SURGERY OF THE THIRD VENTRICLE EDITED BY Michael L. J. Apuzzo, M.D. Professor Department of Neurological Surgery Univer...
35 downloads
2228 Views
96MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
SURGERY OF THE THIRD VENTRICLE EDITED BY Michael L. J. Apuzzo, M.D. Professor Department of Neurological Surgery University of Southern California School of Medicine Los Angeles, California
WILLIAMS & WILKINS Baltimore • London • Los Angeles • Sydney
Foreword
The third ventricle lying at the center of the cerebral hemispheres surrounded by brain, exquisite both in terms of complexity and function, has enticed neurosurgeons for as long as the depths of brain have entered their ken. The ventricle and its neural surround also have interested neuroanatomists, neurophysiologists, endocrinologists, psychiatrists, and psychologists attempting to decipher the anatomical and functional relationships of the area. These functions are as basic as reproduction and emotion and extend to maturation and differentiation of the entire organism. The relationships of this area to cerebral function and differentiation of the organism have resisted study with an almost overwhelming tenacity, in part because of the complexities involved, in part because of the anatomical position of the neural neighborhood, and in part because the techniques used for the studies have, in themselves, altered the functional milieu being investigated. The background of almost unlimited questions, anatomical position, and altered function when the area has been approached, has also caused problems in diagnosing and treating lesions of the third ventricle, as well as being an attraction to basic and clinical neuroscientists. Studies, both basic and clinical, have, however, slowly assembled a matrix of information that allows, at this time, the formation of concepts and hypotheses that can be evaluated and supported or disproved. Not the least of the supporting structures for building the matrix have been changes in diagnostic imaging, surgical techniques, anesthesia, and control of vital functions of the body during surgery that have allowed the clinician both to treat and study problems in the area. As a clinician, I am more aware of the changes in clinical approaches to malfunction of areas adjacent to the third ventricle, particularly neoplasms adjacent to and within the ventricle and how the information gained from this has influenced the study of this central neural area. This manuscript fills a void that has been present in assembling what is known about this area; what experimental approaches are available, what they have revealed, and how modern diagnostic imaging, neuroan-atomical techniques, neurosurgical techniques, endocrine studies and psychological methods have contributed to an understanding of the functional relationships of the neural. The sum of this information is greater than its parts and cannot be appreciated without this collation. When I started in medical school, or even at the time when I finished neurosurgical training, a volume such as this would have been short, inaccurate, and mainly of value in demonstrating what damage was done with surgical approaches, what pathology was found, and perhaps suggesting that other methods of treatment of conditions in this area must evolve. The manuscripts constituting this volume, in contrast, not only summarize what has been accomplished, but also rather than pointing to an area that on the basis of previous work appears fruitful or should be studied, defines those areas and those directions. Lesions in this area are not common in the average physician's practice, are relatively uncommon even in a neurosurgeon's practice, but they are important. Their diagnosis and treatment not only adds to the well-being of the individual suffering from such conditions but also allows investigative clinicians, treating such patients, to add to the general knowledge of the nervous system and neurosurgery. This book defines the bases for such investigations and, in addition, offers the neurosurgeon an up-to-date evaluation and summary of the multiple therapeutic options for treating lesions in and around the third ventricle. William F. Collins, Jr. New Haven, Connecticut
Preface It has been more than 50 years since the publication of Walter Dandy's monograph, Benign Tumors in the Third Ventricle of the Brain, an important work detailing his concepts of pathology and surgery and his experience with the management of tumors affecting the third ventricular chamber. Since that time progress in aspects of surgical techniques, radiological imaging, scientific methodologies related to neurological anesthesia, and comprehension of neurophysiological parameters has developed at a striking pace. As no text or atlas has singularly addressed the topic of third ventricular surgery for over a half century, it seems that a work (text/atlas) focusing on issues attendant to this challenging region is in order. In the development of this project an effort has been made to provide commentary on pertinent topics from recognized authoritative neurologists, neurosurgeons, neuroradiologists, and neuroscientists with special interest in the topic and to develop a broad and practical perspective. The book is particularly addressed to the general neurological surgeon who encounters lesions involving this region infrequently. The work has been arranged in two major portions. The first provides an intellectual substrate related to issues of anatomy, embryology, physiology, pathology, and radiology that is the essential basis for substantive approaches to issues of therapy. The second section focuses upon surgical technique and strategy with a detailed account of the methodology attendant to the multiple surgical corridors applicable for access to the region. Matters associated with intraoperative anatomy, techniques, and decision-making are approached. Issues of preoperative and intraoperative strategies are discussed and developed in detail. To expand perspective, innovative and challenging methodologies are also presented. Considering the number of contributors, each with a wealth of experience in dealing with specific surgical issues, no effort has been made to provide a rigidly unified surgical attitude, and certain elements of controversy related to the individual prejudices of the authors are presented within the work. However, effort has been expended to provide the logic involved in the development of each perspective. Finally, commentary is offered on the role of and expectations for adjuvant therapy (radiotherapy and chemotherapy) for neoplastic lesions in the region. The primary objective of this text/atlas is to provide neurosurgeons, neurologists, and neuroscientists with a single reference that will offer current, substantive, and practical guidance in the comprehension and management of lesions affecting the third ventricular chamber. Michael L. J. Apuzzo Los Angeles
Acknowledgments
Special gratitude must be expressed to Carolyn Soter, who provided a major force in bringing this project to completion. Over a 3-year period she gave attentive and tireless energy to all aspects of the general organization, solicitation of manuscripts, and physical preparation of this work. Janice Jones energetically attended to the complex task of bibliography retrieval and the procurement of resource materials, invaluably contributing to the substance and credibility of the manuscripts. The Department of Photography at the Los Angeles County-University of Southern California Medical Center provided careful attention to the reproduction and generation of many of the illustrations in this volume. In particular Andy Gero, R.B.P., Chief Medical Photographer, Frank Park, Michael Gail, and Tom Meichelbock should be recognized for their patience and care in producing illustrations of the highest quality. The Williams & Wilkins staff has been a flexible, valuable, and understanding partner during the development of this complex work. In particular, Carol-Lynn Brown must be recognized for her guidance, patience, prudence, and support during the evolution of the scope of the volume. Finally, it is most important to recognize our teachers, patients, and predecessors in neurosurgery who have provided the events, stimulation, and experiences generating the intellectual substrate that made this work possible.
Contributors
Diane Abeloff, M.A., A.M.I. Medical Illustrator, Baltimore, Maryland Michael L. J. Apuzzo, M.D. Professor, Department of Neurological Surgery, University of Southern California School of Medicine, Los Angeles, California Erik-Olof Backlund, M.D. Professor and Chairman, Department of Neurological Surgery, University of Bergen, School of Medicine, Bergen, Norway Gilbert A. Block, M.D. Instructor, Department of Neurology, Cornell University Hospital, New York, New York Joseph E. Bogen, M.D. Clinical Professor of Neurological Surgery, University of Southern California School of Medicine, and Adjunct Professor, Department of Psychology, University of California at Los Angeles, Los Angeles, California Adam Borit, M.D. Professor of Pathology, Department of Pathology, The University of Texas System Cancer Center, Houston, Texas Robert E. Breeze. M.D. Clinical Instructor, Department of Neurological Surgery, University of Southern California School of Medicine, Los Angeles, California Parakrama T. Chandrasoma, M.D. Assistant Professor, Department of Pathology, University of Southern California School of Medicine, Los Angeles, California Ivan S. Ciric, M.D. Professor of Clinical Surgery (Neurosurgery), Northwestern University Medical School, Chicago, Illinois
W. Kemp Clark, M.D. Professor, Department of Neurological Surgery, The University of Texas Southwestern Medical School, Dallas, Texas William F. Collins, Jr., M.D. Harvey and Kate Cushing Professor, Division of Neurological Surgery, and Chairman, Department of Surgery, Yale University School of Medicine, New Haven, Connecticut Eric R. Cosman, Ph.D. Professor, Department of Physics, Massachusetts Institute of Technology, Cambridge, Massachusetts Antonio R. Damasio, M.D., Ph.D. Professor and Chief, Division of Behavioral Neurology, The University of Iowa Hospitals and Clinics, Iowa City, Iowa Richard L. Davis, M.D. Professor of Pathology, Department of Pathology, University of California Medical Center, San Francisco, California Michael S. B. Edwards, M.D. Associate Professor, Department of Neurosurgery, University of California Medical Center, San Francisco, California Bruce Ehni, M.D. Clinical Instructor, University of Texas Medical School, Houston, Texas George Ehni, M.D. Professor, Neurological Surgery, University of Texas Medical School, Houston, Texas Craig A. Fredericks, M.D. Clinical Instructor, Department of Neurological Surgery, University of Southern California School of Medicine, Los Angeles, California
Henry D. Garretson, M.D., Ph.D. Professor and Director, Division of Neurosurgery, University of Louisville School of Medicine, Louisville, Kentucky Steven L. Giannotta, M.D. Associate Professor, Department of Neurological Surgery, University of Southern California School of Medicine, Los Angeles, California Philip H. Gutin, M.D. Associate Professor, Department of Neurosurgery, University of California Medical Center, San Francisco, California M. Peter Heilbrun, M.D. Professor and Chairman, Division of Neurosurgery, University of Utah Medical Center, Salt Lake City, Utah Harold J. Hoffman, M.D., F.R.C.S.(C) Professor, Division of Neurological Surgery, University of Toronto School of Medicine, Toronto, Ontario, Canada Kazuhiro Hongo, M.D. Clinical and Research Fellow, Department of Neurosurgery, Shinshu University, School of Medicine, Matsumoto, Japan Patrick J. Kelly, M.D. Associate Professor, Department of Neurosurgery, Mayo School of Medicine, Rochester, Minnesota E. Leon Kier, M.D. Professor and Chief, Division of Neuroradiology, Yale University School of Medicine, New Haven, Connecticut Alexander N. Konovalov, M.D. Professor and Director, N. N. Burdenko Neurosurgical Institute, Moscow, U.S.S.R. Eddie Kwan, M.D. Assistant Professor, Department of Radiology, Tufts University School of Medicine, Boston, Massachusetts Claude Lapras, M.D. Professeur a la Faculte, Neurochirurgien des Hopitaux, Lyon, France Michael H. Lavyne, M.D. Associate Professor of Surgery, Division of Neurological Surgery, Cornell University Medical Center, New York, New York Victor A. Levin, M.D. Professor, Department of Neurosurgery, University of California Medical Center, San Francisco, California F. Miles Little, M.D. Assistant Professor, Department of Neurosurgery, University of Southern California School of Medicine, Los Angeles, California
Masao Matsutani, M.D. Assistant Professor, Department of Neurosurgery, Tokyo University, Tokyo, Japan J. Gordon McComb, M.D. Professor, Department of Neurological Surgery, University of Southern California School of Medicine, Los Angeles, California Michael T. Modic, M.D. Professor, Department of Radiology, University Hospital, Case Western Reserve School of Medicine, Cleveland, Ohio Robert B. Page, M.D. Professor, Division of Neurosurgery, Department of Anatomy, Hershey Medical Center, Hershey, Pennsylvania J. D. Patet, M.D. Neurochirurgien des Hopitaux, Lyon, France Russel H. Patterson, Jr., M.D. Professor and Chairman, Department of Neurological Surgery, Cornell University Medical Center, New York, New York Jerome B. Posner, M.D. Professor and Chairman, Department of Neurology, Cornell University Hospital, New York, New York Albert L. Rhoton, Jr., M.D. R. D. Keene Family Professor and Chairman, Department of Neurosurgery, University of Florida College of Medicine, Gainesville, Florida Peter Roth Scientific Artist, Neurosurgical Department, University Hospital, Zurich, Switzerland Madjid Samii, Prof. Dr. med. Neurochirugische Klinik, Krankaenhaus Nor-stadt, Hannover, West Germany Robert A. Sanford, M.D. Associate Professor, Department of Neurosurgery, University of Tennessee, Memphis, Tennessee Keiji Sano, M.D., D.M.Sc., F.A.C.S.(hon.) Professor and Director, Department of Neurosurgery, Teikyo University Hospital, Tokyo, Japan Henry H. Schmidek, M.D. Professor of Neurosurgery, Department of Neu-robiology, Harvard Medical School, Boston, Massachusetts William Shucart, M.D. Professor and Chairman, Department of Neurosurgery, New England Medical Center, Boston, Massachusetts Robert R. Smith, M.D. Professor and Chairman, Department of Neurological Surgery, University of Mississippi Medical Center, Jackson, Mississippi
Steven P. Smith, M.D. Fellow, Neuroradiology, New England Medical Center, Boston, Massachusetts Bennett M. Stein, M.D. Byron Stookey Professor of Neurosurgery, Neurological Institute, New York, New York Kenichiro Sugita, M.D. Professor and Chairman, Department of Neurosurgery, Shinshu University, Matsumoto, Japan Jiro Suzuki, M.D. Professor and Chairman, Department of Neurosurgery, Tohoku University, Sendai, Japan Kintomo Takakura, M.D. Professor and Chairman, Department of Neurosurgery, Tokyo University, Tokyo, Japan Peter J. Teddy, D.Phil, F.R.C.S. Consultant Neurosurgeon, Department of Neurological Surgery, Oxforshire Area Health Authority, The Radcliffe Infirmary, Oxford, England George T. Tindall, M.D. Professor of Surgery, Chief, Division of Neurosurgery, The Emory University School of Medicine, Atlanta, Georgia Suzie C. Tindall, M.D. Assistant Professor of Surgery, Division of Neurosurgery, The Emory University School of Medicine, Atlanta, Georgia Gary W. Van Hoesen, Ph.D. Professor, Anatomy and Neurology, Department of Anatomy, The University of Iowa Hospitals and Clinics, Iowa City, Iowa Roger I. von Hanwehr, M.D. Clinical Instructor, Department of Neurological Surgery, University of Southern California School of Medicine, Los Angeles, California
Steven L. Wald, M.D. Assistant Professor of Neurosurgery, Division of Neurosurgery, University of Vermont College of Medicine, Burlington, Vermont William M. Wara, M.D. Professor and Vice Chairman, Department of Radiation Oncology, University of California, San Francisco, San Francisco, California Martin H. Weiss, M.D. Professor and Chairman, Department of Neurological Surgery, University of Southern California School of Medicine, Los Angeles, California Trent H. Wells, Jr. President, Trentwells, Inc., South Gate, California Robert H. Wilkins, M.D. Professor and Chief, Division of Neurological Surgery, Duke University Medical Center, Durham, North Carolina Samuel M. Wolpert, M.B., B.Ch. Professor of Radiology and Neurology, Tufts University School of Medicine, and Chief, Neuroradiology, New England Medical Center, Boston, Massachusetts M. Gazi Yasargil, Prof. Dr. med. Director der Neurochirurg, Universitatsklinik Zurich, Zurich, Switzerland Chi-Shing Zee, M.D. Assistant Professor, Department of Radiology, University of Southern California School of Medicine, Los Angeles, California Vladimir Zelman, M.D., Ph.D. Associate Professor, Department of Anesthesiol-ogy and Department of Neurosurgery, University of Southern California School of Medicine, Los Angeles, California
Contents
обложка
Foreword .................................................................................................................. Preface ................................................................................................................... Acknowledgments ................................................................................................. Contributors .............................................................................................................
vii ix xi xiii
I. HISTORICAL PERSPECTIVE _____________________________________ 1 . History of Surgery of the Third Ventricular Region. Robert H. Wilkins, M.D .......................................................................................................
3
II. FUNDAMENTAL CONSIDERATIONS ______________________________ 2. Comparative Anatomy of the Third Ventricular Region. E. Leon Kier, M.D....................................................................................................
37
3. Microsurgical Anatomy of the Third Ventricular Region. Albert L. Rhoton, Jr., M.D .......................................................................................
92
4. Surgery of the Third Ventricle: Regional Embryology. Ivan S. Ciric, M.D .................................................................................................... 167 5. Physiological Consequences of Complete or Partial Commissural Section. Joseph E. Bogen, M.D
175
6. Pathological Correlates of Amnesia and the Anatomical Basis of Memory. Antonio R. Damasio, M.D., Ph.D., and Gary W. Van Hoesen, Ph.D
195
Commentary A. Memory in Man: A Neurosurgeon's Perspective. Henry D. Garretson, M.D., Ph.D
209
7. Anatomy and Physiology of Consciousness: Syndromes of Altered Consciousness Related to Third Ventricular Surgery. Gilbert A. Block, M.D., and Jerome B. Posner, M.D
213
8. Deep Veins. Robert R. Smith, M.D., Robert A. Sanford, M.D., and Henry H. Schmidek, M.D
224
9. Pathological Lesions of the Third Ventricle and Adjacent Structures. Richard L. Davis, M.D
235
10. Tumor Markers in Third Ventricular Neoplasms. Henry H. Schmidek, M.D., Adam Borit, M.D., and Steven L. Wald, M.D. . . 253 1 1 . Radiology of Third Ventricular Lesions. Eddie Kwan, M.D., Samuel M. Wolpert, M.B., B.Ch., Steven P. Smith, M.D., and Michael T. Modic, M.D
262
III. SURGICAL APPROACHES, TECHNIQUES, AND STRATEGIES ANTERIOR APPROACHES 12. Anterior Transcallosal and Transcortical Approaches. William Shucart, M.D
303
13. Considerations in Transforaminal Entry. George Ehni, M.D., and Bruce Ehni, M.D
326
14. Transcallosal Interforniceal Approach. Michael L. J. Apuzzo, M.D., and Steven L. Giannotta, M.D
354
15. Subchoroidal Trans-Velum Interpositum Approach. Michael H. Lavyne, M.D., and Russel H. Patterson, Jr., M.D. . . . 381 16. Subfrontal Transsphenoidal and Trans-Lamina Terminalis Approaches. Russel H. Patterson, Jr., M.D
398
17. Bifrontal Anterior Interhemispheric Approach. Jiro Suzuki, M.D
413
18. Pterional Approach. George T. Tindall, M.D., and Suzie C. Tindall, M.D
440
19. Combined Approaches. M. Gazi Yasargil, Prof. Dr. med., Peter J. Teddy, D. Phil., F.R.C.S., and Peter Roth
462
20. Transnasal Transsphenoidal Approach. Martin H. Weiss, M.D
476
21. Anterior and Mid-Third Ventricular Lesions: A Surgical Overview. Michael L. J. Apuzzo, M.D., Chi-Shing Zee, M.D., and Robert E. Breeze, M.D
495
Commentary B. Technique and Strategies of Direct Surgical Management of Craniopharyngioma. Alexander N. Konovalov, M.D
542
Commentary C. Diencephalic Ventricular Surgery. Robert B. Page, M.D
Structures
at
Risk
in
Third 553
POSTERIOR APPROACHES 22. Posterior Transcortical Approach. Kenichiro Sugita, M.D., and Kazuhiro Hongo, M.D
557
23. Infratentorial Supracerebellar Approach. Bennett M. Stein, M.D
570
24. Occipital Transtentorial Approach. W. Kemp Clark, M.D. 591 25. Posterior Intrahemispheric Retrocallosal and Transcallosal Approaches. J. Gordon McComb, M.D., and Michael L. J. Apuzzo, M.D
611
Commentary D. Operative Management of Malformations of the Vein of Galen. J. Gordon McComb, M.D., and Michael L. J. Apuzzo, M.D
641
26. Controversies, Techniques, and Strategies for Pineal Tumor Surgery. Claude Lapras, M.D., and J. D. Patet, M.D. . . . 649 27. Pineal Region and Posterior Third Ventricular Tumors: A Surgical Overview. Keiji Sano, M.D., D.M.Sc., F.A.C.S.(hon.) . . . 663 SPECIALIZED ISSUES 28. Technical Aspects of Excision of Giant Basal Tumors with Third Ventricular Involvement. Madjid Samii, Prof. Dr. med. . . 684 29. Cerebrospinal Fluid Diversion. J. Gordon McComb, M.D., and F. Miles Little, M.D
699
30. Considerations and Techniques in the Pediatric Age Group. Harold J. Hoffman, M.D., F.R.C.S.(C)
727
31. Applications of Computerized Tomographic Guidance Stereotaxis. Michael L. J. Apuzzo, M.D., Parakrama T. Chandrasoma, M.D., Vladimir Zelman, M.D., Roger I. von Hanwehr, M.D., and Craig A. Fredericks, M.D
751
Commentary E. Contemporary European Contributions to Neurosurgical Stereotaxy. Roger I. von Hanwehr, M.D
793
Commentary F. Role of Stereotaxis in the Management of Midline Cerebral Lesions. Erik-Olof Backlund, M.D., Ph.D
802
Commentary G. Magnetic Resonance Stereotaxy. Eric R. Cosman, Ph.D., M. Peter Heibrun, M.D., and Trent H. Wells, Jr
806
32. Computer-assisted Stereotaxic Laser Microsurgery. Patrick J. Kelly, M.D
811
IV. ADJUVANT THERAPIES 33. Radiotherapy of Pineal and Suprasellar Tumors. William M. Wara, M.D., and Philip H. Gutin, M.D
831
34. Chemotherapy of Tumors of the Third Ventricular Region. Michael S. B. Edwards, M.D., and Victor A. Levin, M.D
838
35. Therapeutic Modality Selection in Management of Germ Cell Tumors. Kintomo Takakura, M.D., and Masao Matsutani, M.D.
843
1 History of Surgery of the Third Ventricular Region
Robert H.Wilkins,M.D.
This chapter deals with the history of surgery of the third ventricular region. I have arbitrarily limited my considerations to the surgical treatment of abnormalities of or in the third ventricle itself, of lesions of the roof of the third ventricle, and of tumors in the region of the pineal gland. I have not included the treatment of sellar or suprasellar lesions (even those that invaginate the floor of the third ventricle), of arteriovenous malformations draining into the vein of Galen, or of tumors of the thalamus, hypothalamus, or brain stem. I have tried to avoid undue emphasis on the establishment of priority because some uncertainty always exists about who was the first to describe a pathological entity or an approach to treatment. As Palmer has noted, "... an author can never tell how many cases . . . have previously been reported: no amount of library research will permit more than a rough guess" (106). The reason for this is that even after a thorough search in a medical library, the author will have missed unindexed cases reported in sources such as textbooks and monographs, governmental reports, doctoral theses, and the letters to the editor section of many journals. Furthermore, "... the author also must understand that, as important as the entity seems to him, others who have dealt with similar cases [may not] . . . have bothered to write them up" (106). With that caveat in mind, I focus attention in this chapter on the contributions of Walter Dandy, who had a unique interest in third ven-
tricular lesions and an unprecedented experience in dealing with them (36, 55, 62). I begin each section of this chapter with an account of his work and then consider the other historical developments, most of which came after Dandy's pioneering efforts. Cerebrospinal Fluid and the Ventricular System; Hydrocephalus During his senior year and then after his graduation from medical school in 1910, Walter Dandy worked in the Hunterian Laboratory of Experimental Medicine of Johns Hopkins University, where he studied the blood supply and nerve supply of the pituitary gland (24, 46). Subsequently, in association with Kenneth Black-fan, he performed a series of animal experiments and human investigations that clarified where cerebrospinal fluid (CSF) is produced, how it circulates, where it is absorbed, and what types of hydrocephalus result from the various abnormalities that alter its production, circulation, and absorption (43-45). Then during his last year of residency (1918) and first year of practice (1919), Dandy introduced ventriculography and pneu-moencephalography (26, 28), thus permitting visualization of the third ventricle for diagnostic purposes. Based on these early investigations, Dandy maintained for the rest of his career an interest in hydrocephalus and its treatment and in the diagnosis and treatment of tumors and other
lesions in and around the third ventricle (36, 55, 62, 158). He introduced the operations of choroid plexectomy (27, 39), cannulation of the aqueduct of Sylvius (29), and third ventriculostomy (33, 41) to treat hydrocephalus and ventriculoscopy (32) for inspection of the ventricular system and as an aid to choroid plexectomy. In his 1918 paper on extirpation of the choroid plexus of the lateral ventricles for treatment of communicating hydrocephalus, Dandy wrote, "A remarkable exposure is obtained during the operation in the ventricle. One can see the third and opposite lateral ventricle and the septum lucidum which is frequently perforated in many places owing to pressure atrophy" (27). In a report in 1922 on ventriculoscopy, he stated, On two occasions, it has seemed advisable to inspect a lateral ventricle. This was done in one instance through a small cystoscope and in a second an attempt was made with the help of a small operating ventriculoscope to remove and fulgurate the
choroid plexus. ... It was possible to see practically the entire extent of the lateral ventricle, the foramen of Monro, the septum lucidum with numerous perforations in it, and the entire extent of the choroid plexus. . . (32). The procedure of excision of the choroid plexus had a high mortality rate and was abandoned early (59, 113, 131, 132). However, endoscopic cauterization of the choroid plexus, which apparently had also been tried by Lespinasse in 1910 (95), was developed and used by Dandy (39) (Fig. 1.1), Putnam (115), Scarff (129, 132), Feld (57, 58), and others for 30 to 40 years, when the combination of the equivocal results of this procedure and the introduction of successful valved shunt systems led to its abandonment as well (113). Likewise, direct cannulation of the aqueduct of Sylvius was tried by Dandy (29) (Fig. 1.2), and subsequently by Leksell (94), Norlen (104), Elvidge (53), and a few other neurosurgeons (145, 148), but the procedure did not prove to be of value.
Figure 1.1. Composite drawing. Center. Sites at which the choroid plexus is cauterized and removed. A and a. The glomus of the choroid plexus on each side. В and b. The choroid plexus in the posterior cranial fossa. С, с, and c'. The choroid plexus from the body of each lateral ventricle. (From Dandy WE: The operative treatment of communicating hydrocephalus. Ann Surg 108:194-202, 1938.)
Figure 1.2. Sagittal view of the brain showing a shunt tube in position in the aqueduct of Sylvius. (From Dandy WE: The diagnosis and treatment of hydro-cephalus resulting from strictures of the aqueduct of Sylvius. Surg Gynecol Obstet 31:340-358, 1920.) In contrast, neurosurgeons have maintained some interest in third ventriculostomy over the years because of its promise of relieving hydro-cephalus without the insertion of a shunt system (59, 113, 131). In 1922, Dandy wrote,
space where the absorption is slightly, if at all, greater than in the ventricles. Moreover, the opening is through cerebral tissue which proliferates and closes it, unless a good deal of the brain has been destroyed. We have employed this method 6 times. No claim is made for its success. Time alone will decide . . . (33).
Strictures of the aqueduct of Sylvius recur after any attempt to restore the lumen. For this reason, if treatment is to be successful, the fluid must be sidetracked into its normal channels. With this in mind, a procedure which apparently is anatomically correct has been devised, to supersede any direct attack on the aqueduct. This consists in removing the floor of the third ventricle. A small opening is made in the skull and dura in the frontal region, the frontal lobe is elevated until the bulging third ventricle is well exposed. . . . This opening in the floor of the third ventricle affords an exit from the dilated ventricles, so that the fluid can now pass directly into the cisterna chiasmatis and interpeduncu-laris—the normal distributing centers for cerebro-spinal fluid.. . . The ventricular wall is a very thin membrane and offers a minimum of glia tissue to repair the defect. This procedure is by no means analogous to making an opening in the roof of the third ventricle. The latter can have no beneficial result because the fluid escapes into the subdural
When the pathogenesis of obstructive hydrocephalus was fully appreciated, operations to reroute the ventricular fluid were devised. Perhaps, the forerunner of these was the "Balkenstich" operation of Anton and von Bramann [ ], which consisted of passing a brain cannula through a midline frontal trephine opening, along the falx through the corpus callosum and into the ... ventricle. It was presumed that the ventricular fluid would pass through the fenestrated corpus callosum to the cal-losal subarachnoid pathways, but the procedure lost its popularity when it was found that the passageway did not remain open (59). As mentioned, Dandy thought that ventriculostomy through the floor of the third ventricle held advantages over ventriculostomy through its roof. His initial subfrontal approach to third ventriculostomy involved the sectioning of one optic
nerve to gain exposure (33), but in 1932 he reported on a lateral (transtemporal) third ven-triculostomy that did not require division of the optic nerve (36). The steps in this procedure, as listed by Dandy, were: 1 . A plaster cast was molded to the infant's head (Fig. 1.3). A defect was made in the tem poral region of the side to be operated upon. 2. Small curved skin and muscle incisions were made in the temporal region (Fig. 1.3). 3. A small area of bone was removed above the base of the skull and a dural flap was reflected toward the base. 4. The head was lowered about 50° and the temporal horn of the lateral ventricle was tapped, with the evacuation of 60 to 80 ml of CSF. A short flanged ventricular needle was left in place during the operation. 5. The temporal lobe was retracted to expose the lateral wall of the interpeduncular cistern and this wall was opened behind the oculomotor nerve or between the carotid artery and the ocu lomotor nerve (Fig. 1.4). 6. The bulging floor of the third ventricle was then identified and opened, with the opening preferably being made just posterior to the hypophyseal stalk (Fig. 1.5). 7. The ventricles were then filled with the pre viously aspirated CSF or with an isotonic solution and the wound was closed in layers (Fig. 1.5). In 1945, Dandy reported 92 patients that he had treated in this fashion, with a 12% mortality rate and with arrest of the hydrocephalus in 50% of the patients for periods averaging more than 7 years (Table 1.1) (131). Reoperation had been necessary in 7 of the 92 patients. After Dandy's 1922 report of subfrontal third ventriculostomy, several surgeons pursued the idea of venting the third ventricle into an adjacent subarachnoid cistern. The following year, Mixter reported treating an infant with obstructive hydrocephalus by approaching the floor of the third ventricle with a urethroscope inserted through the lateral ventricle and foramen of Monro and then using a flexible sound to puncture an opening into the interpeduncular cistern (102). Scarff subsequently performed a similar procedure, which he reported in 1936 (129). In 1936, Stookey and Scarff described a subfrontal approach with puncturing of the lamina terminalis and then the floor of the third ventricle (130, 141) (Fig. 1.6). In 1963, Scarff gave the results in 527 hydrocephalic patients treated by 12 surgical groups by puncture of the lamina terminalis alone or combined with puncture of the floor of the third ventricle, and/or by
puncture of the lateral wall of the hypothalamus (131). The patients were followed for an average of 5 years. The operative mortality was approximately 15% and the initial success rate was about 70%. Various surgeons have pursued the operation of anterior, inferior, lateral, or superoposterior third ventriculostomy by direct or indirect techniques, as detailed in the excellent publications of Scarff (131) and of Pudenz (113). The procedure has been aided by the development of stereotaxic techniques and positive contrast ventric-ulography (Fig. 1.7) (74, 80, 128). In 1980, Hoffman et al. summarized the results of third ventriculostomy by the open technique that had been performed by 13 surgical groups (80). There were 569 patients; the operative mortality was 10.3% and the success rate was 53.6% (Table 1.2). These authors also showed the results of third ventriculostomy by the percutaneous technique as performed by 12 groups of surgeons. The success rate was approximately the same (53%), but the operative mortality was lower (3.5%) (Table 1.3). During the period when the procedures of cauterization of the choroid plexus and of third ventriculostomy were being tried, the techniques of ventriculoscopy and ventriculography were also being developed. Ventriculoscopy was of some help in these operations initially, but despite clever technical advances such as flexibility, fiberoptic lighting, and video attachments, ventriculoscopy never has become an important diagnostic or therapeutic tool (32, 36, 39, 56-58, 64, 65, 75, 80, 95, 102, 115, 129, 132, 144, 155, 156). In comparison, after their introduction in 1918 (26) and 1919 (28), pneumoventriculography and pneumoencephalography rapidly became the keys to the diagnosis of third ventricular lesions and subsequently to the performance of stereo-taxic operations adjacent to the third ventricle. This has changed only in recent years because of the advent of computerized tomographic (CT) scanning and magnetic resonance imaging. For the contrast media, Dandy used ordinary room air. Bingel ... suggested the use of carbon dioxide in 1922, because it was absorbed rapidly. In the same year, Jungling . . . recommended the use of oxygen. . . . Although gaseous media demonstrated the lateral ventricles clearly, the third and fourth were often not well visualized even with special positioning techniques. To obtain better contrast various other media were tried. After the introduction of lipiodol, . . . Jacobaeus and Nord ... in 1924 experimented with it for ventriculography. The following year Schuster . . . used it. About the same
Figure 1.3. Dandy's operative approach for third ventriculostomy. (From Dandy WE: The brain. In Lewis D (ed): Practice of Surgery. Hagerstown, MD, WF Prior, 1932, pp 1-682.)
