversides, MDa, and Rachel M. Wald, MDa,d,* No published studies have evaluated the role of cardiac magnetic resonance (CMR) imaging for the assessment of Ebstein anomaly. Our objective was to evaluate the right heart characteristics in adults with unrepaired Ebstein anomaly using contemporary CMR imaging techniques. Consecutive patients with unrepaired Ebstein anomaly and complete CMR studies from 2004 to 2009 were identified (n ⴝ 32). Volumetric measurements were obtained from the short-axis and axial views, including assessment of the functional right ventricular (RV) end-diastolic volume (EDV) and end-systolic volume. The volume of the atrialized portion of the right ventricle in end-diastole was calculated as the difference between the total RVEDV and the functional RVEDV. The reproducibility of the measurements in the axial and short-axis views was determined within and between observers. The median value derived from the short-axis and axial views was 136 ml/m2 (range 59 to 347) and 136 ml/m2 (range 63 to 342) for the functional RVEDV, 153 ml/m2 (range 64 to 441) and 154 ml/m2 (range 67 to 436) for the total RVEDV, 49% (range 32% to 46%) and 50% (range 40% to 64%) for the functional RV ejection fraction, respectively. The axial measurements demonstrated lower intraobserver and interobserver variability than the short-axis approach for all values, with the exception of the intraobserver functional RVEDV and interobserver total RVEDV for which the limits of agreement and variance were not significantly different between the 2 views. In conclusion, measurements of right heart size and systolic function in patients with Ebstein anomaly can be reliably achieved using CMR imaging. Axial imaging appeared to provide more reproducible data than that obtained from the short-axis views. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:767–773) In recent years, cardiac magnetic resonance (CMR) imaging has emerged as the reference standard for cardiac imaging in patients with many forms of congenital heart disease, in particular for lesions that affect the right ventricle.1,2 Despite the integration of CMR imaging for the initial assessment and subsequent follow-up of adults with congenital heart disease in the management guidelines from groups worldwide,3–5 no data are available to define the role of CMR imaging in the assessment of Ebstein anomaly. This is particularly striking, given the recognized complexity of right ventricular (RV) anatomy in this lesion. The only quantitative imaging parameter used to guide contemporary surgical referral for those with Ebstein anomaly in consensus statements is enlarged cardiothoracic ratio.3– 6 No puba Division of Cardiology, Toronto Congenital Cardiac Centre for Adults, Peter Munk Cardiac Centre, Toronto General Hospital, University Health Network, University of Toronto, Toronto, Ontario, Canada; bDepartment of Medical Imaging, University Health Network, University of Toronto, Toronto, Ontario, Canada; cDepartment of Clinical Radiology, University of Munich Hospitals, University of Munich, Munich, Germany; and dDivision of Cardiology, Labatt Family Heart Centre, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada. Manuscript received August 15, 2010; manuscript received and accepted October 19, 2010. *Corresponding author: Tel: (416) 340-5502; fax: (416) 340-5014. E-mail address:
[email protected] (R.M. Wald).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.058
lished studies have focused on the CMR imaging assessment of Ebstein anomaly. We hypothesized that multiplanar CMR imaging would allow for precise and reproducible measurements of the right heart size and systolic function in those with the Ebstein anomaly. Therefore, we sought to evaluate the role of CMR imaging in the evaluation of right heart disease in adults with an unrepaired Ebstein anomaly, with particular emphasis on the optimal view for quantification of right heart size and function. Methods Consecutive adults with an unrepaired Ebstein anomaly who had been referred for a CMR study from May 2004 to December 2009 at our hospital were retrospectively identified. This initial point was chosen, because it reflected a change in the CMR imaging protocol and sequence acquisition at our institution that continues to be used at present. The patients were included if the CMR studies were technically adequate for image analysis with a complete complement of cine views (4-chamber, axial, and short-axis views). The exclusion criteria included previous tricuspid valve surgery, Ebstein-like tricuspid valve in the context of complex congenital heart disease (e.g., congenitally corrected transposition of the great arteries), and incomplete CMR studies with inadequate image quality for quantitative analysis. Echocardiograms that were contemporaneous with www.ajconline.org
768
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Three-dimensional reproduction of heart with Ebstein anomaly demonstrating views of tricuspid valve from (A) short axis and (B) axial imaging. aRV, atrialized right ventricle; fRV, functional right ventricle; LA, left atrium; LV, left ventricle; RA, right atrium; TVAL, tricuspid valve anterior leaflet; TVPL, tricuspid valve posterior leaflet; TVSL, tricuspid valve septal leaflet.
Figure 2. (A–D) Systolic and diastolic contours of functional right ventricle (fRV) and atrialized portion of right ventricle (aRV) in (A,B) axial and (C,D) short-axis views. (E) Severity index representing ratio of areas of right atrium (RA) and atrialized right ventricle (aRV) in numerator and summation of fRV and left atrium (LA) and left ventricular areas in denominator (i.e., severity index ⫽ [right atrial area ⫹ atrialized right ventricular area]/[functional right ventricular area ⫹ left atrial area ⫹ left ventricular area]). (F) Degree of apical displacement of septal leaflet of tricuspid valve (in millimeters) measured in ventricular diastole.
the CMR studies (most echocardiograms were performed within 2 weeks of the CMR studies, and, with one exception, all echocardiograms were performed within 3 months of the CMR studies) were reviewed specifically for tricuspid valve morphology (displacement of the septal and/or posterior leaflet and a “sail-like” quality of the anterior leaflet) to confirm the diagnosis of the Ebstein anomaly and to assess tricuspid valve function (i.e., incompetence or stenosis).6 All studies were performed using a standard commercially available 1.5T scanner. The technical parameters of our CMR studies have recently been published.7 In brief, steady-state free-precession imaging of the ventricles in the 2- and 4-chamber planes was performed, followed by the prescription of contiguous short-axis slices oriented perpendicular to the long axis of the heart from the base to the apex. Axial imaging was obtained from a point superior to the RV outflow tract and pulmonary valve to a point inferior to the diaphragmatic surface of the right ventricle to cover the heart in its entirety (Figure 1). The typical slice thickness was 6 to 8 mm for the axial stack and 8 to 10 mm for the short-axis view. To fully characterize right heart size and function, the functional RV, atrialized RV, and total RV volumes were measured. The functional RV was defined as the aspect of the ventricle distal to the attachment points of the tricuspid valve leaflets. The border demarcating the functional RV from the atrialized RV was defined as a line connecting the free wall attachment of the anterior tricuspid valve leaflet at the level of the annulus and the septal attachment of the apically displaced leaflet, drawn in end-diastole and endsystole on the axial and short-axis views, respectively (Figure 2). The border between the atrialized RV and the morphologic right atrium was defined as a line connecting the free wall attachment of the anterior tricuspid valve leaflet
Congenital Heart Disease/CMR Imaging and Ebstein Anomaly
769
Table 1 Cardiac magnetic resonance (CMR) imaging-derived ventricular volumes and ejection fraction Variable Total right ventricular end-diastolic volume (mL) Total right ventricular end-diastolic volume indexed (mL/m2) Functional right ventricular end-diastolic volume (mL) Functional right ventricular end-diastolic volume indexed (mL/m2) Atrialized right ventricular end-diastolic volume indexed (mL/m2) Functional right ventricular end-systolic volume (mL) Functional right ventricular end-systolic volume indexed (mL/m2) Functional right ventricular stroke volume (mL) Functional right ventricular ejection fraction (%) Left ventricular end-diastolic volume (mL) Left ventricular end-diastolic volume indexed (mL/m2) Left ventricular end-systolic volume (mL) Left ventricular end-systolic volume indexed (mL/m2) Left ventricular stroke volume (mL) Left ventricular ejection fraction (%)
Short-Axis View
Axial View
p Value
271 (102–719) 153 (64–441) 239 (95–565) 136 (59–347) 19 (2–94) 119 (43–306) 71 (27–157) 113 (19–298) 49 (32–64) 106 (47–170) 61 (27–90) 70 (21–131) 42 (12–68) 66 (33–108) 61 (51–74)
268 (107–711) 154 (67–436) 240 (101–558) 136 (63–342) 16 (4–94) 117 (44–270) 70 (27–140) 118 (57–329) 50 (40–64) — — — — — —
0.5 0.5 0.2 0.1 0.3 0.2 0.2 0.002 0.041 — — — — — —
Data are presented as median (range).
and the tricuspid valve annulus (i.e., the point of presumed septal leaflet attachment, adjacent to the mitral valve insertion point, had there been normal tricuspid valve development). The total RV volume was defined as a summation of the functional RV and atrialized RV volumes. The end-diastolic volume (EDV) and end-systolic volume of the functional right ventricle and the EDV of the total RV were measured in the axial and short-axis views. Most RV trabeculae were excluded to the blood pool. The stroke volume was derived from the difference between the EDV and end-systolic volume. The left ventricular EDV and end-systolic volumes were measured in the short-axis view, according to accepted practice.8 The volumes were derived using a standard commercially available software package (MASS, Medis, Leiden, The Netherlands) and were indexed to the body surface area. From the 4-chamber view, the absolute apical displacement of the septal tricuspid valve leaflet was measured in ventricular diastole (Figure 2) and indexed to the body surface area. The areas of the right atrium, atrialized RV, functional RV, left atrium, and left ventricle were traced in end-diastole on the 4-chamber view and are reported as a severity score, as previously described echocardiographically by Celermajer et al9 (Figure 2): severity index ⫽ (right atrial area ⫹ atrialized RV area)/(functional RV area ⫹ left atrial area ⫹ left ventricular area). The companion measures of septal leaflet displacement and severity score were derived from the corresponding echocardiographic views. The RV volumes of 10 randomly selected patients (31% of the population) were remeasured by the same investigator who was unaware of the original measurements to determine the intraobserver variability. The interval between the first and second measurements for a single patient was 90 to 120 days. Similarly, the RV volume of an additional 10 randomly selected patients (31% of the population) was measured independently by a second investigator who was unaware of the results derived by the first investigator. Statistical analysis was performed using the Statistical Package Social Sciences, version 17.0, software program (SPSS, Chicago, Illinois). The data are described as the mean ⫾ SD or the median and range, as appropriate. The
Pearson or Spearman correlation coefficient, as appropriate, was used to determine the relation between measurements derived from the axial and short-axis views. The betweengroup comparisons (such as axial vs short-axis volumetrics) were performed using Student’s t test or Mann-Whitney U test for continuous variables. The chi-square test or Fisher’s exact test was used to compare the categorical variables. Bland-Altman plots were used to display the variation between the RV volumes calculated from the axial and shortaxis views. For each of the methods (axial and short-axis views), the mean percentage of difference, standard deviation, and variance were also calculated to determine the intra- and interobserver variability.10 Statistical significance was set at p ⬍0.05. The research ethics board at our institution approved the study. Results Clinical data were available for 31 patients (median age at CMR study 39 years, range 21 to 68; 72% women). Most were classified as New York Heart Association (NYHA) functional class I (20 [65%]) or NYHA functional class II (10 [32%]), with 1 as NYHA functional class III (3%) and none as NYHA functional class IV. A significant subset (15 [48%]) of patients had atrial level shunts (patent foramen ovale or secundum atrial septal defect) on the echocardiogram and/or CMR imaging study; moderate or severe tricuspid regurgitation was evident in most patients on the echocardiogram (23 [74%]). At least one episode of sustained atrial arrhythmia (lasting ⱖ30 seconds) was documented in nearly ½ of the patient population (14 [45%]). A total of 41 CMR studies were identified; however, 32 complete CMR studies were analyzed (3 studies were terminated early because of claustrophobia, 4 studies did not have both axial and short-axis cine imaging, and 2 studies had insufficient cardiac gating in the context of persistent arrhythmia with consequent degradation in image quality). A single patient was referred to our institution for CMR analysis alone without any accompanying clinical data. From the CMR measurements, the median apical displacement of the tricuspid valve was 3.0 cm (range 1.7 to 6.7) and
770
The American Journal of Cardiology (www.ajconline.org)
Figure 3. Bland-Altman plots of variability for measurements of functional right ventricular end-diastolic volumes (RVEDV) between axial and short-axis views with absolute measures (A) and as percent difference (B). Intraobserver variability for functional RVEDV in short axis (C) and axial (D) views. Interobserver variability for functional RVEDV shown in short axis (E) and axial (F) views.
the median value indexed to the body surface area was 1.6 cm/m2 (range 0.8 to 4). The median severity score was 0.4 (range 0.2 to 2.2). From the echocardiographic measurements, the median apical displacement of the tricuspid valve was 3.1 cm (range 1.7 to 5.5), and the median value indexed to the body surface area was 1.6 cm/m2 (range 0.9 to 3.4). The median severity score was 0.37 (range 0.18 to 1.87). The differences in the values obtained from echocardiography and CMR imaging were not statistically significant
(measures of absolute tricuspid valve displacement, p ⫽ 0.453; indexed tricuspid valve displacement, p ⫽ 0.544; and severity score, p ⫽ 0.647). The CMR characteristics relating to the heart volumes and ejection fraction for the population studied are summarized in the Table 1. A comparison of the axial and shortaxis measurements using the Bland-Altman method is demonstrated in Figure 3 and Table 2. The intra- and interobserver variabilities for the RV volumes and ejection
Congenital Heart Disease/CMR Imaging and Ebstein Anomaly
771
Table 2 Relation between right ventricular (RV) measurements using axial and short-axis slices in adults with Ebstein anomaly10 Variable
Functional RVEDV
Functional RVESV
Functional RVEF
Total RVEDV
Mean difference (%) Limits of agreement (%) SD (%) Variance (%) Correlation (p value)
0.8 ⫺8.2 and 9.8 4.5 20.3 0.967 (p ⬍0.0001)
0.1 ⫺18.6 and 18.4 9.2 85.6 0.966 (p ⬍0.0001)
1.9 ⫺16.6 and 20.3 9.2 85.1 0.765 (p ⬍0.0001)
0.1 ⫺8.3 and 8.5 4.2 17.8 0.966 (p ⬍0.0001)
EDV ⫽ end-diastolic volume; EF ⫽ ejection fraction; ESV ⫽ end-systolic volume; RV ⫽ right ventricle. Table 3 Intraobserver variability10 of axial and short-axis slices in adults with Ebstein anomaly Variable Short-axis imaging Mean difference (%) Limits of agreement (%) SD (%) Variance (%) Axial imaging Mean difference (%) Limits of agreement (%) SD (%) Variance (%)
Functional RVEDV
Functional RVESV
Functional RVEF
Total RVEDV
1.1 ⫺9.1 and 11.3 5.1 26.0
8.3 ⫺13.3 and 30.0 10.8 116.7
⫺8.3 ⫺25.6 and 9.0 8.7 75.0
⫺2.3 ⫺13.9 and 9.4 5.8 33.9
0.4 ⫺9.3 and 10.1 4.8 23.5
6.4 ⫺10.0 and 22.9 8.2 67.4
⫺6.2 ⫺15.1 and 3.9 4.7 22.5
⫺1.2 ⫺10.3 and 7.9 4.6 20.7
Abbreviations as in Table 2. Table 4 Interobserver variability10 of axial and short-axis slices in adults with Ebstein anomaly Variable Short-axis imaging Mean difference (%) Limits of agreement (%) SD (%) Variance (%) Axial imaging Mean difference (%) Limits of agreement (%) SD (%) Variance (%)
Functional RVEDV
Functional RVESV
Functional RVEF
Total RVEDV
2.6 ⫺18.2 and 23.5 10.6 113.0
⫺1.0 ⫺22.3 and 20.4 10.9 118.9
5.3 ⫺19.0 and 29.5 12.4 152.7
3.1 ⫺12.6 and 18.8 8.0 64.1
⫺4.1 ⫺18.1 and 10.0 7.2 51.3
⫺6.5 ⫺24.9 and 11.9 9.4 88.1
9.0 ⫺2.0 and 19.9 5.6 31.2
⫺2.2 ⫺18.0 and 14.0 8.0 64.0
Abbreviations as in Table 2.
fraction assessed by the Bland-Altman method are listed in Tables 3 and 4. Figure 3 illustrates the intra- and interobserver variability of functional RVEDV measurements. The intraand interobserver variability for the axial measurements were lower than those with the short-axis approach for all values, with the exception of the intraobserver functional RVEDV and interobserver total RVEDV, for which the limits of agreement and variance were not significantly different between the 2 views. Good correlations were found between the atrialized RVEDV from the axial and short-axis views and the following CMR measurements: magnitude of apical displacement of the septal leaflet indexed to the body surface area (r ⫽ 0.459, p ⫽ 0.008, and r ⫽ 0.464, p ⫽ 0.008, respectively) and the severity score (r ⫽ 0.459, p ⫽ 0.008, and r ⫽ 0.464, p ⫽ 0.008, respectively). Discussion The present study is the first to have focused on the CMR features of Ebstein anomaly. Quantitative measures of right
heart size and systolic function were achievable and reproducible using contemporary CMR methods, even in the presence of severe disease and significant distortion in the RV anatomy. Axial imaging appeared to provide more reliable measurements than short-axis imaging in this setting. New expressions of disease severity, such as the atrialized RV volume, have emerged from the present study; however, the clinical significance of these CMR parameters has yet to be determined. Precise morphologic characterization is vitally important in patients with Ebstein anomaly. Specifically, RV size, determined echocardiographically, has been directly related to prognosis and outcome in this population.11–13 Unlike children, in whom permissive acoustic windows allow characterization of right heart anatomy using echocardiography, the limited views in adults have led to the emergence of CMR imaging as the reference standard for the assessment of cardiac size and systolic function, particularly for the right heart.14 –17 Published data related to the imaging char-
772
The American Journal of Cardiology (www.ajconline.org)
acteristics in adults with the Ebstein anomaly have been exclusively echocardiographic. There is a striking paucity of CMR-based data, with the existing publications limited to case studies or very small case series.14 –16 Although the existing consensus guidelines still recommend the use of the cardiothoracic ratio on the chest radiograph to guide the treatment of adults with the Ebstein anomaly, in particular with reference to the timing of tricuspid valve surgery,3–5,18 we suggest that CMR-derived volumetric measurements might be more detailed and perhaps more robust than the cardiothoracic ratio alone. In the absence of any published data to guide our methodology, our aim was to establish a reproducible approach for the delineation of the functional, atrialized, and total RV components. Using a method similar to the widely accepted contouring technique for normal ventricles in which a straight line is drawn between the tricuspid valve insertion points,17 we created a straight line between anterior and septal leaflet insertion points to demarcate the functional from the atrialized portion of the right ventricle. On published echocardiographic images,9 the contouring line connects the leaflet insertion points but then follows the course of the leaflets, forming a nadir at the point of coaptation. We believe this is not a practical approach using CMR imaging, because the valve morphology is not as easily defined using CMR imaging owing to the accepted limitations of spatial and temporal resolution. Therefore, although CMR imaging can provide reliable and reproducible volumetric measures of the right ventricle, a detailed assessment of tricuspid valve morphology would still be best achieved using echocardiography. Our data have suggested that axial imaging of the right ventricle is more reproducible in patients with the Ebstein anomaly, in terms of both intra- and interobserver quantification. These findings are in keeping with previous CMR studies of the normal and diseased right heart.17,19 In patients with repaired tetralogy of Fallot, axial slices were more reproducible than the short-axis slices for measuring the ventricular volumes and therefore has been the preferred imaging plane for serial follow-up at many institutions.17,19 In the setting of the Ebstein anomaly, it is not surprising that the axial plane is more reliable because this view provides more reliable imaging of the tricuspid valve leaflet insertion and points of coaptation, enabling more accurate delineation of the functional and atrialized components of the right ventricle. In the short-axis view, the tricuspid valve will be seen “en face,” because it is apically displaced and rotated toward the outflow tract, making delineation of the hinge points difficult.20 Additionally, visualization of the insertion points will be limited by through-plane motion and partialvolume artifact at the base of the heart, limitations which are more pronounced in the short-axis view. A novel quantitative measure that emerged from our study was the atrialized RVEDV. The atrialized RVEDV correlated well with more traditional measures of disease severity in those with Ebstein anomaly, such as the magnitude of apical displacement of the septal leaflet and severity score, as described by Celermajer et al.9 As a 3-dimensional value, this might provide greater fidelity expression of disease severity, although proof of this concept was beyond the scope of the present study. We propose that future directions
for research should not only incorporate validation of our findings and evaluation of the clinical relevance of these CMR measurements, but should also include additional dimensions such as assessment of tricuspid valve anatomy and integrity, using 2- or 3-dimensional echocardiography. The relation between the atrialized RVEDV and severity of tricuspid valve disease, as well as the timing of tricuspid valve repair, are topics of ongoing study at our center. Our study was limited by its retrospective design and cross-sectional nature with relatively small patient numbers. The slice thicknesses for the steady-state free-precession cine sequences were, on average, 2 mm thicker for the short axis than for the axial stack, which might have enabled more accurate visualization of the borders in the latter view thereby affecting the reproducibility of the measurements between the imaging planes. Another limitation was the lack of quantification of tricuspid valve regurgitation or shunt fraction using CMR imaging, because phase-contrast flow analysis was not routinely prescribed for clinical CMR studies at our institution. Furthermore, a detailed assessment of valve morphology was not possible, given the accepted limitations of spatial and temporal resolution inherent to CMR cine imaging. Finally, in the absence of a recognized reference standard for right heart volumes in those with Ebstein anomaly, it would be difficult to establish the diagnostic accuracy of the method described in the present study. 1. Babar JL, Jones RG, Hudsmith L, Steeds R, Guest P. Application of MR imaging in assessment and follow-up of congenital heart disease in adults. Radiographics 2010;30:1145. 2. Kilner PJ, Geva T, Kaemmerer H, Trindade PT, Schwitter J, Webb GD. Recommendations for cardiovascular magnetic resonance in adults with congenital heart disease from the respective working groups of the European Society of Cardiology. Eur Heart J 2010;31: 794 – 805. 3. Deanfield J, Thaulow E, Warnes C, Webb G, Kolbel F, Hoffman A, Sorenson K, Kaemmer H, Thilen U, Bink-Boelkens M, Iserin L, Daliento L, Silove E, Redington A, Vouhe P, Priori S, Alonso MA, Blanc JJ, Budaj A, Cowie M, Deckers J, Fernandez Burgos E, Lekakis J, Lindahl B, Mazzotta G, Morais J, Oto A, Smiseth O, Trappe HJ, Klein W, Blomstrom-Lundqvist C, de Backer G, Hradec J, Parkhomenko A, Presbitero P, Torbicki A. Management of grown up congenital heart disease. Eur Heart J 2003;24:1035–1084. 4. Warnes CA, Williams RG, Bashore TM, Child JS, Connolly HM, Dearani JA, del Nido P, Fasules JW, Graham TP Jr, Hijazi ZM, Hunt SA, King ME, Landzberg MJ, Miner PD, Radford MJ, Walsh EP, Webb GD, Smith SC Jr, Jacobs AK, Adams CD, Anderson JL, Antman EM, Buller CE, Creager MA, Ettinger SM, Halperin JL, Krumholz HM, Kushner FG, Lytle BW, Nishimura RA, Page RL, Riegel B, Tarkington LG, Yancy CW. ACC/AHA 2008 guidelines for the management of adults with congenital heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Develop Guidelines on the Management of Adults With Congenital Heart Disease). Developed in collaboration with the American Society of Echocardiography, Heart Rhythm Society, International Society for Adult Congenital Heart Disease, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol 2008;52:e1– e121. 5. Silversides CK, Kiess M, Beauchesne L, Bradley T, Connelly M, Niwa K, Mulder B, Webb G, Colman J, Therrien J. Canadian Cardiovascular Society 2009 Consensus Conference on the management of adults with congenital heart disease: outflow tract obstruction, coarctation of the aorta, tetralogy of Fallot, Ebstein anomaly and Marfan’s syndrome. Can J Cardiol 2010;26:e80 – e97.
Congenital Heart Disease/CMR Imaging and Ebstein Anomaly 6. Bonow RO, Carabello BA, Chatterjee K, de Leon AC Jr, Faxon DP, Freed MD, Gaasch WH, Lytle BW, Nishimura RA, O’Gara PT, O’Rourke RA, Otto CM, Shah PM, Shanewise JS. 2008 Focused update incorporated into the ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to revise the 1998 guidelines for the management of patients with valvular heart disease). Endorsed by the Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol 2008;52:e1– e142. 7. Wald RM, Redington AN, Pereira A, Provost YL, Paul NS, Oechslin EN, Silversides CK. Refining the assessment of pulmonary regurgitation in adults after tetralogy of Fallot repair: should we be measuring regurgitant fraction or regurgitant volume? Eur Heart J 2009;30:356 – 361. 8. Alfakih K, Reid S, Jones T, Sivananthan M. Assessment of ventricular function and mass by cardiac magnetic resonance imaging. Eur Radiol 2004;14:1813–1822. 9. Celermajer DS, Cullen S, Sullivan ID, Spiegelhalter DJ, Wyse RK, Deanfield JE. Outcome in neonates with Ebstein’s anomaly. J Am Coll Cardiol 1992;19:1041–1046. 10. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307– 310. 11. Nihoyannopoulos P, McKenna WJ, Smith G, Foale R. Echocardiographic assessment of the right ventricle in Ebstein’s anomaly: relation to clinical outcome. J Am Coll Cardiol 1986;8:627– 635.
773
12. Munoz-Castellanos L, Espinola-Zavaleta N, Kuri-Nivon M, Keirns C. Ebstein’s anomaly: anatomo-echocardiographic correlation. Cardiovasc Ultrasound 2007;5:43. 13. Frescura C, Angelini A, Daliento L, Thiene G. Morphological aspects of Ebstein’s anomaly in adults. J Thorac Cardiovasc Surg 2000;48: 203–208. 14. Choi YH, Park JH, Choe YH, Yoo SJ. Imaging of Ebstein’s anomaly of the tricuspid valve. AJR Am J Roentgenol 1994;163:539 –543. 15. Eustace S, Kruskal JB, Hartnell GG. Ebstein’s anomaly presenting in adulthood: the role of cine magnetic resonance imaging in diagnosis. Clin Radiol 1994;49:690 – 692. 16. Cantinotti M, Bell A, Razavi R. Role of magnetic resonance imaging in different ways of presentation of Ebstein’s anomaly. J Cardiovasc Med 2008;9:628 – 630. 17. Alfakih K, Plein S, Bloomer T, Jones T, Ridgway J, Sivananthan M. Comparison of right ventricular volume measurements between axial and short axis orientation using steady-state free precession magnetic resonance imaging. J Magn Reson Imaging 2003;18:25–32. 18. Warnes CA. Adult congenital heart disease importance of the right ventricle. J Am Coll Cardiol 2009;54:1903–1910. 19. Fratz S, Schuhbaeck A, Buchner C, Busch R, Meierhofer C, Martinoff S, Hess J, Stern H. Comparison of accuracy of axial slices versus short-axis slices for measuring ventricular volumes by cardiac magnetic resonance in patients with corrected tetralogy of Fallot. Am J Cardiol 2009;103:1764 –1769. 20. Bharucha T, Anderson RH, Lim ZS, Vettukattil JJ. Multiplanar review of three-dimensional echocardiography gives new insights into the morphology of Ebstein’s malformation. Cardiol Young 2010;20:49 –53.
Prognosis Based on Creatine Kinase Isoenzyme MB, Cardiac Troponin I, and Right Ventricular Size in Stable Patients With Acute Pulmonary Embolism Paul D. Stein, MDa,c,*, Muhammad Janjua, MDd, Fadi Matta, MDa,c, Pramod K. Pathak, PhDb, Fadel Jaweesh, MDe, Ahmad Alrifai, MDd, and Haroon L. Chughtai, MDd Prognosis of stable patients with acute pulmonary embolism (PE) has been assessed with cardiac troponin I (cTnI) and right ventricular (RV) function or size. Whether creatine kinase-MB isoenzyme (CK-MB) would add to the prognostic assessment is uncertain. We retrospectively assessed in-hospital mortality from PE in 392 stable patients to test the hypothesis that CK-MB would be of greater prognostic value than cTnI or RV size and we assessed whether combinations would increase prognostic value. CK-MB was high in 29 patients (7.4%); cTnI was high in 76 patients (19%) and intermediate in 78 patients (20%). The right ventricle was dilated in 128 patients (33%). Trends showed highest in-hospital mortality from PE in 4 of 29 (14%) with high CK-MB compared to 6 of 76 (7.9%) with high cTnI and 8 of 128 (6.3%) with RV dilatation (differences NS). High CK-MB and high cTnI provided added prognostic information only in patients with RV dilatation. Mortality with high CK-MB plus RV dilatation (4 of 19, 21%) tended to exceed mortality with high cTnI plus RV dilatation (5 of 39, 13%, NS). When CK-MB and cTnI were high and the right ventricle was dilated, PE mortality tended to be highest (4 of 14, 29%, NS). In conclusion, cardiac biomarkers contributed to prognosis only in patients with RV dilatation. CK-MB was the strongest predictor of death from PE but its prevalence was low, thus limiting its value as a single prognostic indicator. The combination of high CK-MB, high cTnI, and RV dilatation tended to indicate the highest mortality. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:774 –777) Cardiac troponin I (cTnI) has been used for risk stratification of stable patients with acute pulmonary embolism (PE)1,2 and has been particularly useful in combination with an echocardiographic assessment of right ventricular (RV) function or size.2– 4 There are sparse data on creatine kinase-MB isoenzyme (CK-MB) for assessment of prognosis. High levels of CK-MB have been observed in patients with acute PE.1,5–9 CK-MB has been shown to be an independent predictor of mortality.8 It seemed to be more specific with regard to prognosis but overall incidence of a high CK-MB was low, limiting its sensitivity.1 We are not aware of investigations in stable patients with PE in which prognosis was assessed from CK-MB in combination with cTnI or RV size. We assessed in-hospital mortality from acute PE in stable patients to test the hypothesis that CK-MB would be of greater prognostic value than cTnI or RV size and whether combinations with CK-MB would increase its prognostic value.
a Departments of Internal Medicine and Research and Advanced Studies Program, College of Osteopathic Medicine, and bDepartment of Statistics, Michigan State University, East Lansing, Michigan; cDepartment of Research, St. Mary Mercy Hospital, Livonia, Michigan; dDepartment of Internal Medicine, St. Joseph Mercy Oakland, Pontiac, Michigan; eDepartment of Internal Medicine, William Beaumont Hospital, Royal Oak, Michigan. Manuscript received August 18, 2010; revised manuscript received and accepted October 19, 2010. *Corresponding author: Tel: 248-858-6772; fax: 248-858-6974. E-mail address:
[email protected] (P.D. Stein).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.061
Methods This was a retrospective study of stable patients hospitalized with acute PE who had measurements of cTnI, CKMB, and estimates of RV size. Patients were studied from March 2007 through May 2010 at St. Joseph Mercy Oakland Hospital, Pontiac Michigan and from January 2004 through June 2008 at William Beaumont Hospital, Royal Oak, Michigan. The investigation was approved by the institutional review boards of the 2 hospitals. Diagnosis of PE was made with computed tomographic pulmonary angiography in 378 of 392 patients (96%) and a high probability interpretation of ventilation/perfusion lung scan in 14 of 392 patients (4%). Exclusions included 573 patients with PE who did not have assessments of each marker (cTnI, CK-MB, and RV size). Patients were also excluded if they were hypotensive (systolic blood pressure ⬍90 mm Hg) or on ventilatory support. Additional exclusions were for possible causes of cTnI increase other than PE such as heart failure10,11 (which we defined as an ejection fraction ⬍40%), chronic obstructive pulmonary disease,12,13 pneumonia,14 hemodialysis,15 coronary artery bypass graft surgery within 1 week, or myocardial infarction or ischemia within 1 week. Exclusions for other known causes of high CK or CK-MB included rhabdomyolysis,16,17 severe muscle trauma,18 and liposuction.19 Cardiac biomarkers (CK-MB and cTnI) were measured within 48 hours before or after diagnosis of PE. CK-MB at www.ajconline.org
Miscellaneous/Biomarkers in Pulmonary Embolism Table 1 Mortality from pulmonary embolism with individual risk factors
Creatine kinase-MB isoenzyme Cardiac troponin I High Intermediate Right ventricular size
was carried out using SAS 9.2 (SAS Institute, Cary, North Carolina).
Mortality Value Normal
Mortality Value Abnormal
p Value (normal vs abnormal)
8/363 (2.2%)
4/29 (14%)
0.008
6/76 (7.9%) 2/78 (2.6%) 8/128 (6.3%)
0.02 NS 0.02
4/238 (1.7%)
4/264 (1.5%)
775
High creatine kinase-MB isoenzyme versus high cardiac troponin I (NS), versus intermediate cardiac troponin I (p ⫽ 0.05), versus dilated right ventricle (NS). High cardiac troponin I versus intermediate cardiac troponin I (NS), versus dilated right ventricle (NS). Intermediate cardiac troponin I versus dilated right ventricle (NS).