Figure 1.4. The bulging third ventricle exposed through the lateral wall of the interpeduncular cistern between the carotid artery and the oculomotor nerve. (From Dandy WE: The brain. In Lewis D (ed): Practice of Surgery. Hagerstown, MD, WF Prior, 1932, pp 1-682.)
a From Scarff JE: Treatment of hydrocephalus: An historical and critical review of methods and results. J Neurol Neurosurg Psychiatry 26:1-26, 1963.
time Balado, Morea and Donovan ... in the Argentine tested the oil in the ventricles of experimental animals and finding no untoward reactions began using it routinely, particularly for demonstration of the third and fourth ventricles (157).
Figure 1.5. The upper composite drawing indicates all of the steps in Dandy's third ventriculostomy. The lower drawing shows the method of filling the ventricles before closure. (From Dandy WE: Diagnosis and treatment of strictures of the aqueduct of Sylvius [causing hydrocephalus]. Arch Sarg 51:1-14, 1945.)
The use of lipiodol waned, and trials of other positive contrast agents were performed (124, 157). These included thorium dioxide, diiodoty-rosine-gelatin, and abrodil. Then, when ethyl iodophenylundecylate was introduced as a mye-lographic agent, it found use in ventriculography as well. More recent attempts to use water-soluble contrast media for ventriculography did not meet with much success until the introduction of metrizamide, and this has been combined with CT scanning to give excellent visualization of the cisterns and ventricles (49).
Table 1.2. Summary of Experiences with Third Ventriculostomy by Open Operationa Operative Success Mortality Rate (%) (%) Wertheimer & Mansuy 1938 3 33 66 Pennybacker 1940 5 20 80 White & Michelsen 1942 11 18 46 Dandy 1945 92 17 39 Tolosa 1948 26 25 50 Guillaume & Mazars 1950 230 3.4 54 Krayenbuhl et al. 1950 17 17 60 Fasiani 1951 72 20 70 Scarff 1951 34 15 56 Morello & Migliavacca 1959 28 3 75 Volkel & Voris 1966 12 0 33 Patterson & Bergland 1968 29 6 34 Brocklehurst 1974 10 0 70 Summary 569 10.3 53.6 a From Hoffman HJ, Harwood-Nash D, Gilday DL: Percutaneous third ventriculostomy in the management of noncommunicating hydrocephalus. Neurosargery 7:313-321, 1980. Author
Year
No. Cases
Table 1.3. Summary of Experiences with Third Ventriculostomy by Percutaneous Techniquea Author McNickle Forjaz et al. Perlman Raimondi Guiot Plerre-Kahn et al. Poblete & Zamboni Hoffman Sayers & Kosnick des Plantes & Crezee Vries Hoffman et al. Summary
Year 1947 1968 1968 1972 1973 1975 1975 1976 1976 1978 1978 1980
No. Cases 7 15 1 3 14 44 10 11 46 61 5 22 228
Operative Mortality (%) 0 20 0 0 0 7 0 0 4 0 0 0 3.5
Success Rate (%) 71 67 100 0 64 64 70 27b 89 15 20 45 53
a From Hoffman HJ, Harwood-Nash D, Gilday DL: Percutaneous third ventriculostomy in the management of noncommunicating hydrocephalus. Neurosurgery 7:313-321, 1980. b Originally reported as 64%; subsequent obstruction largely due to shunt placement has reduced this success rate to 27%.
Figure 1.6. Third ventriculostomy by the subfrontal route, with puncture of both the lamina termlnalis and the floor of the third ventricle. (From Scarff JE: Treatment of obstructive hydrocephalus by puncture of the lamina terminalis and floor of the third ventricle. J Neurosurg 8:204-213, 1951.)
contents of the fifth and sixth ventricles, Dandy then presented two patients with symptomatic cysts of these structures (Fig. 1.8) that he had treated by a transcallosal exposure and fenestra-tion of the cyst into the lateral ventricles (Fig. 1.9). He ended his paper as follows: 1 . Two cases of cysts of the cavum septi pellucidi and cavum vergae are reported. In each case the two cavities were continuous. 2. The cysts acted as tumors and caused compression of the motor tracts on both sides. Men tal symptoms were decided in both cases. One pa tient had peculiar epileptic attacks.... Suggestive evidence of intermittent intracranial pressure ex isted in both instances. . . . 4. The diagnosis is easily made by ventriculography. . . . 5. An operation is offered for cysts of this char acter (35).
Figure 1.7. Diagrammatic sketch of a percutaneous method of third ventriculostomy. (From Hoffman HJ, HarwoodNash D, Gilday DL: Percutaneous third ventriculostomy in the management of noncommunicat-ing hydrocephalus. Neurosurgery 7:313-321, 1980.) Lesions of the Roof of the Third Ventricle In 1931, Walter Dandy published an unusual paper entitled, "Congenital Cerebral Cysts of the Cavum Septi Pellucidi (Fifth Ventricle) and Ca-vum Vergae (Sixth Ventricle): Diagnosis and Treatment" (35). He began the paper as follows: "In the midline of the brain and within the confines of the corpus callosum either or both of the cavum septi pellucidi and cavum vergae are not infrequently found. Neither cavity has excited much interest either anatomically or clinically. The two cases here reported are, I believe, the first instances in which a diagnosis of these spaces, dilated in abnormal degree, has been made during life, the first in which the lesion has been found at or treated by operation and the first in which clinical symptoms are shown to be related to the lesions (35). After discussing the anatomy, occurrence, and
In 1967, Kempe and Busch reported a symptomatic cyst of the cavum septum interpositum that had caused head enlargement during the first 4 months of life (90). This cyst was exposed through the corpus callosum and psalterium, and a communication was established with the cis-terna venae magnae Galeni and ambiens. The patient then developed normally over the follow-up period of 4 years. Symptomatic cysts of this sort have been very rare (152), as have neuroepithelial (colloid) cysts of the septum pellucidum (18) and diencephalic cysts that extend up from the third ventricle through the corpus callosum (11). Somewhat less rare are meningiomas arising from the velum interpositum and confined predominantly to the third ventricle and pineal region (without dural attachment) (111, 123). Among 16 cases of this nature reviewed by Rozario et al. in 1979, 8 were located in the pineal/posterior third ventricular region, whereas only 3 were located in the anterior third ventricle (123). Various surgical approaches were used for the removal of these tumors. Most were operated upon via a transcorti-cal or occipital approach before the advent of the operating microscope. In 50%, total removal was achieved. The operative mortality was 37% (before 1970 it was 60% and after 1970 it was 0%).
Figure 1.8. Position of a congenital cyst of the cavum septi pellucidi and cavum vergae in relationship to the ventricular system. (From Dandy WE: Congenital cerebral cysts of the cavum septi pellucidi [fifth ventricle] and cavum vergae [sixth ventricle): Diagnosis and treatment. Arch Neurol Psychiatry 25:44-66, 1931.)
Figure 1.9. Dandy's method of producing an opening between each lateral ventricle and the cyst of the cavum septi pellucidi. (From Dandy WE: Congenital cerebral cysts of the cavum septi pellucidi [fifth ventricle] and cavum vergae [sixth ventricle]: Diagnosis and treatment. Arch Neurol Psychiatry 25:44-66, 1931.) Third Ventricular Tumors In 1922, Walter Dandy published the following brief account in the Johns Hopkins Hospital Bulletin: In one patient, a woman 24 years of age, the only symptoms were those referable to intracranial pressure. By cerebral pneumography, it was determined that each lateral ventricle was greatly enlarged, but there was no communication between them. Hence we concluded that there must be a tumor in the third ventricle and occluding each foramen of Monro; that the tumor must be small because neither ventricle was dislocated away from it. A large bone flap was turned down as in a pineal approach, and the corpus callosum split posteriorly for about 5 to 6 cm. The right lateral ventricle was then opened through the mesial wall at the septum lucidum. No tumor could be seen; the right foramen of Monro seemed normal on inspection from this lateral ventricle, but a probe would not pass through it, an obstruction being encountered in the third ventricle. The foramen of Monro was then widened and a small encapsulated, spherical tumor, about 1 cm. in diameter, easily shelled out in toto. The tumor was of ependymal origin. The patient recovered. Without cerebral
pneumography localization of the tumor and consequently its removal would have been impossible (31). Then in a monograph published 11 years later, Walter Dandy summarized his experience with the diagnosis and treatment of benign third ventricular tumors (37). He purposely omitted from this work a consideration of cerebral gliomas, pineal tumors, "hypophyseal duct tumors," and large pituitary tumors with upward extension into the third ventricle. The tumors in his 21 patients included 5 colloid cysts, 5 "ependymal gliomata," 1 choroid plexus papilloma, 1 epider-moid tumor, and 1 "dermoid." The other 8 tumors were more difficult to classify. In addition to his own 21 cases, Dandy also considered in his discussion "all cases of similar character that could be assembled from the literature" (37). These latter tumors were from autopsy material, "... there being no instance of a tumor having been disclosed at operation" (37). This last statement by Dandy is of interest be-
cause Harvey Cushing had removed a third ventricular meningioma from one patient in 1927 and from another in 1931 (23, 123) and had resected an astroblastoma of the third ventricle at the first meeting of the Harvey Cushing Society in 1932 (66). The operation for astroblastoma was memorable. At 10 a.m. on May 6, 1932, Dr. Cushing welcomed the members of the newly formed organization that had adopted his name. "He then operated in the large amphitheater before the entire group, exposing a thirdventricle tumor through a transcortical incision and removing a large part of it. [The patient, who was shown at the evening clinic the day of her operation turned out to have a most favorable tumor. . . . She was married a short time later and is still living and well, with a family of two children (Nov. 1945).]" (66). In Dandy's defense, Cushing did not publish the first two cases until 1938 (23), and Dandy was not a member of the Harvey Cushing Society. Dandy began his monograph as follows: A most interesting and important group of tumors lying within the third ventricle has, until the past few years, remained without the fields of diagnosis and treatment. Of benign type, slow growth and
Figure 1.10. Diagram to show how air injected into a dilated lateral ventricle fails to fill the third ventricle, which is occupied by tumor, and the opposite lateral ventricle (unless there are f enestrations in the septum pellucidum). (From Dandy WE: Benign Tumors in the Third Ventricle of the Brain: Diagnosis and Treatment. Springfield, IL, Charles С Thomas, 1933.)
small size, their removal by surgical treatment is relatively simple and affords a permanent cure if only a correct diagnosis and localization can be made. They produce fulminating signs and symptoms of intracranial pressure owing to their strategic position along the channels through which cerebro-spinal fluid must leave the brain, and yet localization of the tumors by neurological signs and symptoms is almost impossible. The advent of ventricu-lography has, however, by its mechanical evidence made the diagnosis and localization one of the greatest accuracy and certainty, and has, therefore, provided the precision that is an absolute prerequisite for their surgical attack (37). Dandy then went on to present his 21 cases individually and then to summarize the various aspects of these cases plus similar ones from the literature. He pointed out the characteristic ven-triculographic findings of third ventricular tumors (Fig. 1.10) and described various operative techniques for their removal (Figs. 1.11 to 1.15). Concerning these operative approaches, Dandy stated: Four operative attacks are used to expose and remove tumors of the third ventricle: (1) Posterior (pineal) approach . . . ; (2) Anterior (a) Frontal (hy-pophyseal) approach with removal of a round or oval area of the frontal lobe (usually right); (b) Frontal (hypophyseal) with transection of the frontal lobe (usually right); (c) Mid-sagittal approach through corpus callosum (used only for cysts of the septum pellucidum). .. . Since tumors within the third ventricle are quite
Figure 1.11. Sketches concerning a posterior trans-callosal operation for the removal of a colloid cyst of the third ventricle. (From Dandy WE: Benign Tumors in the Third Ventricle of the Brain: Diagnosis and Treatment. Springfield, IL, Charles С Thomas, 1933.)
Figure 1.12. Another posterior transcallosal operation for the removal of a colloid cyst, showing the method of incising the roof of the ventricle to expose the tumor. (From Dandy WE: Benign Tumors in the Third Ventricle of the Brain: Diagnosis and Treatment. Springfield, IL, Charles С Thomas, 1933.)
Figure 1.14. Transfrontal operation for the removal of a colloid cyst of the third ventricle. These drawings show the intraventricular exposure and removal of the cyst. (From Dandy WE: Benign Tumors in the Third Ventricle of the Brain: Diagnosis and Treatment. Springfield, IL, Charles С Thomas, 1933.)
Figure 1.13. Transfrontal operation for the removal of a colloid cyst of the third ventricle. These drawings show the approach. (From Dandy WE: Benign Tumors in the Third Ventricle of the Brain: Diagnosis and Treatment. Springfield, IL, Charles С Thomas, 1933.)
Figure 1.15. A posterior transcallosal operation for the removal of a posterior third ventricular tumor. (From Dandy WE: Benign Tumors in the Third Ventricle of the Brain: Diagnosis and Treatment. Springfield, IL, Charles С Thomas, 1933.)
small when compared with other intracranial tumors, it might seem that any approach that exposes the growth would be satisfactory. On the contrary, the determination of the approach is a matter of the utmost importance. The choice of methods is almost entirely dependent upon the ventriculographic information concerning the third ventricle. As previously noted, it is, therefore, necessary to fill with air and roentgenographically disclose, as far as possible, every part of the third ventricle that is not actually filled by tumor. The choice between either anterior and the posterior (pineal) approach is decided entirely upon the position of the tumor, i.e., whether it lies in the anterior or posterior part of the third ventricle. If the tumor occupies only the posterior part of the third ventricle the pineal approach is absolutely necessary. If the tumor occupies only the anterior part of the third ventricle the frontal (hypophyseal) approach is much superior, although the pineal approach is very satisfactory. The greatest diagnostic difficulty lies in separating the small tumors which occupy only the anterior part of the third ventricle from the larger growths that occupy the entire ventricle. Obviously the only instances in which this differential diagnosis can be made by ventriculog-raphy are afforded by tumors causing only a partial block and allowing a small amount of air to fill the small posterior remains of the third ventricle. . .. When the tumor is known to block the foramina of Monro and no information is obtainable, either by ventriculographic or neurological findings, of the posterior extent of the tumor, the choice of approach is difficult. Formerly I used the pineal approach exclusively, but more recently the frontal approach in increasing numbers, because of the possibility of encountering tumors that are not primary in the third ventricle but have grown into the ventricle from without. Then too the colloid cysts which usually give signs of complete obstruction of the foramina of Monro are much easier to remove by the anterior approach. Other things being equal, I prefer the frontal approach with oval resection or transverse section of the frontal lobe to the pineal approach. It is perhaps less dangerous and there is a much better exposure of the tumor area. The danger is less because the attack upon the tumor is at a distance from the veins of Galen. On the other hand, if the growth extends back to the great vein of Galen, the danger is greater by the frontal route, for the tumor must be blindly separated from the venous trunks. There is practically no risk in either type of section of the frontal lobe and no loss of function, mental or physical. One may ask why it is not possible to remove such tumors through a speculum introduced into the anterior horn of the lateral ventricle. Such a procedure is indeed possible with some of the simpler tumors, but should the operator encounter bleeding, he would be greatly handicapped for room and might well lose the patient through the effects of trauma and hemorrhage, when it is too late to provide the adequate exposure except through a heroic effort. There is little to lose and everything to gain in safety of life and function by providing an adequate exposure through the preliminary removal of cerebral tissue. There is, of course, always the
thought in mind that epilepsy may follow in the wake of the cerebral defects created for the purpose of exposing the tumor. As yet no patients so treated have had convulsions following the operation, if they were absent before. It is not always certain that the tumor is confined entirely to the third ventricle. For example, in Case 8, Group II, the tumor had passed through the foramen of Monro and the extension in the lateral ventricle was much larger than the tumor in the third ventricle. For the removal of such a growth, the pineal approach would be very embarrassing, perhaps impossible. In this instance the filling defect in the lateral ventricle left no doubt of the exact location of the tumor and of the character of the operative approach. For the anterior approach either the right or left side may be chosen. Although there is little difference in the choice, the right side is chosen in right-handed individuals, and the left side in those who are lefthanded, unless there are reasons to believe that the tumor may have extended to one side as in the above case. If the tumor is suspected or known to have extended to one side, the operative approach is made upon that side. . . . The size of the lateral ventricle is another very important factor in determining the site of the operative approach. In all of our cases there have been very large lateral ventricles, such as would be expected with tumors obstructing either the foramina of Monro or the aqueduct of Sylvius. Large ventricles are absolutely essential to a posterior (pineal) approach because room for the operative attack is provided only by releasing fluid from one or both lateral ventricles. It is conceivable that in the early stages of primary tumors of the third ventricle the lateral ventricles might not yet be sufficiently enlarged to permit of this approach, but in the present state of our diagnostic ability it is most unlikely that such an early diagnosis would be made. But in the event that a tumor should be diagnosed when the ventricles are small, the anterior approach with resection of the frontal lobe would afford ample room. The larger tumors in the region of the third ventricle and filling it secondarily usually show small lateral ventricles and, therefore, practically preclude the posterior approach, except perhaps to explore a tumor along the falx. At first glance the statement that in the larger tumors the lateral ventricles are smaller may appear paradoxical. Although they cause the same ventricular obstruction and should, therefore, produce the same degree of hydrocephalus, the ventricles are actually smaller because the bulk of the tumor encroaches upon and fills or obliterates both lateral ventricles. The primary tumors of the third ventricle are usually so small that their actual volume is of little or no importance in the creation of intracranial pressure, which is due solely to the obstruction to the channels of exit of the cerebrospinal fluid. Small ventricles with a tumor of the third ventricle, therefore, strongly suggest a large tumor and one arising presumably outside the third ventricle. Moreover, such tumors-usually gliomata—are much more common than the primary tumors in the third ventricle. In the treatment of such large growths the anterior approach, with cerebral resection, is necessary (37).
Dandy ended his monograph with the following statement: "The mortality from the entire series of extirpations, covering a period of eleven years, has been 33.3 per cent. However, in the past two and one-half years fourteen of these tumors have been removed with a mortality of 14.3 per cent. Except for occasional minor effects, these patients make almost perfect recovery of all functions (37). After the publication of Dandy's book, the direct operative removal of colloid cysts of the third ventricle became the standard approach to the treatment of these lesions (3) until the development of CT scanning and of stereotaxic and ven-triculoscopic techniques offered other possibilities for management (depending upon symptoms and ventricular size), such as simply following the patient with serial CT scans, shunting the lateral ventricle(s), shunting the cyst, aspirating the cyst percutaneously, or aspirating or removing the cyst endoscopically (10, 76, 112, 121). Direct operative excision also became the preferred treatment of other types of benign tumors of the third ventricle such as epidermoid and dermoid cysts, meningiomas (23, 111, 123), symptomatic xanthogranulomas (68, 125), cho-roid plexus papillomas (70), intraventricular craniopharyngiomas (69, 91, 96, 97, 99, 107, 137, 143, 149), and the subependymal giant cell astrocytomas of tuberous sclerosis (12, 20, 86). However, the removal of such tumors sometimes proved difficult, especially when they were located primarily in the posterior portion of the third ventricle or were adherent to the hypothal-amus. It was found that operative injury to the hypothalamus might result in diabetes insipidus. (Incidentally, another condition studied by Walter Dandy during his career was diabetes insipidus (40, 117)). Problems of this sort were especially true before the development of microsurgery and prompted Torkildsen in 1948 to publish a paper entitled, "Should Extirpation be Attempted in Cases of Neoplasm in or near the Third Ventricle of the Brain? Experiences with a Palliative Method" (146). Torkildsen concluded as follows: Attempts at the removal of neoplasms in regions of the pineal gland and the 3rd ventricle of the brain are associated with such a grave mortality rate that, if possible, such operations should be avoided. It is possible to treat patients with such lesions by ventriculocisternostomy, with satisfactory results. I have performed the operation in 8 cases of tumor in the pineal region, with 1 postoperative death, and in 11 cases of tumor in or near the 3rd ventricle outside the pineal region, with 2 fatalities. The above results show that the mortality rate by
this operation is far below that following attempts at surgical extirpation of the tumors. The fate of the patients depends primarily on the malignancy of the neoplasm. Tumors of high expanding energy infiltrating this region of the brain will in any case cause the death of the patient within limited time. In such cases no other operation can possibly save the life of the patient. Most of the tumors, however, are of very slow growth, and the patients may live comfortably with the neoplasm if the symptoms due to the obstruction of the flow of the cerebrospinal fluid can be relieved. In my series there are several cases in which no visible growth of the neoplasm has taken place after intervals of 7 or 8 years. In any case of tumor in the pineal region, neoplasm in or near the 3rd ventricle, craniopharyn-gioma, and stenosis of the Sylvian aqueduct whether due to neoplasm or not, ventriculocisternostomy should be carried out before any other surgical procedures are undertaken (146). Most neurosurgeons at that time took a similar approach to third ventricular masses, namely, diagnosis by air study and (unless a colloid cyst, choroid plexus papilloma, epidermoid or dermoid cyst, meningioma, craniopharyngioma, or subependymal giant cell astrocytoma was suspected) treatment by radiation therapy from an external source. Significant hydrocephalus was treated by ventriculocisternostomy. Yet without a tissue diagnosis the neurosurgeon took the chance that the lesion was benign and the radiation therapy inappropriate. With the advent of microneurosurgery, the pendulum then swung toward open operative exposure of the tumor for diagnosis and, frequently, resection. Stein stated, "In our experience with 46 patients encompassing all types of third ventricular tumors, the surgical mortality has been under 5% and approximately 30% of the tumors have been benign, encapsulated, and resectable" (139). But with such a direct approach the surgeon did add to the patient's immediate morbidity and mortality risks and took the chance of opening up a tumor that would be more appropriately treated with radiotherapy or chemotherapy. Fortunately, the subsequent development of CT-guided stereotaxic biopsy techniques (5, 87, 109, 118), the discovery of tumor markers (126, 134), and the improvement of CSF cytodiagnosis (9) have provided the neurosurgeon the means of being relatively certain of the diagnosis before the decision is made about definitive treatment. In addition, if surgical resection is chosen, the neurosurgeon of today has more tools than ever before to accomplish this goal, such as micro-surgical instruments, the ultrasonic aspirator,
and various surgical lasers. Kelly and his colleagues are currently perfecting a CT-guided stereotaxic system that uses a computer to direct and monitor tumor removal by laser energy (88, 89). Cysticercosis of the Third Ventricle Cysticercosis can involve the central nervous system in various ways (54, 98, 100, 140). One of these is by the development of cysticercus cysts within the ventricular system. The larval forms pass through the choroid plexus, develop in the ventricles as cysts, and migrate caudally through the ventricular system. When these cysts are within the ventricles, they usually do not incite an inflammatory response, but when they reach the basilar cisterns they usually evoke a significant adhesive basilar arachnoiditis. Thus, over the years, emphasis has been placed on the surgical removal of such cysts while they are still within the ventricular system. The cysts are ordinarily diagnosed by positive contast ventriculography, now with CT imaging. They frequently are multiple, and the exact route of operative attack is determined by their locations and sizes. When such cysts have been discovered within the third ventricle, the same surgical approaches have been used as for third ventricular tumors. Ventriculoscopy and cyst removal have been described, but the latter has been reported to be associated with breakage of the cyst with intraventric-ular spillage, which can evoke ventricular inflammation. Exposure . . . permits gradual removal of the lesion into the aperture and complete delivery of the cysts... . The lesions seem to conform to the surgical egress route but may need to be assisted in their delivery by the use of hydraulic dissection, a mild Valsalva maneuver by the anesthesiologist, and gentle teasing with nonpenetrating instruments. When the cysts are approached surgically in the lateral and third ventricles, the use of a slender Pasteur pipette with gentle suction is reported to extract the lesions effectively. . . . Patients are best treated intraoperatively by corticosteroid medication. Under these circumstances, rupture has not excited adverse reactions. . . . Results in the management of exclusively intraventricular cysts are excellent, especially when these are isolated or solitary. The surgical management of the other forms appears to be primarily palliative (140). Pineal Tumors In 1915, Walter Dandy published a paper, "Extirpation of the Pineal Body," that dealt with pinealectomy in the dog (25). Dandy concluded: 1. Following the removal of the pineal I have observed no sexual precocity or indolence, no adi-
oposity or emaciation, no somatic or mental precocity or retardation. 2. Our experiments seem to have yielded nothing to sustain the view that the pineal gland has an active endocrine function of importance either in the very young or adult dogs. 3. The pineal is apparently not essential to life and seems to have no influence upon the animal's well being. [25] Of more importance to the subsequent development of the surgical exposure and resection of pineal tumors in humans, Dandy noted, in comparing his new method to his previous methods: To be of practical value the pineal body must be more easily reached and removed with greater certainty and less mortality. Consequently a new and simple method of attack has been evolved. Though more delicate and requiring more painstaking care, it can be done almost as easily as a canine hypoph-ysectomy. The new operation can be done in less than one hour. It differs from the preceding operation in that the pineal is reached from in front through the third ventricle rather than from behind. In this way the extensive bleeding consequent to liberation of the vein of Galen is obviated, sidetracked as it were, and the operation can be performed almost bloodlessly. This is accomplished by dividing the splenium of the corpus callosum in the midline for a distance of about 2 cm. from its posterior terminus. This exposes the transparent roof of the third ventricle which is distended by the contained cerebrospinal fluid. A large anemic area is visible in the midline of the roof of the ventricle, between the two small veins of Galen. This is perforated and the opening enlarged backward to the origin of the vena Galena magna by releasing the blades of the forceps. The entire third ventricle is thus brought in full view and the pineal body is readily seen under the origin of this vein, in the median quadrigeminal groove. The pineal body can easily be grasped in the jaws of the cupped biting forceps and completely removed (Figs. 1.16 and 1.17) (25). Six years later, in 1921, Dandy published an account entitled, "An Operation for the Removal of Pineal Tumors" (30), a subject to which he returned in 1929 (34) and again in 1936 (38). He began his 1921 paper in this way: "Tumors of the pineal body have rarely been diagnosed and substantiated. The total number of authenticated pineal tumors is less than one hundred and almost all have been accidental findings at necropsy" (30). In his thorough review of the history of the surgery of pineal lesions, Pendl (110) gives credit to Victor Horsley (81, 82) as being the first to attempt to remove a pineal tumor by direct surgical attack. Two such procedures (in 1905 and 1909), done through an infratentorial approach, were unsatisfactory, prompting Horsley in 1910
Figure 1.16. Canine pinealectomy. The upper drawing demonstrates the sple-nium divided, exposing the roof of the third ventricle with its veins. The lower drawing shows the method of perforation of the roof between the internal cerebral veins. (From Dandy WE: Extirpation of the pineal body. J Exp Med 22:237-247, 1915.) to recommend a supratentorial exposure with splitting of the tentorium (81). Others also reported operations for palliation of pineal tumors in 1908, 1910, and 1911 (110). * According to Pendl (110), Ludwig Pussep (114) in 1910 was probably the first to reach a pineal tumor by a direct surgical approach. Pussep exposed the cystic tumor of a 10year-old boy through an occipital approach by transecting the transverse sinus and splitting the tentorium. He was able to remove parts of the tumor, but the child died on the third postoperative day. In his neurosurgical textbook, Fedor Krause (92) "... suggested in 1911 an infratentorial and supracerebellar route for surgical exploration of the upper vermis and quadrigeminal region. In 1913 he presented a 10-year-old boy to the Medical Society in Berlin, from whom, 6 weeks prior, he had resected a 4 cm pineal tumor. This was achieved in the sitting position using an inf raten-torialsupracerebellar approach. ... As was pointed out by Zulch. . ., this was the first report
of a successful extirpation of a tumor from the pineal region" (110). Pendl further reports that Brunner unsuccessfully explored the posterior fossa of a patient with a quadrigeminal tumor in 1911 and then the following year decided to approach a pineal tumor in a cousin of the first patient by a transcallosal route. "Because of severe venous hemorrhage and difficulties in exposure, the tumor was not reached and the operation abandoned" (110). Other operations were proposed or attempted by various surgeons, but without success (110). Thus, the stage was set for Dandy's 1921 report of his method of posterior transcallosal exposure of a pineal tumor (Figs. 1.18 to 1.22): The operation has been performed on three patients. In the first instance a silent cerebellar tumor had secondarily involved the region of the pineal body and the corpora quadrigemina; after exposure of the tumor, no attempt was made to remove it, because of its infiltrating character. This case, how-
Figure 1.17. Canine pinealectomy. The upper drawing shows that, after the defect in the roof of the third ventricle is opened and widened, the pineal gland is well exposed. The lower drawing illustrates the way the pineal gland is grasped with the biting forceps, the jaws of which are also shown separately. (From Dandy WE: Extirpation of the pineal body. J Exp Med 22:237-247, 1915.) ever, showed that a good exposure of this region is possible. On two subsequent occasions, tumors of the pineal body have been completely removed. In one case an encapsulated tubercle of the pineal body was extirpated, the patient recovering only to die 8 months later, presumably of the effects of other tubercles of the brain. . . . The results of this case demonstrated not only the feasibility of the removal of tumors of the pineal body but also the absence (in this case at least) of any injurious mental or physical effects due to the operation. The extirpation of the second pineal tumor was very much more difficult. The tumor was much larger and since the vena Galena magna and both small veins of Galen passed directly through the tumor and could not be dissected from it, the removal of practically all of these veins was a prerequisite to enucleation of the tumor. . . . The patient lived 48 hours, dying presumably of the shock due to the magnitude of the operation. His vitality was doubtless impaired by a cerebellar operation which was performed 10 days previously and at which the
tumor was found; but it was entirely inaccessible for removal by this approach. . . . One must also consider whether the complete removal of all the main trunks of the intracranial venous system— the large vein of Galen and both small veins of Galen—would be compatible with life. I know of no previous instance in which these veins have been removed or even ligated. In dogs I have ligated the vena Galena magna without effect when the ligation is distally placed but when proximally placed a mild grade of hydrocephalus has resulted. . . . Whether the collateral venous system in the human brain could be developed to compensate for this tremendous loss can only be surmised; it does seem doubtful, but there was no alternative to the removal of these veins with the tumor. [30] In 1928, Otfrid Foerster published his experience with the exposure of the quadrigeminal plate and pineal area in three patients (60, 61). A large occipital craniotomy, close to the superior
Figure 1.18. Cross sectional diagram showing the various steps in the transcal-losal removal of a pineal tumor. (From Dandy WE: An operation for the removal of pineal tumors. Surg Gynecol Obstet 33:113-119, 1921.) longitudinal sinus and the right transverse sinus, exposed the occipital lobe; the dura was opened and reflected as well over the sagittal and transverse sinuses; the bridging veins were ligated and transected. The posterior horn of the lateral ventricle was punctured by a cannula with sufficient decompression to allow retraction of the occipital lobe from the falx and tentorium. The splenium corporis callosi was thus exposed. For better access to the quadrigeminal area, the tentorium along the sinus rectus could be cut and reflected laterally, and a few centimeters of the splenium split. If the tumor on the left was not sufficiently exposed, the falx could be incised to better visualize the opposite side. Foerster's first case, a 25-year-old man, was operated on in 1927 with a total removal of a glioma and subsequent full recovery. The second reported case, operated on initially in 1923 with only a partial evacuation of a cystic tumor, required a second approach. However, the well-exposed tumor could not be removed from dense adhesions to the midbrain and the patient died on the second postoperative day. In the third case, the clinical signs indicated a pineal tumor; the whole quadrigeminal area was exposed, but no tumor and also no pineal gland was found (110). In 1931, William Van Wagenen introduced an entirely different approach to pineal tumor sur-
gery (150). He had exposed the tumor in his patient through the dilated right lateral ventricle. The lateral cortical incision had extended superiorly and posteriorly from the posterior end of the superior temporal gyrus, and the ventricle had been entered at the atrium. The spongioblas-tic tumor had then been exposed through the thinned medial ventricular wall and had been almost completely removed. Harris and Cairns in 1932 reported an unsuccessful attempt to remove a pineal tumor through a suboccipital exposure in 1930 (77). Six months after the initial attempt, the tumor was successfully removed through a right parietooccipital transcallosal approach. In his 1932 monograph on intracranial tumors, Cushing wrote: "Personally I have never succeeded in exposing a pineal tumour sufficiently well to justify an attempt to remove it. Encouraging reports of such procedures, however, have been made by W. E. Dandy of Baltimore (1921), by Otrid Foers-ter of Breslau (1928), and I have personal knowledge of two highly successful operations per-
Figure 1.19. Exposure of the medial cerebral cortex before exposure of the corpus callosum. A ventricular needle has been inserted to drain CSF from the lateral ventricle. (From Dandy WE: An operation for the removal of pineal tumors. Surg Gynecol Obstet 33:113-119, 1921.) formed by my former assistants, Dr. W. P. Van Wagenen of Rochester, New York, and Mr. Hugh Cairns of London" (22). Pendl cites various sources to indicate that Ernest Sachs removed a large cystic tumor from the pineal region in 1926 and Max Peet removed a pineal tumor through a right parietooccipital transcallosal approach in 1927 (110). By 1936, Dandy was able to comment on 10 personal cases (Figs. 1.23 to 1.25) (38), but he was pessimistic: Although an operative approach to the pineal region was proposed by me in 1921 [30], it was not until a decade later (1931) that the first pineal tumor was successfully extirpated [37]. A disastrous toll of seven fatal issues during this long period seemed almost to indicate the futility of further efforts. Yet the same approach was used successfully in a number of cases of tumor of the third ventricle in which, though the lesion was in the same general region, less serious difficulties were offered. In addition to the aforementioned case, successful extirpation has since been performed in two others, thus indicating that the lesion is not entirely hopeless, although it remains one of the most dangerous of all intracra-nial growths. This seeming success, however, is still further tempered by the fact that in two of these
cases the tumor has since recurred, in one after two and one-half years and in the other after three months. Only the last of the three patients remains well, and the period after operation has been only four months. Since the growth in the first case was solid and encapsulated and was extirpated intact, I expected a permanent cure. However, the growth was teratomatous, containing, in addition to the pineal element, ciliated columnar cells of ependymal origin, cartilage and tissue resembling the salivary gland. The tumor in the second case was much larger; it was so soft and cellular and its capsule so thin that it could not be shelled out intact but had to be removed in fragments. The patient died at her home three months later, but permission for necropsy was not obtained. It is, perhaps, more conceivable that cicatrization at the mouth of the aqueduct caused her death than that such rapid recurrence was responsible. The tumor of the third patient was a pure pinealoma and was perfectly encapsulated. It is unbelievable that it can recur, although I felt equally secure in such a prediction for the first case. In that instance, however, the tumor was a teratoma, whereas in this case the growth was a pure pinealoma. Reports of three cases of pineal tumor in which the patient survived the operation are appended. Particularly of interest are the remarkable transient postoperative sequelae and the absence of recognizable effects arising from the removal of part or all
Figure 1.20. After the bridging veins have been ligated and divided, the cerebral hemisphere is retracted to expose the falx and the corpus callosum. (From Dandy WE: An operation for the removal of pineal tumors. Sura Gynecol Obstet 33-113-119, 1921.)