William Beaumont Hospital was assayed in 382 patients with a 2-site sandwich immunoassay (ADVIA Centaur CKMB, Deerfield, Illinois). Levels ⬎5.0 ng/ml were defined as high. CK-MB at St. Joseph Mercy Oakland Hospital was assayed in 10 patients with a modified 2-site immunoenzymatic (sandwich) assay (Access Immunoassay System, Beckman Coulter, Fullerton, California). Levels ⬎6.3 ng/mL were defined as high. Cardiac TnI before October 2006 at William Beaumont Hospital was measured in 73 patients by a Bayer Diagnostics TnI 2-site sandwich immunoassay (Siemens ADVIA Centaur TnL-Ultra, Deerfield, Illinois). A normal level was ⬍0.3 ng/ml, an intermediate or indeterminate level was 0.4 to 1.4 ng/mL, and highly suggestive of myocardial infarction (high) or critical was a level ⬎1.5 ng/ml. Cardiac TnI after October 2006 at William Beaumont Hospital was measured in 309 patients by a Bayer Diagnostics TnI UltraTM troponin 3-site sandwich immunoassay (Siemens ADVIA Centaur TnL-Ultra). A normal level was ⱕ0.05 ng/ml, an intermediate or indeterminate level was 0.06 to 0.19 ng/ml, and highly suggestive of myocardial infarction (high) or critical was a level ⱖ0.20 ng/ml. Cardiac TnI at St. Joseph Mercy Oakland Hospital was measured in 10 patients by a 2-site immunoenzymatic (sandwich) assay (Access Immunoassay System, Beckman Coulter, Brea, California). A normal level was ⬍0.05 ng/mL, an intermediate or indeterminate level was 0.05 to 0.49 ng/mL, and highly suggestive of myocardial infarction (high) or critical was a level ⱖ0.50 ng/ml. RV dilatation by echocardiography was determined by qualitative impression of reports in 325 of 392 (83%). RV/ left ventricular ratios were measured in 67 of 392 (17%). An RV/left ventricular ratio ⬎1 was defined as RV dilatation.20 Data were analyzed using SPSS 11.5 for Windows (SPSS, Inc., Chicago, Illinois). Significant tests of equality of 2 proportions were carried out using 2-tailed Fisher’s exact test (http://www.graphpad.com/quickcalcs/contingency2.cfm). Although multiple comparisons were made, we did not use Bonferroni correction and considered a p value ⱕ0.05 as statistically significant. Odds ratios and 95% confidence intervals were calculated using online software (http://www.hutchon. net/ConfidOR.htm). Multiple logistic regression analysis
Results The mean age of 392 stable patients with acute PE was 68 ⫾ 16 years (mean ⫾ SD); 172 patients (44%) were men. CK-MB was high in 29 patients (7.4%), cTnI was high in 76 patients (19%) and intermediate in 78 patients (20%), and the right ventricle was dilated in 128 patients (33%). Death from PE in patients with high CK-MB, high cTnI, or RV dilatation was higher than in patients with normal values, but mortality with intermediate cTnI was not higher (Table 1). Mortality with these markers did not significantly differ from each other due to the small sample. Mortality with high CK-MB plus RV dilatation (21%) tended to exceed mortality with high cTnI plus RV dilatation (13%, NS; Table 2). Mortality tended to be highest with all 3 markers in combination (high CK-MB, high cTnI, and RV dilatation, 29%; Table 2). Cardiac biomarkers provided added prognostic information only in patients with RV dilatation. As with individual markers, combinations did not differ significantly from each other due to the small sample. Sensitivity analysis showed no statistically significant difference in mortality with cTnI using all data (392 patients) compared to only the 3-site sandwich technique (309 patients) and there was no difference of mortality with CK-MB using all data compared to 382 patients in whom measurements were with ADVIA Centaur. There was no correlation in mortality from PE and levels of CK-MB (r ⫽ ⫺0.3157, NS) using measurements with the ADVIA Centaur or levels of cTnI using the 3-site sandwich technique (r ⫽ 0.1783, NS). Logistic regression showed that high CK-MB, high cTnI, and RV dilatation were each important predictors of PE death (Table 3). Prognosis with intermediate cTnI did not differ significantly from normal. Adjustment for age and gender decreased the odds ratios a small amount. Age was almost significant. Only patients ⬎60 years of age died. When all 3 markers were included in the model and stepwise logistic regression was run, CK-MB was chosen as the best predictor of death from PE. High CK-MB had the largest proportion of death from PE. With respect to percentage, 14% mortality with high CK-MB was ⬎7.9% with high cTnI or 6.3% with RV dilatation. However, absolute number of deaths (4 for high CK-MB, 6 for high cTnI, and 8 for RV dilatation) was higher for high cTnI and RV dilatation because the prevalence of RV dilatation and high cTnI was higher than that of high CK-MB. Therefore, in absolute terms, RV dilatation predicted the largest number of deaths, but the proportion of patients with a given abnormal marker at risk for death from PE was predicted best by a high CK-MB. Discussion High CK-MB, high cTnI, and RV dilatation were each associated with an increased mortality from PE in initially stable patients. Due to the small sample, mortality rates did not differ significantly among these markers. However, a trend suggested that high CK-MB was associated with
776
The American Journal of Cardiology (www.ajconline.org)
Table 2 Death from pulmonary embolism according to cardiac troponin I, creatine kinase-MB isoenzyme, and right ventricular size Variable
Right ventricle dilated Right ventricle not dilated Total
CK-MB High
cTnI High
CK-MB High and cTnI High
cTnI Intermediate or High
CK-MB High and cTnI Intermediate or High
CK-MB Normal
cTnI Normal
CK-MB Normal and cTnI Normal
4/19 (21%)*
5/39 (13%)
4/14 (29%)
6/71 (8.5%)
4/18 (22%)
4/109 (3.7%)
2/57 (3.5%)
2/56 (3.6%)
0/10 (0%)
1/37 (2.7%)
0/5 (0%)
2/83 (2.4%)
0/7 (0%)
4/254 (1.6%)
2/181 (1.1%)
2/178 (1.1%)
4/29 (14%)†
6/76 (7.9%)
4/19 (21%)#
8/154 (5.2%)
4/25 (16%)‡
8/363 (2.2%)
4/238 (1.7%)
4/234 (1.7%)
* p ⫽ 0.02, creatine kinase-MB isoenzyme high and dilated right ventricle versus creatine kinase-MB isoenzyme normal and dilated right ventricle. p ⫽ 0.008, creatine kinase-MB isoenzyme high versus normal. ‡ p ⫽ 0.004, creatine kinase-MB isoenzyme high and cardiac troponin I intermediate or high versus creatine kinase-MB isoenzyme normal and cardiac troponin I normal. p ⫽ 0.01, creatine kinase-MB isoenzyme high and cardiac troponin I high and dilated right ventricle versus creatine kinase-MB isoenzyme normal and cardiac troponin I normal and dilated right ventricle. p ⫽ 0.03, creatine kinase-MB isoenzyme high and cardiac troponin I intermediate and dilated right ventricle versus creatine kinase-MB isoenzyme normal and cardiac troponin I normal and dilated right ventricle. p ⫽ 0.02, creatine kinase-MB isoenzyme high and cardiac troponin I high or intermediate and dilated right ventricle versus creatine kinase-MB isoenzyme high and cardiac troponin I high or intermediate and normal right ventricle. p ⫽ 0.02, cardiac troponin I high versus cardiac troponin I normal. p ⫽ 0.01, creatine kinase-MB isoenzyme high and cardiac troponin I high versus creatine kinase-MB isoenzyme normal and cardiac troponin I normal. # p ⫽ 0.03, creatine kinase-MB isoenzyme high and cardiac troponin I high verses cardiac troponin I intermediate or high. p ⫽ 0.002, creatine kinase-MB isoenzyme high and cardiac troponin I high verses creatine kinase-MB isoenzyme normal. p ⫽ 0.001, creatine kinase-MB isoenzyme high and cardiac troponin I high verses cardiac troponin I normal. p ⫽ 0.001, creatine kinase-MB isoenzyme high and cardiac troponin I high verses creatine kinase-MB isoenzyme normal and cardiac troponin I normal. All other comparisons are not significant. †
Table 3 Odds ratio estimates Parameter Creatine kinase-MB isoenzyme high versus normal Cardiac troponin I high versus normal Cardiac troponin I intermediate versus normal Right ventricle dilated versus normal
Odds Ratio
95% Confidence Interval
7.1
2.0–25.2
5.0 1.6
1.4–18.3 0.3–8.6
4.3
1.3–14.7
higher PE mortality than high levels of cTnI or RV dilatation. When all 3 markers were abnormal, the proportion of deaths due to PE tended to be highest (29%). Others have recognized that the proportion of patients with PE who have an abnormal CK-MB is much smaller than the proportion with an abnormal cTnI.1,2,6 Only a relatively small proportion of patients with PE would have an abnormal CK-MB, which limits its value if used as the only indicator of prognosis. In-hospital mortality from PE in stable patients who had an abnormal cTnI was reported by Bova et al21 in 1 of 68 (1.5%) compared to 0 of 133 (0%) with a normal cTnI. Jiménez et al22 showed a 30-day mortality from PE in stable patients with an abnormal cTnI in 8 of 102 (7.8%) compared to 4 of 216 (1.9%) in those with a normal cTnI. Vuilleumier et al23 in patients with nonmassive PE found a 3.5-fold higher risk for death or hospitalization for PE-related complications within 3 months in patients with cTnI ⬎0.09 ng/ml. Kline et al,24 however, showed an insignificant difference in all-cause mortality comparing normotensive patients with PE and cTnI ⱕ0.1 ng/ml to those with PE and cTnI ⬎0.1 ng/ml.
A strength of this investigation is the uniqueness of combined measurements of biomarkers and RV size. Weaknesses are that it was retrospective, RV size was usually based on reader impression rather than objective measurement, levels of cTnI and CK-MB were measured by different techniques, and a wide (48-hour) time window between imaging and tests was allowed. 1. Konstantinides S, Geibel A, Olschewski M, Kasper W, Hruska N, Jäckle S, Binder L. Importance of cardiac troponins I and T in risk stratification of patients with acute pulmonary embolism. Circulation 2002;106:1263–1268. 2. Goldhaber SZ. Cardiac biomarkers in pulmonary embolism. Chest 2003;123:1782–1784. 3. Stein PD, Matta F, Janjua M, Yaekoub AY, Jaweesh F, Alrifai A. Outcome in stable patients with acute pulmonary embolism who had right ventricular enlargement and/or elevated levels of troponin I. Am J Cardiol 2010;106:558 –563. 4. Scridon T, Scridon C, Skali H, Alvarez A, Goldhaber SZ, Solomon SD. Prognostic significance of troponin elevation and right ventricular enlargement in acute pulmonary embolism. Am J Cardiol 2005;96: 303–305. 5. Punukollu G, Khan IA, Gowda RM, Lakhanpal G, Vasavada BC, Sacchi TJ. Cardiac troponin I release in acute pulmonary embolism in relation to the duration of symptoms. Int J Cardiol 2005;99:207–211. 6. Pruszczyk P, Bochowicz A, Torbicki A, Szulc M, Kurzyna M, Fijałkowska A, Kuch-Wocial A. Cardiac troponin T monitoring identifies high-risk group of normotensive patients with acute pulmonary embolism. Chest 2003;123:1947–1952. 7. Adams JE 3rd, Siegel BA, Goldstein JA, Jaffe AS. Elevations of CK-MB following pulmonary embolism. A manifestation of occult right ventricular infarction. Chest 1992;101:1203–1206. 8. Gallotta G, Palmieri V, Piedimonte V, Rendina D, De Bonis S, Russo V, Celentano A, Di Minno MN, Postiglione A, Di Minno G. Increased troponin I predicts in-hospital occurrence of hemodynamic instability in patients with sub-massive or non-massive pulmonary embolism independent to clinical, echocardiographic and laboratory information. Int J Cardiol 2008;124:351–357.
Miscellaneous/Biomarkers in Pulmonary Embolism 9. Giannitsis E, Müller-Bardorff M, Kurowski V, Weidtmann B, Wiegand U, Kampmann M, Katus HA. Independent prognostic value of cardiac troponin T in patients with confirmed pulmonary embolism. Circulation 2000;102:211–217. 10. Missov E, Calzolari C, Pau B. Circulating cardiac troponin I in severe congestive heart failure. Circulation 1997;96:2953–2958. 11. Horwich TB, Patel J, MacLellan WR, Fonarow GC. Cardiac troponin I is associated with impaired hemodynamics, progressive left ventricular dysfunction, and increased mortality rates in advanced heart failure. Circulation 2003;108:833– 838. 12. Martins CS, Rodrigues MJ, Miranda VP, Nunes JP. Prognostic value of cardiac troponin I in patients with COPD acute exacerbation. Neth J Med 2009;67:341–349. 13. Baillard C, Boussarsar M, Fosse JP, Girou E, Le Toumelin P, Cracco C, Jaber S, Cohen Y, Brochard L. Cardiac troponin I in patients with severe exacerbation of chronic obstructive pulmonary disease. Intensive Care Med 2003;29:584 –589. 14. Ilva TJ, Eskola MJ, Nikus KC, Voipio-Pulkki LM, Lund J, Pulkki K, Mustonen H, Niemelä KO, Karhunen PJ, Porela P. The etiology and prognostic significance of cardiac troponin I elevation in unselected emergency department patients. J Emerg Med 2010;38:1–5. 15. Petrovic D, Stojimirovic BB. Cardiac troponins: outcome predictors in hemodialysis patients. J Artif Organs 2009;12:258 –263. 16. Lee ME, Sethna DH, Conklin CM, Shell WE, Matloff JM, Gray RJ. CK-MB release following coronary artery bypass grafting in the absence of myocardial infarction. Ann Thorac Surg 1983;35:277–279. 17. Giannoglou GD, Chatzizisis YS, Misirli G. The syndrome of rhabdomyolysis: pathophysiology and diagnosis. Eur J Intern Med 2007;18:90 –100.
777
18. Schwartz JG, Prihoda TJ, Stuckey JH, Gage CL, Darnell ML. Creatine kinase MB in cases of skeletal muscle trauma. Clin Chem 1988;34: 898 –901. 19. Lipschitz AH, Kenkel JM, Luby M, Sorokin E, Rohrich RJ, Brown SA. Electrolyte and plasma enzyme analyses during large-volume liposuction. Plast Reconstr Surg 2004;114:766 –775. 20. ten Wolde M, Sohne M, Quak E, Mac Gillavry MR, Büller HR. Prognostic value of echocardiographically assessed right ventricular dysfunction in patients with pulmonary embolism. Arch Intern Med 2004;164:1685–1689. 21. Bova C, Pesavento R, Marchiori A, Palla A, Enea I, Pengo V, Visonà A, Noto A, Prandoni P; TELESIO Study Group. Risk stratification and outcomes in hemodynamically stable patients with acute pulmonary embolism: a prospective, multicentre, cohort study with three months of follow-up. J Thromb Haemost 2009;7:938 –944. 22. Jiménez D, Díaz G, Molina J, Martí D, Del Rey J, García-Rull S, Escobar C, Vidal R, Sueiro A, Yusen RD. Troponin I and risk stratification of patients with acute nonmassive pulmonary embolism. Eur Respir J 2008;31:847– 853. 23. Vuilleumier N, Righini M, Perrier A, Rosset A, Turck N, Sanchez JC, Bounameaux H, Le Gal G, Mensi N, Hochstrasser D. Correlation between cardiac biomarkers and right ventricular enlargement on chest CT in non massive pulmonary embolism. Thromb Res 2008;121:617– 624. 24. Kline JA, Zeitouni R, Marchick MR, Hernandez-Nino J, Rose GA. Comparison of 8 biomarkers for prediction of right ventricular hypokinesis 6 months after submassive pulmonary embolism. Am Heart J 2008;156:308 –314.
Usefulness of Postexercise Ankle-Brachial Index to Predict All-Cause Mortality Mobeen A. Sheikh, MDa,*, Deepak L. Bhatt, MDb, Jianbo Li, PhDc, Songhua Lin, MSc, and John R. Bartholomew, MDd Peripheral arterial disease predicts future cardiovascular events and all-cause mortality. Conventional methods of assessment might underestimate its true prevalence. We sought to determine whether a postexercise ankle-brachial index (ABI), not only improved peripheral arterial disease detection, but also independently predicted death. This was an observational study of consecutive patients referred for ABI measurement before and after the fixed-grade treadmill or symptom-limited exercise component to a noninvasive vascular laboratory from January 1990 to December 2000. The subjects were classified into 2 groups. Group 1 included patients with an ABI of >0.85 before and after exercise, and group 2 included patients with a normal ABI at rest but <0.85 after exercise. A total of 6,292 patients underwent ABI measurements with exercise during the study period. Propensity score matching of the groups was performed to minimize observational bias. Overall mortality, as determined using the United States Social Security death index, was the end point. The 10-year mortality rate of groups 1 and 2 was 32.7% and 41.2%, respectively. An abnormal postexercise ABI result independently predicted mortality (hazard ratio 1.3, 95% confidence interval 1.07 to 1.58, p ⴝ 0.008). Additional independent predictors of mortality were age, male gender, diabetes, and hypertension. After the exclusion of patients with a history of cardiovascular events, the predictive value of an abnormal postexercise ABI remained statistically significant (hazard ratio 1.67, 95% confidence interval 1.29 to 2.17, p <0.0001). In conclusion, our results have shown that the postexercise ABI is a powerful independent predictor of all-cause mortality and provides additional risk stratification beyond the ABI at rest. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:778 –782) A diagnosis of peripheral arterial disease (PAD) serves as a marker for systemic atherosclerosis. The National Cholesterol Education Panel guidelines1 have recognized PAD as a disease equivalent to the presence of coronary artery disease (CAD). This serves to highlight the newfound appreciation for a PAD diagnosis in the cardiovascular risk assessment and determining the vigor of preventive strategies. The ankle-brachial index (ABI) is a useful tool to screen for PAD; however, conventional ABI measurements have been obtained in the at rest state and might underestimate the true prevalence of PAD. Although it has been recognized in clinical practice that obtaining a measurement of the ABI after exercise might improve the sensitivity of PAD detection,2 the postexercise ABI has not thus far been validated as a screening test nor has the association with all-cause mortality been determined in a United Statesbased population. In the present study, we assessed the
a Interventional Cardiac and Vascular Service, The Medical Group, Beverly, Massachusetts; bDepartment of Cardiology, Veterans Affairs Boston Healthcare System, Integrated Interventional Cardiovascular Program, Brigham and Women’s Hospital, and Thrombolysis In Myocardial Infarction Study Group, Harvard Medical School, Boston, Massachusetts; and Departments of cBiostatistics and Epidemiology and dCardiovascular Medicine, Cleveland Clinic, Cleveland, Ohio. Manuscript received August 15, 2010; manuscript received and accepted October 19, 2010. *Corresponding author: Tel: (617) 860-3536; fax: (978) 232-7027. E-mail address:
[email protected] (M.A. Sheikh).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.060
prognostic value of the postexercise ABI among patients referred for testing with exercise for clinical reasons. We also sought to determine whether an abnormal postexercise ABI results was an independent predictor of death and could potentially identify a population at greater mortality risk that would otherwise have been missed using the conventional at rest ABI measurements. Methods Consecutive patients referred to the Cleveland Clinic noninvasive vascular laboratory from January 1990 to December 2000 for either complete pulse volume recordings of the lower extremities or standard ABI measurement before and after a fixed-grade treadmill protocol or symptom-limited exercise test were considered for analysis. The patients had to be ⬎40 years old and residents of the United States with a valid Social Security number. The indication for testing was at the sole discretion of the referring physician. For the purposes of the present study, the patients were divided into 2 groups according to their ABI result: group 1, a normal ABI of ⱖ0.85 both at rest and after exercise; and group 2, a normal ABI at rest that was abnormal (⬍0.85) in either limb after exercise. The exclusion criteria were a history of lower limb revascularization, either surgical or percutaneous, the lack of pressure values secondary to completely or partially calcified arteries or an inaudible Doppler signal, unilateral studies, and nonatherosclerotic etiologies for lower extremity occlusive vascular disease. www.ajconline.org
Miscellaneous/Postexercise ABI
At testing, the patient information, including demographic data, pulse pressure measurements, and risk factor profile, were stored in an institutional review board-approved vascular laboratory database. Additional data were collected, if required, by reviewing the electronic and paper medical records and adding to the database in a manner that protected patient confidentiality. The criteria for the various risk factors were as follows. CAD was defined as present if the patient had a history of angina, congestive heart failure, or myocardial infarction or had undergone percutaneous coronary intervention or coronary artery bypass grafting. Similarly, a history of a cerebrovascular accident or abdominal aortic aneurysm was defined as a history of stroke, transient ischemic attack, or previous carotid endarterectomy and previous identification on suitable imaging studies or previous surgical repair, respectively. The assessment for hypertension, diabetes, and hyperlipidemia was determined by a combination of questioning at testing, chart review, and medication use. Smoking history was determined by selfreported current use at testing. Doppler-derived systolic pressures at the level of the brachial artery and posterior tibial and dorsalis pedis arteries were obtained by recording the level at which the first Doppler sound was heard after deflation of the cuff placed over the upper arm and ankle, respectively. The index was calculated using the greater of the 2 ankle pressures (posterior tibial or dorsalis pedis) obtained on each leg, with the greater of the brachial pressures. The patients exercised on a treadmill for either 5 minutes or until the onset of leg pain that limited them from continuing. The exercise was also terminated by the onset of any cardiopulmonary symptoms, especially if these were the principal restriction to exercise. Once the exercise had been completed, the patient returned to the bed in the supine position and both ankle pressures were obtained (first in the symptomatic extremity or the extremity with the lower pressure at rest) followed by the brachial pressure in the arm with the greater pressure at rest. The ABI thus obtained was referred to as the postexercise ABI in the present study. When the patients had undergone more than one test during the study period, only the first was taken. The United States Social Security death index was used to match all subjects to their records according to name and Social Security number. The vital status was determined as of May 2007. The continuous data are displayed as the mean ⫾ SD. The categorical data are displayed as frequencies and percentages within the population. The continuous data were analyzed using the nonparametric Wilcoxon test and categorical data using the chi-square test. Survival analysis was done using Cox proportional hazard modeling, adjusted for confounding factors. The proportionality assumption was tested with all time-dependent covariates simultaneously. The relative risks (hazard ratio) were estimated and 95% confidence intervals given. Patient age at testing, gender, race, and history of CAD, cerebrovascular accident, smoking, diabetes, hypertension, hyperlipidemia, abdominal aortic aneurysm, and appropriate transformations were included in the modeling. The patients from group 2 were matched one-to-one to the patients from group 1 using propensity scores before survival analysis to minimize any bias introduced by the
779
Table 1 Demographic and risk factor profile of patients in groups 1 and 2 after propensity score matching Variable Men Race White Black Age (years) 40–49 50–59 60–69 ⱖ70 Smoker Hypertension Diabetes mellitus Coronary artery disease Cerebrovascular accident Abdominal aortic aneurysm Hyperlipidemia
Group 1 (n ⫽ 713)
Group 2 (n ⫽ 713)
p Value
484 (67.9%)
467 (65.5%)
0.3395
659 (92.4%) 47 (6.6%) 64.5 (⫾9.2) 58 (8.1%) 149 (20.9%) 302 (42.4%) 204 (28.5%) 21 (3.0%) 230 (32.3%) 87 (12.2%) 236 (33.1%) 69 (9.7%) 39 (5.5%) 126 (17.7%)
664 (93.1%) 44 (6.2%) 64.7 (⫾9.5) 59 (8.3%) 159 (22.3%) 259 (36.3%) 236 (33.1%) 25 (3.5%) 234 (32.8%) 91 (12.8%) 243 (34.1%) 798 (10.9%) 41 (5.8%) 129 (18.1%)
0.6090 0.7452 0.6745
0.5488 0.8211 0.7486 0.6947 0.4332 0.8180 0.8358
Group 1, patients with normal ankle-brachial index at rest and after exercise, group 2: patients with normal ankle-brachial index at rest but abnormal after exercise.
Table 2 Demographic and risk factor profile of patients in groups 1 and 2 Variable Men Race White Black Age (years) 40–49 50–59 60–69 ⱖ70 Smoker Hypertension Diabetes mellitus Coronary artery disease Cerebrovascular accident Abdominal aortic aneurysm Hyperlipidemia
Group 1 (n ⫽ 1,700) 960 (56.5%) 1,453 (85.5%) 216 (12.7%) 63.9 (⫾10.6) 213 (12.5%) 362 (21.3%) 601 (35.4%) 524 (30.8%) 27 (1.6%) 451 (26.5%) 161 (9.5%) 392 (23.1%) 101 (5.9%) 64 (3.8%) 194 (11.4%)
Group 2 (n ⫽ 716)
p Value
470 (65.36%)
⬍0.0001
667 (93.3%) 44 (6.2%) 64.7 (⫾9.5) 59 (8.2%) 159 (22.2%) 262 (36.6%) 236 (33.0%) 28 (3.9%) 237 (33.1%) 92 (12.9%) 246 (34.4%) 79 (11.0%) 43 (6.0%) 131 (18.3%)
⬍0.0001 ⬍0.0001 0.0729
0.0005 0.0011 0.0133 ⬍0.0001 ⬍0.0001 0.0145 ⬍0.0001
Group 1, patients with normal ankle-brachial index at rest and after exercise, group 2: patients with normal ankle-brachial index at rest but abnormal after exercise.
observational nature of the present study. The propensity score matching was done in all aspects, except for their postexercise ABI results, using a greedy algorithm.3 This was accomplished using a nonparsimonious logistic regression model, with all available covariates included to derive a propensity score for patients with an abnormal postexercise ABI result, and then using the propensity scores to match those from group 2 to the patients in group 1. Patients with missing values were excluded from matching and additional analysis. The resulting matched groups were compared for each covariate to confirm the similarity between the 2 groups (Table 1). The propensity score was also
780
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Kaplan-Meier curves of mortality rates for matched groups 1 and 2 from testing. Patients in both groups had normal ABI values at rest. Patients in group 2 had an ABI of ⬍0.85 after exercise, and patients in group 1 had maintained a normal ABI value even after exercise and served as the control group. Table 3 Statistically significant independent predictors among propensity score matched patients Predictor Postexercise ankle-brachial index in patients with no history of cardiovascular events Hypertension Age per decade Male gender Postexercise ABI Diabetes mellitus
HR (95% CI)
p Value
1.67 (1.29–2.17)
⬍0.0001
1.55 (1.23–1.96) 1.54 (1.38–1.71) 1.37 (1.06–1.75) 1.34 (1.07–1.58) 1.34 (1.01–1.77)
0.0002 ⬍0.0001 0.014 0.008 0.044
Group 1, patients with normal ankle-brachial index at rest and after exercise, group 2: patients with normal ankle-brachial index at rest but abnormal after exercise. CI ⫽ confidence interval; HR ⫽ hazard ratio.
included in the Cox proportional hazard model in the analysis. All the analyses were done using the SAS, version 8.2, statistical package (SAS, Cary, North Carolina). Results From January 1990 to December 2000, 11,295 patients underwent either complete or limited lower extremity pulse volume recording studies with ABI measurement at the Cleveland Clinic noninvasive vascular laboratory. Of these, 6,292 were performed with a fixed-grade treadmill or symptom-limited exercise component. For the purposes of the present study, the patients with an abnormal ABI at rest were not considered. A total of 2,416 patients met the inclusion criteria according to the postexercise ABI results.
After matching, 1,426 patients (713 each in group 1 and 2), as defined in the “Methods” section, were included in the present study. The baseline demographic and risk factor profiles of the matched and unmatched patient groups are listed in Tables 1 and 2. The mean patient age was 64.1 ⫾ 10.3 years, and 40.8% were women. Notable differences were present between the 2 groups before matching. The subjects in group 2 had a greater percentage of white men and were more likely to be older, have a history of CAD, cerebrovascular accident, or abdominal aortic aneurysm, take medications for hyperlipidemia, hypertension, or diabetes, and be current smokers at testing. This bias was minimized by matching the patients from group 2 with those from group 1 using propensity scores (Table 1). During a mean duration of 6.9 ⫾ 3.2 years from the date of testing, a total of 416 deaths occurred in both matched groups, 189 in group 1 and 227 in group 2. In the propensity score-matched patients, an abnormal postexercise ABI result was a strong independent predictor of mortality from all causes, with a hazard ratio of 1.3 (95% confidence interval 1.07 to 1.58, p ⫽ 0.008). The 5- and 10-year mortality rate for group 1 and 2 was 15.4% and 17% and 32.7% and 41.2%, respectively. The Kaplan-Meier curve graphically demonstrated the increased mortality of patients in group 2 (Figure 1). The independent clinical predictors of death are summarized in Table 3. An abnormal postexercise ABI result was as strong a predictor of mortality as diabetes in the present analysis. Similar results were obtained when the patients with any history of cardiovascular events (CAD, cerebrovascular accident, or abdominal aortic aneurysm) were excluded from
Miscellaneous/Postexercise ABI
both groups and subjected to the same analysis (hazard ratio 1.67, 95% confidence interval 1.29 to 2.17, p ⬍0.0001). Thus, for the prediction of mortality, adding the exercise component to the ABI measurement added incremental information in both patients with and without a history of cardiovascular event. Discussion The results from the present study have demonstrated that the postexercise ABI might not simply help in diagnosing PAD in more patients but could independently identify patients at a greater mortality risk who would have remained unidentified using conventional testing. It must be re-emphasized that the diagnosis of PAD is important, not only from a lower extremity standpoint, but also as a marker of systemic atherosclerosis. It has been demonstrated that PAD is a marker of future cardiovascular-related mortality, independent of both conventional risk factors and baseline cardiovascular disease.4,5 The reason for this greater mortality rate is poorly understood. Various hypotheses have been investigated, some of which included systemic endothelial dysfunction,6,7 elevated levels of C-reactive protein,8 other related inflammatory markers,9 increased levels of white blood cells,10 enhanced platelet activation,11–13 and hemostatic factors, such as plasma fibrinogen and von Willebrand factor.14 In an age in which increasing emphasis has been placed on estimating the risk of future cardiovascular events, it was surprising that a predictor as potent as PAD has not been routinely included in risk factor analyses. A part of the continued underrecognition could have been because only a fraction of patients with PAD are symptomatic,15 and, in most cases, their symptoms will not be that of classic intermittent claudication.15,16 However, even patients with symptomatic PAD have been treated less aggressively in terms of risk factor modification and medical therapy than those with other manifestations of atherosclerotic disease, such as CAD.17 Additionally, if we were to wait for the development of symptoms, it might be already too late. This fact has been demonstrated in an earlier study at our institution in which all patients presenting for lower extremity revascularization surgery also underwent diagnostic coronary angiography. Only 10% of these patients had angiographic evidence of normal coronary arteries.18 The measurement of the ABI does add a fair degree of objectivity to the diagnosis of PAD; however, it was our hypothesis that an ABI measured in the at rest state would be analogous to ruling out CAD using an at rest electrocardiogram. Although abnormal findings on an electrocardiogram at rest would be suggestive of CAD, normal findings by no means rule it out. The findings from an exercise electrocardiogram will be much more sensitive. In light of the impressive statistics supporting the predictive role of PAD in terms of cardiovascular morbidity and all-cause mortality, it was our contention that a more sensitive measure of PAD detection, such as that afforded by a postexercise ABI, could identify an additional population of patients worthy of secondary preventive measures that might otherwise be missed. The clinical effect of the ABI has been considered limited owing to the low prevalence of abnormal
781
at rest ABI values in patients ⬍60 years old. Very limited epidemiologic information has been available on what percentage of patients with a normal ABI value at rest would have abnormal findings after exercise, although 2 recent studies have reported data that ranged from 31%19 to 86.2%.20 However, neither of these studies was population based. A normal subject should maintain or increase the ankle systolic pressure even with moderate exercise levels. The key is early detection, and when an area of narrowing exceeds 50% in terms of diameter reduction, the systolic pressure will decrease beyond the site of involvement. Stress testing has commonly been used for the evaluation of cardiac performance, because it has been recognized to provide an enhanced index of myocardial perfusion. The principles are similar for the legs. The findings of our study could also have important public health and economic implications. The National Cholesterol Education Panel (in establishing the Adult Treatment Guidelines) has taken the important step to include the presence of PAD as a CAD equivalent, similar to diabetes. It is important to note that no conditions in these guidelines were set forth in terms of the presence or absence of symptoms. Currently, reimbursement for ABI testing under Medicare guidelines is only allowed if the patient had a history of symptoms secondary to lower limb ischemia. However, symptoms have been poorly predictive, and one needs to use maximum objective strategies to improve PAD detection, considering its mortality implications, such as were demonstrated in our study using the postexercise ABI. Furthermore, the exercise portion for diagnostic purposes does not necessarily have to be a treadmill protocol but could just as easily be performed by having patients walk down the office hallway or even perform active pedal plantar flexion (the details of the latter testing technique have been previously published21). Only a handful of studies have examined the prognostic role of postexercise ABI measurements in the clinical setting. Two such studies demonstrated it to be of value in identifying failing lower extremity angioplasty,22,23 and another study showed that a postexercise ABI did not add prognostic information24 It is imperative to note the patients in these studies already had previously diagnosed PAD, determined by either abnormal ABI values at rest or percutaneous revascularization, both of which were exclusion criteria in our study. A more contemporary publication from Europe, consisting of patients with both abnormal and normal at rest ABI values, also demonstrated increased mortality in the abnormal postexercise ABI cohort.20 Although the patients with abnormal postexercise ABI values in that study were a part of a subgroup analysis, it was also their contention that adding the exercise portion to the ABI measurement adds incremental prognostic information in terms of mortality prediction. We sought to prove that in patients with appropriate risk factors and a high index of suspicion for PAD (determined by typical or atypical symptoms), an at rest ABI value alone would not conclusively exclude PAD. The use of postexercise ABI testing might help identify a subgroup of patients with a normal ABI value at rest who have a CAD risk equivalent and subsequent greater mortality, in a noninvasive and cost-effective manner. How this approach compares with other modalities for the early detection of atherosclerotic
782
The American Journal of Cardiology (www.ajconline.org)
vascular disease such as brachial artery reactivity testing, carotid intima medial thickness, or electron beam computed tomography is an area of potential future investigation. The limitations of our study were the same as those traditionally associated with any observational study. An unrecognized selection bias might have been present for patients referred for ABI testing with exercise. However, it is unlikely that this bias would have had a substantial effect on the association between postexercise ABI and mortality. Other limitations included a lack of information pertaining to the use of statins and antithrombotic medications at testing. Also, most of the risk factors noted were determined by self-reported rates, which perhaps best explains the low rates of smoking in the patient groups. 1. Third Report of the National Cholesterol Education Program (NCEP). Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III) final report. Circulation 2002;106:3143–3421. 2. Ouriel K, McDonnell AE, Metz CE, Zarins CK. Critical evaluation of stress testing in the diagnosis of peripheral vascular disease. Surgery 1982;91:686 – 693. 3. Joffe MM, Rosenbaum PR. Invited commentary: propensity scores. Am J Epidemiol 1999;150:327–333. 4. Eagle KA, Rihal CS, Foster ED, Mickel MC, Gersh BJ; Coronary Artery Surgery Study (CASS) Investigators. Long-term survival in patients with coronary artery disease: importance of peripheral vascular disease. J Am Coll Cardiol 1994;23:1091–1095. 5. Newman AB, Shemanski L, Manolio TA, Cushman M, Mittelmark M, Polak JF, Powe NR, Siscovick D; Cardiovascular Health Study Group. Ankle-arm index as a predictor of cardiovascular disease and mortality in the Cardiovascular Health Study. Arterioscler Thromb Vasc Biol 1999;19:538 –545. 6. Silvestro A, Scopacasa F, Ruocco A, Olivia G, Schiano V, Zincarelli C, Brevetti G. Inflammatory status and endothelial function in asymptomatic and symptomatic peripheral arterial disease. Vasc Med 2003; 8:225–232. 7. Brevetti G, Silvestro A, Di Giacomo S, Bucur R, Di Donato A, Schiano V, Scopacasa F. Endothelial dysfunction in peripheral arterial disease is related to increase in plasma markers of inflammation and severity of peripheral circulatory impairment but not to classic risk factors and atherosclerotic burden. J Vasc Surg 2003;38:374 –379. 8. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Plasma concentration of C-reactive protein and risk of developing peripheral vascular disease. Circulation 1998;97:425– 428. 9. Brevetti G, Piscione F, Silvestro A, Galasso G, Di Donato A, Oliva G, Scopacasa F, Chiarello M. Increased inflammatory status and higher prevalence of three-vessel coronary artery disease in patients with concomitant coronary and peripheral atherosclerosis. Thromb Haemost 2003;89:1058 –1063.