Figure 1.21. The inferior longitudinal sinus has been clipped and the falx has been divided inferiorly. The corpus callosum has been split longitudinally to expose the pineal tumor, the internal cerebral veins, and the vein of Galen. (From Dandy WE: An operation for the removal of pineal tumors. Surg Gynecol Obstet 33:113-119, 1921.)
Figure 1.22. After removal of the pineal tumor, the intact roof of the third ventricle is seen. (From Dandy WE: An operation for the removal of pineal tumors. Surg Gynecol Obstet 33:113-119, 1921.)
Figure 1.23. Drawings showing the operative procedure in the removal of a teratoma from the pineal region. (From Dandy WE: Operative experience in cases of pineal tumor. Arch Surg 33:19-46, 1936.)
Figure 1.24. Drawings illustrating the method of producing more adequate exposure of a large tumor in the pineal region by resecting the occipital lobe and, when necessary, the lower part of the falx, parts of the right and left sides of the tentorium, the vein of Galen, and the straight sinus. (From Dandy WE: Operative experience in cases of pineal tumor. Arch Surg 33:19-46, 1936.)
Figure 1.25. Operative sketches of the removal of a pineal tumor, showing that the right internal cerebral vein had to be divided. (From Dandy WE: Operative experience in cases of pineal tumor. Arch Surg 33:19-46, 1936.) of the main trunks of the internal venous system of the brain (38). Many other surgeons have been involved in the development of the surgical treatment of pineal tumors since 1930, as mentioned in the publications of Thomson (145), Schmidek (133), Zulch (162), Sano (126), and Pendl (110). As with tumors of the third ventricle, the high operative morbidity and mortality of direct exposure and resection of pineal tumors led most neurosur-geons for the next 40 years to use the more conservative course of CSF shunting and radiation therapy (1, 48, 52, 79, 116, 146). This ap-
proach to treatment was strengthened by the development of better techniques of CSF shunting and of radiotherapy (7, 59, 113, 131). Then, after the advent of microneurosurgery, interest in direct surgical treatment began to return (122, 138, 153, 160). However, the need for operative exposure and resection of these lesions has been modified by the development of CT-guided stereotaxic biopsy techniques, the increase in reliance on biochemical tumor markers and CSF cytology for the diagnosis of pineal tumors, and a revival of interest in the implantation of radiation sources for the treatment of intracranial neoplasms (5, 9, 63, 85, 103, 109, 118, 126, 134).
Operative Approaches to the Third Ventricle As noted previously, Dandy exposed the third ventricle for ventriculostomy by a subfrontal and later by a transtemporal approach. For anterior third ventricular tumors, he preferred a frontal transcortical route, and for posterior third ventricular and pineal tumors, a parietooccipital transcallosal approach. Dandy thought that no neurological deficit occurred as a result of either approach (37, 71). However, "if the splenium of the corpus callosum is completely sectioned, appropriate testing will reveal that most postoperative patients cannot read in their left visual fields, even when all other neurological functions have returned to normal. This phenomenon, appropriately called hemialexia, was first reported by Trescher and Ford in 1937 [147]. Their patient had the posterior corpus callosum split by Walter Dandy for removal of a colloid cyst of the 3rd ventricle. . . . Since that time the reality of the deficit has been proven repeatedly. . ." (71).
These and various other operative approaches to the region of the third ventricle have been developed, primarily since Dandy's time. Many of these have been discussed by Rhoton and Yamamoto (Fig. 1.26) (119, 120, 161) and by Antunes et al. (4), as well as at a recent international symposium organized by Madjid Samii and held in Hannover, West Germany (83). The transsphenoidal and certain subfrontal (i.e., subfrontal subchiasmatic or subfrontal transsphenoidal) approaches permit access to the sella turcica and chiasmatic area but do not ordinarily permit visualization within the third ventricle. The same is true of the more lateral subfrontal and subtemporal approaches. Tumors can be drawn down from the third ventricle but entry into the ventricle is quite limited by the angle of entry and the important structures that are interposed (optic nerves, chiasm, and tracts; arteries of the circle of Willis and their branches; hypothalamus; pituitary stalk and gland; etc.). The occipital transtentorial and infratentorial supracerebellar approaches have been used pri-
Figure 1.26. A midsagittal view of the head, showing the basic operative approaches to the third ventricle. (From Rhoton AL Jr, Yamamoto I, Peace DA: Microsurgery of the third ventricle: Part 2. Operative approaches. Neurosurgery 8:357-373, 1981.)
marily to expose tumors in the region of the pineal gland. The early history of these approaches has been considered, and the reader is referred to the publications of Thomson (145), Schmidek (133), Zulch (162), Sano (126), and Pendl (110) for details about the subsequent development of these procedures. Among the subf rental approaches, the one that has been somewhat helpful in dealing with low anterior third ventricular tumors has been the one that involves entry into the ventricle through the lamina terminalis. Such entry was used for third ventriculostomy, as previously discussed. King (91) and Suzuki et al. (142) have outlined the history of its use for tumors such as intra-ventricular craniopharyngiomas. The posterior transventricular approach has been of limited value. However, the anterior transventricular route through a dilated lateral ventricle to high anterior third ventricular tumors such as colloid cysts has remained useful (52). Transcallosal operations were done occasionally after Dandy developed the posterior transcallosal approach (16, 17, 50-52, 72, 73, 108, 151), but it was not until the introduction of microsurgical techniques (101) that the transcallosal route became popular (52, 96, 97, 136). The anterior transcallosal approach has become a preferred method of access to the third ventricle (Sano even uses it for pineal region neoplasms (126, 127)), especially since it has been found that anterior or middle transcallosal operations may not cause behavioral or neurological abnormalities (67, 159). The emphasis in recent years has been on the refinement of techniques for opening the roof of the third ventricle (4, 6, 13, 21,47,78,83,93, 154). Concluding Remarks The central location of the third ventricle and the pineal gland within the brain have lent them a certain mystery. Descartes thought of the pineal gland as "the seat of the soul" (19). Surgeons dealing with lesions of the third ventricle have noted that such lesions (or injuries resulting from the attempts to remove them) may cause alterations in consciousness (14, 15, 42). In fact, Walter Dandy's last paper, published 3 months after his death, contained his speculations about the location of the "conscious center" in the brain, based on a study of 10 of his patients, 2 of whom were rendered unconscious by the removal of tumors from the third ventricle (42). The mysteries surrounding lesions of the third
ventricular region are slowly being solved. The pioneering work of Walter Dandy and others has laid the foundation for our present methods of diagnosing and treating these conditions. These current methods and the future advancements they may permit are detailed in the rest of this book. References 1 . Albay EO II, Laws ER Jr, Grado GL, Bruckman JE, Forbes GS, Gomez MR, Scott M: Pineal tu mors in children and adolescents: Treatment by CSF shunting and radiotherapy. J Neurosurg 55:889-895, 1981. 2. Anton, von Bramann: Balkenstich bei Hydrozephalien, Tumoren und bei Epilepsie. Munch Med Wochenschr 55:1673-1677, 1908. 3. Antunes JL, Louis KM, Ganti SR: Colloid cysts of the third ventricle. Neurosurgery 7:450-455, 1980. 4. Antunes JL, Muraszko K, Quest DO, Carmel PW: Surgical strategies in the management of tu mours of the anterior third ventricle, in Brock M (ed): Modern Neurosurgery 1. Berlin, SpringerVerlag, 1982, pp 215-224. 5. Apuzzo MLJ, Chandrasoma PT, Zelman V, Giannotta SL, Weiss MH: Computed tomographic guidance stereotaxis in the management of le sions of the third ventricular region. Neurosur gery 15:502-508, 1984. 6. Apuzzo MLJ, Chikovani OK, Gott PS, Teng EL, Zee CS, Giannotta SL, Weiss MH: Transcallo sal, interfornicial approaches for lesions affect ing the third ventricle: Surgical considerations and consequences. Neurosurgery 10:547-554, 1982. 7. Backlund EO, Rahn T, Sarby B: Treatment of pinealomas by stereotaxic radiation surgery. ActaRadiol Ther 13:368-376, 1974. 8. Balado M, Morea R, Donovan C: La radiografia del tercer ventriculo. Bol Inst Clin Quirurg 2:603-610, 1926 (cited by Walker (157)). 9. Bigner SH, Johnston WW: Cytopathology of the Central Nervous System. New York, Masson, 1983. 10. Bosch DA, Rahn T, Backlund EO: Treatment of colloid cysts of the third ventricle by stereotactic aspiration. Surg Neurol 9:15-18, 1978. 1 1 . Brocklehurst G: Diencephalic cysts. J Neurosurg 38:47-51, 1973. 12. Burger PC, Vogel FS: Surgical Pathology of the Nervous System and Its Coverings. New York, John Wiley & Sons, 1982, ed 2, p 279. 13. Busch E: A new approach for the removal of tumors of the third ventricle. Acta Psychiatr Neurol 19:57-60, 1944. 14. Cairns H: Disturbances of consciousness with lesions of the brain-stem and diencephalon. Brain 75:109-146, 1952. 15. Cairns H, Oldfield RC, Pennybacker JB, Whitteridge D: Akinetic mutism with an epidermoid cyst of the 3rd ventricle. (With a report on the asso ciated disturbance of brain potentials.) Brain 64:273-290, 1941.
16. Cassinari V, Bernasconi V: Tumori della parte anteriore del terzo ventricolo. Acta Neurochir (Wien) 11:236-271, 1964. 17. Christensen JC: Third ventricle tumours: Clini cal findings and surgical results in 28 consecu tive cases. Excerpta Med Int Congr Series 36:E61-E62, 1961. 18. Ciric I, Zivin I: Neuroepithelial (colloid) cysts of the septum pellucidum. J Neurosurg 43:69-73, 1975. 19. Clarke E, O'Malley CD: The Human Brain and Spinal Cord: A Historical Study Illustrated by Writings from Antiquity to the Twentieth Cen tury. Berkeley, CA, University of California Press, 1968, pp 469-472. 20. Cooper JR: Brain tumors in hereditary multiple system hamartomatosis (tuberous sclerosis). J Neurosurg 34:194-202, 1971. 21. Cossu M, Lubinu F, Orunesu G, Pau A, Sehrbundt Viale E, Sini MG, Turtas S: Subchoroidal approach to the third ventricle: Microsurgical anatomy. Surg Neurol 21:325-331, 1984. 22. Cushing H: Intracranial Tumours: Notes Upon a Series of Two Thousand Verified Cases with Surgical-Mortality Percentages Pertaining Thereto. Springfield, IL, Charles С Thomas, 1932, p 64. 23. Cushing H, Eisenhardt L: Meningiomas: Their Classification, Regional Behaviour, Life His tory, and Surgical End Results. Springfield, IL, Charles С Thomas, 1938 (New York, Hafner, 1969), pp 125-126, 136-138. 24. Dandy WE: The nerve supply to the pituitary body. Am J Anat 15:333-343, 1913. » 25. Dandy WE: Extirpation of the pineal body. J Exp Med 22:237-247, 1915. 26. Dandy WE: Ventriculography following the in jection of air into the cerebral ventricles. Ann Surg 68:5-11, 1918. 27. Dandy WE: Extirpation of the choroid plexus of the lateral ventricles in communicating hydrocephalus. Ann Surg 68:569-579, 1918. 28. Dandy WE: Rontgenography of the brain after the injection of air into the spinal canal. Ann Surg 70:397-403, 1919. 29. Dandy WE: The diagnosis and treatment of hydrocephalus resulting from strictures of the aq ueduct of Sylvius. Surg Gynecol Obstet 31:340358, 1920. 30. Dandy WE: An operation for the removal of pineal tumors. Surg Gynecol Obstet 33:113119, 1921. 31. Dandy WE: II. Diagnosis, localization and re moval of tumors of the third ventricle. Johns Hopkins Hosp Bull 33:188-189, 1922. 32. Dandy WE: III. Cerebral ventriculoscopy. Johns Hopkins Hosp Bull 33:189, 1922. 33. Dandy WE: V. An operative procedure for hydrocephalus. Johns Hopkins Hosp Bull 33:189190, 1922. 34. Dandy WE: Mechanism and symptoms of tumors of the third ventricle and pineal body. Res Publ Assoc Res Nerv Ment Dis 8:375-385, 1929. 35. Dandy WE: Congenital cerebral cysts of the cavum septi pellucidi (fifth ventricle) and cavum vergae (sixth ventricle): Diagnosis and treat ment. Arch Neurol Psychiatry 22:44-66, 1931.
36. Dandy WE: The brain, in Lewis D (ed): Practice of Surgery. Hagerstown, MD, WF Prior, 1932, pp 1-682. 37. Dandy WE: Benign Tumors in the Third Ven tricle of the Brain: Diagnosis and Treatment. Springfield, IL, Charles С Thomas, 1933. 38. Dandy WE: Operative experience in cases of pineal tumor. Arch Surg 33:19-46, 1936. 39. Dandy WE: The operative treatment of commu nicating hydrocephalus. Ann Surg 108:194202, 1938. 40. Dandy WE: Section of the human hypophysial stalk: Its relation to diabetes insipidus and hy pophysial functions. JAMA 114:312-314, 1940. 41. Dandy WE: Diagnosis and treatment of stric tures of the aqueduct of Sylvius (causing hydro cephalus). Arch Surg 51:1-14, 1945. 42. Dandy WE: The location of the conscious center in the brain—the corpus striatum. Bull Johns Hopkins Hosp 79:34-58, 1946. 43. Dandy WE, Blackfan KD: An experimental and clinical study of internal hydrocephalus. JAMA 61:2216-2217, 1913. 44. Dandy WE, Blackfan KD: Internal hydrocepha lus: An experimental, clinical and pathological study. Am J Dis Child 8:406-482, 1914. 45. Dandy WE, Blackfan KD: Internal hydrocepha lus: Second paper. Am J Dis Child 14:424-443, 1917. 46. Dandy WE, Goetsch E: The blood supply of the pituitary body. Am J Anat 11:137-150, 1911. 47. Delandsheer JM, Guyot JF, Jomin M, Scherpereel B, Laine E: Acces au troisieme ventricule par voie inter-thalamo-trigonale. Neurochirurgie 24:419-422, 1978. 48. Donat JF, Okazaki H, Gomez MR, Reagan TJ, Baker HL Jr, Laws ER Jr: Pineal tumors: A 53year experience. Arch Neurol 35:736-740, 1978. 49. Drayer BP, Rosenbaum AE: Studies of the third circulation; Amipaque CT cisternography and ventriculography. J Neurosurg 48:946-956, 1978. 50. Ehni G: Pineal teratoma: Report of case. J Neu rosurg 3:86-94, 1946. 51. Ehni G: Transcallosal Removal of Brain Tumors. Motion picture available from American College of Surgeons Film Library or Baylor College of Medicine, 1969. 52. Ehni G: Interhemispheric and percallosal (transcallosal) approach to the cingulate gyri, intraventricular shunt tubes, and certain deeply placed brain lesions. Neurosurgery 14:99-110, 1984. 53. Elvidge AR: Treatment of obstructive lesions of the aqueduct of Sylvius and the fourth ventricle by interventriculostomy. J Neurosurg 24:1123, 1966. 54. Estanol B, Kleriga E, Loyo M, Mateos H, Lombardo L, Gordon F, Saguchi AF: Mechanisms of hydrocephalus in cerebral cysticercosis: Impli cations for therapy. Neurosurgery 13:119-123, 1983. 55. Fairman D: Evolution of neurosurgery through Walter E. Dandy's work. Surgery 19:581-604, 1946. 56. Fay T, Grant FC: Ventriculoscopy and intraven-
57. 58.
59.
60. 61. 62. 63.
64. 65.
66. 67.
68. 69. 70.
71. 72. 73. 74.
75.
tricular photography in internal hydrocephalus: Report of case. JAMA 80:461-463, 1923. Feld M: Ventriculoscope coagulant utilise dans les hydrocephalies communicantes. Presse Med 64:632, 1956. Feld M: La coagulation des plexus choroides par ventriculoscopie directe dans l'hydrocephalie non obstructive du nourrisson. Neurochirurgie 3:70-79, 1957. Fisher RG: Surgery of the congenital anomalies, in Walker AE (ed): A History of Neurological Surgery. Baltimore, Williams & Wilkins, 1951, pp 334-361. Foerster O: Das operative Vorgehen bei Tumor der Vierhugelgegend. Wien Klin Wochenschr 41:986-990, 1928. Foerster O: Uber das operative Vorgehen bei Operationen der Vierhugelgegend. Zbl Ges Neurol Psychiat 61:457-459, 1932. Fox WL: Dandy of Johns Hopkins. Baltimore, Williams & Wilkins, 1984. Frank F, Gaist G, Piazza G, Ricci RF, Sturiale C, Galassi E: Stereotaxic biopsy and radioactive im plantation for interstitial therapy of tumors of the pineal region. Surg Neurol 23:275-280, 1985. Fukushima T: Endoscopic biopsy of intraventricular tumors with the use of a ventriculof iberscope. Neurosurgery 2:110-113, 1978. Fukushima T, Ishijima B, Hirakawa K, Nakamura N, Sano K: Ventriculofiberscope: A new technique for endoscopic diagnosis and opera tion. Technical note. J Neurosurg 38:251-256, 1973. Fulton JF: Harvey Cushing: A Biography. Springfield, IL, Charles С Thomas, 1946, p 618. Geffen G, Walsh A, Simpson D, Jeeves M: Com parison of the effects of transcortical and transcallosal removal of intraventricular tu mours. Brain 103:773-788, 1980. Godersky JC, Rockswold G, Larson DA: Xanthogranuloma of the third ventricle producing hy drocephalus. Neurosurgery 7:68-70, 1980. Goldstein SJ, Wilson DD, Young AB, Guidry GJ: Craniopharyngioma intrinsic to the third ventri cle. Surg Neurol 20:249-253, 1983. Gradin WC, Taylon C, Fruin AH: Choroid plexus papilloma of the third ventricle: Case report and review of the literature. Neurosurgery 12:217220, 1983. Greenblatt SH: Neurosurgery and the anatomy of reading: A practical review. Neurosurgery 1:6-15, 1977. Greenwood J Jr: Paraphyseal cysts of the third ventricle: With report of eight cases. J Neuro surg 6:153-159, 1949. Greenwood J Jr: Radical surgery of tumors of the thalamus, hypothalamus, and third ventricle area. Surg Neurol 1:29-33, 1973. Guiot G, Derome P, Hertzog E, Bamberger C, Akerman M: Ventriculo-cisternostomie sous controle radioscopique pour hydrocephalie ob structive. Presse Med 76:1923-1926, 1968. Guiot G, Rougerie J, Fourestier M, Fournier A, Comoy C, Vulmiere J, Groux R: Explorations endoscopiques intracraniennes. Presse Med 71:1225-1228, 1963.
76. Gutierrez-Lara F, Patino R, Hakim S: Treatment of tumors of the third ventricle: A new and sim ple technique. Surg Neurol 3:323-325, 1975. 77. Harris W, Cairns H: Diagnosis and treatment of pineal tumours: With report of a case. Lancet 1:3-9, 1932. 78. Hirsch JF, Zouaoui A, Renier D, Pierre-Kahn A: A new surgical approach to the third ventricle with interruption of the striothalamic vein. Acta Neurochir (Wien) 47: 135-147, 1979. 79. Hitchon PW, Abu-Yousef MM, Graf CJ, Turner DM, VanGilder JC: Management and outcome of pineal region tumors. Neurosurgery 13:248253, 1983. 80. Hoffman HJ, Harwood-Nash D, Gilday DL: Per cutaneous third ventriculostomy in the manage ment of noncommunicating hydrocephalus. Neurosurgery 7:313-321, 1980. 81. Horsley V: Discussion of paper by Howell (82). Proc R Soc Med 3:77, 1910. 82. Howell CMH: Tumours of the pineal body. Proc R Soc Med 3:65-77, 1910. 83. International Symposium on Surgery in and around the Brain Stem and the Third Ventricle. Hannover, West Germany, February 18-23, 1985. 84. Jacobaeus HC, Nord F: Air and lipiodol as con trast agents for roentgen diagnosis within the central nervous system. Acta Radiol 3:367-382, 1924. 85. Jooma R, Kendal BE: Diagnosis and manage ment of pineal tumors. J Neurosurg 58:654665, 1983. 86. Kapp JP, Paulson GW, Odom GL: Brain tumors with tuberous sclerosis. J Neurosurg 26:191202, 1967. 87. Kelly PJ, Earnest F IV, Kall BA, Goerss SJ, Scheithauer B: Surgical options for patients with deepseated brain tumors: Computer-assisted stereotactic biopsy. Mayo Clin Proc 60:223-229, 1985. 88. Kelly PJ, Kall BA, Goerss S: Transposition of volumetric information derived from computed tomography scanning into sterotactic space. Surg Neurol 21:465-471, 1984. 89. Kelly PJ, Kall B, Goerss S, Alker GJ Jr: Precision resection of intra-axial CNS lesions by CT-based stereotactic craniotomy and computer monitored CO2 laser. Acta Neurochir (Wien) 68:1-9, 1983. 90. Kempe LG, Busch E: Clinical significance of the cisterna veli interpositi. Acta Neurochir (Wien) 16:241-248, 1967. 91. King TT: Removal of intraventricular craniopharyngiomas through the lamina terminalis. Acta Neurochir (Wien) 45:277-286, 1979. 92. Krause F: Chirurgie des Gehirns und Ruckenmarks, Vol 1. Berlin, Urban and Schwarzenberg, 1911 (cited by Pendl (110)). 93. Lavyne MH, Patterson RH Jr: Subchoroidal trans-velum interpositum approach to mid-third ventricular tumors. Neurosurgery 12:86-94, 1983. 94. Leksell L: A surgical procedure for atresia of the aqueduct of Sylvius. Acta Psychiatr Neurol 24:559-568, 1949. 95. Lespinasse VL: Cited by Davis L: Neurological
Surgery. Philadelphia, Lea & Febiger, 1936, p 405. 96. Long DM, Chou SN: Transcallosal removal of craniopharyngiomas within the third ventricle. J Neurosurg 39:563-567, 1973. 97. Long DM, Leibrock L: The transcallosal ap proach to the anterior ventricular system and its application in the therapy of craniopharyngioma. Clin Neurosurg 27:160-168, 1980. 98. Madrazo NI, Garcia-Renteria JA, Sandoval BM, Lopez Vega FJ: Intraventricular cysticercosis. Neurosurgery 12:148-152, 1983. 99. Matthews FD: Intraventricular craniopharyngioma. AJNR 4:984-985, 1983. 100. McCormick GF, Zee CS, Heiden J: Cysticercosis cerebri: Review of 127 cases. Arch Neurol 39: 534-539, 1982. 101. Milhorat TH, Baldwin M: Technique for surgical exposure of the cerebral midline: Experimental transcallosal microdissection. J Neurosurg 24:687-691, 1966. 102. Mixter WJ: Ventriculoscopy and puncture of the floor of the third ventricle. Boston Med Surg J 188:277-278, 1923. 103. Neuwelt EA (ed): Diagnosis and Treatment of Pineal Region Tumors. Baltimore, Williams & Wilkins, 1984. 104. Norlen G: Contribution to the surgical treatment of inoperable tumours causing obstruction of the sylvian aqueduct. Acta Psychiatr Neurol 24: 629-637, 1949. 105. Oppenheim H, Krause F: Operative Erfolge bei Geschwulsten der Sehhugel-und Vierhugelgegend. Berl Klin Wochenschr 50:2316-2322, 1913. 106. Palmer ED: Who was first? JAMA 239:16091610, 1978. 107. Papo I, Scarpelli M, Caruselli G: Intrinsic third ventricle craniopharyngiomas with normal pres sure hydrocephalus. Acta Neurochir (Wien) 23: 80-88, 1980. 108. Pecker J, Ferrand B, Javalet A: Tumeurs du troisieme ventricule. Neurochirurgie 12:7-136, 1966. 109. Pecker J, Scarabin JM, Vallee B, Brucher JM: Treatment in tumours of the pineal region: Value of stereotaxic biopsy. Surg Neurol 12:341- 348, 1979. 110. Pendl G: The surgery of pineal lesions—histori cal perspective, in Neuwelt EA (ed): Diagnosis and Treatment of Pineal Region Tumors. Bal timore, Williams & Wilkins, 1984, pp 139-154. 1 1 1 . Piatt JH Jr, Campbell GA: Pineal region meningioma: Report of two cases and literature review. Neurosurgery 12:369-376, 1983. 112. Powell MP, Torrens MJ, Thomson JLG, Horgan JG: Isodense colloid cysts of the third ventricle: A diagnostic and therapeutic problem resolved by ventriculoscopy. Neurosurgery 13:234-237, 1983. 113. Pudenz RH: The surgical treatment of hydro cephalus—an historical review. Surg Neurol 15:15-26, 1981. 114. Pussep L: Die operative Entfernung einer Zyste der Glandula pinealis. Neurol Zentralbl 33:560563, 1914. 115. Putnam TJ: Treatment of hydrocephalus by endoscopic coagulation of the choroid plexus: De-
116. 117.