10. Kirk G, Hickman P, McLaren M, Stonebridge PA, Belch JJ. Interleukin-8 (IL-8) may contribute to the activation of neutrophils in patients with peripheral arterial occlusive disease (PAOD). Eur J Vasc Endovasc Surg 1999;18:434 – 438. 11. Robless PA, Okonko D, Lintott P, Mansfield AO, Mikhailidis DP, Stansby GP. Increased platelet aggregation and activation in peripheral arterial disease. Eur J Vasc Endovasc Surg 2003;25:16 –22. 12. Reininger CB, Boeger CA, Steckmeier B, Spannagl M, Scweiberer L. Mechanisms underlying increased platelet reactivity in patients with peripheral arterial disease. Preliminary results. Int Angiol 1999;18: 163–170. 13. Cassar K, Bachoo P, Brittenden J. The role of platelets in peripheral vascular disease. Eur J Vasc Endovasc Surg 2003;25:6 –15. 14. Smith FB, Lee AJ, Hau CM, Rumley A, Lowe GD, Fowkes FG. Plasma fibrinogen, haemostatic factors and prediction of peripheral arterial disease in the Edinburgh Artery Study. Blood Coagul Fibrinolysis 2000;11:43–50. 15. Criqui MH, Denenberg JO, Bird CE, Fronek A, Klauber MR, Langer RD. The correlation between symptoms and non-invasive test results in patients referred for peripheral arterial disease testing. Vasc Med 1996;1:65–71. 16. McDermott MM, Mehta S, Greenland P. Exertional leg symptoms other than intermittent claudication are common in peripheral arterial disease. Arch Intern Med 1999;159:387–392. 17. McDermott MM, Mehta S, Ahn H, Greenland P. Atherosclerotic risk factors are less intensively treated in patients with peripheral arterial disease than in patients with coronary artery disease. J Gen Intern Med 1997;12:209 –215. 18. Hertzer NR, Beven EG, Young JR, O’Hara PJ, Ruschhaupt WF, Graor RA, Dewolfe VG, Maljovec LC. Coronary artery disease in peripheral vascular patients: a classification of 1000 coronary angiograms and results of surgical management. Ann Surg 1984;199:223–233. 19. Stein R, Hriljac I, Halperin JL, Gustavson SM, Teodorescu V, Olin JW. Limitation of the resting ankle-brachial index in symptomatic patients with peripheral arterial disease. Vasc Med 2006;11:29 –33. 20. Feringa HH, Bax JJ, van Waning VH, Boersma E, Elhendy A, Schouten O, Tangelder MJ, van Sambeek MH, van den Meiraceker AH, Poldermans D. The long-term prognostic value of the resting and postexercise ankle-brachial index. Arch Intern Med 2006;166:529 – 535. 21. McPhail IR, Spittell PC, Weston SA, Bailey KR. Intermittent claudication: an objective office-based assessment. J Am Coll Cardiol 2001; 37:1381–1385. 22. Hartmann A, Gehring A, Vallbracht C, Landgraf H, Liermann D, Kollath J, Kaltenbach M. Noninvasive methods in the early detection of restenosis after percutaneous transluminal angioplasty in peripheral arteries. Cardiology 1994;84:25–32. 23. Tong Y, Somjen G, Teeuwsen W, Royle JP. Percutaneous transluminal angioplasty: follow-up with treadmill exercise testing. Cardiovasc Surg 1994;2:503–507. 24. Sikkink CJ, van Asten WN, Van’t Hof MA, van Langen H, van der Vliet JA. Decreased ankle/brachial indices in relation to morbidity and mortality in patients with peripheral arterial disease. Vasc Med 1997; 2:169 –173.
Comparison of Central Artery Elasticity in Swimmers, Runners, and the Sedentary Nantinee Nualnim, MS, Jill N. Barnes, PhD, Takashi Tarumi, MA, Christopher P. Renzi, MA, and Hirofumi Tanaka, PhD* Although swimming is one of the most popular, most practiced, and most recommended forms of physical activity, little information is available regarding the influence of regular swimming on vascular disease risks. Using a cross-sectional study design, key measurements of vascular function were performed in middle-aged and older swimmers, runners, and sedentary controls. There were no group differences in age, height, dietary intake, and fasting plasma concentrations of glucose, total cholesterol, and low-density lipoprotein cholesterol. Runners and swimmers were not different in their weekly training volume. Brachial systolic blood pressure and pulse pressure were higher (p <0.05) in swimmers than in sedentary controls and runners. Runners and swimmers had lower (p <0.05) carotid systolic blood pressure and carotid pulse pressure than sedentary controls. Carotid arterial compliance was higher (p <0.05) and -stiffness index was lower (p <0.05) in runners and swimmers than in sedentary controls. There were no significant group differences between runners and swimmers. Cardiovagal baroreflex sensitivity was greater (p <0.05) in runners than in sedentary controls and swimmers and baroreflex sensitivity tended to be higher in swimmers than in sedentary controls (p ⴝ 0.07). Brachial artery flow-mediated dilation was significant greater (p <0.05) in runners compared with sedentary controls and swimmers. In conclusion, our present findings are consistent with the notion that habitual swimming exercise may be an effective endurance exercise for preventing loss in central arterial compliance. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:783–787) Swimming can be an ideal mode of exercise for those at increased risk of vascular disease including elderly patients, obese patients, and patients with arthritis. Although swimming is widely promoted and recommended as a mode of aerobic exercise by national and international organizations,1–3 research focusing on influences of swimming on vascular disease risks is lacking.4 The aims of the present investigation were to determine (1) whether swimmers would demonstrate higher levels of key vascular function measurements (i.e., central artery compliance, arterial baroreflex sensitivity, and endothelial-dependent vasodilation) than sedentary controls and (2) if levels of such vascular measurements are different from runners who are matched for age and exercise training status. Methods We studied 75 apparently healthy middle-aged and older adults (37 to 75 years of age). They were swimmers, runners, or sedentary controls. All subjects were nonobese, nonsmoking, normotensive (⬍140/90 mm Hg), normolipi-
Cardiovascular Aging Research Laboratory, Department of Kinesiology and Health Education, The University of Texas at Austin, Austin, Texas. Manuscript received September 2, 2010; revised manuscript received and accepted October 19, 2010. This study was in part supported by Grant GRNT2010136 from the American Heart Association. *Corresponding author: Tel: 512-232-4801; fax: 512-471-0946. E-mail address:
[email protected] (H. Tanaka). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.062
demic, and free of overt cardiovascular and other chronic diseases as assessed by medical history questionnaire, blood chemistry, and hematologic evaluation. None of the subjects were taking cardiovascular-acting medications including hormone replacement therapy. Physical activity status was verified by a modified physical activity questionnaire5 and maximal oxygen consumption. In average, runners and swimmers had been exercising 4.1 ⫾ 2.2 times/week for 9 ⫾ 2 years and 4.8 ⫾ 1.1 times/week for 9 ⫾ 2 years, respectively. Sedentary participants had been sedentary at least for the previous 12 months. All subjects gave their written informed consent to participate. The study was reviewed and approved by the institutional review board. Before they were tested, subjects abstained from food, alcohol, and caffeine for ⱖ4 hours (overnight 12-hour fast for metabolic risk factors). Premenopausal women were tested during the early follicular phase of the menstrual cycle. All testing was performed 24 to 48 hours after the last exercise bout.6 Body composition was assessed using dual-energy x-ray absorptiometry (Lunar DPX, GE Medical, Fairfield, Connecticut). A 3-day dietary intake record was obtained and analyzed by a registered dietitian using Nutritionist Pro software (Axxya Systems, Stafford, Texas). A blood sample was collected by venipuncture after an overnight fast. Fasting plasma concentrations of glucose, lipids, and lipoproteins were determined using a Vitros DT60 analyzer (OrthoClinical Diagnostics, Raritan, New Jersey). Graded exercise testing was undertaken using a metabolic cart during a www.ajconline.org
784
The American Journal of Cardiology (www.ajconline.org)
modified Bruce protocol. After a 5-minute warm-up, subjects walked or ran while the treadmill slope was gradually increased 2% every 2 minutes until volitional exhaustion. Bilateral brachial and ankle blood pressures, carotid and femoral pulse waves, and heart rate were measured by an automated vascular testing device (VP-2000, Omron Healthcare Bannockburn, Illinois)7 after a subject had been lying in a supine position for ⱖ15 minutes. Ankle– brachial pressure index was calculated as ankle systolic blood pressure divided by brachial systolic blood pressure. Carotid and femoral artery pulse waves were recorded by arterial applanation tonometry incorporating an array of 15 micropiezoresistive transducers placed on the carotid and femoral arteries. Time delay was measured automatically with the foot-to-foot method, and pulse wave velocity was subsequently calculated. Augmentation index, an index of arterial wave reflection, was obtained using an arterial tonometry placed on the carotid artery as previously described.8 Arterial compliance and -stiffness index were measured noninvasively by a combination of ultrasound imaging on the carotid artery and simultaneous applanation tonometry on the contralateral artery.9 A longitudinal image of the common carotid and femoral artery were acquired 1 to 2 cm proximal to the bifurcation using an ultrasound machine equipped with a high-resolution linear array transducer (Philips iE33 Ultrasound System, Philips, Bothel, Washington). All data were analyzed by the same investigator who was blinded to group assignment and used image analysis software (Vascular Research Tool Carotid Analyzer, Medical Imaging Applications, Coralville, Iowa). Pressure waveform and amplitude were obtained from the contralateral artery using arterial applanation tonometry (VP-2000, Omron Healthcare) and analyzed by waveform browsing software (WinDaq 2000, Dataq Instruments, Akron, Ohio). Brachial artery flow-mediated dilation (FMD) was measured using a standard procedure as described previously.10 Briefly, a pneumatic blood pressure cuff was positioned 2 inches below the antecubital fossa. Brachial diameter and blood flow velocity were acquired from a Doppler ultrasound machine equipped with a high-resolution linear array transducer (Philips iE33 Ultrasound). After baseline images were obtained, the cuff was inflated to 100 mm Hg above a subject’s systolic blood pressure for 5 minutes. All ultrasound-derived blood flow and diameter data were analyzed by the same investigator using image analysis software (Brachial Analyzer, Medical Imaging Applications). FMD was calculated as (maximal artery diameter minus baseline artery diameter)/baseline artery diameter ⫻ 100.11 Cardiovagal baroreflex sensitivity (BRS) was determined using the Valsalva maneuver as previously described.12,13 Subjects performed 3 Valsalva maneuvers by forcibly exhaling against a closed airway. Subjects were asked to maintain an expiratory mouth pressure of 40 mm Hg for 10 seconds. Data for cardiovagal BRS were recorded and analyzed by waveform browsing software (WinDaq 2000) during the phase IV overshoot. Systolic blood pressure values were linearly regressed against corresponding RR intervals from the point where the RR intervals began to lengthen to the point of maximal systolic blood pressure increase.14
Table 1 Selected subject characteristics Variable
Sedentary
Runners
Swimmers
Men/women Age (years) Height (cm) Body mass (kg) Body mass index (kg/m2) Body fat (%) Maximal oxygen consumption (ml/kg/ min) Physical activity score (units) Total caloric intake (kcal/ day) Carbohydrate intake (%) Fat intake (%) Protein intake (%) Alcohol intake (%) Sodium intake (mg/day) Total cholesterol (mg/dl) Low-density lipoprotein cholesterol (mg/dl) High-density lipoprotein cholesterol (mg/dl) Triglyceride (mg/dl) Plasma glucose (mg/dl)
16/9 54 ⫾ 2 170 ⫾ 2 74 ⫾ 2 26 ⫾ 1 30 ⫾ 2 31 ⫾ 2
17/8 52 ⫾ 2 173 ⫾ 2 67 ⫾ 2* 22 ⫾ 1* 18 ⫾ 2* 50 ⫾ 2*
17/8 56 ⫾ 2 173 ⫾ 2 76 ⫾ 2† 25 ⫾ 1† 24 ⫾ 2*† 41 ⫾ 2*†
11 ⫾ 4
58 ⫾ 3*
57 ⫾ 4*
2,370 ⫾ 212
2,343 ⫾ 169
2,160 ⫾ 150
46 ⫾ 3 36 ⫾ 2 13 ⫾ 1 5⫾2 3,472 ⫾ 348 193 ⫾ 9 120 ⫾ 8
49 ⫾ 3 32 ⫾ 2 14 ⫾ 1 4⫾1 2,968 ⫾ 278 179 ⫾ 9 105 ⫾ 8
47 ⫾ 2 33 ⫾ 2 16 ⫾ 1 3⫾1 3,014 ⫾ 246 194 ⫾ 11 123 ⫾ 9
47 ⫾ 3
61 ⫾ 3*
52 ⫾ 4†
125 ⫾ 14 94 ⫾ 3
70 ⫾ 14* 95 ⫾ 3
97 ⫾ 15 99 ⫾ 3
Values are means ⫾ SEMs. * p ⬍0.05 versus sedentary. † p ⬍0.05 versus runners. Table 2 Hemodynamic measurements at rest Variable
Sedentary
Heart rate (beats/min) Systolic blood pressure (mm Hg) Mean blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Pulse pressure (mm Hg) Carotid systolic pressure (mm Hg) Carotid pulse pressure (mm Hg) Brachial-ankle pulse wave velocity (cm/s) Carotid augmentation index (%) Carotid artery diameter at end-diastole (mm) Brachial artery diameter at end-diastole (mm)
60 ⫾ 2 119 ⫾ 3
50 ⫾ 2* 119 ⫾ 3
58 ⫾ 2† 128 ⫾ 3*†
89 ⫾ 2
88 ⫾ 2
93 ⫾ 2
71 ⫾ 1
72 ⫾ 1
74 ⫾ 2
48 ⫾ 2 116 ⫾ 3
47 ⫾ 2 104 ⫾ 3*
54 ⫾ 2*† 104 ⫾ 3*
47 ⫾ 2
36 ⫾ 2*
37 ⫾ 2*
1,336 ⫾ 36 15 ⫾ 5 6.77 ⫾ 0.16 4.2 ⫾ 1.9
Runners
Swimmers
1,230 ⫾ 35*
1,334 ⫾ 36†
13 ⫾ 4
13 ⫾ 5
6.0 ⫾ 0.15* 4.2 ⫾ 1.8
6.67 ⫾ 0.15† 4.6 ⫾ 1.8
Values are means ⫾ SEMs. * p ⬍0.05 versus sedentary. † p ⬍0.05 versus runners.
Analysis of covariance was used for statistical analysis to determine significant group differences. Correlation and regression analyses were performed to determine the relation
Miscellaneous/Swimming and Arterial Stiffness
785
Figure 3. Cardiovagal baroreflex sensitivity. Values are means ⫾ SEMs. * p ⬍0.05 versus sedentary.
Figure 1. Carotid artery compliance and -stiffness index. Values are means ⫾ SEMs. *p ⬍0.05 versus sedentary.
Figure 4. Flow-mediated dilation. Values are means ⫾ SEMs. *p ⬍0.05 versus sedentary; †p ⬍0.05 versus runners.
Figure 2. Femoral artery compliance and -stiffness index. Values are means ⫾ SEMs.
between carotid arterial compliance and cardiovagal BRS. All variables were expressed as mean ⫾ SEM. Results As presented in Table 1, there were no group differences in age, height, and dietary intakes. Body mass and body
mass index were lower (p ⬍0.05) in runners than in sedentary controls and swimmers. Percent body fat of swimmers was lower (p ⬍0.05) than of sedentary controls but higher (p ⬍0.05) than of runners. Maximal oxygen consumption of swimmers was greater than of sedentary controls but lower than of runners. Fasting plasma concentrations of glucose, total cholesterol, and low-density lipoprotein cholesterol were not different among groups. Runners had significantly lower plasma triglyceride and higher high-density lipoprotein cholesterol concentrations than sedentary controls. Heart rate at rest was lower (p ⬍0.05) in runners than in sedentary controls and swimmers (Table 2). Brachial systolic blood pressure and pulse pressure were higher (p ⬍0.05) in swimmers than in sedentary controls and runners. Runners and swimmers had lower (p ⬍0.05) carotid systolic blood pressure and carotid pulse pressure than sedentary controls. Brachial ankle pulse wave velocity was significantly lower in runners than in swimmers and sedentary controls. There were no group differences in carotid artery augmentation index. Carotid arterial compliance was higher (p ⬍0.05) and -stiffness index was lower (p ⬍0.05) in runners and swimmers than in sedentary controls (Figure 1). Unlike measurements of central artery stiffness, measurements of peripheral artery stiffness, femoral artery compliance, and femoral -stiffness index were not different among the 3 groups (Figure 2). Cardiovagal BRS was greater (p ⬍0.05) in
786
The American Journal of Cardiology (www.ajconline.org)
runners than in sedentary controls and swimmers (Figure 3). Cardiovagal BRS of swimmers tended to be higher than in sedentary controls but this did not achieve statistical significance (p ⫽ 0.07). Cardiovagal BRS was positively associated with carotid arterial compliance in the pooled population (r ⫽ 0.44, p ⬍0.01). Brachial artery FMD was significant greater (p ⬍0.05) in runners compared with sedentary controls and swimmers (Figure 4). Peak brachial artery blood flow and calculated shear rate were not different among the 3 groups so FMD values were not adjusted for shear rate. Discussion This is the first study, to our knowledge, to determine whether swimmers would exhibit a similar phenotype of vascular function to runners. The salient finding of the present study is that central artery compliance was greater in swimmers than in age-matched sedentary controls, and the level of arterial compliance was not different from runners, suggesting that high levels of regular swimming exercise may prevent arterial stiffening similar to land-based exercises. In the present study, middle-aged and older swimmers demonstrated higher levels of brachial systolic and pulse pressures than runners and sedentary controls. Our findings are consistent with previous cross-sectional studies showing that cardiovascular risk profiles, in particular arterial blood pressure, of swimmers are less favorable than those of runners.15,16 Previous exercise intervention studies have also shown that swimming exercise intervention may induce effects on blood pressure that are smaller in magnitude than running exercise17 and may even increase blood pressure.18 Interestingly, in the present study, “central” blood pressures were significantly lower in swimmers than in sedentary controls. Central blood pressure is determined by several factors including aortic diameter, arterial wave reflection, and left ventricular ejection characteristic19 and is a better predictor of cardiovascular disease risks than brachial blood pressure.20 Lower central blood pressure in swimmers was associated with greater arterial compliance. Taken together, these results are consistent with the notion that regular swimming exercise plays an important role in preventing arterial stiffening. It remains unclear how regular aerobic exercise improves arterial compliance. One possibility is that regular physical activity may act on the elasticity of the artery through endothelium-dependent vasodilation.21 FMD serves as an index of nitric oxide–mediated endothelium-dependent vasodilator function in humans. In the present study, greater arterial compliance in swimmers was not associated with a higher FMD. These results in swimmers are consistent with our previous pharmacologic study showing that nitric oxide does not appear to play a role in increasing arterial compliance through regular walking exercise.22 Other possibilities to explain the beneficial effects of regular exercise on macrovascular function include decreases in vascular vasoconstrictor tone,22 endothelin-1,23 and collagen crosslinking. In contrast to the central arteries, compliance of peripheral arteries does not appear to change much with different interventions or states including aging and endurance train-
ing.8,9 Consistent with this, we found that femoral arterial compliance was not different among the 3 groups. A lack of influence of regular exercise on peripheral arterial compliance is attributed to the fact that peripheral arteries do not exhibit the same extent of pulsatile changes in diameter compared with central (cardiothoracic) arteries. Accumulating evidence has indicated that habitual aerobic exercise favorably modulates age-associated decreases in cardiovagal BRS.12,14 Consistent with previous findings, results from the present study showed that cardiovagal BRS is increased in middle-aged and older endurance-trained runners compared with sedentary controls. Cardiovagal BRS was ⬃25% greater in swimmers than in sedentary controls, and there was a trend (p ⫽ 0.07) for the difference to be significant. Moreover, cardiovagal BRS was significantly associated with arterial compliance in the pooled population. Thus, regular swimming appears to be associated with a greater level of cardiovagal BRS. Is regular swimming associated with a lower risk of cardiovascular and all-cause mortalities? Only 2 studies are available to answer this question. In 1 epidemiologic study, swimming was not associated with a lower risk of cardiovascular disease, although walking and running examined in the same study demonstrated significant associations.24 A more recent epidemiologic study, however, reported a smaller relative risk of developing cardiovascular disease in swimmers than in sedentary populations,25 and the relative risk of swimmers were even lower than those of walkers and runners. Thus, at present, it remains highly controversial as to whether swimming is equally cardioprotective to landbased exercise. Similar to these epidemiologic studies, our findings are somewhat divergent. Central arterial compliance was greater in swimmers than in sedentary controls. However, endothelium-dependent vasodilation as assessed by FMD was not different between swimmers and sedentary controls. Clearly, more research effort should be directed toward the influence of swimming exercise on vascular disease risks. Swimming is an attractive form of exercise because it is easily accessible, inexpensive, and isotonic.26 Because of the buoyancy of water, compressive stress on joints is low and orthopedic injury rate is low.27 Due to cold temperature and increased thermoconductivity of surrounding water, heat-related illness is extremely low.28 Thus, swimming can be an ideal form of exercise for those at increased risk of vascular disease including elderly patients, obese patients, and patients with arthritis. Results of the present study indicate that swimming is also associated with destiffening effects on the central artery. However, it should be noted that the beneficial impact of regular swimming may be smaller than land-based exercises as summarized in a recent review.26 Even within our subject sample, swimmers did not exhibit phenotypes in lipid profiles and brachial blood pressure similar to runners. In addition to use of a cross-sectional study design, the present study has other limitations that should be discussed. Although swimmers and runners were matched well for exercise training volume, maximal oxygen consumption was significantly lower in swimmers than in runners. However, based on the principles of specificity of training, this is an expected finding because maximal oxygen consumption
Miscellaneous/Swimming and Arterial Stiffness
was assessed on a treadmill. Transfer of cardiovascular training benefits is very limited for maximal oxygen consumption on a treadmill when swimming exercise is performed as a training method.29 1. Fletcher GF, Blair SN, Blumenthal J, Caspersen C, Chaitman B, Epstein S, Falls H, Froelicher ES, Froelicher VF, Pina IL. Benefits and recommendations for physical activity programs for all Americans: a statement for health professionals by the committee on exercise and cardiac rehabilitation of the council cardiology. Circulation 1992;86: 340 –344. 2. World Hypertension League. Physical exercise in the management of hypertension: a consensus statement by the World Hypertension League. J Hypertens 1991;9:283–287. 3. Guideline Subcommittee. 1993 Guidelines for the management of mild hypertension: memorandum from a World Health Organization/International Society of Hypertension meeting. J Hypertens 1993;11:905– 918. 4. Tanaka H. Swimming exercise. Impact of aquatic exercise on cardiovascular health. Sports Med 2009;39:377–387. 5. Godin G Sr. A simple method to assess exercise behavior in the community. Can J Appl Sport Sci 1985;10:141–146. 6. Tanaka H, Sommerlad SM, Renzi CP, Barnes JN, Nualnim N. Postexercise hypotension and blood lipoprotein changes following swimming exercise. In: Kjendlie PL, Stallman RK, Cabri J, eds. Biomechanics and Medicine in Swimming. XI. Oslo, Norway: Norwegian School of Sport Science, 2010:381–383. 7. Cortez-Cooper MY, Tanaka H. A new device for automatic measurements of arterial stiffness and ankle-brachial index. Am J Cardiol 2003;91:1519 –1522. 8. Tanaka H, Seals D. Absence of age-related increase in central arterial stiffness in physically active women. Arterioscler Thromb Vasc Biol 1998;18:127–132. 9. Tanaka H, Monahan K, Clevenger C, DeSouza C, Seals D. Aging, habitual exercise, and dynamic arterial compliance. Circulation 2000; 102:1270 –1275. 10. Corretti MC, Benjamin EJ, Celermajer D, Charbonneau F, Creager MA, Deanfield J, Drexler H, Gerhard-Herman M, Herrington D, Vallance P, Vita J, Vogel R. Guidelines for the ultrasound assessment of endothelial-dependent flow-mediated vasodilation of the brachial artery. A report of the International Brachial Artery Reactivity Task Force. J Am Coll Cardiol 2002;39:257–265. 11. Corretti MC, Vogel RA. Technical aspects of evaluating brachial artery vasodilation using high frequency ultrasound. Am J Physiol Heart Circ Physiol 1995;268:H1397–H1404. 12. Monahan KD, Dinenno FA, Seals DR. Central arterial compliance is associated with age- and habitual exercise-related differences in cardiovagal baroreflex sensitivity. Circulation 2001;104:1627–1632. 13. Cook JN, Schleifer JL, Anton MM, Cortez-Cooper MY, Tanaka H. Arterial compliance of rowers: implications for combined aerobic and strength training on arterial elasticity Am J Physiol Heart Circ Physiol 2006;290:H1596 –H1600.
787
14. Monahan KD, Tanaka H, Clevenger CM, DeSouza CA, Seals DR. Regular aerobic exercise modulates age-associated declines in cardiovagal baroreflex sensitivity in healthy men. J Physiol 2000;529:263– 271. 15. Jost J, Weicker H. Sympathoadrenergic regulation and the adrenoceptor system. J Appl Physiol 1990;68:897–904. 16. Marconnet P, Gastaud M, Ardisson JL. Preexercise, exercise and early post-exercise arterial blood pressure in young competitive swimmers versus non swimmers. J Sports Med 1984;24:252–258. 17. Tanaka H, Bassett DR, Howley ET, Thompson DL, Ashraf M, Rawson FL. Swimming training lowers the resting blood pressure in individuals with hypertension. J Hypertens 1997;15:651– 657. 18. Cox KL, Burke V, Beilin LJ, Grove JR, Blanksby BA, Puddey IB. Blood pressure rise with swimming versus walking in older women: the Sedentary Women Exercise Adherence Trial 2 (SWEAT 2). J Hypertens 2006;24:307–314. 19. O’Rourke MF. Aortic diameter, aortic stiffness, and wave reflection increase with age and isolated systolic hypertension. Hypertension 2005;45:652– 658. 20. Protogerou AD, Blacher J, Papamichael CM, Lekakis JP, Safar ME. Central blood pressures: do we need them in the management of cardiovascular disease? Is it feasible therapeutic target? J Hypertens 2007;25:265–272. 21. DeSouza CA, Clevenger CM, Dinenno FA, Monahan KD, Tanaka H, Seals DR. Regular aerobic exercise prevents and restores age-related declines in endothelium-dependent vasodilation in healthy men Circulation 2000;102:1351–1357. 22. Sugawara J, Hayashi K, Yoshizawa M, Otsuki T, Shimojo N, Miyauchi T, Yokoi T, Maeda S, Tanaka H. Reduction in alpha-adrenergic receptor-mediated vascular tone contributes to improved arterial compliance with endurance training. Int J Cardiol 2009;135:346 –352. 23. Maeda S, Yoshizawa M, Otsuki T, Shimojo N, Jesmin S, Ajisaka R, Miyauchi T, Tanaka H. Involvement of endothelin-1 in habitual exercise-induced increase in arterial compliance. Acta Physiol 2009;196: 223–229. 24. Tanasescu M, Rimm EB, Willett WC, Stanpfer MJ, Hu FB. Exercise type and intensity in relation to coronary heart disease in men. JAMA 2002;288:1994 –2000. 25. Chase NL, Blair SN. Swimming and all-cause mortality risk compared with running, walking, and sedentary habits in men. Int J Aquatic Res Educ 2008;2:213–223. 26. Tanaka H. Swimming exercise: impact of aquatic exercise on cardiovascular health. Sports Med 2009;39:377–387. 27. Levy CM, Kolin E, Berson BL. Cross training: risk or benefit? an evaluation of injuries in four athlete populations. Sports Med Clin Forum 1986;3:1– 8. 28. Sheldahl LM, Buskirk ER, Loomis JL, Hodgson JL, Mendez J. Effects of exercise in cool water on body weight loss. Int J Obes 1982;6:29 – 42. 29. Tanaka H. Effects of cross-training. Transfer of training effects on VO2max between cycling, running and swimming. Sports Med 1994; 18:330 –339.
Carcinoid Heart Disease Without the Carcinoid Syndrome but With Quadrivalvular Regurgitation and Unsuccessful Operative Intervention William Clifford Roberts, MDa,b,d,*, Cyril Abie Varughese, DOe, Jong Mi Ko, BAd, Paul A. Grayburn, MDb,d, Robert Frederick Hebeler, Jr., MDc, and Elizabeth C. Burton, MDa A 53-year-old woman is described who underwent mitral and aortic valve replacement and tricuspid valve annuloplasty for pure regurgitation at all 3 valve sites for unrecognized carcinoid heart disease without the carcinoid syndrome 22 days before death. Metastatic carcinoid was not recognized until necropsy, which disclosed a probable ovarian primary but with large hepatic metastases and left-sided cardiac involvement either greater than or equal to the right-sided involvement. Pulmonary hypertension, very unusual in carcinoid heart disease, persisted postoperatively and probably played a role in the patient’s early death. Hepatic metastasis with ovarian primary is most unusual in this circumstance. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:788 –792) It was in 1930, 80 years ago, when the first patient with a metastasizing carcinoid neoplasm associated with fibrous lesions on the right side of the heart was described.1 In 1931, the first patient with metastasizing carcinoid syndrome (head and upper chest flushes and diarrhea) associated with fibrous lesions not only on the tricuspid and pulmonic valves but also on the anterior mitral leaflet and left ventricular mural endocardium was described.2 Subsequently, of course, numerous reports have appeared describing clinical and morphologic features of the carcinoid syndrome and carcinoid heart disease.2,3 Most patients have the primary carcinoid in the small intestine, widespread metastases, and specific carcinoid plaques limited to the right side of the heart.2,3 The present report was prompted by study of a patient with severe mitral and aortic regurgitation leading to double valve replacement but without symptoms of the carcinoid syndrome but with metastasizing carcinoid. Case Description A 53-year-old mother of 6, who was born May 5, 1955, and died February 18, 2009, had been well until November 2008, when she noted exertional dyspnea, subcutaneous peripheral edema, and recurring palpitations, which proved to be runs of atrial fibrillation. The symptoms gradually worsened, with episodes of rapid heart rate, each lasting several minutes. On January 26, 2009, cardiac catheterization disclosed the following pressures in mm Hg: pulmonary arterial wedge mean 22, a wave 30, v wave 28; pulmonary trunk 56/10; right
a
Departments of Pathology, bInternal Medicine (Cardiology), and cCardiothoracic Surgery; dBaylor Heart and Vascular Institute, Baylor University Medical Center; and eDepartment of Internal Medicine, Methodist Dallas Medical Center, Dallas, Texas, USA. Manuscript received September 1, 2010; revised manuscript received and accepted October 11, 2010. *Corresponding author: Tel: 214-820-7911; fax: 214-820-7533. E-mail address:
[email protected] (W.C. Roberts). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.064
ventricle 56/18; right atrial mean 9, a wave 15, v wave 10; left ventricle 130/26; and aorta 127/58. Left ventriculography disclosed a normal-sized left ventricular cavity, with an ejection fraction of 60% and 4⫹/4⫹ mitral regurgitation. Aortography disclosed 4⫹/4⫹ aortic regurgitation. The coronary arteries were angiographically normal. Echocardiography showed thickened mitral and aortic valve cusps and a normal-sized left ventricle (Figure 1). On January 27, 2009, the mitral and aortic valves were replaced with mechanical prostheses (#20 ATS Medical [Minneapolis, Minnesota] in the aortic position and #29 St. Jude Medical [St. Paul, Minnesota] in the mitral position). A Maze procedure was also performed as well as tricuspid valve annuloplasty. The operatively excised valves are shown in Figures 2 and 3. The 6-day postoperative hospital course was characterized by a weight gain of 6.6 kg (from 76.7 to 83.3 kg), sinus rhythm, and gradual ambulation. On February 13, 2009, the patient was rehospitalized because of increasing weakness and dyspnea, evidence of gastrointestinal bleeding (on warfarin with an international normalized ratio of 4.7), and large pleural and pericardial effusions. The blood hemoglobin level was 11.1 g/dl, and the hematocrit was 35%. The serum bilirubin level was 1.4 mg/dl; alkaline phosphatase, 295 U/L; aspartate aminotransferase, 210 U/L; and alanine aminotransferase, 147 U/L. The blood urea nitrogen level was 17 mg/dl and glucose 96 mg/dl. The brain natriuretic peptide level was 430 pg/ml (normal range ⬍100). Echocardiography now showed a normal-sized, normally functioning left ventricle, a very dilated and dysfunctional right ventricle, and large pericardial and right pleural effusions (Figure 4). Approximately 1,800 ml of serous fluid was drained from the right pleural space. Shortly thereafter, the patient had a cardiac arrest and died. At necropsy, classic carcinoid neoplasms were present in the right ovary (2.3 ⫻ 2.1 ⫻ 1.9 cm), the liver (2 hemorrhagic and necrotic carcinoid masses, 10 ⫻ 9 ⫻ 9 and 3 ⫻ 3 ⫻ 3 cm; www.ajconline.org
Case Report/Carcinoid Heart Disease
789
atria were dilated, the right more than the left. The pulmonic valve cusps were thickened, and the anterior cusp was rigid and immobile. The foramen ovale was closed. Comments
Figure 1. Preoperative (1 day) echocardiograms in the patient are described. (A) Four-chamber view showing the thickened mitral and tricuspid leaflets and the normal-sized right ventricular and left ventricular cavities. (B) Long-axis view again showing the thickened mitral leaflets and also a thickened aortic valve cusp. AO ⫽ aorta; LA ⫽ left atrium; LV ⫽ left ventricle; RA ⫽ right atrium; RV ⫽ right ventricle.