118.
119.
120. 121. 122.
123. 124.
125. 126. 127. 128. 129.
130.
131.
132.
133. 134.
scription of a new instrument and preliminary report of results. N Engl J Med 210:1373-1376, 1934. Rand RW, Lemmen LJ: Tumors of the posterior portion of the third ventricle. J Neurosurg 10:118, 1953. Reichert FL, Dandy WE: Polyuria and polydipsia (diabetes insipidus) and glycosuria resulting from animal experiments on the hypophysis and its environs. Bull Johns Hopkins Hosp 58:418427, 1936. Rekate HL, Ruch T, Nulsen FE, Roessmann U, Spence J: Needle biopsy of tumors in the region of the third ventricle. J Neurosurg 54:338-341, 1981. Rhoton AL Jr, Yamamoto I: Operative ap proaches to the third ventricle, in Wilkins RH, Rengachary SS (eds): Neurosurgery. New York, McGraw-Hill, 1985, pp 938-951. Rhoton AL Jr, Yamamoto I, Peace DA: Microsur gery of the third ventricle: Part 2. Operative approaches. Neurosurgery 8:357-373, 1981. Rivas JJ, Lobato RD: CT-assisted stereotaxic as piration of colloid cysts of the third ventricle. J Neurosurg 62:238-242, 1985. Rout D, Sharma A, Radhakrishnan VV, Rao VRK: Exploration of the pineal region: Observa tions and results. Surg Neurol 21:135-140, 1984. Rozario R, Adelman L, Prager RJ, Stein BM: Meningiomas of the pineal region and third ven tricle. Neurosurgery 5:489-495, 1979. Ruggiero G, Bories J, Calabro A, Cristi G, Scialfa G, Smaltino F, Thibaut A: Radiological Explo ration of the Ventricles and Subarachnoid Space. New York, Springer-Verlag, 1974. Rush JL, Kusske JA, Porter RW, Pribram HW: Xanthogranulomas of the third ventricle. Neu rosurgery 4:329-333, 1979. Sano K: Pineal region tumors: Problems in pa thology and treatment. Clin Neurosurg 30:5991, 1983. Sano K: Microsurgery of tumors in the pineal region, in Rand RW (ed): Microneurosurgery. St Louis, CV Mosby, 1985, ed 3, pp 391-397. Sayers MP, Kosnik EJ: Percutaneous third ventriculostomy: Experience and technique. Childs Brain 2:24-30, 1976. Scarff JE: Endoscopic treatment of hydroceph alus: Description of a ventriculoscope and pre liminary report of cases. Arch Neurol Psychia try 35:853-860, 1936. Scarff JE: Treatment of obstructive hydroceph alus by puncture of the lamina terminalis and floor of the third ventricle. J Neurosurg 8:204213, 1951. Scarff JE: Treatment of hydrocephalus: An his torical and critical review of methods and re sults. J Neurol Neurosurg Psychiatry 26:1-26, 1963. Scarff JE: The treatment of nonobstructive (communicating) hydrocephalus by endoscopic cauterization of the choroid plexuses. J Neuro surg 38:1-18, 1970. Schmidek HH: Surgical management of pineal region tumors, in Schmidek HH (ed): Pineal Tu mors. New York, Masson, 1977, pp 99-113. Schmidek HH, Waters A: Pineal masses: Clinical
135. 136. 137.
138. 139. 140. 141.
142.
143. 144.
145.
146.
147.
features and management, in Wilkins RH, Rengachary SS (eds): Neurosurgery. New York, McGraw-Hill, 1985, pp 688-693. Schuster J: Ventriculographie mit Lipiodol Ascendens und Descendens. Klin Wochenschr 4:2064-2068, 1925. Shucart WA, Stein BM: Transcallosal approach to the anterior ventricular system. Neurosur gery 3:339-343, 1978. Sole-Llenas J, Royo Salvador M, Llovet J, Sanchez-Larrea R, Rovira RR: Craniopharyngioma of the third ventricle. Neurochirurgia 26:93-94, 1983. Stein BM: Surgical treatment of pineal tumors, Clin Neurosurg 26:490-510, 1979. Stein BM: Third ventricular tumors, din Neurosurg 27:315-331, 1980. Stern WE: Parasitic infestations, in Wilkins RH, Rengachary SS (eds): Neurosurgery. New York, McGraw-Hill, 1985, pp 2010-2015. Stookey B, Scarff J: Occlusion of the aqueduct of Sylvius by neoplastic and non-neoplastic processes with a rational surgical treatment for relief of the resultant obstructive hydrocephalus. Bull Neurol Inst NY 5:348-377, 1936. Suzuki J, Katakura R, Mori T: Interhemispheric approach through the lamina terminalis to tu mors of the anterior part of the third ventricle. Surg Neurol 22:157-163, 1984. Sweet WH: Radical surgical treatment of craniopharyngioma. Clin Neurosurg 23:52-79, 1976. Takei Y, Pearl GS, Fresh CB, Fleischer AS, Payne NS, Tindall GT: Intraventricular fiberscopic ob servations in adult hydrocephalic patients. Neu rosurgery 5:437-440, 1979. Thomson AF: Surgery of the third ventricle, in Walker AE (ed): A History of Neurological Sur gery. Baltimore, Williams & Wilkins, 1951, pp 136-149. Torkildsen A: Should extirpation be attempted in cases of neoplasm in or near the third ventri cle of the brain? Experiences with a palliative method. J Neurosurg 5:249-275, 1948. Trescher JH, Ford FR: Colloid cyst of the third ventricle. Arch Neurol Psychiatry 37:959-973, 1937.
148. Turnbull IM, Drake CG: Membranous occlusion of the aqueduct of Sylvius. J Neurosurg 24:2433, 1966. 149. Van Den Bergh R, Brucher JM: L'abord transventriculaire dans les cranio-pharyngiomes du troisieme ventricule: Aspects neuro-chirurgicaux et neuro-pathologiques. Neurochirurgie 16:51-65, 1970. 150. Van Wagenen WP: A surgical approach for the removal of certain pineal tumors. Surg Gynecol Obstet 53:216-220, 1931. 151. Van Wagenen WP: Surgery of the hypothalamic region. Res Publ Assoc Res Nerv Ment Dis 20:841-853, 1940. 152. Van Wagenen WP, Aird RB: Dilatations of the cavity of the septum pellucidum and cavum vergae. Am J Cancer 20:539-557, 1934. 153. Ventureyra ECG: Pineal region: Surgical man agement of tumours and vascular malforma tions. Surg Neurol 16:77-84, 1981. 154. Viale GL, Turtas S: The subchoroid approach to the third ventricle. Surg Neurol 14:71-76, 1980. 155. Volkmann J: Ueber Versuche zur unmittelbaren Besichtigung der Gehirnkammern (Enzephaloskopie). Munch Med Wochenschr 70:1382, 1923. 156. Volkmann J: Das Encephaloskop. Zbl Chir 51:1233-1234, 1924. 157. Walker AE: Diagnostic procedures, in Walker AE (ed): A History of Neurological Surgery. Balti more, Williams & Wilkins, 1951, pp 23-39. 158. Wilkins RH: Neurosurgical classic—XX. J Neu rosurg 21:586-635, 1964. 159. Winston KR, Cavazzuti V, Arkins T: Absence of neurological and behavioral abnormalities after anterior transcallosal operation for third ven tricular lesions. Neurosurgery 4:386-393, 1979. 160. Wood JH, Zimmerman RA, Bruce DA, Bilaniuk LT, Norris DG, Schut L: Assessment and man agement of pineal-region and related tumors. Surg Neurol 16:192-210, 1981. 161. Yamamoto I, Rhoton AL Jr, Peace DA: Microsurgical anatomy of the third ventricle, in Brock M (ed): Modern Neurosurgery 1. Berlin, Sprin ger-Verlag, 1982, pp 205-214. 162. Zulch KJ: Reflections on the surgery of the pineal gland (a glimpse into the past). Neurosurg Rev 4:159-162, 1981.
2 Comparative Anatomy of the Third Ventricular Region E.Leon Kier,M.D. What is often more condensed or more concealed in one species Nature displays more clearly and openly in another. De Graaf (1664) quoted by Cole (1944)
Lateral Walls Hypothalamus Thalamus (Massa Intermedia) Intraventricular Foramen (Monro) Floor Optic Chiasm Optic Recess Infundibular Recess Saccus Vasculosus Tuber Cinereum Hypophyseal Recess Hypophysis Mammillary Bodies Premamillary Recess Postmamillary Recess Midbrain Roof and Posterior Wall Paraphysis Velum Transversum Dorsal Sac (Saccus Dorsalis) Suprapineal Recess Parencephalon Habenulae Pineal (epiphysis) Parietal Eye Pineal Recess Parapineal Posterior Commissure Tela choroidea (Velum Interpositum) Optic Lobes (Corpora Bigemina) Auditory Lobes Superior Colliculi Inferior Colliculi Quadrigeminal Plate (Corpora Quadrigemina) Mesencephalic Ventricle Optic Ventricle Aqueduct Aqueductal Ampulla
Commissures Limbic System Lamina Terminalis Anterior Commissure Paraterminal Body Septal Area Hippocampal Commissure Hippocampus Fornix Corpus Callosum Subiculum Indusium Griseum Longitudinal striae Septum Pellucidum Cavum Septi Pellucidi Arteries Principles of Cerebral Vascular Development Delivering Arteries End Arteries Recurrent Artery of Heubner Perforating end arteries of Hypothalamus and Thalamus Arterial Circle (Willis) Anterior Cerebral Anterior Choroidal Posterior Cerebral Posterior Choroidal Veins
Comparative anatomical knowledge can enhance and enliven the study of neuroanatomy and neuroradiological anatomy (11). The study of developmental and adult neuroanatomy is thus transformed from memorizing a large number of anatomical structures to a meaningful understanding of many features of the human brain (32, 33). This meaningful understanding is based on the slowly evolving needs of different vertebrates manifested in various evolutionary modifications of the central nervous system. The study of comparative anatomy permits a relatively slow longitudinal study of the developing human central nervous system. This slower pace is not available in human embryos that "recapitulate" many stages very quickly and transiently (13). In addition, many congenital, metabolic, toxic, and degenerative disorders may be viewed through the perspective of evolution (67). Certain disorders such as developmental hypoplasias involve new phylogenetic structures while sparing the older phylogenetic section of the nervous
Figure 2.2. Midsagittal section of the third ventricle. The floor extends from the optic chiasm to the aqueduct. The anterior wall extends from the optic chiasm to the foramen of Monro (interventricular foramen). The roof extends from the foramen of Monro to the suprapineal recess. The posterior wall extends from the suprapineal recess to the aqueduct. (From Yama-moto I, Rhoton AL, Pierce RA: Microsurgery of the third ventricle: Part 1. Microsurgical anatomy. Neu-rosurgery 8:333-356, 1981.)
Figure 2.1. Diagram of the medial aspect of the left cerebral hemisphere demonstrating the majority of the structures of the limbic system. Note that many of these complex structures are part of the anatomy of the third ventricular region. From Warwick R, Williams PL (eds): Gray's Anatomy. Philadelphia, WB Saunders Co, 1980, British ed 36.
system. In contrast, anomalies of the central nervous system such as holoprosencephaly and Dandy-Walker selectively affect the oldest phy-logenetic structures while sparing the newer ones. The merit of a comparative anatomical study is particularly apparent in the elucidation of the complex anatomy of the commissures in the region of the third ventricle (Fig. 2.1). In addition the various forms of corpus callosum dysgenesis are clarified by the knowledge of its evolutionary changes. In this chapter an effort was made to group the various structures according to the usual anatomical and neurosurgical landmarks of the third ventricle (83, 89) (Fig. 2.2). However, the roof and posterior wall structures are grouped together because in many vertebrates the posterior wall cannot be defined as a separate structure. The complex evolution of the commissures and cer-
tain principles governing vascular evolutionary changes are described in separate sections. Lateral Walls Phylogenetically the hypothalamus is probably the oldest cerebral structure with remarkably similar nuclear differentiation and fiber connections in all vertebrates (67). Evidence of neurose-cretory activity is present even in the simplest invertebrates. In certain worms the entire brain could be considered to be a form of the hypothalamus. Internal neurosecretory control preceded the evolution of special sensory and locomotor systems relating the organism to the outside world. Thus the phylogenetic variability of the evolving hypothalamus is manifested in the variability of the floor of the third ventricle. The evolving thalamus affects the topography of the phylogenetically changing lateral ventricles and commissures (33). In addition, the evolv-
Figure 2.3. Ventrodorsal roentgenograms of a shark (A) and a frog (B) and the brain of an iguana (C), rabbit (D), dog (E), and monkey (F) with barium-cast ventricles. Except the shark, all other specimens demonstrate a small interven-tricular foramen. The wider thalamic part is superimposed on the narrower hypothalamic part of the third ventricle. In the shark and the frog the lateral ventricles are anterior to the interventricular foramen. In the iguana the posterior pole of the lateral ventricle is now posterior to the interventricular foramen (arrow). In the rabbit the major part of the lateral ventricle is posterior to the interventricular foramen. From Kier EL: The cerebral ventricles: a phylogenetic and ontogenetic study. In Newton TH, Potts DG (eds): Radiology of the Skull and Brain, vol. 3, Anatomy and Pathology, St. Louis, The CV Mosby Co, 1977, pp 2787-2914.
ing thalamus and lateral geniculate body contribute to the phylogenetic changes involving the anterior and posterior choroidal arteries (32). The interthalamic adhesion (massa intermedia) varies in size, and the hypothalamus and thalamus create minor differences in the lateral walls of the third ventricle. Except for the frog and iguana, the vertebrates studied demonstrate that the hypothalamic part of the third ventricle is narrower than the thalamic part (Fig. 2.3). In addition, the evolution of the cerebrum changes the topographic relationship of the lateral ventricles and the intraventricular foramen (Fig. 2.3).
In the shark and the frog the lateral ventricles are anterior to the intraventricular foramen. In the iguana the posterior pole of the lateral ventricle is posterior to the intraventricular foramen. In the rabbit the major part of the lateral ventricle is posterior to the intraventricular foramen. The comparative study demonstrated that all specimens except the shark had small interven-ticular foramina (Fig. 2.3). It seems that the corpus striatum is the major structure reducing the size of the interventricular foramen and not the evolving fornix and thalamus. The latter structures are not present in the frog or iguana.
Floor An optic chiasm is present in the floor of the diencephalon of all vertebrates except bony fish. In the latter the decussation of the optic nerves is between the eyes and the diencephalon (25). The optic chiasm is a major modifier of the floor of the third ventricle because its size and location affect the size and configuration of the optic and infundibular recesses (Figs. 2.4 to 2.11). The number of optic nerve fibers, the extent of myelinization, and the degree of decussation in the optic chiasm vary among the different vertebrates (59, 67). The following examples demonstrate the extensive variability that exists in the number of optic nerve fibers present and the importance and development of the visual system: lamprey, 5,217; frog, 29,000; turtle and alligator, 105,000; shark, 113,000; cat, 119,000; dog, 154,000; rabbit, 265,000; sheep, 649,000. In some primates and birds in which sight is the dominant sense, the number of optic nerve fibers reaches or exceeds 1 million. In the less developed vertebrates the optic nerve fibers are poorly myelinated. In advanced teleosts, birds, and most
terrestrial vertebrates the majority of fibers are heavily myelinated. The presence of myelin seems to be associated with improved visual function. Nearly all optic nerve fibers in nonmammalian vertebrates and in some lower vertebrates decussate in the optic chiasm. The crossing of the optic fibers at a primoridal chiasm was an early evolutionary development. The presence of axons that crossed between the epithelium of the primitive photoreceptor—the precursor of the eye— and the effector system of the opposite side of the body permitted the primitive prevertebrate to move away from a threatening photic stimulus. The evolution of the two eyes in the non-mammalian vertebrates, each with its own monocular field of vision, is postulated to have necessitated the decussation of the optic fibers. Without the decussation the segments of a visualized object would be incompatibly recombined in the brain. In mammals the number of crossing optic nerve fibers is reduced progressively. This reduction is associated with the changing axis of vision and the development of binocular fields of view. In rodents, with laterally placed eyes and
Figure 2.4. Sagittal section of the brain of a shark with barium-cast ventricles. The roof of the third ventricle and its recesses are not covered by the cerebrum. The anterior commissure is present. The dorsal sac is the homologue of the suprapineal recess in the human. The infundibular recess in the shark is the ventricular extension into the inferior lobes of the hypothalamus. Note the large optic ventricle. Only the proximal end of the pineal recess is seen.
minimal overlapping of the monocular fields of vision, only a small number of fibers do not decussate. At least one-third of the fibers are nondecussating in carnivores and primates, with frontally placed eyes and overlapping fields of view. In primates nearly half the fibers do not decussate. This degree of nondecussation is one of the important factors that allow two identical retinal images to be transferred to the brain, producing the highly evolved stereoscopic vision of primates. All of these factors affect the size and shape of the optic chiasm and result in variation of the optic and infundibular recess. The optic recess seems to be a residual space on the floor of the
third ventricle whose formation, shape, and size are the result of the varying indentation of the ventricular floor by the chiasm (Figs. 2.5, 2.7 to 2.10, and 2.13 to 2.17). The infundibular recess, however, is a true extension of the third ventricle that undergoes evolutionary modifications, ranging from that seen in the shark to that seen in mammals. The infundibular recess is a funnelshaped extension of the third ventricle extending through the hypothalamus to end in the hypophysis. The infundibular portion of the hypothalamus and its lumen, the infundibular recess, vary greatly among the different vertebrates (Figs. 2.4 to 2.20). In the lamprey, which is without an infun-
Figure 2.5. Sagittal section of the brain of a frog with barium-cast ventricles. The roof of the third ventricle and its recesses are not covered by the cerebrum. The anterior and hippocampal commissures have developed in the paraterminal body. The infundibular recess is incompletely filled, and its full extent is seen in Figure 2.13. The mesencephalic and optic ventricles are no longer a single large cavity.
Figure 2.6. Sagittal section of the brain of an iguana with barium-cast ventricles. The posterior pole of the cerebrum overlaps the roof of the third ventricle except for the dorsal sac—the homologue of the suprapineal recess. As a result the paraphysis, velum transversum, and dorsal sac are angled posteriorly before reaching the dorsal surface of the brain. The pineal body is large. Note the hippocampal commissure dorsal to the anterior commissure. The cast of the large infundibular recess was partially damaged during dissection. It extended to fill the area outlined by the arrows. 45
Figure 2.7. Sagittal section of the brain of a rabbit with barium-cast ventricles. The enlarged cerebrum covers the entire third ventricle and superior col-liculus. A new commissure, the corpus callosum, is present. Note the junctional zone (crossed arrow) where the posterior margin of the corpus blends into the subiculum dorsal to the hippocampus. The septum pellucidum area is still small. The large suprapineal recess was damaged during dissection. The direction of the long axis of the suprapineal recess changes from the iguana to the rabbit. Note the recesses of the superior and inferior colliculi, the remnants of the optic ventricle. dibulum, the homologue of the neural lobe of the hypophysis is a simple plate of tissue that lies on the floor of the diencephalon (64). In the various fishes the hypothalamus is highly developed and extends ventrally to the midbrain to form two distinct structures, the inferior lobes of the hypothalamus (Fig. 2.21). The prominent inferior lobes of the hypothalamus in the shark surround a large ependymal-linked infundibular recess (Figs. 2.4 and 2.12). An additional unique structure in fishes is the saccus vasculosus, a highly vascularized vesicle with neurosensory cells whose lumen is continuous with the infundibular recess (Figs. 2.4 and 2.12). The saccus is largest in deep sea fishes,
small in shallow water fishes, and absent in amphibians and other terrestrial vertebrates. Its function probably relates to perception of the pressure of intraventricular fluid relative to the depth of water and influences the size of the swim bladder (67). In amphibians and reptiles the hypothalamus is a relatively small structure consisting largely of the tuber cinereum (55) (Figs. 2.22 and 2.23). The small size is a result of the absence of the saccus vasculosus and the limited development of gustatory impulses (30). The largest hypothalamus is present in teleosts and mammals. In mammals the tuber cinereum may be divided into median and lateral eminences. Of note is
Figure 2.8. Sagittal section of the brain of a cat with barium-cast ventricles. The corpus callosum is further developed, especially posteriorly, where the sple-nium overlaps the suprapineal recess. Note the anteroposterior axis of the supra-pineal recess and its curvature around the splenium (arrouj). The fornix is elongated and arched under the corpus callosum. The corpus callosum is thin. The presence of the premamillary and postmamillary recesses relates to the variable development of the mamillary bodies.
the prominent size of the infundibular recess in many nonhuman vertebrates (Figs. 2.4 to 2.7). The hypophyseal recess in the cat (Fig. 2.16) is of interest because it represents the residual cavity of the infundibular recess within the developing neurohypophysis. The hypophyseal recess has been observed by Levinger and Edery (42) and apparently is a frequent finding in adult cats (35). Only two cases of persistence of the infundibular recess in the adult human have been reported (10, 35). The varied appearance of the mammalian infundibular recess and the presence of premamillary and postmamillary recesses are manifestations of the varied development of the components of the hypophysis and the size of the mamillary bodies (Fig. 2.9). In the cat and the dog the pars distalis of the hypo-
physis, originating from Rathke's pouch, extends much more caudally around the pars neur-alis than in the human (56). The pars intermedia, which in the human is limited to the rostral aspect, surrounds the pars neuralis completely in the dog and cat. The mamillary body is probably premordial in teleosts and amphibians. It can be identified in reptiles, birds, and mammals. In the rat it is a band of gray matter continuous in the midline. In humans the mamillary body is relatively larger than in other primates. The posterior inferior recess—demonstrated by Westergaard (84) in the guinea pig, the hamster (85), and the rat (86)—seems to be a manifestation of the previously mentioned varied development of the hypophysis and mamillary body.
Figure 2.9. Sagittal section of the brain of a dog with barium-cast ventricles. The corpus callosum is larger. Note the hippocampus, which is present ventral to the splenium and fornix. The suprapineal recess, the pineal body, and the colliculi are in a more ventral position.
Figure 2.10. Sagittal section of brain of a monkey with barium-cast ventricles. The corpus callosum is increased in size. The septum pellucidum is enlarged. The cavum septi pellucidi is partially filled with barium. The pineal body is wedged between the splenium and the corpora quadrigemina. The smaller suprapineal recess is associated with the larger splenium. The optic chiasm is large.
Figure 2.11. Sagittal section of the brain of adult human. The corpus callosum is markedly enlarged and arched. The splenium is posterior to the thalamus. Note the relative reduction in the size of the quadrigeminal plate.
Figure 2.12. Lateral roentgenograms of a shark (A) and of the brain of a shark (B) with barium-cast ventricles. The ventricular system has a linear arrangement. The intraventricular foramen is large. Contrast (arrow) has leaked into the ventral aspect of the medulla.
Figure 2.13. Lateral roentgenograms of a frog (A) and of the brain of a frog (B) with barium-cast ventricles. The ventricular system is still linear with the entire lateral ventricle anterior to the interventricular foramen, which is small.
Figure 2.14. Lateral roentgenograms of an iguana (A) and of the brain of an iguana (B) with barium-cast ventricles. The linearity of the ventricular system is no longer present and the lateral ventricles are dorsal to the third ventricle. Note the molding of the lateral ventricle with the presence of frontal and temporal horns.
Figure 2.15. Lateral roentgenograms of a rabbit (A) and of the brain of a rabbit (B) with barium-cast ventricles. Note the large suprapineal recess and massa intermedia. In the midbrain of the shark, an acoustico-lateral line and trigeminal fibers, nuclei of the trochlear nerve, and a few fiber tracts are present. In amphibians the midbrain is less devel-
oped than in bony fish. In reptiles, midbrain structures are demonstrated, suggestive of mammalian differentiation. The reticular formation in reptiles has changed to a nuclear mass that
Figure 2.16. Lateral roentgenograms of a cat (A) and of the brain of a cat (B) with barium-cast ventricles. The suprapineal recess is in a more caudal position as a result of enlargement of the splenium of the corpus callosum. The central cavity of the hypophysis is filled via the infundibular recess.
Figure 2.17. Lateral tomographic roentgenogram of a monkey (A) and a roent-genogram of the brain of a monkey (B) with barium-cast ventricles. Note the small size and caudal position of the suprapineal recess. The optic recess is very small because of the large chiasm and was visualized only by tomography.
Figure 2.18. Ventral views of the brains of shark (A), frog (B), rabbit (C), cat (D), and sheep (£) demonstrating evolutionary modifications in the relationship of the chiasm, hypothalamus, and hypophysis. The hypothalamus is no longer visible in the cat and sheep.
constitutes a primordial red nucleus. In mammals the red nucleus becomes a prominent structure. The cerebral peduncles increase progressively as the corticopontine, corticobulbar, and corticospinal tracts enlarge during cerebral evolution. The substantia nigra also enlarges progressively in mammals. The foregoing evolutionary changes are maniFigure 2.20. Diagram of the third ventricle and neighboring structures of a lizard to show the various structures of the
Figure 2.19. Schema to show probable evolutionary changes in the vertebrate pituitary gland. Arrow extending from the median eminence to the pars distalis represents the hypophyseal portal system. CAU, caudal division of the avian interior lobe; CEP, cephalic division of the same; PPD, proximal pars distalis; RPD, rostral pars distalis; VL, ventral lobe of the elasmo-branch pituitary; SV, saccus vasculosus. (From Romer AS: The Vertebrate Body. Philadelphia, WB Saunders Co, 1970, ed 4.)
roof including the dorsal (parietal) eye. (From Romer AS: The Vertebrate Body. Philadelphia, WB Saunders Co, 1970, ed 4).
Figure 2.21. Lateral view of the brain of a shark demonstrates the major subdivisions of the brain. The roof of the diencephalon is visible and is interlaced by a prominent venous plexus.
Figure 2.22. Lateral view of the brain of a frog shows the linearity of the major subdivisions of the brain. The forebrain (cerebrum) demonstrates a small cerebral hemisphere (arrow), which begins to overlap the diencephalon. The relatively small hypothalami of the amphibians and reptiles consist largely of the tuber cinereum.
Figure 2.23. Lateral view of the brain of an iguana. The linear arrangement of the subdivisions of the brain is no longer present. The forebrain (cerebrum) is increased in size and dorsally overlaps the diencephalon. Only the ventral hypo-thalamus and hypophysis are still visible.
fested in the increased size of the midbrain ventral to the mesencephalic ventricle and aqueduct. In the shark and the frog the floor of the midbrain is sagittally smaller than either the optic lobes or the spinal cord (Figs. 2.4 and 2.5). In the iguana
the floor of the midbrain is enlarged and is greater than the optic lobes in sagittal height (Fig. 2.6). The mammalian midbrain is greatly increased in size and is even larger sagittally than the spinal cord (Figs. 2.7 to 2.11).
Roof and Posterior Wall In the shark (Figs. 2.4 and 2.12) the following structures are identifiable: paraphysis, velum transversum, dorsal sac, pineal recess, and ha-benular and posterior commissures. The paraphysis and velum transversum are choroid plexus structures (73). The paraphysis is an extraventricular choroid plexus. In the shark it is a dorsal evagination with origin in the anterior end of the roof of the third ventricle (Fig. 2.4). The paraphysis curves around the posterior pole of the cerebral hemisphere and its cavity communicates freely with the third ventricle (Fig. 2.4 and 2.12). The velum transversum is a modified choroid plexus that originates in the roof of the third ventricle, projects into the ventricle, and serves as an arbitrary border between the diencephalon and the telencephalon (Fig. 2.4). In the frog the paraphysis is a prominent extraventricular structure (Fig. 2.5). It is cone-shaped and protrudes dorsally from the anterior end of the roof of the third ventricle (73). The numerous villi of the intraventricular choroid plexus are attached to the roof of the third ventricle around the narrow orifice of the paraphysis. These villi are seen as filling defects in casts of the ventricles (Figs. 2.5 and 2.13). A distinct velum transversum does not exist in the frog (28). The dorsal sac (saccus dorsalis) is the thin-walled roof of the third ventricle in front of the habenulae. It is also called the parencephalon (30). Whether the term parencephalon includes the velum transversum and paraphysis is not clear. The dorsal sac is a large, distensible structure in the shark (Figs. 2.4 and 2.12). The present investigation did not reveal a dorsal sac in the frog. This absence may relate to the relatively thick roof of the third ventricle, which seems to be secondary to the habenular commissure. Herrick (24) demonstrated the habenular commissure as forming the roof of the third ventricle between the paraphysis and the posterior commissure. Comparative investigation suggested that the dorsal sac in the shark is the homologue of the human suprapineal recess (33). The pineal and the dorsal sac—suprapineal recess—maintain their adjacent relation from the shark to the human. What does change is the direction of the long axis of these two structures because of overlapping of the third ventricle by the cerebrum and corpus callosum. In the shark the dorsal sac has a ventrodorsal axis (Fig. 2.4). In the iguana
the proximal end of the dorsal sac is angled posteriorly by the posterior pole of the enlarged cerebrum (Fig. 2.6). The axis of the large supra-pineal recess in the rabbit is more posteriorly directed (Fig. 2.7 and 2.15). In the cat and the dog, because of the great enlargement of the corpus callosum, the long axis of the suprapineal recess has an anteroposterior direction (Figs. 2.8, 2.9, and 2.16). The shape and size of the suprapineal recess are related to the size of the corpus callosum. The increasing size of the splenium, as seen from the cat to the monkey, is associated with a concomitant reduction of the suprapineal recess (Figs. 2.8 to 2.10, 2.16, and 2.17). The habenulae are relatively large, paired structures in vertebrates with a small forebrain (67). The habenulae are especially developed in vertebrates that rely heavily on olfaction, such as sharks and bloodhounds. In humans, because of the development of the neocortex and the relative unimportance of olf action, the habenulae are the smallest. The habenulae may have an important function in neonatal feeding reactions. An unusual feature in the central nervous system is the asymmetry of the habenulae. In sharks and frogs the left habenula is larger. In birds and mammals, habenular symmetry is variable. The pineal gland in the shark is a long and slender structure whose distal end is situated in a depression in the cartilaginous brain case (Fig. 2.4). In the frog the pineal is a long slender nerve structure extending from the roof of the diencephalon through the parietal foramen in the roof of the skull. It terminates in the frontal organ, situated under the calvarial skin (80). The frontal organ in the frog is a poorly differentiated parietal eye (67). Jolie (28) and van de Kamer (80) demonstrated a small pineal cavity situated within the diencephalon of the frog. This location may explain why the pineal recess was not seen in the studied specimens. In the iguana the striking evolutionary changes of the reptilian brain are demonstrated. The cerebrum has increased to such a degree that it completely overlaps the roof of the third ventricle (Figs. 2.6, 2.24, and 2.25). Only the distal end of the dorsal sac and the dorsal tip of the pineal gland remain visible on the dorsal aspect of the brain. In the shark the paraphysis curves anteriorly and the dorsal sac has a straight dorsal course (Fig. 2.4). In the iguana these two structures are angled posteriorly around the posterior
Figure 2.24. Lateral view of the brain of a rabbit. In the rabbit the neopallium of the cerebrum has enlarged to such a degree that it dorsally overlaps the diencephalon and mesencephalon. None of the diencephalic structures are visible. pole of the cerebral hemisphere (Fig. 2.6). Of note is the massive size of the pineal. The pineal undergoes complex evolutionary changes closely related to changes of the parietal eye (also called the median or dorsal eye). A median eye situated on the forehead and directed dorsally was apparently present in ancestral bony fishes, amphibians, and reptiles (64) (Fig. 2.20). Some of these ancestral forms may have had two median eyes. In modern vertebrates the parietal eye is present in the lamprey, some fishes and lizards, and the sphenodon (a surviving Mesozoic lizard) (64). Although a miniature cornea and retina may be present, the parietal eye—covered by connective tissue and epithelium—is capable of little except detection of light. Romer postulated that the ancestral vertebrates were presumably bottomdwelling mud strainers in which the parietal eye provided vital protective information (64). Vertical vision lost its importance for the more active modern vertebrates. It also has been theorized that the parietal eye-pineal complex was associated with thermoregulation (63).
The parietal eye, when present, develops from the distal end of the pineal body. Since the pineal originates as an evagination from the roof of the diencephalon, the parietal eye, like the lateral eyes, is essentially a structure of the brain. Occasionally the parietal eye develops from another vesicular structure adjacent and similar to the pineal, the parapineal organ. In the sphenodon and modern lizards the parapineal organ is the functional eye. In the lamprey both the pineal and the parapineal form parietal eyes, of which the pineal eye is dominant. The complex and controversial anatomy and terminology of the pineal and parapineal bodies and the parietal eye are reviewed by Oksche (50). The lumen of the pineal gland, the pineal recess, becomes progressively reduced in the ascending vertebrate series. This reduction results from cellular proliferation and leads to the compact glandlike structure in mammals. In some bony fish and turtles the lumen of the pineal does not communicate with the third ventricle. The pineal is absent in certain vertebrates, including the
Figure 2.25. Dorsal views of brains demonstrate evolutionary changes at the roof of the third ventricle. A. Shark. The highly vascular roof of the third ventricle (diencephalon) is uncovered. Note that the veins of the cerebrum are anterior to the diencephalic veins.
crocodile and some whales and dolphins. The external form of the pineal body varies greatly among the different mammals (30, 50, 63, 81, 82). It is elongated in guinea pigs, club-shaped in rabbits, round in goats, conical in carnivores, and pear-shaped in cattle and humans (Figs. 2.6 to 2.10). The mechanism of formation of the double-layered tela choroidea (velum interpositum) in the human (7, 17) is diagrammed in Figure 2.26. This process is clarified when considered in evolutionary terms, as demonstrated by the shark, frog, iguana, and rabbit (Figs. 2.21 to 2.25). In the shark the linear arrangement of the
major subdivisions of the brain is demonstrated (Fig. 2.21). The cerebrum is anterior to the diencephalon. The entire roof of the diencephalon is visible and is comprised of a single layer of pia mater (Figs. 2.21 and 2.25A). The linear arrangement of the brain is still present in the frog (Figs. 2.22 and 2.25B). A rudimentary cerebral hemisphere, however, is beginning to overlap the diencephalon (Fig. 2.22). The pronounced changes of the reptilian brain are shown in Figure 2.23. The linear arrangement of the subdivision of the brain is no longer present. The cerebrum has increased to such a degree that it overlaps the diencephalon, which is no longer
Figure 2.25B. Frog. The roof of the third ventricle is still uncovered. visible (Figs. 2.23 and 2.25C). In the rabbit the cerebrum overlaps the diencephalon and mes-encephalon (Fig. 2.24). This overlapping produces a double-layered tela choroidea. The human embryological changes that result in the formation of a double-layered tela choroidea are an ontogenic "recapitulation" of the phylogenetic process previously described. In their anatomical and radiographic studies, Larroche (36), Larroche and Baudey (37), and Baudey-Pasquier and Larroche (6) considered the
cavum vergae, which is situated below the corpus callosum and above the fornix, as the posterior continuation of the cavum septi pellucidi. In the material of Larroche, a cavum vergae was never present without a cavum septi pellucidi (6, 37). The cavum vergae seemed to develop ontogenetically later than and disappeared before the cavum septi pellucidi. The disappearance of the cavum vergae by the time of birth may be explained by several factors demonstrated in a comparative study (33). The progressive enlargement of the splenium as seen in the
Figure 2.25C. Iguana. The forebrain (cerebrum) is increased in size and overlaps the roof of third ventricle, which is no longer visible. The only diencephalic structures visible are the distal ends of the dorsal sac and pineal gland. rabbit, cat, dog, and monkey (Figs. 2.7 to 2.10) is "recapitulated" in the human fetus (Fig. 2.27). Kappers et al. (30) also noted that the posterior thalamus enlarges progressively in the ascending phylogenetic series. The thalamic enlargement increases the dorsal arching of the fornix. All of these factors would contribute to the obliteration of the cavum vergae in the older human fetus while the cavum septi pellucidi is maintained.
A comparative study demonstrated that the human aqueduct is a residual cavity that results from phylogenetic modifications of the walls surrounding the ventricular system of the midbrain (33). The roof (tectum) of the midbrain is the primary visual center in nonmammalian vertebrates. The presence of the visual center in the tectum would seem to have been the reason for the evolution of this region into a major associa-
Figure 2.25D. Rabbit. The cerebrum covers both the diencephalon and mesen-cephalon.
tion center (64). The tectum in fishes and amphibians is the dominant brain center; it associates visual, olfactory, lateral line, and other somatic sensory stimuli with the motor columns of the brain stem and cord. In the lamprey the nerve structures of the eye are developed and the optic tract ends in a discernible tectum. Although the tectum is still partly ependymal, it already has enlarged into a pair of recognizable optic lobes. The two optic lobes (corpora bigemina) are the homologues of the superior colliculi. In the shark and the frog the optic lobes are prominent structures (Figs. 2.4, 2.5, 2.21, and 2.22). In the iguana the cerebrum is beginning to
become the dominant structure and the optic lobes are relatively reduced in size (Figs. 2.6 and 2.23). In some reptiles a pair of auditory lobes, the homologues of the inferior colliculi, are present caudal to the optic lobes. The auditory lobes, which receive fibers from the cochlea, are not present in fishes that lack a cochlea. In amphibians the cochlea is small and the auditory center is not large enough to form a bulge on the surface of the mesencephalon (31). At some critical evolutionary stage a supraseg-mental area, more elaborate than the optic lobes, became necessary for the integration of refined visual perceptions (79). The evolution of the cerebrum provided a more elaborate suprasegmen-
Figure 2.26. Diagram to demonstrate the formation of the double layer of the tela choroidea (velum inter-positum) in the human. 1. The developing neural tube includes the forebrain (FB), which is surrounded by primitive pia mater (broken line). 2 and 3. The expanding vesicle of the cerebral hemisphere (HV) carries its own pia mater (b), which overlaps the pia mater (a) of the third ventricle. IF, interventricular foramen. 4. A double layer of pia is interposed between the two cerebral vesicles and the roof of the third ventricle (3V). E, ependyma. 5. The two layers of pia (b, a) over the roof of the third ventricle persist after connection of the hemispheres by the commissures. This process "recapitulates" the phylogenetic overlapping of the diencephalon by the cerebrum demonstrated in Figures 2.12 to 2.17 and 2.21 to 2.25. (From Brash JC (ed): Cunningham's Textbook of Anatomy. New York, Oxford University Press, 1951, ed 9. Modified from Frazer JE: A Manual of Embryology. New York, William Wood & Co, 1932.) tal region capable of integrating refined visual perceptions such as color, form, size, and detection of motion and distance. In mammals most of the visual, auditory, and other somatic sensations—instead of being integrated in the mid-brain—are relayed via the thalamus to the cerebral hemispheres. The thalamus in nonmammalian vertebrates is a smaller anterior extension of the sensory association area of the mid-brain. It reaches its maximal development in mammals as a result of its function as a relay center to the association centers in the cerebral hemispheres. In mammals the optic lobes differentiate into the superior and inferior colliculi (corpora quadrigemina). The cephalic shift in mammals is demonstrated in the inverse relationship in the sizes of the cerebrum and corpora quadrigemina. The increase in the size of the cerebrum from the rabbit to the human is associated with a relative reduction in the size of the corpora quadrigemina (Figs. 2.7 to 2.11). Furthermore, the relative sizes of the superior and inferior colliculi also change. In the rabbit the superior colliculi are much larger than the
inferior colliculi (Fig. 2.7), in the cat the difference between collicular sizes is diminished (Fig. 2.8), and in the dog and monkey the superior and inferior colliculi are of approximately equal size. Previously the superior colliculi were considered to be exclusively an optic structure and thus the homologue of the optic lobes. The inferior colliculi were considered to be auditory centers that developed as a result of the evolution of the hearing function of the ear. This interpretation of collicular function has been modified as other connections than purely visual and auditory have been demonstrated in the colliculi (83). The human fetus demonstrates a "recapitulation" of the interrelationship in the sizes of the tectum and cerebrum (Fig. 2.27). As the cerebrum enlarges, particularly the occipital lobe, the large tectum is transformed into the relatively smaller corpora quadrigemina. The ventricular system of the midbrain, from the fish to the primate, undergoes a pronounced transformation. In the shark the walls of each optic lobe surround a large optic ventricle, a dorsal expansion of the mesencephalic ventricle (Figs. 2.4 and 2.12). The mesencephalic and optic ventricles form a large continuous cavity. In the frog the ventricular system of the mid-brain is smaller as a result of increased thickness of the midbrain walls (tegmentum) (Figs. 2.5 and 2.13). A constricted communication demarcates the dorsolateral optic ventricles from the mesencephalic ventricle. In the iguana, as a result of the enlarged tegmentum and peduncles, the mesencephalic ventricle joins the third ventricle at a more acute angle than in the frog (Figs. 2.6 and 2.14). In the rabbit the mesencephalic ventricle demonstrates some of the features of the definitive aqueduct. As a result of the enlargement of the tegmentum and peduncles, the floor assumes a dorsally convex curve. The rostral end of the ventricle, molded by the superior colliculis and tegmentum, has the configuration of the anterior part of the aqueduct (Figs. 2.7 and 2.15). The recesses of the superior and inferior colliculi are the remnants of the optic ventricles. The definitive aqueductal configuration may be seen in the cat (Figs. 2.8 and 2.16). The aqueductal ampulla, the widest segment of the mammalian aqueduct, is the remnant of the mesencephalic and optic ventricles. Further minor modifications in the contours of the mammalian aqueduct are the result of enlargement of the cerebellum, tegmentum, and peduncles associated with reduction in size of the corpora quadrigemina and increased angulation of the brain (46).
Figure 2.27. Sagittal sections of human fetal heads demonstrate the "recapitulating" change in the relative size of the midbrain tectum and cerebrum. As the cerebrum enlarges, particularly the occipital lobe, the large tectum is transformed into the relatively smaller corpora quadrigemina. Also note the enlarging corpus callosum. A. 11 weeks. The posterior margin of the thin occipital lobe does not extend posteriorly to the relatively large midbrain. Note the size of the tectum relative to cerebellar size. From Kier EL: The cerebral ventricles: A phylogenetic and ontogenetic study. In Newton TH, Potts DG (eds): Radiology of the Skull and Brain, Vol. 3, Anatomy and Pathology. St. Louis, The CV Mosby Co, 1977, pp 2787-2914.
Figure 2.27B. 15 weeks. The occipital pole of the cerebrum extends posteriorly between the tectum and the skull. The tectum is still larger than the cerebellum.
Figure 2.27C. 24 weeks. The occipital pole is behind the cerebellum, which is now larger than the tectum.
Figure 2.27D. 30 weeks. Note the relatively small corpora quadrigemina. The splenium is well developed. A cavum vergae is present between the corpus callosum and the fornix. Commissures The literature dealing with the evolution of the commissures is not always easy to understand. Occasionally the descriptions and terminology are conflicting. In this section, I describe aspects relevant to the third ventricle. The commissures are part of the limbic system (lobe). Many of the structures of the limbic system (Fig. 2.1) are phylogenetically old and are topographically interposed between the dien-cephalon and the neocortex of the cerebral hemispheres (83). The fibers of the anterior commissure in the human connect with the olfactory tracts and the piriform, prepiriform, and amygdaloid bodies. The fornix is the efferent fiber system of the hippocampi, and the corpus callosum joins the neopallial regions of the cerebral hemispheres. The location, size, and interrelation of the anterior commissure, fornix, and corpus callosum are understood best in light of the evolutionary modifications of the cerebrum.
The cerebrum first evolved as an olfactory organ. In the lamprey the primordia of the cerebral hemispheres, manifested by the presence of bilateral vesicular outpouchings of the forebrain, are united by the lamina terminalis. Cells from the two adjacent medial portions of the olfactory lobes invaded the lamina terminalis. Thus the two olfactory nuclei were connected by a thickened lamina terminalis through which passed the anterior commissure (27). The term para-terminal body was introduced by Smith (75) to include the septal area and the medial olfactory ganglionic mass. The septal area consists mainly of nuclear masses of gray matter anterior and superior to the lamina terminalis and anterior commissure (83). The septal area superior to the anterior commissure (supracommissural) corresponds to the area of the septum pelluci-dum. During the evolutionary growth of the cerebrum, the hippocampal formation (archipallium) and the neopallium evolved dorsal to the olfactory cortex (Fig. 2.5). At the same time the fornix
(the commissures of the hippocampus) and corpus callosum evolved dorsal to the anterior commissure (Figs. 2.6 and 2.7). In the lamprey the lamina terminalis delineates the anterior end of the brain. An anterior commissure connecting the striatal areas and a primordial hippocampus are already present (26). The rudimentary hippocampus is situated above the interventricular foramen and receives second order olfactory fibers and ascending fibers from the hypothalamus. The entrance of these fiber tracts into the rudimentary hippocampus establishes it as a correlating center for olfactory and visceral impulses. In the shark the primordial hippocampus is larger, occupies the roof of the lateral ventricle, and extends along the upper border of the wall of the third ventricle (Fig. 2.4). According to Johnston (26), the fiber tracts, homologous to the fornix, descend from the primordial hippocampus through the mediorostral and ventral wall of the cerebrum and pass between the lateral ventricles to the hypothalamus.
In the frog the primordial hippocampus forms the entire length of the dorsomedial wall of the cerebral hemisphere (23). Johnston (26) and Her-rick (24) described a hippocampal commissure, dorsal to the anterior commissure, that projected into the floor of the third ventricle of the amphibian salamander (Fig. 2.5). The hippocampal commissure is the homologue of the fornix and in the amphibian passes behind the interventricular foramen. In reptiles most of the primordial hippocampus has been transformed into a specialized cerebral cortex that occupies the dorsomedial aspect of the entire cerebral hemisphere (23, 26, 27). Within the hippocampus, fibers arise to form a tract on its ventricular surface. This tract, the homologue of the mammalian fornix, passes rostral to the interventricular foramen and forms a commissure that crosses the lamina terminalis dorsal to the anterior commissure (Fig. 2.6). In monotremes such as the platypus and spiny anteater, which belong to the lowest subclass of mammals, the hippocampus is still large and
Figure 2.28. Midsagittal sections of the head of a rabbit (A), dog (B), and monkey (C) demonstrating the relationship between the increasing size of the cerebrum and corpus callosum. The genu and splenium are present in the dog and are increased in size in the monkey. Note the changing facial-neurocranium ratio from the rabbit to the monkey. Compare with the predominant neurocranium in the human fetus (Fig. 2.27).
Figure 2.28B
Figure 2.28C
assumes the elongated and arched contours of the cerebral hemisphere. In the platypus the posterior limb of the fornix assumes the curved shape of the hippocampus (74). In the spiny anteater the posterior limb of the fornix is better developed and corresponds to the larger temporal segment of the hippocampus. Thus already in monotremes the fornix extends from the temporal lobe to the mamillary body and is curved by the large mammalian thalamus. The arch of the fornix increases in the ascending mammalian scale as a result of the progressive enlargement of the thalamus. A major transformation of the mammalian brain is the progressive enlargement of the cerebrum. The increase in size of the mammalian cortical neopallium in placental mammals is accompanied by a new commissure, the corpus callosum (Figs. 2.7 to 2.11). Nieto et al. (48) demonstrated a linear correlation between surface area of the corpus callosum and the brain weight of different animals (Fig. 2.28). The only exception was in the dolphin. In this mammal the corpus callosum is comparatively small in spite of relatively massive development of the cerebral hemispheres. They postulated that the lack of extremities in cetaceans may account for the relative smallness of their corpus callosum. The appearance of the corpus callosum first in placental mammals may relate to expansion of the associated areas of the neocortex (67). Smith (74), Johnston (27), and Abbie (4) demonstrated that the evolutionary increase of the corpus is associated with a concomitant decrease of the hippocampal formation in the roof and medial wall of the lateral ventricle (Fig. 2.29). To state that the corpus callosum replaces the hippocampus in the roof of the body of the lateral ventricle is not strictly accurate. The concept of replacement, substitution, or suppression, however, facilitates understanding of the topographic relation of the corpus callosum, fornix, and hippocampus. An example of the "replacement" is seen in the rabbit, where the posterior margin of the evolving corpus blends with the anterior margin of the remaining hippocampus (Fig. 2.7). Smith (75) stated that, "in all mammals the cerebral cortex is fringed with some structure representing a hippocampal formation along the whole extent of its mesial edge—from the olfactory peduncle in front to the region of the uncus behind and below" (Fig. 2.30). Smith thought that the indusium griseum and the longitudinal striae were vestigial hippocampal structures. Kappers et al. (30) further elaborated this theory. Abbie (4) demonstrated that the indusium gri-
seum is not a remnant of the hippocampus but is formed by eversion of a lip of the subiculum. The subiculum is transitional (mesocortex) cerebral tissue at the junction of the hippocampus and the neocortex (3). The evolution of the corpus callosum was divided by Abbie (4) into the following stages (Fig. 2.29): (a) reduction in the size of the hippocampal segment adjacent the lamina terminalis, (b) piercing of the subiculum adjacent to the lamina terminalis by fibers of the corpus callosum, (c) enlargement of the corpus callosum adjacent to the lamina terminalis, (d) formation of the sple-nium by linear caudal expansion, and (e) formation of the genu and septum pellucidum by frontal arciform expansion. These stages are diagrammed in Figure 2.29 and are demonstrated in dissected specimens in Figures 2.7 to 2.10. In monotremes, such as the platypus and the spiny anteater, the paraterminal body contains the lamina terminalis, a large anterior commissure, and dorsal to it a smaller round hippocampal commissure. In the marsupial a large hippocampus on the medial aspect of the cerebral hemisphere extends anterior to the lamina terminalis (precommissural hippocampus) and posteriorly (postcommissural hippocampus) into the temporal lobe (Fig. 2.29A). The anterior commissure is still a large structure. The commissure of the fornix (inferior fornix) has a primordial elongated shape. A postulated intermediate phase demonstrates the first stage in the formation of the corpus callosum (Fig. 2.29B). The hippocampus adjacent to the lamina terminalis is reduced in size secondary to infolding. The hippocampal infolding results in approximation of the subiculum and lamina terminalis. The broken line in the subiculum indicates the point at which the fibers of the corpus callosum pierce the subiculum. In the hedgehog and the bat the fibers of the corpus callosum break through the subiculum (Fig. 2.29C). With the appearance of the corpus callosum, the hippocampus disappears almost completely from the region dorsal to the commissures in the lamina terminalis. The anterior commissure is reduced, the inferior fornix is enlarged, and a new hippocampal fiber tract (the superior fornix) appears in the region of the lamina terminalis. In the rat the corpus callosum is enlarged dorsal to the lamina terminalis (Fig. 2.29D). This enlargement is associated with a further reduction of the precommissural and postcommissural hippocampi, which are ventral to the corpus cal-
Figure 2.29. Diagram of the medial aspect of a cerebral hemisphere to demonstrate the evolution of the corpus callosum and septum pellucidum. A, marsupial; B, postulated intermediate phase; C, hedgehog and bat; D, rat; E, rodent; F, primate. Expansion of the corpus callosum (c.c.) is associated with concomitant reduction in size of the hippocamus (hi.). Elongation of the superior (f.s.) and inferior (fi.) fibers of the fornix is related to growth of the splenium (s.) of the corpus callosum. The septum pellucidum (si.) results from thinning of the lamina terminalis (lt.) secondary to the arched expansion of the genu of the corpus callosum; ca., anterior commissure; pt. b., paraterminal body; hi. p., precommis-sural hippocampus; r., rostrum; su., subiculum. (From Abbie AA: J Comp Neural 70:9-44, 1939.)
Figure 2.30. Diagrammatic across sections of the left cerebral hemisphere show the evolutionary changes in the position of the hippocampus. A. In the primitive stage the cerebrum is essentially an olfactory lobe composed only of the paleopallium. The hippocampus (archipallium) is not present. B. In the amphibian the hippocampus forms the medial wall of the hemisphere. С and D. In early and late reptilian stages the hippocampus is medial and superior to the lateral ventricle. In the late reptilian stage (D), the neopallium is still rudimentary. E. In the primitive mamallian stage, the neopallium has expanded and the hippocampus is rolled in on the medial surface of the hemisphere, adjacent to the lateral ventricles. F. In the higher mammals the neopallium has markedly expanded. The rolled-in hippocampus is separated from the ventricle by the corpus callosum. a, archipallium (hippocampus); b, basal nuclei; cc, corpus callosum; n, neopallium; p, paleopallium; v, ventricle. (From Romer AS: The Vertebrate Body. Philadelphia, WB Saunders Co, 1970, ed 4.) losum. The inferior and superior fornices are elongated. Similar changes are seen in the rabbit (Fig. 2.7). In another rodent the stage of caudal expansion of the corpus callosum is demonstrated (Fig. 2.29E). This expansion leads to formation of the splenium. The caudal expansion of the corpus callosum is a simple linear extension and carries with it the fibers of the fornix, which become increasingly elongated. Both the splenium and the commissure of the fornix remain dorsal to
the hippocampus. These changes are demonstrated in the cat and dog (Figs. 2.8 and 2.9). The last stage of development of the corpus callosum involves the region of the genu. Progressive expansion of the genu associated with formation of a septum pellucidum is seen in the cat and the dog (Figs. 2.8 and 2.9). The frontal expansion of the corpus callosum reaches its maximal size in primates (Figs. 2.10 and 2.11). The development of large frontal lobes in primates is associated with the development of the large arched genu and a large septum pellucidum (Figs. 2.29F). Thompson (78) reviewed the controversial literature dealing with the presence of a cavum septi pellucidi and its open status in a large number of mammals such as ungulates, carnivores, and primates. Johnston (27) postulated that the cavum septi pellucidi was the result of trapping and roofing by the genu of a cavum formed by lamina of the paraterminal body and open dorsally into the superior longitudinal fissure (Fig. 2.31). Abbie (4) pointed out that a cavum opening dorsally into the superior longitudinal fissure has never been demonstrated in vertebrates without a corpus callosum. Abbie (4) attributed the formation of the septum pellucidum to the arched expansion of the genu (Fig. 2.31). The upward expansion of the genu draws up the lamina terminalis and the superior fornix into the concavity of the arch, thus forming the septum pellucidum. In the cat the septum is still solid. The tension produced by further arched expansion of the genu results in condensation of the septum into two laminae with a cavum between them. Abbie (4) postulated that similar upward expansion in the region of the splenium cannot take place because of the restraining fibers of the columns of the fornix (62). According to Sarnat and Netsky (67) the leaves of the septum pellucidum are composed of neural tissues and the cavum septi pellucidum is an ependyma-linked cavity deriving from the lamina terminalis. Rakic and Yakovlev (62) considered the cavum septi pellucidi as a dorsally open pocket that is open at birth. Later it is sealed by the rostrum of the corpus callosum. The pocket stays open in rodents, carnivores, and monkeys (rhesus). Loeser and Alvord (43, 44), Probst (60, 61), and Lemire et al. (41) reviewed the contro-versial development of the human septum pellucidum.
Figure 2.31. Diagrams of the brain of the rat to illustrate two different theories of the formation of the septum pellucidum, the indusium griseum, and the longitudinal striae. A. Johnston demonstrates a hippocampal remnant above the corpus callosum (c.c.). The septum pellucidum is formed within the hippocampal primordium (spindle-shaped spots). The paraterminal body is shaded with horizontal lines. c. h., hippocampal commissure; c.a., anterior commissure. B. Abbie demonstrates no hippocampal (black) tissue above the corpus callosum (c.c.). Instead, the subiculum (su), a mesocortical tissue, lies above the corpus callosum. The septum pellucidum area is formed by the drawing up of the lamina terminalis (l.t.).f.s., superior fornix;f.i., inferior fornix; hi.p., precommissural hippocampus; c.a., anterior commissure;f.l. (f.s.), long fornix; pt.b., paraterminal body; ol.t., olfactory tubercle. (A. From Johnston; J Comp Neurol 23:371478, 1913.) (B. From Abbie: J Comp Neurol 70:9-44, 1939.)
Arteries Neuroradiologists and neurosurgeons are constantly aware of the marked variability in angio-graphic and neurosurgical vascular anatomy. A
study of the general principles that govern the vascularization of the brain and of certain evolutionary modifications of the arterial system is helpful in understanding the vascular variability of the arteries supplying the hypothalamus and the thalamus (32). Streeter (76, 77) pointed out the following important principles of cerebrovascular develop-
ment. (a) The development of the cerebrovascu-lar system is not an independent phenomenon. (b) The vascular system constantly adapts to the changing brain. In comparative anatomical studies Shellshear (69-72) repeatedly stressed the unity of the brain and its vascular supply. He expressed the belief that earlier workers had been misled by overemphasis on the variation of the larger vessels. Because the size of the arteries served as an index of their importance, the fundamental characteristic of arterial distribution was overlooked. An analysis of arteries, in terms of their final areas of supply, showed a remarkable constancy. Shellshear divided the cerebral arteries into two major divisions: the end arteries and the blood vessels that lead to the end arteries (32). Abbie (1, 2), in studies of the comparative anatomy of the circulation of the brain, confirmed the postulates of Shellshear in regard to constancy of the cerebral arteries in their terminal supply. In his studies he reemphasized the concept that the brain and its blood supply are not independent and that both evolve simultaneously. Abbie, as well as Shellshear, pointed out that the variation in the origin and course of the vascular trunks leading to the functional end arteries is of secondary importance. This variation depends on the evolutionary direction of growth of an organ and on hemodynamic factors such as the closeness, economy of distribution, and convenience of the source of blood (Fig. 2.38). This last concept was alluded to by Evans (15), who stated, "The larger vessels are to be considered in the light of servants of the capillaries, for which they are but the delivering and draining pipes." Roofe (65), in his comprehensive study of the endocranial circulation of the salamander, emphasized the great variability in the supply of various territories of the brain. Areas of the brain that are well defined morphologically and functionally are supplied by delivering arteries whose course varies greatly. The terms nutrient, applied by Gillilan (19) to the end arteries, and conducting, applied to the major trunks that lead to the end arteries, convey the physiological aspects of the cerebral circulation. The end arteries supply a constant functional territory. This constancy is mediated through elements of the sympathetic nerves that accompany the vessels. These end arteries develop synchronously with the structures that they supply, and their evolution parallels the evolutionary changes of their territories of supply.
The central artery of the retina is the best example of an end artery. The cerebral perforating arteries of the brain communicate with each other only through the capillary bed and from a practical point of view are end arteries (83). Neuroangiography and microneurosurgery visualize only the variable "delivering (conducting) pipes" of the hypothalamus and thalamus. These are the circle of Willis and the anterior cerebral, anterior communicating, internal carotid, posterior cerebral, posterior communicating, and anterior and posterior choroidal arteries (38, 89). The perforating branches of these arteries are the end arteries supplying fairly specific territories in the third ventricular region (Fig. 2.32) (18, 34, 39, 40, 57, 58, 66). Although the hypo-thalamic branches of the superior hypophyseal artery are end arteries, the branches supplying the hypophysis and its stalk are not end arteries (47). Padget (51, 52), in her studies of the early fetal arterial system, noted that, in accordance with the five stages of cerebral blood vessel development of Streeter (77), the early process of vascular development involves changes in certain channels of the primitive vascular plexus. These changes depend on the specific needs of areas to be supplied. The needs may be permanent or transitory and result in the elaboration of permanent or transitory arteries. The great variability in the region of the anterior communicating artery is a function of these factors (38). The recurrent artery of Heubner is an example of evolutionary change of source of supply for reasons of economy of blood distribution (2). Because of the bending of the rhinal fissure, the anterior cerebral artery takes over some of the territory previously supplied by the middle cerebral artery. This territory includes a segment of the basal ganglia. The terminal arterioles of Heubner's artery never change their territory of supply; rather the conducting arteries to these arterioles are changed (32). Earlier in the evolutionary phase, the terminal arterioles of Heub-ner's artery are supplied by the middle cerebral artery. In primates they are supplied by a new conducting artery, which, because of its origin from the anterior cerebral artery, must pass laterally to reach its arterioles. The great variability in the origin of the artery of Heubner depends on the particular anastomotic channel that becomes the stem of the artery (Fig. 2.33). Other examples of variability in the vascular trunks supplying the same end arterioles are the inverse relationship between the size of the anterior and posterior choroidal arteries and the
Figure 2.32. Diagrams of the arterial territories in the third ventricular region. A. Sagittal: 1, anterial cerebral; 2, posterior cerebral; 3, anterior and posterior choroidal; 4, posterior communicating; 5, internal carotid. B. Floor: 1, anterior cerebral; 2, posterior cerebral; 3, posterior communicating; 4, internal carotid. С and D. Transverse and coronal: 1, anterior cerebral; 2, middle cerebral; 3, posterior cerebral; 4, anterior choroidal; 5, posterior choroidal; 6, posterior communicating. (From Krayenbuhl HA, Yasargil MG: Cerebral Angiography. Philadelphia, JB Lippincott, 1968. After Lazorthes G: Vascularisation et Circulation Cere-brales. Paris, Masson & Cie, 1961.)