Figure 2. Photographs of the operatively excised mitral (A,B) and aortic (C,D) valves. (A) View of the atrial aspect of the anterior mitral leaflet. (B) View of the ventricular aspect of the anterior mitral leaflet with marked fibrous thickening of the chordae tendineae. A fragment of posterior leaflet with attached papillary muscle is also visible. (C) Focally thickened ventricular aspect of the tricuspid aortic valve. (D) Aortic aspect.
liver weight 2,100 g); the serosa of the bowel (1 nodule) and uterus (1 nodule); the adrenal glands (multiple small nodules); the pancreas (1 nodule, 0.3 cm); and the heart (multiple microscopic-sized nodules). Sections of the tumors in the liver, ovary, and adrenal glands were positive for neuron-specific enolase. The heart (Figures 5 to 8) weighed 370 g. The coronary arteries were free of atherosclerotic plaque. The left ventricular cavity was small, and the right ventricular cavity was very dilated. No myocardial foci of fibrosis or necrosis were present, except in the posteromedial left ventricular papillary muscle, which was fibrotic. Both
Our patient underwent mitral and aortic valve replacement and tricuspid valve annuloplasty for pure mitral, aortic, and tricuspid valve regurgitation. Examination of the operatively excised valves suggested an appearance similar to that described in patients having taken fenfluraminephentermine for weight reduction. The patient, however, denied having ever taken that medication. Postoperatively, the patient’s condition worsened, and she died 22 days after the operation. Necropsy disclosed carcinoid tumors in 1 ovary, both adrenal glands, the liver, the pancreas, the serosal surfaces of the uterus and bowel, and the heart (intramyocardial). There was never evidence of the carcinoid syndrome (flushing, diarrhea), and the presence of the carcinoid neoplasm was not diagnosed until necropsy, which also disclosed evidence of carcinoid heart disease involving both right-sided cardiac valves as well as both left-sided valves. The primary in the patient described was not certain, but the ovary appeared to be the most logical site. Only 1 of the 2 ovaries was involved, and the cancerous nodule was ⬎2 cm in diameter. Hepatic metastasis of carcinoid with ovarian primary, however, is quite unusual.4,5 Chatterjee and Heather6 found hepatic carcinoid metastases in only 1 of 35 reported cases with primary carcinoid in the ovary. Although our patient had only 2 carcinoid metastases in the liver, both were large. In our patient, carcinoid heart disease involved all 4 cardiac valves: the pulmonic valve to a worse extent than the aortic valve, but the mitral valve to a worse extent than the tricuspid valve. Although the pulmonic valve was heavily involved by the carcinoid process, there was no pressure gradient across the valve, although the enddiastolic pressures in both the pulmonary trunk and the right ventricle were similar, indicating pure pulmonic regurgitation. Is the electrocardiogram helpful in patients with malignant carcinoid in diagnosing carcinoid heart disease? No. Ross and Roberts3 examined electrocardiograms in 34 patients with the carcinoid syndrome: the total 12-lead QRS voltages7 in the 19 patients with carcinoid heart disease ranged from 58 to 227 mm (mean 105; standardization 10 mm) and in the 15 patients without carcinoid heart disease from 89 to 192 mm (mean 132). Twelve-lead QRS voltages in 16 men aged 44 to 74 years (mean 55) without cardiovascular disease ranged from 84 to 159 mm (mean 124) and heart weight from 288 to 392 g (mean 351).8 Thus, in groups of patients with the carcinoid syndrome, those with carcinoid heart disease have lower total 12-lead QRS voltages, but much overlap occurred between the 2 groups. Our patient had a total 12-lead QRS voltage of 127 mm. The duration of the P-R, QRS, and Q-T intervals and heart rates at rest were similar in the 2 groups with and without carcinoid heart disease. The presence of pulmonary hypertension in our patient is most unusual in carcinoid heart disease and reasonably
790
The American Journal of Cardiology (www.ajconline.org)
Figure 3. Photomicrograph of anterior mitral leaflet (A), chordae tendineae (B), and aortic valve cusp (C). The underlying leaflet and chordae are normal, but both are quite thickened by superimposed cellular fibrous tissue devoid of elastic fibrils. The sinus portion of the aortic valve is filled with cellular fibrous tissue devoid of elastic fibers (C). Elastic van Gieson’s stains (100⫻).
Figure 4. Four-chamber echocardiogram shortly before death in the patient described. The right ventricular cavity is severely dilated, whereas the left ventricle (LV) is not dilated. LA ⫽ left atrium; RA ⫽ right atrium; RV ⫽ right ventricle.
Figure 6. View of heart after removing the atrial walls and most of the ascending aorta and pulmonary trunk. The prosthesis in the mitral position fills the entire “floor” of the left atrium. The cloth-covered ring is visible in the tricuspid valve annulus.
Figure 5. X-ray of the heart at necropsy. Prostheses are present in the mitral and aortic valve positions, and a ring is in the tricuspid valve annular position. The left ventricular cavity is not dilated, whereas the right ventricular cavity is considerably dilated.
can be attributed to the left-sided mitral disease.9 The reason for its persistence postoperatively is unclear. The patient’s symptoms worsened considerably postoperatively. The discs of the prostheses in both mitral and aortic positions moved without interference. Cardiac valve replacement and/or “repair” for carcinoid heart disease is becoming more accepted, but nevertheless, outcomes are not always favorable.9 In an early
Case Report/Carcinoid Heart Disease
791
Figure 7. View of various “slices” of the cardiac ventricles showing the very small left ventricular cavity and the very dilated right ventricular cavity.
Figure 8. View of the pulmonic valve from above. The anterior cusp is very thick and immobile. The other 2 cusps are only mildly thickened by fibrous tissue.
report from the Mayo Clinic, 9 of 26 patients died in the early perioperative period and 9 others a mean of 19 months postoperatively. Only 4 of their 26 patients had mitral or aortic valve replacement, and none had both valves replaced, as did our patient. Although the mortality was higher, late operative survival (8 of 26 patients) resulted in considerable decrease (2 patients) or elimination (6 patients) of symptoms.9 A later report from the same institution summarized results in 11 patients with carcinoid heart disease who underwent operation for leftand right-sided valve disease10: the tricuspid valve was replaced in all 11, the pulmonic in 3 (valvectomy in 7), the mitral valve in 6 (repair in 1), and the aortic valve in 4 (repair in 2). There were 2 perioperative deaths and 4 additional deaths in a mean follow-up period of 41 months. All but 1 operative survivor improved by ⱖ1 functional class. In retrospect, had the presence of carcinoid been recognized preoperatively or at operation in
792
The American Journal of Cardiology (www.ajconline.org)
our patient, both tricuspid and pulmonic valves probably also would have been replaced (quadruple valve replacement).11 Acknowledgment: We thank Brad J. Roberts, BS, RCS, RDCS, for his help in preparing the echocardiogram. 1. Cassidy MA. Abdominal carcinomatosis with probable adrenal involvement. Proc Roy Soc Med 1930;24:139 –141. 2. Roberts WC, Sjoerdsma A. The cardiac disease associated with the carcinoid syndrome (carcinoid heart disease). Am J Med 1964;36: 5–34. 3. Ross EM, Roberts WC. The carcinoid syndrome: comparison of 21 necropsy subjects with carcinoid heart disease to 15 necropsy subjects without carcinoid heart disease. Am J Med 1985;79:339 – 353. 4. Chaoqalit N, Connolly HM, Schaff HV, Webb MJ, Pellikka PA. Carcinoid heart disease associated with primary ovarian carcinoid tumor. Am J Cardiol 2004;93:1314 –1315. 5. Rabban JT, Lerwill MF, McCluggage WG, Grenert JP, Zaloudek CJ. Primary ovarian carcinoid tumors may express CDX-2: a potential
6. 7.
8.
9.
10.
11.
pitfall in distinction from metastatic intestinal carcinoid tumors involving the ovary. Int J Gynecol Pathol 2009;28:41– 48. Chatterjee K, Heather JC. Carcinoid heart disease from primary ovarian carcinoid tumors. Am J Med 1968;45:643– 648. Siegel RJ, Roberts WC. Electrocardiographic observations in severe aortic valve stenosis: correlative necropsy study to clinical, hemodynamic, and ECG variables demonstrating relation of 12-lead QRS amplitude to peak systolic transaortic pressure gradient. Am Heart J 1982;103:210 –221. Odom H II, Davis L, Dinh HA, Baker BJ, Roberts WC, Murphy ML. QRS voltage measurements in autopsied men free of cardiopulmonary disease: a basis for evaluating total QRS voltage as an index of left ventricular hypertrophy. Am J Cardiol 1986;58:801– 804. Connolly HM, Nishimura RA, Smith HC, Pellikka PA, Mullany CJ, Kvols LK. Outcome of cardiac surgery for carcinoid heart disease. J Am Coll Cardiol 1995;25:410 – 416. Connolly HM, Schaff HV, Mullany CJ, Rubin J, Abel MD, Pellikka PA. Surgical management of left-sided carcinoid heart disease. Circulation 2001;104:I36 –I40. Arghami A, Connolly HM, Abel MD, Schaff HV. Quadruple valve replacement in patients with carcinoid heart disease. J Thorac Cardiovasc Surg 2010;140:1432–1434.
Disappearance of Angina Pectoris by Lipid-Lowering in Type III Hyperlipoproteinemia Eun Jeung Cho, MD, Yun Joo Min, MD, Min Seok Oh, MD, Jee Eun Kwon, MD, Jeung Eun Kim, MD, and Chee Jeong Kim, MD* Type III hyperlipoproteinemia is a rare familial disease characterized by marked elevations of serum cholesterol and triglyceride levels caused by an accumulation of remnant lipoproteins in apolipoprotein E2/E2 homozygotes. It is associated with an increased risk for premature atherosclerotic vascular disease. A 55-year-old woman was diagnosed as having type III hyperlipoproteinemia on the basis of skin lesions, serum lipid levels, lipid electrophoresis, and apolipoprotein E genotyping and stable angina pectoris on the basis of typical symptoms and treadmill exercise electrocardiographic results. After 1 year of combination therapy with atorvastatin and fenofibrate, skin xanthomata disappeared, leaving minimal remnants. In addition, there was no exertional chest pain, and treadmill exercise electrocardiographic results were negative. This finding was confirmed by coronary computed tomographic angiography. This case suggests that proper medical therapy can induce the regression of uncomplicated coronary lesions in type III hyperlipoproteinemia. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;107:793–796) Case Description A 55-year-old woman visited the hospital because of progressive systemic skin lesions for 2 years and chest pain for 4 months. Anterior chest pain without radiation appeared during moderate exercise and was reduced with several minutes of rest. The patient’s body mass index was 25.4 kg/m2 (height 160 cm, weight 65 kg). She exhibited extensive eruptive and tuberous xanthomata over the elbow and buttock areas and yellowish orange discoloration of the palmar creases (Figure 1). Her mother and brother had experienced acute myocardial infarctions. Fasting cholesterol and triglyceride levels were 747 and 1,315 mg/dl, respectively. Lipid electrophoresis revealed a broad -band pattern (Figure 2), and polymerase chain reaction multiplex amplification refractory mutation system showed apolipoprotein E2/E2 (Arg158 ¡ Cys) genotype. Treadmill exercise electrocardiography showed 1.5-mm downsloping depression of the ST segment in leads II, III, aVF, and V4 to V6, with T-wave inversion at 8.9 METs (Figure 3), and the patient described chest pain. She refused any further invasive procedures and was managed with aspirin, nitrate, a calcium channel blocker, and a  blocker. In addition, fenofibrate 160 mg and atorvastatin 10 mg were administered. After 1 month of medications and diet control, fasting cholesterol and triglyceride levels decreased to 157 mg/dl and 108 mg/dl, respectively (Figure 4). During follow-up, aspartate aminotransferase and alanine aminotransferase levels increased to 178 and 232 IU/L, respectively, and Division of Cardiology, Department of Internal Medicine, College of Medicine, Chung-Ang University, Seoul, Korea. Manuscript received August 28, 2010; revised manuscript received and accepted October 19, 2010. This research was supported by Chung-Ang University Research Grants in 2009. *Corresponding author: Tel: 82-2-2260-2382; fax: 82-2-822-2769. E-mail address:
[email protected] (C.J. Kim). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.10.063
lipid-lowering drugs were discontinued. For 3 months, repeated administrations of fenofibrate increased aspartate aminotransferase and alanine aminotransferase levels. Hence afterward, atorvastatin was initially administered and fenofibrate was added again later. After that, fasting cholesterol (164 to 237 mg/dl) and triglyceride (105 to 193 mg/dl) levels were well controlled. After 3 months of repeat administration of lipid-lowering drugs, the hard consistency of the xanthomata was softened. After 1 year, the skin lesions disappeared, leaving minimal remnants (Figure 1). Treadmill exercise electrocardiographic results were negative (Figure 3). On coronary computed tomographic angiography, no stenotic lesions of the coronary arteries were observed (Figure 5). After the discontinuation of antianginal drugs, she did not experience exertional chest pain. Comments This case demonstrates that the proper medical management of hyperlipidemia can reverse stable angina pectoris occurring in a patient with type III hyperlipoproteinemia. To our knowledge, this is the first case showing that angina pectoris can be cured by the use of lipidlowering drugs only. Previous studies have reported that lipid-lowering drugs retarded the progression of coronary artery disease in type III hyperlipoproteinemia.1,2 The association of type III hyperlipoproteinemia with premature atherosclerotic vascular disease is well established.3,4 We recommended further invasive procedures in our patient, but she refused for personal reasons. This situation provided an opportunity to observe the effect of lipid-lowering drugs on angina pectoris. Although pretreatment coronary lesions were not documented, a history of typical exertional chest pain and positive results on treadmill exercise electrocardiography made it clear for this patient to have significant lesions of coronary arteries. Just as the skin xanthomata disappeared, stenotic lesions of coronary arteries mimicking the skin lesions www.ajconline.org
794
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Yellowish orange discoloration in palmar creases (A) and eruptive and tuberous xanthomata over the elbow (C) disappeared after medication (B,D).
Figure 2. Lipoprotein agarose gel electrophoresis. Broad -band pattern (A) disappeared 1 month after medication (B).
might be also regressed. Statins have been shown to decrease cholesterol and triglyceride levels by 39% to 52% and by 22% to 56%, respectively.5 Fibrates have been shown to reduce cholesterol and triglyceride levels by 40% to 45% and by 56% to 70%, respectively.6 However, neither was enough to meet the criteria for high-risk patients of published guidelines.7 A combination of a statin and a fibrate was shown to be more effective than monotherapy.8,9 In the present case, ator-
Figure 3. Treadmill exercise electrocardiogram before (A) and after (B) lipid-lowering drugs. T-wave inversion and 1.5-mm downsloping depression of ST segment in lead V6 were observed before therapy and were not detected after 1 year of medication.
Case Report/Angina Pectoris and Type III Hyperlipoproteinemia
795
Figure 4. Changes in cholesterol, triglyceride, and liver function with medication. AST ⫽ aspartate aminotransferase.
Figure 5. Coronary computed tomographic angiography revealed no significant stenotic lesions in the left (A) and right (B) coronary arteries.
vastatin and fenofibrate, along with diet control, decreased cholesterol and triglyceride levels by 78% and 92%, respectively. Combination therapy with a statin and a fibrate is known to be relatively safe.10 In the present case, aspartate aminotransferase and alanine aminotransferase levels were intermittently elevated, so these drugs were discontinued repeatedly and restarted later. There were no elevations of aspartate aminotransferase and alanine aminotransferase levels during 1-year follow-up
with a strict control of diet and intake of other medications. 1. Kuo PT, Wilson AC, Kostis JB, Moreyra AB, Dodge HT. Treatment of type III hyperlipoproteinemia with gemfibrozil to retard progression of coronary artery disease. Am Heart J 1988;116:85–90. 2. Kawashiri MA, Higashikata T, Takata M, Katsuda S, Miwa K, Nohara A, Inazu A, Kobayashi J, Shimizu M, Koizumi J, Mabuchi H. Type III hyperlipoproteinemia exaggerated by Sheehan’s syndrome with advanced systemic atherosclerosis: a 28-year clinical course. Circ J 2005;69:746–751.
796
The American Journal of Cardiology (www.ajconline.org)
3. Brewer HB Jr, Zech LA, Gregg RE, Schwartz D, Schaefer EJ. NIH conference. Type III hyperlipoproteinemia: diagnosis, molecular defects, pathology, and treatment. Ann Intern Med 1983;98:623– 640. 4. Hopkins PN, Wu LL, Hunt SC, Brinton EA. Plasma triglycerides and type III hyperlipidemia are independently associated with premature familial coronary artery disease. J Am Coll Cardiol 2005;45:1003–1012. 5. van Dam M, Zwart M, de Beer F, Smelt AHM, Prins MH, Trip MD, Havekes LM, Lansberg PJ, Kastelein JJP. Long term efficacy and safety of atorvastatin in the treatment of severe type III and combined dyslipidaemia. Heart 2002;88:234 –238 6. Lussier-Cacan S, Bard JM, Boulet L, Nestruck AC, Grothé AM, Fruchart JC, Davignon J. Lipoprotein composition changes induced by fenofibrate in dysbetalipoproteinemia type III. Atherosclerosis 1989; 78:167–182.
7. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001;285: 2486 –2497. 8. Illingworth DR, O’Malley JP. The hypolipidemic effects of lovastatin and clofibrate alone and in combination in patients with type III hyperlipoproteinemia. Metabolism 1990;39:403– 409. 9. Feussner G, Eichinger M, Ziegler R. The influence of simvastatin alone or in combination with gemfibrozil on plasma lipids and lipoproteins in patients with type III hyperlipidemia. Clin Invest 1992;70:1027–1035. 10. Ellen RL, McPherson R. Long-term efficacy and safety of fenofibrate and a statin in the treatment of combined hyperlipidemia. Am J Cardiol 1998;81:60B– 65B.