Figure 2.33. Dissected human fetal arteries: 16-week fetus (A); 24-week fetus (B). Note the variability in the origin, size, and course of the recurrent arteries of Heubner. From Kier EL: The fetal cerebral arteries: a phylogenetic and ontogenetic study. In Newton TH, Potts DG (eds): Radiology of the Skull and Brain, Vol. 2, Book 1, Angiography. St. Louis, The CV Mosby Co, 1974. large anterior inferior cerebellar artery supplying the territory of the posterior inferior cerebellar artery or the reverse anomaly. The human circle of Willis is but one of many cerebral arterial circles that evolved in the various vertebrate groups to assure an adequate supply of blood to the brain (87). An additional possible explanation of its function in the evolve-ment of the cerebral arterial circle is to assure equal pressure in the three major cerebral arteries (88). In fish, amphibians, reptiles, and birds, the internal carotid arteries are the sole source of blood to the brain (14). In these four classes of vertebrates, the internal carotid arteries divide into cranial and caudal divisions. The caudal division unites to form the basilar artery in all four classes. The cranial divisions remain separate in fish, amphibians, and birds. In most rep-
tiles the cranial division unites to form a single artery that passes anteriorly to supply the olfactory lobes. Thus in fish, amphibians, and birds the arterial circle is open anteriorly, whereas in the majority of reptiles it is completely closed. Occasionally an anterior communicating artery may exist in certain amphibians (14). The monotremes possess definite mammalian features, such as the presence of an internal capsule and a crus cerebri and pons (22). The cerebral arterial circle of monotremes is the earliest true homologue of the human circle of Willis, for it unites the carotid and vertebral circulations and is completely closed. The term circle of Willis should be limited to the human cerebral arterial circle (22). The cerebral arterial circle of the different mammals varies greatly and can be grouped into three fundamental groups (14). These groups are
distinguished by the relative contribution of the carotid and vertebral arteries to the arterial circle. Group I The carotid and vertebral circulations participate in approximately equal manner. To this group belong humans, anthropoid apes, monkeys, insectivores (shrew, mole, hedgehog), and certain carnivores, rodents, and toothless mammals. Group II The carotid arteries are the major source of supply to the cerebral arterial circle. In this group the vertebral and basilar arteries are greatly attenuated. The monotremes, marsupials, whales, odd-toed ungulates (horse, zebra), even-toed ungulates (cattle, deer, camel, giraffe), and many land and marine (seal, sea lion, walrus) carnivores belong to this group. Group III The vertebral and basilar arteries are the major source of supply to the arterial circle. The inter-
nal carotid arteries are attenuated. To this group belong many rodents and some lemurs, bats, and toothless mammals. According to Nilges (49) the arterial circle of the guinea pig is derived entirely from the vertebral arteries. The internal carotid artery leaves the cranial cavity as the ophthalmic artery without participating in the formation of the cerebral arterial circle. In some ruminants (ox, goat, sheep, pig) and in the cat, the internal carotid artery has been replaced by a vascular plexus—the rete mirabile— which connects the cerebral arterial circle with branches of the external carotid artery (12, 16, 20). The fish brain is supplied entirely by the internal carotid artery. The internal carotid artery divides into two branches, a rostral and a caudal division, whose sizes relate to the functional preponderance of the various brain components. The rostral branch is small and divides into a medial and a lateral olfactory artery. The medial olfactory artery supplies the medial aspect of the primitive forebrain, the septal region, and the medial part of the primordial hippocampi. It is
Figure 2.34. Diagrams demonstrating the variable evolutionary origin of the posterior cerebral artery. A. In the chicken the posterior cerebral artery (a. cer. post.) originates from the rostral branch (r. rost.) of the internal carotid artery. B. In the marsupial the origin of the posterior cerebral artery is now more posterior, originating from the basilar artery. (A. From Gillilan LA: J Comp Neurol 130:175-196, 1967. B. From Abbie AA: J Anat 67:491-521, 1933.)
the homologue of the anterior cerebral artery of the higher forms. The lateral olfactory artery supplies the region of the primordia of the hy-popallium, piriform lobe, most of the hippocampus, and the paleostriatum. The caudal branch is larger, and its branches supply the dien-cephalon, optic lobes, cerebellum, and medullary centers. In amphibians the internal carotid artery divides into a cranial and a caudal division of approximately equal calibers. The medial olfactory artery of amphibians is similar to that of the fish in its course and territory of supply and is the homologue of the anterior cerebral artery. The anterior cerebral artery of reptiles is the homologue of the medial olfactory artery of fish and amphibians. In reptiles the main stream of blood has been deflected to the medial wall of the telencephalic hemisphere, probably in response to the increased demands of the new reptilian cortex. The anterior cerebral artery in birds contributes little to the supply of the forebrain. A few small branches to the medial and central aspect of the forebrain originate from the anterior cerebral artery before this vessel leaves the cranial cavity to become the ethmoidal artery. The original course of the anterior cerebral artery homologue in the lower forms is indicated in humans by a few small branches that run forward toward the olfactory bulb. The homologue of the anterior choroldal artery first appears as an entity in the reptilian class. This vessel supplies branches to the optic tract, lateral geniculate body, and amygdaloid. In marsupials the anterior choroidal artery has evolved into a complete homologue of the primate anterior choroidal artery. It arises from the internal carotid artery, courses posteriorly along the optic tract and lateral geniculate body, and terminates in the posterior cerebral artery. Along its course it sends a few small branches that reach the choroid plexus through the choroid fissure. In the sheep the course of the artery is similar to that in marsupials. The choroidal branches in the sheep, however, are much more prominent than in the lower forms. The branches of the anterior choroidal artery that supply the choroid plexus in the ape equal the size of those that supply the optic tract and lateral geniculate body. From reptiles to humans, the anterior choroidal artery maintains a close topographic relation with the optic tract and lateral geniculate body. The artery undergoes a functional change, however. From mainly a visual function in the
reptiles, its contribution to the choroid plexus becomes greater as the cerebrum progressively enlarges in the ascending evolutionary scale. The posterior cerebral artery and its homo-logues undergo significant evolutionary changes in their origin, course, and territory of supply. The diencephalic and tectal arteries of fish and amphibians (21) are homologues of the posterior cerebral artery of higher forms. They arise from the caudal division of the internal carotid artery and supply the optic lobes and the dorsomedial aspect of the primitive forebrain. The reptilian posterior cerebral artery is the true forerunner of that artery in mammals. In the various reptiles it may arise as a branch of either the cranial division of the internal carotid artery (21) or the caudal division (8); it may also arise at the bifurcation of the internal carotid artery (9), or it may represent the entire caudal division of lower forms (2). In the mammalian class the origin or source of blood for the posterior cerebral artery shifts from the carotid to the basilar artery. As the cerebral hemisphere enlarges in the ascending mammalian series, the cerebral territory supplied by the posterior cerebral artery is gradually shifted more posteriorly and overhangs the midbrain. The distance between the origin of the posterior cerebral artery and the final areas of supply is increased. For economy of distribution, the posterior cerebral artery shifts its stem progressively posteriorly (Fig. 2.34). Successive posterior anastomotic channels enlarge to convey the main stream of blood to the posterior cerebral artery. These anastomotic channels, which successively enlarge to form the stem of the posterior cerebral artery, revert to their original small size when their need as the main conduit of blood to the posterior cerebral artery disappears. In its evolutionary course posteriorly the main trunk of the posterior cerebral artery incorporates the territory of the various vessels that have become its temporary or permanent stem. In the higher mammals the posterior cerebral artery uses the anterior midbrain channels as a stem. These channels are supplied by a greatly enlarged basilar artery; eventually the posterior cerebral artery seems to arise directly from the basilar artery. The posterior choroldal arteries join the anterior choroidal artery in a rich network that supplies the optic tract, lateral geniculate body, and choroid plexus of the third ventricle. The posterior choroidal vessels increase in number and complexity in the ascending phylo-genetic scale. According to Abbie (1) the variabil-
Figure 2.35. Lateral view of arterially injected brain of an iguana. The posterior cerebral artery is seen on the surface of the brain supplying the optic lobe. Note the artery supplying the chiasm and hypophysis (arrow).
Figure 2.36. Lateral view of a dissected brain of a rabbit. The posterior cerebral artery can only be demonstrated by dissection because of the marked expansion of the neopallium. Note the relatively large size of the superior colliculus and the large homologues of the medial and lateral posterior choroidal arteries. The posterior pericallosal artery is curving around the corpus callosum to supply the roof of the third ventricle.
Figure 2.37. Lateral view of dissected brain of a cat. The colliculi and posterior choroidal arteries are relatively reduced in size. The enlarged occipital lobe is supplied by a large branch of the posterior cerebral artery (crossed arrow).
Figure 2.38. Dorsal view of the roof of the third ventricle of a dissected 24-week human fetus. Note the large posterior pericallosal artery curving around the rudimentary splenium of the corpus callosum.
ity of the major posterior choroidal branches is a function of the rich vascular network that supplies the choroid plexus. The functional end arteries of the posterior choroidal arteries remain constant; only their larger conducting trunks vary. The constancy of the end arteries is manifested in the precise correlation of their distribution in the various laminae of the lateral ge-niculate body and in the areas of the functional representation of the retina. In the iguana the posterior cerebral artery supplies the optic lobe (Fig. 2.35). In the rabbit significant evolutionary change has occurred. The posterior cerebral artery supplies not only the optic lobe but the new visual centers, the geniculate body and pulvinar. The homologues of the medial and lateral posterior choroidal arteries can be seen only by dissection because of the marked enlargement of the cerebrum (Fig. 2.36). In the cat there are further changes. The optic lobe is reduced in relative size. A large posterior cerebral artery branch to the occopital lobe is seen (Fig. 2.37). In higher vertebrates the choroidal branches will relatively diminish in size, while the cortical branches increase in size. The posterior growth of the cerebrum and corpus callosum elongates the course of the anterior cerebral artery to the roof of the third ventricle. This long course results in the inconsistent presence of the posterior pericallosal artery in the human adult. The medial posterior choroidal artery has a shorter course to the roof of the third ventricle and is rarely absent (Fig. 2.38).
Veins There are few studies dealing with the cerebral veins in nonhuman vertebrates. In the shark the forebrain veins are anterior to the diencephalic veins of the uncovered roof of the third ventricle (Fig. 2.25A). In the salamander (65) the veins of the brain are superficial and their course is exceedingly variable. The veins of the choroid plexus are wide sinusoidal channels that drain into a rete of the nodus vasculoses that permeates the paraphysis and dorsal sac. The dorsal hemispheric veins are pial and also drain into the nodus vasculoses. In the turtle (68) the venous channels are situated on the superior surface of the brain. The largest one is a median dorsal longitudinal vein that commences anteriorly near the interventricular foramen and passes posteriorly on the surface of the brain to exit from the skull near the fourth ventricle. There are no venous channels within the brain. In the dog the great cerebral vein is the main vessel of the velum interpositum. It passes ven-
tral to the corpus callosum (5). Review of their Figure 5 demonstrates a straight great cerebral vein secondary to the small splenium in the dog. A diagram of the intraventricular venous drainage of the capuchin monkey also demonstrates a straight great cerebral vein (45). These features seem to be "recapitulated" in the human embryo (29, 54). In the 40-mm embryo (21/2 months) the arterial circle of Willis and its branches are clearly defined. However, the deep venous system is still rudimentary, consisting of the primitive great cerebral vein (Galen), which leaves the membranous diencephalic roof near the pineal primordium (53). The primitive internal cerebral vein is still extracer-ebral, lying in dorsal contact with the diencephalic roof. The internal cerebral vein will become enclosed within the two leaves of the tela choroidea of the third ventricle during further growth of the cerebrum. The great cerebral vein derived its curvature during the growth of the splenium of the corpus callosum. References 1 . Abbie AA: The blood supply of the lateral geniculate body, with a note on the morphology of the choroidal arteries. J Anat 67:491-521, 1933. 2. Abbie A A: The morphology of the fore-brain ar teries, with special reference to the evolution of the basal ganglia. J Anat 68:433-470, 1934. 3. Abbie AA: The relations of the fascia dentata, hippocampus and neocortex and the nature of the subiculum. J Comp Neurol 68:307-323, 1938. 4. Abbie AA: The origin of the corpus callosum and the fate of the structures related to it. J Comp Neurol 70:9-44, 1939. 5. Armstrong LD, Horowitz A: The brain venous sys tem of the dog. Am J Anat 132:479-490, 1971. 6. Baudey-Pasquier J, Larroche JC: Les cavites cerebrales supplementaires pendant le developpement foetal et neofoetal. Acta Radiol [Diagn] (Stockh) 1:533-540, 1963. 7. Brash JC (ed): Cunningham's Textbook of Anat omy. New York, Oxford University Press, Inc, 1951, ed 9. 8. Burda DJ: Development of intracranial arterial patterns in turtles. J Morphol 116:171-188, 1965. 9. Burda DJ: Developmental aspects of intracranial arterial supply in the alligator brain. J Comp Neu rol 135:369-380, 1969. 10. Cabanes MD: Asymptomatic persistence of indundibularis recessus. J Neurosurg 49:769-772, 1978. 1 1 . Cole FJ: A History of Comparative Anatomy. London, Macmillan & Co, Ltd, 1944. 12. Daniel PM, Dawes JDK, Prichard ML: Studies of the carotid rete and its associated arteries. Philos Trans R Sac Lond Ser В 237:173-208, 1953. 13. De Beer GR: Embryos and Ancestors. New York, Oxford University, 1958, ed 3. 14. De Vriese B: Sur la signification morphologique des arteres cerebrales. Arch Biol 21:357-457, 1905.
15. Evans HM: The development of the vascular sys tem. In Kiebel F, Mall FP (eds): Manual of Human Embryology. Philadalphia, JB Lippincott Co, 1912. 16. Francke JP, Clarisse J, Dhellemmes P, Bousquet C, Francke-Mauroy B: The arterial circle of the base of the brain and its feeding vessels in certain mammals. J Neuroradiol 4:271-289, 1977. 17. Frazer JE: A Manual of Embryology. New York, William Wood & Co, 1932. 18. George AE, Raybaud C, Salamon G, Kricheff I: Anatomy of the thalamoperforating arteries with special emphasis on arteriography of the third ventricle: Part 1. AJR 124:220-230, 1975. 19. Gillilan LA: Significant superficial anastomoses in the arterial blood supply to the human brain. J Comp Neurol 112:55-74, 1959. 20. Gillilan LA, Markesbery WR: Arteriovenous shunts in the blood supply to the brains of some common laboratory animals—with special atten tion to the rete mirabile conjugatum in the cat. J Comp Neurol 121:305-311, 1963. 21. Gillilan LA: A comparative study of the extrinsic and intrinsic arterial blood supply to brains of submammalian vertebrates. J Comp Neurol 130:175-196, 1967. 22. Gillilan LA: Blood supply to primitive mammalian brains. J Comp Neurol 145:209-221, 1972. 23. Herrick CJ: The morphology of the forebrain in amphibia and reptilia. J Comp Neurol 20:413547, 1910. 24. Herrick CJ: The Brain of the Tiger Salamander. Chicago, University of Chicago Press, 1948. 25. Johnston JB: The Nervous System of Verte brates. Philadelphia, P Blakiston's Son & Co, 1906. 26. Johnston JB: The evolution of the cerebral cortex. Anat Rec 4:143-166, 1910. 27. Johnston JB: The morphology of the septum, hip pocampus, and pallial commissures in reptiles and mammals. J Comp Neurol 23:371-478, 1913. 28. Jollie MT: Chordate morphology. New York, Reinhold Publishing Corp, 1962. 29. Kaplan HA, Ford DH: The Brain Vascular Sys tem. New York, American Elsevier Publishing Co, Inc. 1966. 30. Kappers CUA, Huber GC, Crosby EC: The Com parative Anatomy of the Nervous System of Ver tebrates, including Man. New York, The Macmillan Co, 1936. (Reprinted by Hafner Publishing Co, Inc, New York, 1967.) 31. Kent GC: Comparative Anatomy of the Verte brates. St Louis, The CV Mosby Co, 1973, ed 3. 32. Kier EL: The fetal cerebral arteries, a phylogenetic and ontogenetic study. In Newton TH, Potts DG (eds): Radiology of the Skull and Brain, vol 2, book 1, Angiography. St Louis, The CV Mosby Co, 1974, pp 1089-1130. 33. Kier EL: The cerebral ventricles: A phylogenetic and ontogenetic study. In Newton TH, Potts DG (eds): Radiology of the Skull and Brain, vol 3, Anatomy and Pathology. St Louis, The CV Mosby Co, 1977, pp 2787-2914. 34. Krayenbuhl HA, Yasargil MG: Cerebral Angiog raphy. Philadelphia, JB Lippincott, 1968.
35. Kuhne D, Schwartz RB: Persisting intrapituitary recessus infundibuli. Neuroradiology 10:177178, 1975. 36. Larroche JC: The development of the central nerv ous system during intrauterine life. In Falkner F (ed): Human development. Philadelphia, WB Saunders Co, 1966. 37. Larroche JC, Baudey J: Cavum septi lucidi, cavum vergae, cavum veli interpositi: Cavites de la linge mediane. Biol Neonat 3:193-236, 1961. 38. Lazorthes G: Vascularisation et Circulation Cerebrales. Paris, Masson & Cie, 1961. 39. Lazorthes G, Salamon G: Etudes anatomique et radio-anatomique de la vascularisation arterielle du thalamus. Ann Radiol 14:11-12, 905-924, 1971. 40. Lazorthes G, Salamon G: The arteries of the thal amus: An anatomical and radiological study. J Neurosurg 34:23-26, 1971. 41. Lemire RJ, Loeser JD, Leech RW, Alvord EC Jr: Normal and Abnormal Development of the Hu man Nervous System. Hagerstown, MD, Harper & Row, 1975. 42. Levinger IM, Edery H: Casts of the cat cerebroventricular system. Brain Res 11:294-304, 1968. 43. Loeser JD, Alvord EC Jr: Agenesis of the corpus callosum. Brain 91:553-570, 1968. 44. Loeser JD, Alvord EC Jr: Clinicopathological cor relations in agenesis of the corpus callosum. Neu rology 18:745-756, 1968. 45. Madeira MC, Watanbe S: Anatomical data on the intracranial venous drainage of the tufted capu chin. Anat Rec 183:589-598, 1975. 46. Millen JW, Wollam DHM: The Anatomy of the Cerebrospinal Fluid. London, Oxford University Press, 1962. 47. Netter FH, Crelin ES: Nervous System, vol 1, part 1, Anatomy and Physiology. West Caldwell, NJ, CIBA, 1983. 48. Nieto A, Nieto D, Pacheco A: Possible phylogenetical significance of the corpus callosum with spe cial reference to the dolphin brain (Stenella graffmani). Acta Anat 94:397-402, 1976. 49. Nilges RG: The arteries of the mammalian cornu ammonis. J Comp Neurol 89:177-190, 1944. 50. Oksche A: Survey of the development and com parative morphology of the pineal organ. Prog Brain Res 10:3-29, 1965. 51. Padget DH: The circle of Willis, its embryology and anatomy. In Dandy WE (ed): Intracranial Ar terial Aneurysms. Ithaca, NY, Comstock Publish ing Co, Inc, 1944, pp 67-147. 52. Padget DH: The development of the cranial arter ies in the human embryo. Contrib Embryol 32:205-262, 1948. 53. Padget DH: The cranial venous system in man in reference to development, adult configuration, and relation to the arteries. Am J Anat 98:307356, 1956. 54. Padget DH: The development of the cranial venous system in man, from the viewpoint of comparative anatomy. Contrib Embryol 36:79-140, 1957. 55. Papez JW: Comparative Neurology. New York, Thomas Y Crowell Co, 1929 (Reprinted by Hafner Publishing Co, Inc, New York, 1967.) 56. Patten BM: Human Embryology. New York,
McGraw-Hill Book Co, 1968, ed 3. 57. Percheron G: Les arteres du thalamus humain. Rev Neurol (Paris) 132:297-307, 1976. 58. Plets C: The arterial blood supply and angioarchitecture of the posterior wall of the third ventricle. Acta Neurochir (Wien) 21:309-317, 1969. 59. Polyak S: The Vertebrate Visual System. Chi cago, University of Chicago Press, 1957. 60. Probst FP: Congenital defects of the corpus callosum: Morphology and encephalographic appear ances. Acta Radiol (Suppl) 331:1-152, 1973. 61. Probst FP: The Prosencephalies: Morphology, Neuroradiological Appearances and Differen tial Diagnosis. New York, Springer, 1979. 62. Rakic P, Yakovlev PI: Development of the corpus callosum and cavum septi in man. J Comp Neurol 132:45-72, 1968. 63. Reiter RR: The mammalian pineal gland, struc ture and function. Am J Anat 162:287-313, 1981. 64. Romer AS: The Vertebrate Body. Philadelphia, WB Saunders Co, 1970, ed 4. 65. Roofe PG: The endocranial blood vessels of Amblystoma tigrinum. J Comp Neurol 61:257-293, 1935. 66. Rosner SS, Rhoton AL Jr, Ono M, Barry M: Microsurgical anatomy of the anterior perforating ar teries. J Neurosurg 61:468-485, 1984. 67. Sarnat HB, Netsky MG: Evolution of the Nervous System. New York, Oxford University Press, Inc, 1981, ed 2. 68. Schepers GWH: Evolution of the Forebrain. Cape Town, Maskew, Miller Ltd., 1948. 69. Shellshear JL: The basal arteries of the forebrain and their functional significance. J Anat 55:2735, 1920. 70. Shellshear JL: The arteries of the brain of the orangutan. J Anat 61:167-197, 1927. 71. Shellshear JL: The arterial supply of the cerebral cortex in the chimpanzee (Anthropopithecus troglodytes). J Anat 65:45-87, 1930. 72. Shellshear JL: A study of arteries of the brain of the spiny anteater (Echidna aculeata) to illus trate the principle of arterial distribution. Philos Trans R Soc Lond Ser В 218:1036, 1930. 73. Shuangshoti S, Netsky MG: Choroid plexus and paraphysis in lower vertebrates. J Morphol 120:157-188, 1966.
74. Smith GE: The morphology of the true "limbic lobe" corpus callosum, septum pellucidum, and fornix: A preliminary communication. J Anat Physiol 30:157-205, 1895. 75. Smith GE: Some problems relating to the evolu tion of the brain. Lancet 1-6, 147-155, 221-227, 1910. 76. Streeter GL: The development of the nervous sys tem. In Keibal F, Mall FP (eds): Manual of Human Embryology. Philadelphia, JB Lippincott Co, 1912. 77. Streeter GL: The developmental alterations in the vascular system of the brain of the human em bryo. Contrib Embryol 8:5-38, 1918. 78. Thompson IM: On the cavum septi pellucidi. J Anat 67:59-77, 1932. 79. Tilney F, Riley HA: The form and function of the central nervous system. New York, Paul В Hoeber, Inc, 1938, ed 3. 80. van de Kamer JC: Histological structure and cy tology of the pineal complex in fishes, amphibi ans, and reptiles. Prog Brain Res 10:30-48, 1965. 81. Vollrath L: Comparative morphology of the verte brate pineal complex. In Kappers JA, Peret P (eds): The Pineal Gland of Vertebrates Including Man. Prog Brain Res 52: 1979, pp 25-38. 82. Vollrath L: The Pineal Organ. Berlin, Springer, 1981. 83. Warwick R, Williams PL (eds): Gray's Anatomy. Philadelphia, WB Saunders Co, 1980, British ed 36. 84. Westergaard E: The cerebral ventricles of the guinea pig during growth. Acta Anat 70:382-402, 1958. 85. Westergaard E: The cerebral ventricles of the golden hamster during growth. Acta Anat 72:533-548, 1969. 86. Westergaard E: The cerebral ventricles of the rat during growth. Acta Anat 74:405-423, 1969. 87. Wolff HG: The cerebral blood vessels—anatomi cal principles. Assoc Res Nerv Ment Dis Proc 18:29-68, 1938. 88. Woollard HH, Weddell G: Arterial vascular pat terns. J Anat 69:25-37, 1934. 89. Yamamoto I, Rhoton AL, Peace RA: Microsurgery of the third venticle: Part 1. Microsurgical anat omy. Neurosurgery 8:333-356, 1981.
3 Microsurgical Anatomy of the Third Ventricular Region Albert L. Rhoton, Jr., M.D.
The third ventricle is located in the center of the head, below the corpus callosum and the body of the lateral ventricle; above the sella turcica, pituitary gland, and midbrain; and between the cerebral hemispheres, thalami, and the walls of the hypothalamus. It is intimately related to the circle of Willis and its branches and the vein of Galen and its tributaries. Selecting an operative approach to a lesion involving the third ventricle requires an understanding of the anatomy of the sella turcica, sphenoid sinus, lateral ventricles, and basal cisterns. This chapter is divided into four sections. The first three sections deal with the neural, arterial, and venous relationships around the third ventricle. The fourth section on the sellar region deals with the structures in the cranial base that are important in performing the subcranial operative approaches to lesions around the third ventricle. The section on neural relationships includes a review of the anatomy of the lateral ventricles and the cisterns surrounding the ten-torium incisura because many lesions involving the third ventricle are approached through these spaces (8, 25, 26, 43, 55). Neural Relationships Third Ventricle The third ventricle is a narrow, funnel-shaped, unilocular, midline cavity. It communicates at its anterosuperior margin with each lateral ventricle through the foramen of Monro and posteriorly with the fourth ventricle through the aqueduct
of Sylvius. It has a roof, a floor, and an anterior, posterior, and two lateral walls (Fig. 3.1). Roof The roof of the third ventricle forms a gentle upward arch, extending from the foramen of Monro anteriorly to the suprapineal recess posteriorly. The roof has four layers: one neural layer formed by the fornix, two thin membranous layers of tela choroidea, and a layer of blood vessels between the sheets of tela choroidea (Figs. 3.2 to 3.4). The upper, or neural, layer is formed by the fornix. The fornix is formed by axons that arise in the floor of the temporal horns from the hippocampi and extend around the thalami to reach the mamillary bodies. The fornix is composed of a body, two anterior limbs called columns, two posterior limbs called crura, and two inferior limbs called fimbria (Fig. 3.2). The initial part of the fornix, the fimbria, arises in the floor of the temporal horn on the ventricular surface of the hippocampal formation and passes posteriorly to become the crus of the fornix. The crus wraps around the posterior surface of the pulvinar of the thalamus and arches superomedial toward the lower surface of the splenium of the corpus callosum. At the junction of the atrium and the body of the lateral ventricle, the paired crura meet to form the body of the fornix, which runs forward along the superomedial border of the thalami in the medial wall of the body of the lateral ventricle. The body forms a gentle arch located between the roof of the third ventricle
Figure 3.1. Midsagittal section of the third ventricle. The floor extends from the optic chiasm (O.Ch.) to the aqueduct of Sylvius and includes the lower surface of the optic chiasm, the infundibulum (Infund.), the infundibular recess (Infund. Recess), the pituitary gland (Pit. Gland), the tuber cinereum (Tuber Cin.), the mamillary bodies (Mam. В.), the posterior perforated substance (Post. Perf. Subst.), and the part of the midbrain anterior to the aqueduct. The anterior wall extends from the optic chiasm to the foramen of Monro (F. Monro) and includes the upper surface of the optic chiasm, the optic recess (O. Recess), the lamina terminalis (Lam. Ter.), the anterior commissure (Ant. Comm.), and the foramen of Monro. The roof extends from the foramen of Monro to the suprapineal recess and is formed by the fornix and the layers of the tela choroidea (Tela), between which course the internal cerebral vein and the medial posterior choroidal artery. The hippocampal commissure (Hippo. Comm.), corpus callosum (Corp. Call.), and septum pellucidum (Sept. Pel.) are above the roof. The posterior wall extends from the suprapineal recess to the aqueduct and includes the habenular commissure (Hab. Comm.), pineal gland, pineal recess, and posterior commissure (Post. Comm.). The oculomotor nerve (HI) exits from the midbrain. The hypothalamic sulcus (Hypothal. Sulc.) forms a groove between the thalamic and hypothalamic (Hypothal.) surfaces of the third ventricle. (From Yamamoto I, Rhoton AL Jr, Peace DA: Microsurgery of the third ventricle: Part 1. Microsurgical anatomy. Neurosurgery 8:334-356, 1981.)
and the floor of the body of each lateral ventricle. The body of the fornix splits into two columns at the anterior margin of the opening of each foramen of Monro into the ipsilateral lateral ventricle. The columns terminate in the mamillary bodies. The crura, just posterior to where they join to form the body of the fornix, are interconnected by a sheet of white matter called the hippocampal commissure. The upper layer of the anterior part of the roof of the third ventricle is formed by the body of the fornix, and the posterior part of the roof is formed by the crura and the hippocampal commissure. The septum pellucidum is attached to the upper surface of the body of the fornix. The septum pellucidum is tallest anteriorly and shortest posteriorly, disappearing near the junction of the body and crura of the fornix. In our studies, the anteroposterior length of the septum pellucidum varied from 28 to 50 mm (55). At the posterior end of the septum pellucidum, the crura and hippocampal commissure fuse with the lower surface of the corpus callosum. The tela choroidea forms two of the three layers in the roof below the layer formed by the fornix (Fig. 3.3). The tela choroidea consists of two thin, semiopaque membranes derived from pia mater, which are interconnected by loosely organized trabeculae. The final layer in the roof is a vascular layer located between the two layers of tela choroidea. The vascular layer consists of the medial posterior choroidal arteries and their branches and the internal cerebral veins and their tributaries. Parallel strands of choroid plexus project downward on each side of the midline from the inferior layer of tela choroidea into the superior part of the third ventricle. The velum interpositum is the space between the two layers of the tela choroidea in the roof of the third ventricle. The velum interpositum is usually a closed space that tapers to a narrow apex just behind the foramen of Monro, but it may infrequently have an opening situated between the splenium and the pineal body that communicates with the quadrigeminal cistern to form the cisterna velum interpositum. The upper layer of tela choroidea is attached to the lower surface of the fornix and the hippocampal commissure (Figs. 3.2 and 3.3). The anterior part of the lower wall is attached to the teniae thalami, small ridges on the free edge of a fiber tract, the striae medullaris thalami, which extends along the superomedial border of the thalamus from the foramen of Monro to the habenular commissure. The posterior part of the lower wall is attached to the superior surface of the pineal body. The internal cerebral veins arise in the anterior part of the velum interpositum, just behind the foramen of Monro, and they exit the velum inter-
positum above the pineal body to enter the quadrigeminal cistern and join the great vein. The lateral margin of the roof is formed by the cleft between the lateral edge of the fornix and the superomedial surface of the thalamus (Fig. 3.2). This narrow cleft between the fornix and the thalamus, which in its entirety is C-shaped, is called the choroidal fissure. The choroid plexus of the lateral ventricle is attached along this fissure. The fornix forms the outer margin of the C-shaped fissure, and the thalamus forms the inner margin. The choroidal fissure is limited in the body of the ventricle by the body of the fornix superiorly and by the thalamus inferiorly, in the atrium by the crus of the fornix posteriorly and the pulvinar anteriorly, and in the inferior horn by the fimbria of the fornix below and the striae terminalis and the thalamus above. The tela cho-roidea forming the pedicle of the choroid plexus of the third ventricle is continuous through the choroidal fissure with the choroid plexus in the lateral ventricle.
Figure 3.2. A-D. Neural relationships. A. Superior view. The upper part of the cerebral hemispheres has been removed to expose the lateral ventricles. The frontal horn (Front. Horn) extends into the frontal lobe (Front. Lobe) anterior to the foramen of Monro (For. Monro). The body of the lateral ventricle (Body Lat. Vent.) is behind the foramen of Monro and above the thalamus. The atrium is located posterior to the thalamus deep to the parietal lobe (Par. Lobe). The occipital horn (Occip. Horn) extends backward from the atrium into the occipital lobe (Occip. Lobe). The temporal horn extends forward from the atrium into the temporal lobe (Temp. Lobe). The lateral wall of the frontal horn is formed by the head of the caudate nucleus (Caudate Nucl. Head); the floor medial to the caudate nucleus is formed by the rostrum of the corpus callosum; the medial wall is formed by the septum pellucidum (Sept. Pell.); the anterior wall is formed by the forceps minor, a bundle composed of fibers crossing in the genu of the corpus callosum; and the roof is formed by the anterior part of the body and the posterior part of the genu of the corpus callosum. The lateral wall of the body of the lateral ventricle is formed by the body of the caudate nucleus: the floor is formed by the superior surface of the thalamus; the roof is formed by the body of the corpus callosum; and the medial wall is formed by the septum pellucidum above and the body of the fornix below. The fornix passes along and is separated from the superomedial margin of the thalamus by a narrow cleft, the choroidal fissure (Chor. Fiss.), along which the choroid plexus (Chor. Plex.) is attached. The choroid plexus has been removed along the posterior part of the choroidal fissure on the right side. The atrium and occipital horn have the same structures in their walls. The lateral part of the anterior wall of the atrium is formed by the pulvinar, and the medial part is formed by the cms of the fornix. The superior part of the medial wall is formed by a large fiber tract from the splenium, the forceps major, which interconnects the occipital lobes and produces a prominent bulge in the medial wall called the bulb of the corpus callosum. The inferior part of the medial wall is formed by the calcar avis, a prominence overlying the depths of the calcarine sulcus. The medial part of the floor of the atrium is formed by the posterior part of the hippocampus, the smooth prominence overlying the hippocampal formation, and the lateral part of the floor is formed by the collateral trigone (Collat. Trig.), a prominence that overlies the collateral sulcus on the inferior surface of the temporal lobe. The lateral wall just posterior to the pulvinar is formed by the tail of the caudate nucleus as it wraps around the lateral edge of the pulvinar, and the part behind the caudate nucleus is formed by the tapetum of the corpus callosum, a fiber tract from the body and splenium that passes around the lateral margin of the atrium and
the temporal horn. The parietooccipital sulcus (Par. Occip. Sufc.) extends deeply into the medial surface of the hemisphere above the cuneus. The superior sagittal sinus (Sup. Sag. Sinus) courses in the outer periphery of the falx. B. Enlarged view of the region of the foramen of Monro. The choroid plexus has been removed from its attachment along the choroidal fissure on the right side. The anterior thalamic tubercle (Ant. Thal. Tuber.), which overlies the anterior nucleus of the thalamus, bulges upward at the posterior margin of the foramen of Monro. The columns of the fornix pass anterior and superior to the foramen of Monro. The striothalamic sulcus (Str. Thal. Sufc.) separates the caudate nucleus and the thalamus. C. The level of the cross section has been extended deeper into the right cerebral hemisphere through the caudate nucleus and the thalamus. The posterior part of the body and the superior part of the crus of the fornix have been removed on the right side to expose the tela choroidea (Tela) in the roof of the third ventricle. The anterior limb of the internal capsule (Int. Cap. Ant. Limb) is located between the head of the caudate nucleus and the lentiform nuclei (Lent. Nucl.). The posterior limb of the internal capsule (Int. Cap. Post. Limb) is located between the thalamus and the lentiform nucleus. The genu of the internal capsule (Genu Int. Cap.) touches the lateral wall of the ventricle between the caudate nucleus and the thalamus. The calcar avis and the bulb of the corpus callosum (Bulb Corp. Call.) bulge into the medial wall of the atrium and the occipital horn. The collateral trigone bulges in the floor of the atrium. The glomus, a large tuft of choroid plexus, is attached in the atrium between the fornix and the pulvinar. D. The fornix has been divided at the junction of its body and the columns above the foramen of Monro and reflected backward to expose the velum interpositum (Vel. Interpos.) located between the layers of tela choroidea in the roof of the third ventricle. The medial part of the floor of the temporal horn (Temp. Horn) is formed by the hippocampus, the smooth prominence overlying the hippocampal formation, and the lateral part of the floor is formed by the collateral eminence (Coff. Eminence), which overlies the deep part of the collateral sulcus. The pes hippocampus (Pes Hipp.) is the bulbous, digitated anterior part of the hippocampal formation. The hippocampal commissure (Hipp. Comm.) interconnects the medial margins of the crura of the fornix.
Figure 3.2. E-H. Neural relationships (continued). E. The body and anterior part of the crura of the fornix have been removed to expose the third ventricle (3 Vent.) and the quadrigeminal plate and pineal regions. The massa intermedia (Massa Inter.) extends into the third ventricle anteriorly, and the habenular (Hab. Comm.) and posterior (Post. Comm.) commissures cross the posterior part of the third ventricle. The habenular commissure is connected anteriorly with the habenular trigones (Hab. Trig.) and the striae medullaris thalami (Str. Med. Thal.). The pineal body projects posteriorly above the superior (Sup. Coll.) and inferior colliculi (Inf. Coll.). The subiculum, the rounded medial edge of the parahippocampal gyrus, is exposed lateral to the colliculi. The inferior sagittal sinus (Inf. Sag. Sinus) courses in the inferior margin of the falx and joins the straight sinus at the tentorial apex (Apex. Tent.). F. Right superolateral view. The frontal and parietal operculae have been removed to expose the sylvian fissure (Sylvian Fiss.) and the transverse temporal gyri (Transv. Temp. Gyr.). G. Further removal of deep cerebral tissue exposes the hippocampal formation and the collateral eminence in the floor of the temporal horn. The collateral eminence is continuous posteriorly with the collateral trigone in the floor of the atrium. H. Enlarged view to show structures in the medial wall of the atrium and the floor of the temporal horn. The fimbria of the fornix arises on the surface of the hippocampus and is continuous posteriorly with the crus of the fornix. (From Ono M, Rhoton AL Jr, Peace DA, Rodriguez R: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984.)
Figure 3.3. Superior view of stepwise dissection of layers in the roof of the third ventricle. A. Superior part of cerebral hemispheres, with the corpus callosum (Corp. Call.) including a part of the splenium removed to expose the choroid plexus (Ch. PL) of the lateral ventricle. A superior choroidal vein (S. Ch. V.) passes along the free edge of the choroid plexus. The glomus of the choroid plexus projects into the atrium. The caudate nucleus projects into the anterior horn and body of the lateral ventricle. The septum pellucidum (Sept. Pel.) forms the medial walls of the lateral ventricles. The bodies of the fornix are medial to the choroid plexus in the bodies of the lateral ventricle. B. The body of the fornix and septum pellucidum are divided and lifted up to expose the tela choroidea (Tela) in the roof of the third ventricle. C. The body of the fornix removed at the junction with the columns of the fornix (Columns). The medial posterior choroidal arteries (M. P. Ch. A., broken line) course between the upper and lower layers of the tela choroidea. D. The tela choroidea was removed to expose the parallel strands of choroid plexus in the roof of the third ventricle (Ch. PI.-3V) and the interior third ventricle (3V). Part of the splenium of the corpus callosum was removed to show the pineal body. Right and left thalami form the lateral walls of the third ventricle. (From Fujii K, Lenkey C, Rhoton AL Jr: Microsurgical anatomy of the choroidal arteries: Lateral and third ventricles. J Neurosurg 52:165-188, 1980.)
Figure 3.4. Inferior views of the roof of the third ventricle. A. The portion of the third ventricle and cerebral hemispheres below the roof of the third ventricle has been removed to provide an inferior view of the roof. The choroid plexus (Ch. Pl.) hangs downward from the tela choroidea into the superior part of the third ventricle. The internal cerebral veins (Int. Сет. V) and the medial posterior choroidal arteries (Med. Post. Ch. A.) course below the fornix and between the two halves of the thalamus. The choroid plexus in the roof of the third ventricle is continuous through the foramen of Monro (F. Monro) with the choroid plexus in the lateral ventricle. The body of the fornix splits into two columns at the anterior margin of the foramen of Monro. The anterior commissure (Ant. Comm.) is anterior to the columns of the fornix. The massa intermedia (Massa Inter.) projects into the upper partof the third ventricle. The internal cerebral veins join to form the vein of Galen (V. Galen) in the posterior part of the roof. B. The medial edges of the thalamus on each side have been removed to provide a wider view of the body and crura of the fornix, the medial posterior choroidal arteries, and the internal cerebral veins. The crura of the fornix pass posteriorly below the sple-nium of the corpus callosum (Splenium). The internal cerebral veins pass posteriorly above the pineal gland (Pineal) and join to form the vein of Galen. The choroidal fissure (Chor. Fiss.) is a groove on each side between the fornix and the thalamus. The upper part of the stalk of the pineal gland formed by the habenular commissure (Hab. Comm.) has been preserved. C. The choroidal fissure on each side has been opened wider by the removal of more of the medial surface of the thalamus. The lateral ventricles (Lat. Vent.) are seen through the enlarged choroidal fissures. The choroid plexus of the lateral ventricles is exposed and branches of the lateral posterior choroidal arteries (Lat. Post. Ch. A.) and the superior choroidal veins (Sup. Ch. V.) are seen on the surface of the choroid plexus. The hippocampal commissure (Hippo. Comm.) interconnects the medial margins of the crura of the fornix above the pineal gland. The pineal gland is below the splenium and the hippocampal commissure. The posterior part of the crura of the fornix and the hippocampal commissure are adherent to the lower margin of the splenium of the corpus callosum. (From Yamamoto I, Rhoton AL Jr, Peace DA: Microsurgery of the third ventricle: Part 1. Microsurgical anatomy. Neurosurgery 8:334-356, 1981.)
Floor The floor extends from the optic chiasm anteriorly to the orifice of the aqueduct of Sylvius posteriorly (Figs. 3.1 and 3.5). The anterior half of the floor is formed by diencephalic structures and the posterior half is formed by mesence-phalic structures. When viewed from inferiorly, the structures forming the floor from anterior to posterior include the optic chiasm, the infundibulum of the hypothalamus, the tuber cinereum, the mamil-lary bodies, the posterior perforated substance, and (most posteriorly) the part of the tegmentum of the midbrain located above the medial aspect of the cerebral peduncles. The optic chiasm is located at the junction of the floor and the anterior wall of the third ventricle. The chiasm slopes posteriorly and superiorly from its junction with the optic nerves. The inferior surface of the chiasm forms the anterior part of the floor and the superior surface forms the lower part of the anterior wall. The optic tracts arise from the posterolateral margin of the chiasm and course obliquely away from the floor toward the lateral margin of the midbrain. The infundibulum, tuber cinereum, mamillary bodies, and posterior perforated substance are located in the space limited anteriorly and laterally by the optic chiasm and tracts and posteriorly by the cerebral peduncles. The infundibulum of the hypothalamus is a hollow, funnel-shaped structure located between the optic chiasm and the tuber cinereum. The pituitary gland (hypophysis) is attached to the infundibulum, and the axons in the infundibulum extend to the posterior lobe of the hypophysis. The tuber cinereum is a prominent mass of hypothlamic gray matter located anterior to the mamillary bodies. The tuber cinereum merges anteriorly into the infundibulum. The tuber cinereum around the base of the infundibulum is raised to form a prominence called the median eminence. The mamillary bodies form paired, round prominences posterior to the tuber cinereum. The posterior perforated substance is a depressed, punctuated area of gray matter located in the interval between the mamillary bodies anteriorly and the medial surface of the cerebral peduncles posteriorly. The posterior part of the floor extends posterior and superior to the medial part of the cerebral peduncles and superior to the tegmentum of the midbrain. When viewed from above and inside the third ventricle, the optic chiasm forms a prominence at the anterior margin of the floor (Fig. 3.5). The infundibular recess extends into the infundibu-lum behind the optic chiasm. The mamillary bodies form paired prominences on the inner surface of the floor posterior to the infundibular recess. The part of the floor between the mamillary bodies and the aqueduct of Sylvius has a smooth surface, which is concave from side to side. This smooth surface lies above the posterior perforated substance anteriorly and the medial part of the cerebral peduncles and the tegmentum of the midbrain posteriorly. Anterior Wall The anterior margin of the third ventricle extends from the foramina of Monro above to the optic chiasm below (Figs. 3.1, 3.5, and 3.6). Only the lower two-thirds of the anterior surface is seen on the external surface of the brain;
the upper one-third is hidden posterior to the rostrum of the corpus callosum. The part of the anterior wall visible on the surface is formed by the optic chiasm and the lamina terminalis. The lamina terminalis is a thin sheet of gray matter and pia mater that attaches to the upper surface of the chiasm and stretches upward to fill the interval between the optic chiasm and the rostrum of the corpus callosum. When viewed from within, the boundaries of the anterior wall from superiorly to inferiorly are formed by the columns of the fornix, foramina of Monro, anterior commissure, lamina terminalis, optic recess, and optic chiasm. The foramen of Monro on each side is located at the junction of the roof and the anterior wall (Figs. 3.1 and 3.5). The foramen is a ductlike canal that opens between the fornix and the thalamus into the lateral ventricle and extends inf eriorly below the fornix into the third ventricle as a single channel. The foramen of Monro is bounded anteriorly by the junction of the body and the columns of the fornix and posteriorly by the anterior pole of the thalamus (Figs. 3.1, 3.2, and 3.5). The size and shape of the foramina of Monro depend on the size of the ventricles: if the ventricles are small, each foramen is a crescent-shaped opening bounded anteriorly by the concave curve of the fornix and posteriorly by the convex anterior tubercle of the thalamus. As the ventricles enlarge, the foramen on each side becomes rounder. The structures that pass through the foramen are the choroid plexus, the distal branches of the medial posterior choroidal arteries, and the internal cerebral, thalamostriate, superior choroidal, and septal veins. The anterior commissure is a compact bundle of myelinated nerve fibers that cross the midline in front of the columns of the fornix. The anter-oposterior diameter of the anterior commissure varies from 1.5 to 6.0 mm (55). The distance from the posterior end of the anterior commissure to the anterior border of the foramen of Monro ranges from 1.0 to 3.5 mm (average, 2.2 mm). The distance from the upper edge of the optic chiasm to the anterior border of the anterior commissure ranges from 8 to 12 mm (average, 10 mm). The lamina terminalis fills the interval between the anterior commissure and the optic chiasm (Fig. 3.1). The lamina attaches to the midportion of the superior surface of the chiasm, leaving a small cleft between the upper half of the chiasm and the lamina, called the optic recess.
Figure 3.5. Anterosuperior views of the third ventricle. A. Part of the anterior wall of the third ventricle and the anterior part of the cerebral hemispheres have been removed. The septum pellucidum (Sept. Pel.) is attached to the upper margin of the fornix. The optic chiasm (O. Ch.) and nerves (O. N.) are at the lower margin of the anterior wall, and the optic tracts (O. Tr.) extend laterally below the floor of the third ventricle (3V). The choroid plexus (Ch. Pi.) is attached over the surface of the thalamus on each side. The thalamostriate vein on each side (Thal. Sir. V.) courses forward between the caudate nucleus (Caudate Nucl.) and the thalamus. The infundibular recess (Infund. Recess) extends downward posterior to the optic chiasm and anterior to the mamillary bodies (Mam. В.). The midportion of the anterior commissure (Ant. Comm.) has been removed to expose the columns of the fornix anterior to the foramina of Monro. The body and columns of the fornix join anterior to the foramina of Monro. The corpus callosum (Corp. Call.) is above the lateral ventricles. B. The septum pellucidum and the medial part of the body of the fornix have been removed to expose each foramen of Monro (F. Monro) and the full length of the floor of the third ventricle. The floor extends from the aqueduct of Sylvius (Aqueduct) posteriorly to the optic chiasm anteriorly. The habenular commissure (Hab. Comm.) forms the upper margin of the stalk of the pineal gland (Pineal), and the posterior commissure (Post. Comm.) forms the lower stalk of the pineal gland. The pineal recess is between the two commissures. A branch of the medial posterior choroidal artery (Med. Post. Ch. A.) passes under the fornix to join the choroid plexus. The superior choroidal veins (Sup. Ch. V.) course on the surface of the choroid plexus. The crus of the fornix is above the pineal gland. The optic recess (O. Recess) extends anterior to the upper one-half of the optic chiasm. C. The body of the fornix has been removed to show the internal cerebral veins (Int. Сет. V.) and the medial posterior choroidal arteries in the roof of the third ventricle. A ventricular vein drains into the internal cerebral vein. The internal cerebral veins join to form the vein of Galen (V. Galen). D. Enlarged view shows the habenular commissure above the posterior commissure below the pineal recess. The internal cerebral veins and the medial posterior choroidal arteries pass forward and intermingle along the superolateral margin of the pineal gland. The columns of the fornix are anterior to the foramen of Monro. The crura of the fornix are above the pineal gland and below the splenium (Splenium) of the corpus callosum. (From Yamamoto I, Rhoton AL Jr, Peace DA: Microsurgery of the third ventricle: Part 1. Microsurgical anatomy. Neurosurgery 8:334-356, 1981.)
Figure 3.6. Neural relationships. Posterior and midsagittal views. A. Posterior view with orientation as shown in the insert. The right occipital lobe (Occip. Lobe) has been retracted away from the tentorium (Tent.) and the straight (Str. Sinus) and inferior sagittal (Inf. Sag. Sinus) sinuses to expose the tentorial apex (Apex Tent.). The tentorium behind its free edge (Tent. Edge) has been opened to expose the superior surface of the cerebellum. The pineal body projects backward above the superior colliculi (Sup. Coll.). The habenular commissure (Hab. Comm.) crosses the posterior end of the third ventricle. The suprapineal recess (Sup. Pineal Recess) of the third ventricle extends posteriorly between the superior surface of the pineal body and the tela choroidea (Tela) in the roof of the third ventricle. The isthmus of the cingulate gyrus (Isthmus) is a narrow cortical bridge between the cingulate (Cing. Gyr.) and parahippocampal (Parahippo. Gyr.) gyri. The medial geniculate body (Med. Gen. Body) and brachium of the inferior colliculus (Brach. Inf. Coll.) are posterior to the cerebral peduncle (Ped.). B. Posterior view. The posterior part of the cerebral hemispheres has been removed to expose the posterior part of the third ventricle (3rd Vent.) and the pineal region, where the deep veins converge on the vein of Galen. The atria and the bodies of the lateral ventricles (Body Lat. Vent.) are deep within the parietal lobe (Par. Lobe), below the corpus callosum (Corp. Call.). The caudate nuclei (Caudate Nucl.) are lateral to the thalami. The crura of the fornix pass around the pulvinar. Choroid plexus (Chor. Plex.) is attached along the choroidal fissure, the cleft between the fornix and the thalamus. The subiculum is the rounded medial edge of the parahippocampal gyrus of the temporal lobe (Temp. Lobe). The collateral sulcus (Coll. Sulc.) is lateral to the parahippocampal gyrus. C. Posterior view below the tentorium. The cerebellum was removed by dividing the superior (Sup. Cer. Ped.), middle (Mid. Cer. Ped.), and inferior (Inf. Сет. Ped.) cerebellar peduncles just dorsal to the floor of the fourth ventricle (4 Vent.}. The structures just above the posterior part of the incisura include the splenium, the crura of the fornix, the pineal body, and the superior and inferior colliculi (Inf. Coll.). The trochlear nerves (V) arise below the inferior colliculi and pass around the brain stem. The superior medullary velum (Sup. Med. Velum) forms the medial part of the roof of the fourth ventricle. The flocculus and the rhomboid lip border the lateral recess (Lat. Recess) of the fourth ventricle and extend laterally above the glossopharyngeal (IX), vagus (X), and accessory (XI) nerves. The hippocampal commissure (Hippo. Comm.) is a fiber tract connecting the medial margins of the crura of the fornix. D. Midsagittal section. The rostrum of the corpus callosum is continuous below the anterior commissure (Ant. Comm.) with the lamina termin-alis (Lam. Term.). The massa intermedia (Massa Inter.) protrudes into the central part of the third ventricle. The pineal body is attached to the posterior part of the third ventricle by a stalk composed of the habenular and posterior commissures (Post. Comm.). The striae medullaris thalami (Str. Med. Thal.) course on the medial surface of the thalamus from the foramen of Monro to the habenular commissure. The septum pellucidum and the part of the fornix above the thalamus have been removed. The choroid plexus is attached along the choroidal fissure (Chor. Fiss.), the cleft between the fornix and thalamus. The caudate nucleus (Caudate Nucl.) is exposed in the far wall of the frontal horn (Front. Horn) and the body of the right lateral ventricle. The cms of the fornix is fused for a short distance to the lower surface of the splenium. The brain stem has been sectioned at the level of the cerebral peduncle. The dentate gyrus (Dentate Gyr.) forms a beaded longitudinal cortical strip above the parahippocampal gyrus. The hippocampal sulcus (Hippo. Sulc.) separates the dentate and parahippocampal gyri. The fimbria of the fornix arises on the surface of the hippocampal formation in the floor of the temporal horn and forms a longitudinal strip that is separated from the dentate gyrus by the fimbriodentate sulcus (Fimb. Dentate Sulc.). The fimbria is continuous posteriorly with the crus of the fornix. The column of the fornix passes anterior and superior to the foramen of Monro (For. Monro) to terminate in the mamillary bodies (Mam. Body). The lateral geniculate body (Lat. Gen. Body) protrudes from the inferior surface of the thalamus. The infundibular recess (Infund. Recess) extends inferiorly into the base of the infundibulum. The hypothalamic sulcus (Hypothal. Sulc.) separates the thalamus and hypothalamus. (From Ono M, Rhoton AL Jr, Peace DA, Rodriguez R: Micro-surgical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984.)
Figure 3.7. A-D. Tentorial incisura: stepwise dissection, anterior-superior view. A. The anterior part of the frontal lobe has been removed to expose the anterior incisural space. The section of the frontal lobe passes adjacent to the septum pellucidum (Sept. Pell.) and through the rostrum and genu of the corpus callosum (Corp. Call.), the caudate nucleus (Caudate Nucl.), the putamen, and the frontal horn of the lateral ventricle (Lat. Vent.). The choroid plexus (Chor. Plex.) extends through the foramen of Monro (For. Monro). The anterior incisural space is located anterior to the midbrain and extends upward around the optic chiasm, lamina terminalis (Lam. Term.), and anterior part of the third ventricle (3 Vent). The optic tract (Optic Tr.) extends posteriorly above the oculomotor nerve (HI). The infundibulum (Infund.) of the pituitary gland passes through the diaphragma sellae (Diaph.). The carotid artery (Car. A.) has been divided below the optic nerve (Optic N.). The cingulate gyri (Cing. Gyr.) lie above the corpus callosum. A septal vein (Septal V.) courses on the septum pellucidum. The paraterminal (Paraterm. Gyr.) and paraolfactory gyri (Paraolf. Gyr.) are above the gyrus rectus (Gyr. Rectus). The semilunar (Semilunar Gyr.) and ambient gyri (Ambient Gyr.) are on the superior surface of the temporal lobe bordering the sylvian fissure (Sylvian Fiss.). The anterior limb of the internal capsule (Int. Cap. Ant. Limb) is situated between the caudate nucleus and the putamen. The oculomotor nerve enters the dura mater posterior to the anterior clinoid process (Ant. Ciinoid). B. The transverse section has been extended behind the foramen of Monro to include part of the cerebral peduncle (Ped.). The posterior part of the right optic nerve and the right half of the optic chiasm have been removed to expose the posterior part of the anterior incisural space. The substantia nigra (Subst. Nigra) and red nucleus (Red. NucL) are at the level of the transection of the midbrain. The thalamus and internal capsule are located directly above the cerebral peduncle. The middle incisural space is between the uncus and the midbrain. The parahippocampal gyrus (Parahippo. Gyr.) and uncus protrude medial to the tentorial edge (Tent. Edge). The caudate nucleus extends from the lateral part of the frontal horn around the atrium into the roof of the temporal horn (Temp. Horn), where the tail of the caudate nucleus is located. The lentiform nucleus formed by the globus pallidus (Globus Pall.) and the putamen are lateral to the internal capsule. The collateral sulcus (Coll. Sulc.) extends into the temporal lobe near the lateral part of the temporal horn. The anterior hippocampal sulcus (Ant. Hippo. Sulc.) separates the uncus and parahippocampal gyrus. The anterior commissure (Ant. Comm.) is in the anterior, the mamillary bodies (Mam. Body) are in the inferior, and the massa intermedia (Massa Int.) is in the central part of the third ventricle. The choroid plexus is attached in the lateral and third ventricle along the choroidal fissure (Chor. Fiss.) located between the fornix and the thalamus. The callosal sulcus (Call. Sulc.) separates the corpus callosum and the cingulate gyrus. C. Enlarged view. The anterior choroidal (Ant. Chor. A.) and posterior cerebral arteries (P. C. A.) pass around the cerebral peduncle medial to the uncus and the parahippocampal gyrus. The tela choroidea and choroid plexus in the temporal horn are attached to the tenia choroidea and the tenia fimbriae along the choroidal fissure in the medial part of the temporal horn. The pontomesencephalic sulcus (Pon. Mes. Sulc.) marks the junction of the midbrain and pons. The oculomotor nerve enters the dura lateral to the posterior clinoid process (Post. Clinoid). The trochlear nerve (IV) passes below the uncus. D. The coronal section extends just behind the posterior (Post. Comm.) and habenular (Hab. Comm.) commissures into the base of the pineal body (Pineal). The internal cerebral veins (Int. Cer. V.) course in the velum interpositum located between the two layers of the tela choroidea (Tela) in the roof of the third ventricle. The upper layer of tela is attached to the lower surface of the fornix. The mesial temporal structures that form the lateral wall of the middle incisural space include the subiculum, the rounded medial surface of the parahippocampal gyrus, which blends into the Ammon's horn (a curled horn-shaped collection of gray matter in the hippocampal formation), the dentate gyrus (Dentate Gyr.), a strip of cortex located above the subiculum, and the fimbria of the fornix, a fiber tract that arises in the floor of the temporal horn on the surface of the hippocampal formation. The pineal body projects into the posterior incisural space below the suprapineal recess and the internal cerebral veins. The medial (Med. Gen. Body) and lateral geniculate bodies (Lat. Gen. Body) are dorsolateral to the cerebral peduncle.
midbrain anteriorly, the splenium above, and the cerebellum below. The vein of Galen (V. of Galen) enters the straight sinus. The trochlear nerve arises below the inferior colliculus (Inf. Coll.) and passes forward in the cerebellomesencephalic fissure (Cer. Mes. Fiss.) situated between the cerebellum and midbrain. The paraterminal and paraolfactory gyri, the gyrus rectus, and the anterior (Ant. Paraolf. Sulc.) and posterior paraolfactory (Post. Paraolf. Sulc.) sulci are below the rostrum. The optic recess extends inferiorly between the optic chiasm and the lamina terminalis, and the infundibular recess (Infund. Recess) extends into the infundibulum behind the chiasm. The layer of tela choroidea that forms the upper wall of the velum interpositum is adherent to the lower margin of the body and cms of the fornix. The layer of tela choroidea that forms the lower wall of the velum interpositum is attached anteriorly to the striae medullaris thalami (Str. Med. Thal.) and posteriorly it adheres to the superior margin of the pineal body. The striae medullaris thalami extend forward from the habenular commissure along the superomedial margin of the thalamus. The lateral mesencephalic sulcus (Lat. Mes. Sulc.) extends along the posterior edge of the cerebral peduncle, and the interpeduncular fossa (Interped. Fossa) is between the cerebral peduncles. G. Enlarged view. The tela choroidea and the deep veins have been removed. The choroidal fissure is located between the fornix and the thalamus. The isthmus of the cingulate gyrus (Isthmus) is a cortical bridge between the parahippocampal and cingulate gyri. The fasciolar gyrus and the gyri Andreae Retzii form a beaded collection of gray matter that is continuous below with the dentate gyrus and above with the indusium griseum, a thin layer of gray matter on the surface of the corpus callosum. The quadrangular lobule (Quad. Lobule) of the cerebellum is behind the colliculi. (From Ono M, Ono M, Rhoton AL Jr, Barry M: Microsurgical anatomy of the region of the tentorial incisura. J Neurosurg 60:365-399, 1984.)