READERS COMMENTS Authors’ Reply We appreciate the thoughtful comments from Dr. Madias. We agree that digital measurements of QRS duration provide greater precision than physician measurements, although the former were unavailable for all but a small minority of participants in our study sample. Participants with shorter QRS durations at the earliest examination were more likely than those with longer QRS durations to experience lengthening of the QRS duration over time; however, the rate of progressing from the lowest (QRS duration ⬍100 ms) or the middle category (QRS duration 100 to ⬍120 ms) to a higher category over the entire follow-up period was 3.7% and 5.5%, respectively. Because rates of progression to a higher QRS category were relatively low, an analysis of its relation to outcomes would have limited power in our sample even if performed over the entire follow-up period. We agree that presence of fascicular block and degree of QRS axis could also be contributors to risk for permanent pacemaker; although coding of these variables was not available in our study sample, this could be the subject of future research. As pointed out, variation in heart rate is a likely precursor to sick sinus rhythm and eventual pacemaker placement. Given the high variability of heart rate as a single clinical measure, we reported the results of secondary analyses that adjusted for PR interval as well as heart rate, whereby PR interval is inversely correlated to heart rate but demonstrates greater temporal stability. We also have previously reported on the independent relation of PR interval with incident pacemaker placement.1 Susan Cheng, MD Thomas J. Wang, MD Boston, Massachusetts 24 October 2010
1. Cheng S, Keyes MJ, Larson MG, McCabe EL, Newton-Cheh C, Levy D, Benjamin EJ, Vasan RS, Wang TJ. Long-term outcomes in individuals with prolonged PR interval or first-degree atrioventricular block. JAMA 2009;301:2571– 2577. doi:10.1016/j.amjcard.2010.10.078
Paul Dudley White on Echocardiography. “Nobody Is Perfect” I enjoyed reading Dr. Blackburn’s1 brief report about Paul Dudley White in the October 15, 2010, issue of The American Journal of Cardiology. Another possible subtitle could be “Nobody Is Perfect.” I have a third-hand story about Dr. White, which I have reason to believe is true. In the early 1950s, Drs. Inge Edler (a cardiologist) and Helmuth Hertz (a physicist) used an ultrasound device borrowed from Siemens (Dr. Hertz’s father’s employer) to examine the heart. After only a few years, Dr. Hertz lost interest and left the field. He advised Siemens that there was no future in cardiac ultrasound (Sven Effert, personal communication). Siemens apparently wanted a second opinion. They asked Drs. Paul Dudley White and Andre Cournand to visit Dr. Edler and review his work. They too agreed that there was no future in cardiac ultrasound. Siemens then lost whatever interest it had in the field, and Edler soon thereafter also left the field. By the time I saw my first ultrasound instrument in 1963 (because of an erroneous advertisement2), there was no activity in cardiac ultrasound in Europe. The only person I could find using cardiac ultrasound was Claude Joyner in Philadelphia, who was recruited by Jack Reid, who had worked with ultrasound previously and was pursuing his PhD in physics at the University of Pennsylvania. They were trying to duplicate 1 of Edler’s findings, which ultimately proved to be clinically unreliable. The field essentially remained dead until we published our report on the use of cardiac ultrasound to detect pericardial effusion in 1965.3 Thus, Dr. White helped temporarily bury cardiac ultrasound, at least for a few years. “Cardiac ultrasound” is now known as “echocardiography.” We could not use “echocardiography” in the early days, because at the time, the only clinical use of diagnostic ultrasound was “echoencephalography.” As a result we had no usable abbreviation for “echocardiography.” We could not use
Am J Cardiol 2011;107:797–798 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved.
“ECG” because it was already taken for electrocardiography. We could not use “echo,” because it did not differentiate from echoencephalography. “Echocardiography” became feasible only when echoencephalography died. No other diagnostic ultrasound technique used “echo,” leaving it available as the abbreviation for echocardiography. Thus, the birth of “echocardiography” as the name for cardiac ultrasound. Harvey Feigenbaum, MD Indianapolis, Indiana 22 November 2010
1. Blackburn H. Paul Dudley White: “our hearts were young and gay.” Am J Cardiol 2010;106: 1193. 2. Feigenbaum H. The first long-lasting use of echocardiography: a result of a false advertisement. JACC Cardiovasc Imaging 2008;1:522– 524. 3. Feigenbaum H, Waldhausen JA, Hyde LP. Ultrasound diagnosis of pericardial effusion. JAMA 1965;191:711–714. doi:10.1016/j.amjcard.2010.11.015
Usefulness of Serial of B-Type Natriuretic Peptide and Troponin-I Levels to Predict Left Ventricular Remodeling After Primary Coronary Angioplasty We read with interest the work by Fertin et al,1 who reported the important finding that despite significant associations on univariate analyses, none of the biomarkers B-type natriuretic peptide (BNP), cardiac troponin I (TnI), and C-reactive protein at baseline (serial blood sampling from days 3 to 7) was retained as an independent predictor of left ventricular (LV) remodeling when the ejection fraction was entered into the multivariate model in Remodelage Ventriculaire 2 (REVE-2). In contrast, Hallén et al2 found that a single-point measurement of TnI 24 or 48 hours after primary angioplasty for those with ST-segment elevation myocardial infarctions in addition to the LV ejection fraction could provide prognostic information on LV remodeling from the results of FX06 in Ischemia and Reperfusion (FIRE). Actually, these results may all be correct, because the times of blood sampling at baseline in www.ajconline.org
READERS COMMENTS Authors’ Reply We appreciate the thoughtful comments from Dr. Madias. We agree that digital measurements of QRS duration provide greater precision than physician measurements, although the former were unavailable for all but a small minority of participants in our study sample. Participants with shorter QRS durations at the earliest examination were more likely than those with longer QRS durations to experience lengthening of the QRS duration over time; however, the rate of progressing from the lowest (QRS duration ⬍100 ms) or the middle category (QRS duration 100 to ⬍120 ms) to a higher category over the entire follow-up period was 3.7% and 5.5%, respectively. Because rates of progression to a higher QRS category were relatively low, an analysis of its relation to outcomes would have limited power in our sample even if performed over the entire follow-up period. We agree that presence of fascicular block and degree of QRS axis could also be contributors to risk for permanent pacemaker; although coding of these variables was not available in our study sample, this could be the subject of future research. As pointed out, variation in heart rate is a likely precursor to sick sinus rhythm and eventual pacemaker placement. Given the high variability of heart rate as a single clinical measure, we reported the results of secondary analyses that adjusted for PR interval as well as heart rate, whereby PR interval is inversely correlated to heart rate but demonstrates greater temporal stability. We also have previously reported on the independent relation of PR interval with incident pacemaker placement.1 Susan Cheng, MD Thomas J. Wang, MD Boston, Massachusetts 24 October 2010
1. Cheng S, Keyes MJ, Larson MG, McCabe EL, Newton-Cheh C, Levy D, Benjamin EJ, Vasan RS, Wang TJ. Long-term outcomes in individuals with prolonged PR interval or first-degree atrioventricular block. JAMA 2009;301:2571– 2577. doi:10.1016/j.amjcard.2010.10.078
Paul Dudley White on Echocardiography. “Nobody Is Perfect” I enjoyed reading Dr. Blackburn’s1 brief report about Paul Dudley White in the October 15, 2010, issue of The American Journal of Cardiology. Another possible subtitle could be “Nobody Is Perfect.” I have a third-hand story about Dr. White, which I have reason to believe is true. In the early 1950s, Drs. Inge Edler (a cardiologist) and Helmuth Hertz (a physicist) used an ultrasound device borrowed from Siemens (Dr. Hertz’s father’s employer) to examine the heart. After only a few years, Dr. Hertz lost interest and left the field. He advised Siemens that there was no future in cardiac ultrasound (Sven Effert, personal communication). Siemens apparently wanted a second opinion. They asked Drs. Paul Dudley White and Andre Cournand to visit Dr. Edler and review his work. They too agreed that there was no future in cardiac ultrasound. Siemens then lost whatever interest it had in the field, and Edler soon thereafter also left the field. By the time I saw my first ultrasound instrument in 1963 (because of an erroneous advertisement2), there was no activity in cardiac ultrasound in Europe. The only person I could find using cardiac ultrasound was Claude Joyner in Philadelphia, who was recruited by Jack Reid, who had worked with ultrasound previously and was pursuing his PhD in physics at the University of Pennsylvania. They were trying to duplicate 1 of Edler’s findings, which ultimately proved to be clinically unreliable. The field essentially remained dead until we published our report on the use of cardiac ultrasound to detect pericardial effusion in 1965.3 Thus, Dr. White helped temporarily bury cardiac ultrasound, at least for a few years. “Cardiac ultrasound” is now known as “echocardiography.” We could not use “echocardiography” in the early days, because at the time, the only clinical use of diagnostic ultrasound was “echoencephalography.” As a result we had no usable abbreviation for “echocardiography.” We could not use
Am J Cardiol 2011;107:797–798 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved.
“ECG” because it was already taken for electrocardiography. We could not use “echo,” because it did not differentiate from echoencephalography. “Echocardiography” became feasible only when echoencephalography died. No other diagnostic ultrasound technique used “echo,” leaving it available as the abbreviation for echocardiography. Thus, the birth of “echocardiography” as the name for cardiac ultrasound. Harvey Feigenbaum, MD Indianapolis, Indiana 22 November 2010
1. Blackburn H. Paul Dudley White: “our hearts were young and gay.” Am J Cardiol 2010;106: 1193. 2. Feigenbaum H. The first long-lasting use of echocardiography: a result of a false advertisement. JACC Cardiovasc Imaging 2008;1:522– 524. 3. Feigenbaum H, Waldhausen JA, Hyde LP. Ultrasound diagnosis of pericardial effusion. JAMA 1965;191:711–714. doi:10.1016/j.amjcard.2010.11.015
Usefulness of Serial of B-Type Natriuretic Peptide and Troponin-I Levels to Predict Left Ventricular Remodeling After Primary Coronary Angioplasty We read with interest the work by Fertin et al,1 who reported the important finding that despite significant associations on univariate analyses, none of the biomarkers B-type natriuretic peptide (BNP), cardiac troponin I (TnI), and C-reactive protein at baseline (serial blood sampling from days 3 to 7) was retained as an independent predictor of left ventricular (LV) remodeling when the ejection fraction was entered into the multivariate model in Remodelage Ventriculaire 2 (REVE-2). In contrast, Hallén et al2 found that a single-point measurement of TnI 24 or 48 hours after primary angioplasty for those with ST-segment elevation myocardial infarctions in addition to the LV ejection fraction could provide prognostic information on LV remodeling from the results of FX06 in Ischemia and Reperfusion (FIRE). Actually, these results may all be correct, because the times of blood sampling at baseline in www.ajconline.org
READERS COMMENTS Authors’ Reply We appreciate the thoughtful comments from Dr. Madias. We agree that digital measurements of QRS duration provide greater precision than physician measurements, although the former were unavailable for all but a small minority of participants in our study sample. Participants with shorter QRS durations at the earliest examination were more likely than those with longer QRS durations to experience lengthening of the QRS duration over time; however, the rate of progressing from the lowest (QRS duration ⬍100 ms) or the middle category (QRS duration 100 to ⬍120 ms) to a higher category over the entire follow-up period was 3.7% and 5.5%, respectively. Because rates of progression to a higher QRS category were relatively low, an analysis of its relation to outcomes would have limited power in our sample even if performed over the entire follow-up period. We agree that presence of fascicular block and degree of QRS axis could also be contributors to risk for permanent pacemaker; although coding of these variables was not available in our study sample, this could be the subject of future research. As pointed out, variation in heart rate is a likely precursor to sick sinus rhythm and eventual pacemaker placement. Given the high variability of heart rate as a single clinical measure, we reported the results of secondary analyses that adjusted for PR interval as well as heart rate, whereby PR interval is inversely correlated to heart rate but demonstrates greater temporal stability. We also have previously reported on the independent relation of PR interval with incident pacemaker placement.1 Susan Cheng, MD Thomas J. Wang, MD Boston, Massachusetts 24 October 2010
1. Cheng S, Keyes MJ, Larson MG, McCabe EL, Newton-Cheh C, Levy D, Benjamin EJ, Vasan RS, Wang TJ. Long-term outcomes in individuals with prolonged PR interval or first-degree atrioventricular block. JAMA 2009;301:2571– 2577. doi:10.1016/j.amjcard.2010.10.078
Paul Dudley White on Echocardiography. “Nobody Is Perfect” I enjoyed reading Dr. Blackburn’s1 brief report about Paul Dudley White in the October 15, 2010, issue of The American Journal of Cardiology. Another possible subtitle could be “Nobody Is Perfect.” I have a third-hand story about Dr. White, which I have reason to believe is true. In the early 1950s, Drs. Inge Edler (a cardiologist) and Helmuth Hertz (a physicist) used an ultrasound device borrowed from Siemens (Dr. Hertz’s father’s employer) to examine the heart. After only a few years, Dr. Hertz lost interest and left the field. He advised Siemens that there was no future in cardiac ultrasound (Sven Effert, personal communication). Siemens apparently wanted a second opinion. They asked Drs. Paul Dudley White and Andre Cournand to visit Dr. Edler and review his work. They too agreed that there was no future in cardiac ultrasound. Siemens then lost whatever interest it had in the field, and Edler soon thereafter also left the field. By the time I saw my first ultrasound instrument in 1963 (because of an erroneous advertisement2), there was no activity in cardiac ultrasound in Europe. The only person I could find using cardiac ultrasound was Claude Joyner in Philadelphia, who was recruited by Jack Reid, who had worked with ultrasound previously and was pursuing his PhD in physics at the University of Pennsylvania. They were trying to duplicate 1 of Edler’s findings, which ultimately proved to be clinically unreliable. The field essentially remained dead until we published our report on the use of cardiac ultrasound to detect pericardial effusion in 1965.3 Thus, Dr. White helped temporarily bury cardiac ultrasound, at least for a few years. “Cardiac ultrasound” is now known as “echocardiography.” We could not use “echocardiography” in the early days, because at the time, the only clinical use of diagnostic ultrasound was “echoencephalography.” As a result we had no usable abbreviation for “echocardiography.” We could not use
Am J Cardiol 2011;107:797–798 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved.
“ECG” because it was already taken for electrocardiography. We could not use “echo,” because it did not differentiate from echoencephalography. “Echocardiography” became feasible only when echoencephalography died. No other diagnostic ultrasound technique used “echo,” leaving it available as the abbreviation for echocardiography. Thus, the birth of “echocardiography” as the name for cardiac ultrasound. Harvey Feigenbaum, MD Indianapolis, Indiana 22 November 2010
1. Blackburn H. Paul Dudley White: “our hearts were young and gay.” Am J Cardiol 2010;106: 1193. 2. Feigenbaum H. The first long-lasting use of echocardiography: a result of a false advertisement. JACC Cardiovasc Imaging 2008;1:522– 524. 3. Feigenbaum H, Waldhausen JA, Hyde LP. Ultrasound diagnosis of pericardial effusion. JAMA 1965;191:711–714. doi:10.1016/j.amjcard.2010.11.015
Usefulness of Serial of B-Type Natriuretic Peptide and Troponin-I Levels to Predict Left Ventricular Remodeling After Primary Coronary Angioplasty We read with interest the work by Fertin et al,1 who reported the important finding that despite significant associations on univariate analyses, none of the biomarkers B-type natriuretic peptide (BNP), cardiac troponin I (TnI), and C-reactive protein at baseline (serial blood sampling from days 3 to 7) was retained as an independent predictor of left ventricular (LV) remodeling when the ejection fraction was entered into the multivariate model in Remodelage Ventriculaire 2 (REVE-2). In contrast, Hallén et al2 found that a single-point measurement of TnI 24 or 48 hours after primary angioplasty for those with ST-segment elevation myocardial infarctions in addition to the LV ejection fraction could provide prognostic information on LV remodeling from the results of FX06 in Ischemia and Reperfusion (FIRE). Actually, these results may all be correct, because the times of blood sampling at baseline in www.ajconline.org
798
The American Journal of Cardiology (www.ajconline.org)
the 2 trials were different. However, there existed critical concerns regarding some patients in the gray zone that the 2 models the investigators provided could not explain clearly enough. For example, Hallén et al’s2 study, most patients with initial ejection fractions ⬍40% who experienced recovered LV ejection fractions (45%) at 3 months had overlapping TnI levels with those with mildly impaired LV systolic function (40% to 45%) and initially high levels of TnI (49.76 ng/ml ⬍TnI ⬍91 ng/ml). Besides, Fertin et al’s1 study, patients with BNP levels ⱖ95 pg/ml but TnI levels ⬍0.05 ng/ml at 1 month had higher LV remodeling rates at 1 year than those with BNP levels ⬍95 pg/ml but TnI levels ⱖ0.05 ng/ml. In our opinion, the rapid decrease in BNP
within 1 month after primary angioplasty could be seen as a predictor for patients with initially significant LV dysfunction who recovered later. As we know, myocardial stunning after myocardial infarction can persist for months. Patients with large myocardial ischemic areas may present with acute LV systolic dysfunction and advanced Killip classification, which are associated with higher BNP and TnI levels during myocardial infarction. Given that they received rapid coronary revascularization (ischemic time ⬍3 hours), most viable myocardium would be saved and cardiac function returned to nearly normal, accompanied by the decrease in BNP. Therefore, we emphasize the prognostic value of the decrease in BNP ⬍1 month after primary angioplasty, which may provide additional
information regarding future LV remodeling rather than a single value of BNP or TnI for patients in this category. Gen-Min Lin, MD Yi-Hwei Li, PhD Hualien, Taiwan 22 November 2010
1. Fertin M, Hennache B, Hamon M, Ennezat PV, Biausque F, Elkohen M, Nugue O, Tricot O, Lamblin N, Pinet F, Bauters C. Usefulness of serial assessment of B-type natriuretic peptide, troponin I, and C-Reactive protein to predict left ventricular remodeling after acute myocardial infarction (from the REVE-2 Study). Am J Cardiol 2010;106:1410 –1406. 2. Hallén J, Jensen JK, Fagerland MW, Jaffe AS, Atar D. Cardiac troponin I for the prediction of functional recovery and left ventricular remodelling following primary percutaneous coronary intervention for ST-elevation myocardial infarction. Heart 2010;96:1892–1897. doi:10.1016/j.amjcard.2010.11.016