Figure 3.7. E-G. Tentorial incisura (continued). E. The level of the section has been extended behind the pineal body to expose the posterior incisural space and the vein of Galen. The pineal body projects into the posterior incisural space. The parahippocampal gyrus overlies the free edge and forms part of the lateral wall of the posterior incisural space. The fibers arising on the surface of the hippocampal formation pass through the fimbria to the crus of the fornix, which extends upward toward the splenium. The parahippocampal gyrus is separated from the dentate gyrus by the hippocampal sulcus (Hipp. Sulc.), and the fornix is separated from the dentate gyrus by the fim-briodentate sulcus (Fimb. Dent. Sulc.). F. The right cerebral hemisphere has been removed to expose all of the posterior incisural space located between the apex of the tentorium (Apex. Tent.) posteriorly, the
Posterior Wall The posterior wall of the third ventricle extends from the suprapineal recess above to the aqueduct of Sylvius below (Figs. 3.1, 3.5, and 3.6). When viewed from anteriorly and within the third ventricle, it consists, from above to below, of the suprapineal recess, the habenular commissure, the pineal body and its recess, the posterior commissure, and the aqueduct of Sylvius (Fig. 3.5). The suprapineal recess projects posteriorly between the upper surface of the pineal gland and the lower layer of tela choroidea in the roof. The pineal gland extends posteriorly into the quadrigeminal cistern from its stalk. The stalk of the pineal gland has a cranial and a caudal lamina. The habenular commissure, which interconnects the habenulae, crosses the midline in the cranial lamina, and the posterior commissure crosses in the caudal lamina. The pineal recess projects posteriorly into the pineal body between the two laminae. The shape of the orifice of the aqueduct of Sylvius is triangular: the base of the triangle is on the posterior commissure and the other two limbs are formed by the central gray matter of the midbrain. When viewed from posteriorly, the only structure in the posterior wall is the pineal body (Fig. 3.6). The pineal gland projects posteriorly into the quadrigeminal cisterns and is concealed by the splenium of the corpus callosum above, the thalamus laterally, and the quadrigeminal plate and the vermis of the cerebellum inferiorly. Lateral Wall The lateral walls are not visible on the external surface of the brain, but are hidden between the cerebral hemispheres (Figs. 3.1, 3.6, and 3.7). They are formed by the hypothalamus inferiorly and the thalamus superiorly. The lateral walls have an outline like the lateral silhouette of a bird's head with an open beak. The head is formed by the oval medial surface of the thalamus; the open beaks, which project anteriorly and inferiorly, are represented by the recesses in the hypothalamus: the pointed upper beak is formed by the optic recess and the lower beak is formed by the infundibular recess. The hypo-thalamic and thalamic surfaces are separated by the hypothalamic sulcus, a groove that is often
ill-defined and extends from the foramen of Monro to the aqueduct of Sylvius. The superior limit of the thalamic surfaces of the third ventricle is marked by narrow, raised ridges, known as the striae medullaris thalami. These striae extend forward from the habenulae along the superomedial surface of the thalamus at the site of the attachment of the lower layer of the tela choroidea. The habenulae are small eminences on the dorsomedial surfaces of the thalamus just in front of the pineal gland. The habenulae are connected across the midline in the rostral half of the stalk of the pineal gland by the habenular commissure. The massa intermedia projects into the upper one-half of the third ventricle and often connects the opposing surfaces of the thalamus. The massa intermedia was present in 76% of the brains examined and was located 2.5 to 6.0 mm (average, 3.9 mm) posterior to the foramen of Monro (55). The columns of the fornix form distinct prominences in the lateral walls of the third ventricle just below the foramen of Monro, but inferiorly they sink below the surface. Lateral Ventricles An understanding of the anatomy of the lateral ventricles is important because third ventricular tumors are commonly approached through the lateral ventricles. Each of the paired lateral ventricles is a C-shaped cavity that wraps around the thalamus and is situated deep within the cerebrum (Figs. 3.2, 3.3, and 3.7). Each ventricle has five parts: the frontal, temporal, and occipital horns and the body and atrium. Each of these five parts has medial and lateral walls, a roof, and a floor; in addition, the frontal and temporal horns and the atrium have anterior walls. These walls are formed predominantly by the thalamus, septum pellucidum, deep cerebral white matter, corpus callosum, and two C-shaped structures, the caudate nucleus and fornix, which wrap around the thalamus. The thalamus is located in the center of the lateral ventricle. Each lateral ventricle wraps around the superior, inferior, and posterior surfaces of the thalamus. The body of the lateral ventricle is above the thalamus, the atrium and occipital horn are posterior to the thalamus, and the temporal horn is inferior to the thalamus. The superior surface of the thalamus forms the floor of the body, the posterior surface of the thalamus forms the anterior wall of the atrium, and the inferior surface of the thalamus forms the medial part of the roof of the temporal horn. The caudate nucleus is an arched, C-shaped,
cellular mass that wraps around the thalamus and constitutes an important part of the wall of the lateral ventricle. It has a head, body, and tail. The head bulges into the lateral wall of the frontal horn and the body of the lateral ventricle. Its body forms part of the lateral wall of the atrium, and its tail extends from the atrium into the roof of the temporal horn and is continuous with the amygdaloid nucleus near the tip of the temporal horn. In the body of the lateral ventricle, the caudate nucleus is superolateral to the thalamus; in the atrium, it is posterolateral to the thalamus; and, in the temporal horn, it is inferolateral to the thalamus. The striae terminalis, a fiber tract that runs parallel and deep to the thalamostriate vein, arises in the amygdaloid nucleus and courses along the border between the caudate nucleus and the thalamus in the wall of the lateral ventricle. The fornix is another C-shaped structure that wraps around the thalamus in the wall of the ventricle. The fornix has four parts (the fimbria, crus, body, and columns) and extends from the floor of the temporal horn around the thalamus to the mamillary bodies. In the body of the lateral ventricle, the body of the fornix is in the lower part of the medial wall; in the atrium, the crus of the fornix is in the medial part of the anterior wall; and, in the temporal horn, the fimbria of the fornix is in the medial part of the floor. The corpus callosum, which forms the largest part of the ventricular walls, contributes to the wall of each of the five parts of the lateral ventricle. The corpus callosum has two anterior parts, the rostrum and genu; a central part, the body; and a posterior part, the splenium. The rostrum is situated below and forms the floor of the frontal horn. The genu gives rise to a large fiber tract, the forceps minor, which forms the anterior wall of the frontal horn as it sweeps obliquely forward and lateral to connect the frontal lobes. The genu and the body of the corpus callosum form the roof of both the frontal horn and the body of the lateral ventricle. The splenium gives rise to a large tract, the forceps major, which forms a prominence, the bulb, in the upper medial wall of the atrium and the occipital horn as it sweeps posteriorly to connect the occipital lobes. Another fiber tract, the tapetum, which arises in the posterior part of the body of the corpus callosum, sweeps laterally and inferiorly to form the roof and lateral walls of the atrium and the temporal and occipital horns. The septum pellucidum, which is composed of paired laminae, separates the frontal horns and bodies of the lateral ventricles in the midline. In
the frontal horn, the septum pellucidum is attached to the rostrum of the corpus callosum below, the genu anteriorly, and the body above. In the body of the lateral ventricle, the septum is attached to the body of the corpus callosum above and the body of the fornix below. The septum pellucidum disappears posteriorly where the body of the fornix meets the splenium. There may be a cavity, the cavum septum pellucidum, in the midline between the laminae of the septum pellucidum. The choroid plexus in the lateral ventricle has a Cshaped configuration that parallels the fornix (Figs. 3.1 to 3.3) (8). It is attached along the choroidal fissure, a narrow cleft between the fornix and the thalamus, in the medial part of the body, atrium, and temporal horn. The choroid plexus extends through the foramen of Monro into the roof of the third ventricle. In the atrium, the choroid plexus has a prominent triangular tuft called the glomus. The edges of the thalamus and fornix bordering this fissure have small ridges, the teniae, along which the tela choroidea, the membrane in which the choroid plexus arises, is attached. The choroidal fissure extends from the foramen of Monro along the medial wall of the body, atrium, and temporal horn to its inferior termination, the inferior choroidal point, located just behind the tip of the temporal lobe and uncus. The veins coursing in the walls of the ventricles exit the ventricles by passing, in a subependymal location, through the margin of this fissure to reach the internal cerebral, basal, or great veins. The frontal horn, the part of the lateral ventricle located anterior to the foramen of Monro, has a medial wall formed by the septum pellucidum, an anterior wall formed by the genu of the corpus callosum, a lateral wall composed of the head of the caudate nucleus, and a narrow floor formed by the rostrum of the corpus callosum (Fig. 3.2). The columns of the fornix, as they pass anterior to the foramen of Monro, are in the posteroinfer-ior part of the medial wall. The body of the lateral ventricle extends from the posterior edge of the foramen of Monro to the point where the septum pellucidum disappears and the corpus callosum and fornix meet. The roof is formed by the body of the corpus callosum, the medial wall by the septum pellucidum above and the body of the fornix below, the lateral wall by the body of the caudate nucleus, and the floor by the thalamus (Fig. 3.2). The atrium and occipital horn together form a roughly triangular cavity, with the apex posteriorly in the occipital lobe and the base anteriorly
on the pulvinar (Figs. 3.2, 3.6, and 3.7). The atrium opens anteriorly above the thalamus into the body, anteriorly below the thalamus into the temporal horn, and posteriorly into the occipital horn. The roof of the atrium is formed by the body and tapetum of the corpus callosum. The medial wall has an upper part formed by the bulb of the corpus callosum and a lower part formed by the calcar avis, a prominence overlying the deepest part of the calcarine sulcus. The lateral wall has an anterior part formed by the caudate nucleus as it wraps around the lateral margin of the pulvinar and a posterior part formed by the fibers of the tapetum as they sweep anteroinfer-iorly along the lateral margin of the ventricle. The anterior wall has a medial part composed of the crus of the fornix as it wraps around the posterior part of the pulvinar and a lateral part formed by the pulvinar of the thalamus. The floor has a medial part composed of the posterior part of the hippocampus, the prominence overlying the hippocampal formation, and a lateral part formed by the collateral trigone, the triangular prominence deep to the posterior end of the collateral sulcus. In the atrium, the choroid plexus has a prominent tuft called the glomus. The occipital horn extends posteriorly into the occipital lobe from the atrium. Its medial wall is composed of the bulb and the calcar avis, the roof and lateral wall are formed by the tapetum, and the floor is formed by the collateral trigone. The temporal horn extends forward from the atrium below the pulvinar into the medial part of the temporal lobe and ends blindly in the anterior wall situated immediately behind the amygdaloid nucleus (Figs. 3.2 and 3.7). The floor of the temporal horn is formed medially by the hippocampus and laterally by the smooth white prominence overlying the collateral sulcus. The roof is formed medially by the inferior surface of the thalamus and the tail of the caudate nucleus, which are separated by the striothalamic sulcus, and laterally by the tapetum of the corpus callosum, which also sweeps inferiorly to form the lateral wall of the temporal horn. The medial wall is little more than a narrow cleft, the choroidal fissure, between the inferolateral part of the thalamus and the fimbria of the fornix. The inferior end of the choroid fissure, the inferior choroidal point, is located just behind the amygdaloid nucleus and the uncus. Basal Cisterns in Tentorial Incisura The third ventricle is commonly approached through the cisterns surrounding the tentorial incisura (25). These cisterns, which are inti-
mately related to the third ventricle, contain the bifurcation of the carotid and basilar arteries, the circle of Willis, the convergence of the deep intracranial venous system on the great vein, and parts of the first six cranial nerves. The tentorial incisura is a triangular space situated between the dorsum sellae and the free edges of the tentorium. Its apex is dorsal to the midbrain, just posterior to the pineal gland. The upper brain stem sits in the center of the incisura. The area between the brain stem and the free edges is divided into: (a) an anterior incisural space located in front of the brain stem; (b) paired middle incisural spaces situated lateral to the brain stem; and (c) a posterior incisural space located behind the brain stem (Figs. 3.7 to 3.9). Anterior Incisural Space The anterior incisural space is located anterior to the midbrain. It extends obliquely forward and upward around the optic chiasm to the subcal-losal area. It opens laterally into the medial part of the sylvian fissure and posteriorly between the uncus and the brain stem into the middle incisural space (Figs. 3.7 to 3.9). The part of the anterior incisural space located below the optic chiasm has posterolateral and posterior walls. The posterolateral wall is formed by the anterior one-third of the uncus, which hangs over the free edge above the oculomotor trigone. The posterior wall is formed by the cerebral peduncles. The infundibulum of the pituitary gland crosses the anterior incisural space to reach the opening in the diaphragma sellae. The part of the anterior incisural space situated above the optic chiasm is limited superiorly by the rostrum of the corpus callosum, posteriorly by the lamina terminalis, and laterally by the part of the medial surfaces of the frontal lobes located below the rostrum. The anterior incisural space opens laterally into the part of the sylvian fissure situated below the anterior perforated substance. The anterior limb of the internal capsule, the head of the caudate nucleus, and the anterior part of the lentiform nucleus are located above the anterior perforated substance. The interpeduncular cistern, which sits in the posterior part of the anterior incisural space between the cerebral peduncles and the dorsum sellae, communicates laterally with the sylvian cistern below the anterior perforated substance and anteriorly with the chiasmatic cistern located below the optic chiasm. The interpeduncular and chiasmatic cisterns are separated by Liliequist's membrane, an arachnoidal sheet ex-
Figure 3.8. A-C. Tentorium and inclsura: stepwise dissection. A. Right super-olateral view. The right cerebral hemisphere has been removed to expose the tentorium (Tent.), falx, and medial surface of the right cerebral hemisphere. The tentorial edge (Tent. Edge) slopes downward from the apex (Apex Tent.) along the side of the brain stem to the area lateral to the dorsum sellae (Dorsum). The third ventricle (3 Vent.) has been divided in the midline and the right half of the upper midbrain has been divided transversely to expose the cerebral peduncle (Ped.), red nucleus (Red Nucl.), substantia nigra (Subst. Nigra), and inferior colliculi (Inf. Coll.). The midbrain is in the center of the incisura. The anterior incisural space is located anterior to the midbrain and extends downward between the dorsum sellae and brain stem and upward around the optic chiasm to the area anterior to the lamina terminalis (Lam. Term.). The optic nerves (Optic N.), carotid arteries (Car. A.), and infundibulum (Infund.) of the pituitary gland cross the anterior incisural space. The middle incisural space is located lateral to the midbrain. It extends into the supratentorial area between the temporal lobe and the midbrain and into the infratentorial area between the cerebellum and the brain stem. The posterior incisural space is located between the apex of the tentorium posteriorly, the midbrain anteriorly, the splenium above, and the cerebellum below. The corpus callosum (Corp. Call.), septum pellucidum (Sept. Pell.), and fornix are located above the third ventricle. The cingulate gyrus (Cing. Gyr.) wraps around the corpus callosum and has the cingulate sulcus (Cing. Sulc.) and medial frontal gyrus (Med. Front. Gyr.) on its outer margin. The tentorium is attached anteriorly to the petrous ridge and temporal bone and posteriorly to the occipital bone. The posterior part of the falx has been removed to expose the medial surface of the parietal and occipital lobes, including the precuneus and cuneus, the parietooccipital (Par. Occip. Sulc.), and calcarine (Calc. Sulc.) sulci and the lingual gyrus (Lingual Gyr.). The olfactory tract (Olf. Tr.) passes posteriorly along the lateral margin of the gyrus rectus (Gyr. Rectus). The oculomotor nerve (III) arises on the medial surface of the cerebral peduncle and enters the dura mater lateral to the posterior clinoid process. The trochlear nerve (IV) arises below the inferior colliculus and passes forward in the cerebellomesencephalic fissure (Cer. Mes. Fiss.) between the cerebellum and midbrain. B. Superior view. The falx and the dura covering the left hemisphere have been removed. The left insula has been exposed by removing the operculum of the frontal lobe (Front. Lobe). The central sulcus separates the precentral and postcentral gyri and the frontal (Front. Lobe) and parietal (Par. Lobe) lobes. The straight sinus (Str. Sinus) passes backward between the occipital lobes (Occip. Lobe). The cerebellum forms the floor of the posterior incisural space. The trochlear nerve enters the anterior part of the free edge. The oculomotor nerve enters the roof of the cavernous sinus. The trigeminal nerve (V) arises in the infratentorial part of the middle incisural space. The frontal sinus (Front Sinus) is anterior to the frontal lobe and the floor of the anterior cranial fossa (Ant. Fossa). C. The superior part of the left cerebral hemisphere has been removed to expose the lateral ventricle (Lat. Vent.). The thalamus and the body of the lateral ventricle (Body) are located above the central part of the incisura; the frontal horn (Front. Horn) is above the anterior incisural space; and the atrium is above the posterior incisural space. The culmen and the quadrangular lobule (Quad. Lobule) of the cerebellum form the floor of the posterior incisural space. The forceps major, a bundle of white matter emanating from the splenium of the corpus callosum, sweeps posteriorly in the medial wall of the atrium and occipital horn (Occip. Horn) deep to the parietooccipital sulcus. The forceps minor, another bundle of white matter, radiates from the genu of the corpus callosum along the medial margin of the frontal horn. The caudate nucleus (Caudate Nucl.) and corona radiata are lateral to the frontal horn and body of the lateral ventricle. The choroid plexus (Chor. Plex.) is attached along the choroidal fissure, the cleft between the fornix and thalamus. The oculomotor nerve enters the roof of the cavernous sinus through the oculomotor trigone (Oculomotor Trig.), which is situated between three dural folds: the interclinoid fold (Interclin. Fold), which extends from the posterior (Post. Clinoid) to the anterior (Ant. Ca'noid) clinoid process, forms the medial margin; the anterior petroclinoid fold (Ant. Petroclin. Fold), which extends from the petrous apex to the anterior clinoid process, forms the lateral margin; and the posterior petroclinoid fold (Post. Petroclin. Fold), which extends from the petrous apex to the posterior clinoid process, forms the posterior margin. The septum pellucidum is in the medial wall of the lateral ventricle. The infundibulum passes through the diaphragma sellae (Diaph.).
Figure 3.8. D and E. Tentorium and incisura (continued). D. Another layer of the left cerebral hemisphere has been removed. The anterior limb (Ant. Limb) of the internal capsule (Int. Cap.) is between the head of the caudate nucleus and the putamen, and the posterior limb (Post. Limb) is between the thalamus and the putamen. The thalamostriate vein (Thal. Str. V.) courses between the caudate nucleus and the thalamus. The column of the fornix passes anterior to the foramen of Monro (For. Monro). Structures in the anterior wall of the atrium include the crus of the fornix, the pulvinar of the thalamus, and the tail of the caudate nucleus. The choroid plexus in the atrium is attached between the thalamus and the fornix. E. Deeper section through the left cerebral hemisphere. The caudate and lentiform nuclei, the thalamus, the internal capsule, and the body, frontal horn, and atrium of the lateral ventricle are above the tentorial incisura. The junction of the anterior and posterior limbs of the internal capsule abuts on the wall of the lateral ventricle between the caudate nucleus and thalamus in the area lateral to the foramen of Monro. The thalamus lies directly above the midbrain and the central part of the incisura. The head of the caudate nucleus, the anterior limb of the internal capsule, and the anterior part of the lentiform nucleus, formed by the putamen and globus pallidus (Globus Pall.), are above the anterior incisural space. The posterior limb of the internal capsule and the posterior part of the lentiform nucleus are above the middle incisural space. The atrium and splenium are directly above the posterior incisural space. (From Ono M, Ono M, Rhoton AL Jr, Barry M: Microsurgical anatomy of the region of the tentorial incisura. J Neurosurg 60:365-399, 1984.) tending from the dorsurn sellae to the anterior edge of the mamillary bodies. The chiasmatic cistern communicates around the optic chiasm with the cisterna laminae terminalis, which lies anterior to the lamina terminalis. The optic and oculomotor nerves and the posterior part of the olfactory tracts pass through the anterior incisural space (Figs. 3.7 to 3.9). Each olfactory tract runs posteriorly and splits just above the anterior clinoid process to form the medial and the lateral olfactory striae, which course along the anterior margin of the anterior perforated substance. The optic nerves and chiasm and the anterior part of the optic tracts cross the anterior incisural space. The optic nerves emerge from the optic canal medial to the attachment of the free edge to the anterior clinoid processes and are directed posterior, superior, and medial toward
the optic chiasm. From the chiasm, the optic tract continues in a posterolateral direction around the cerebral peduncle to enter the middle incisural space. The optic nerve proximal to its entrance into the optic canal is covered by a reflected leaf of dura mater, the falciform process, which extends medially from the anterior clinoid process across the top of the optic nerve (36). The length of nerve covered by dura only at the intracranial end of the optic canal may vary from less than 1 mm to as great as 1 cm. Coagulation of the dura above the optic nerve just proximal to the optic canal on the assumption that bone separates the dura mater from the nerve could lead to nerve injury. Compression of the optic nerves against the sharp edge of the falciform process may result in a visual field deficit even if the compressing lesion does not damage the nerve enough to cause visual loss.
The ophthalmic artery is found inferolateral to the optic nerve when the periosteum lining the optic canal is opened. Normally the optic nerve is separated medially from the sphenoid sinus by a thin plate of bone, but in a few cases this bone is absent and the optic nerves may protrude directly into the sphenoid sinus, separated from the sinus by only mucosa and the dural sheath of the nerve (7, 36). The relationship of the chiasm to the sella (Fig. 3.10) is an important determinant of the ease with which the pituitary fossa may be exposed by the transfrontal surgical route. The normal chiasm overlies the diaphragma sellae and the pituitary, the prefixed chiasm overlies the tuberculum sellae, and the postfixed chiasm overlies the dorsum sellae. In approximately 70% of cases the chiasm is in the normal position. Of the remaining 30%, about half are "prefixed" and half "postfixed" (36). A prominent tuberculum sellae may restrict access to the sellae even in the presence of a normal chiasm. The tuberculum may vary from being almost flat to protruding upward as much as 3 mm, and it may project posteriorly to the margin of a normal chiasm (36). The limited transfrontal approach exposure provided by a prefixed chiasm, a normal chiasm with a small area between the tuberculum and the chiasm, and a superior protruding tuberculum sellae may be enlarged by using the trans-frontal-transsphenoidal approach advocated by Rand (35). In this approach, access is obtained by opening into the sphenoid sinus from above by drilling through the tuberculum and planum sphenoidale, thus converting the approach to a "transfrontal-transsphenoidal" one. If the chiasm is normal or "postfixed," a subchiasmatic approach below the chiasm is possible. If the chiasm is prefixed and the tumor is seen through a thin, stretched anterior wall of the third ventricle, the lamina terminalis may be opened to expose the tumor. If the space between the carotid artery and the optic nerve has been enlarged (e.g., by a lateral or parasellar extension of tumor), the tumor may removed through this space (43). If more space is needed in the interval between the carotid artery and optic nerve, the carotid artery, rather than the optic nerve, may be gently retracted. An understanding of the relationship of the carotid artery, optic nerve, and anterior clinoid process is fundamental to all surgical approaches to the sellar and parasellar areas (Fig. 3.11). The carotid artery and the optic nerve are medial to the anterior clinoid process. The artery exits the
cavernous sinus beneath and slightly lateral to the optic nerve. The optic nerve pursues a pos-teromedial course toward the chiasm and the carotid artery pursues a posterolateral course toward its bifurcation into the anterior and middle cerebral arteries. The oculomotor nerve emerges from the mid-brain on the medial surface of the cerebral peduncle. It crosses the anterior incisural space between the posterior cerebral and the superior cerebellar arteries and passes inferomedial to the uncus to enter the roof of the cavernous sinus. Middle Incisural Space The middle incisural space is located lateral to the brain stem (Figs. 3.6 to 3.9). This narrow space, which extends upward between the temporal lobe and the midbrain, has medial and lateral walls and a roof. The medial wall is formed by the lateral surface of the midbrain. This surface of the midbrain is divided into a larger anterior part formed by the cerebral peduncle and a smaller posterior part formed by the tegmental surface. The tegmental surface is formed by the lemniscal trigone, a triangular area situated just posterior to the peduncle, and by the brachium of the inferior col-liculus, a prominence posterior to the lemniscal triangle that is directed upward from the inferior colliculus toward the medial geniculate body. The roof of the middle incisural space has a narrow anterior part formed by the posterior part of the optic tract, which is flattened between the cerebral peduncle and the uncus, and a wider posterior part formed by the inferior surface of the pulvinar. The lateral geniculate body protrudes from the inferior surface of the pulvinar just lateral to the posterior edge of the cerebral peduncle. The medial geniculate body bulges into the roof posteromedial to the lateral geniculate body just behind the lateral mesencephalic sul-cus. The lateral wall of the supratentorial part of the middle incisural space is composed of the hippocampal formation on the medial surface of the temporal lobe. The uncus and parahippocam-pal gyri, the most inferior structures in this part of the lateral wall, form a curved border around the middle incisural space. The uncus bulges medially at the anterior end of the parahippo-campal gyrus. Posterior to the uncus, the surface of the temporal lobe facing the middle incisural space is formed by three longitudinal strips of neural tissue, one located above the other, which are interlocked with the hippocampal formation to
Figure 3.9. A-C. Neural relationships. A. Lateral view. The temporal lobe (Temp. Lobe) has been elevated from the tentorium (Tent.) and the floor of the middle cranial fossa. The free edge (Tent. Edge) has grooved (Tent. Groove) the lower surface of the uncus. The rhinal (Rhinal Sulc.) and collateral (Coll. Sulc.) sulci separate the occipitotemporal (Occip. Temp. Gyr.) and the parahippocampal (Parahippo. Gyr.) gyri. The gyrus rectus (Gyr. Rectus) is medial and the orbital gyri (Orb. Gyr.) are lateral to the olfactory tract (Off. Tr.). The anterior perforated substance (Ant. Perf. Subst.) extends into the medial part of the sylvian fissure (Sylvian Fiss.). The calcarine sulcus (Calc. Sulc.) divides the medial surface of the occipital lobe. The optic tract (Optic Tr.) passes backward between the uncus and the cerebral peduncle (Ped.). The brachium of the inferior colliculus (Brach. Inf. Coll.) extends from the inferior colliculus (Inf. Coll.) along the side of the superior colliculus (Sup. Coll.) toward the medial geniculate body (Med. Gen. Body). The lemniscal trigone (Lemniscal Trig.) is located between the brachium of the inferior colliculus and the lateral mesencephalic sulcus (Lat. Mes. Sulc.). The oculomotor nerve (III) arises below the mamillary bodies (Mam. Body) just above the pontomesencephalic sulcus (Pon. Mes. Sulc.). B. Enlarged view. The temporal lobe has been elevated further to show the lower surface of the pulvinar and the termination of the optic tract in the lateral geniculate body (Lat. Gen. Body). The optic nerve (Optic N.) passes above the divided end of the carotid artery (Car. A.). The anterior hippocampal sulcus (Ant. Hippo. Sulc.) separates the uncus and the parahippocampal gyrus. The calcarine sulcus separates the lingual gyrus (Lingual Gyr.) and the cuneus. C. The anteromedial part of the temporal lobe has been removed to expose the temporal horn (Temp. Horn). The removal crosses the amygdaloid nucleus (Amygd. Nucl.), which lies deep to the uncus. The choroid plexus (Chor. Plex.) is attached along the choroidal fissure (Chor. Fiss.), a cleft between the temporal lobe and pulvinar. The olfactory tract splits into the medial (Med. Olf. Stria) and lateral olfactory striae (Lat. Olf. Stria).
Figure 3.9. D and E. Neural relationships (continued). D. Inferior view of basal structures surrounding the course of the basal vein. The cerebrum was removed by dividing the midbrain at the level of the red nucleus (Red Nad), substantia nigra (Subst. Nigra), and posterior perforated substance (Post. Perf. Subst.). The infundibulum (Infund.) arises from the tuber cinereum (Tuber). The uncus extends medial to and is grooved by the tentorial edge in the area anterolateral to the cerebral peduncle. The subiculum, the rounded medial edge of the parahip-pocampal gyrus, has been retracted to expose the dentate gyrus (Dentate Gyr.) and the fimbria of the fornix on the medial surface of the temporal lobe. The fimbria and dentate gyrus are separated by the fimbriodentate sulcus (Fimb. Dentate Sulc.), and the dentate gyrus and the subiculum are separated by the hippocampal sulcus (Hippo. Sulc.). The choroidal fissure opens into the temporal horn between the pulvinar and fimbria. The fibers in the fimbria are continuous with those in the crus of the fornix. The dentate gyrus is continuous posteriorly with the fasciolar gyrus (Fasciolar Gyr.), located below the splenium. E. The temporal tip has been retracted to expose the anterior perforated substance, which is bounded anteriorly by the medial and lateral olfactory striae, medially by the optic tract, posteriorly by the temporal lobe, and laterally by the limen insula. The semilunar (Semilunar Gyr.) and ambient (Ambient Gyr.) gyri, which are continuous posteriorly with the uncus, are on the superior surface of the temporal tip immediately posterior to the anterior perforated substance. (From Ono M, Rhoton AL Jr, Peace DA, Rodriguez R: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984.)
Figure 3.10. Three sagittal sections of the sellar region showing neural structures, chiasm, posterior lobe pituitary and hypothalamus, anterior lobe pituitary, dura, and intracavernous venous connections. Prefixed chiasm above tuberculum (left), normal chiasm above diaphragma (center), and postfixed chiasm above dorsum (right). (From Rhoton AL Jr, Maniscalco J: Microsurgery of the sellar region. Neuroophthalmology 9:106-127, 1977.)
Figure 3.11. Right anterior and lateral views of the suprasellar area. A. Right anterolateral view. Anterior clinoid process (Ant. Clinoid), lateral to optic nerve (O.N.) and carotid artery (C.A.). Perforating arteries pass from the carotid to terminate in the optic chiasm and tract (O.Tr.) arid the hypothalamus anterior to the mamillary body (Mam. В.). Right posterior communicating artery (P.Co.A.) is medial to the carotid artery. Anterior choroidal artery (A.Ch.A.) arises from the carotid and passes above the posterior cerebral artery (P.C.A.). The third nerve (HI) lies below the right posterior cerebral artery. The right recurrent artery of Huebner (Rec. A.) arises from the anterior cerebral artery (A.C.A.) proximal to the anterior communicating artery (A.Co.A.). M.C.A., middle cerebral artery; A-2, anterior cerebral artery distal to the anterior communicating artery. B. Right lateral view with temporal lobe removed. Anterior clinoid (din.) is lateral to the carotid artery. Premamillary (Premam. A.) arises from the posterior communicating artery. Thalamoperfor-ating arteries (Th. Pe. A.) arise from the proximal posterior cerebral artery. The optic tract passes around the peduncle to the lateral geniculate body. Arterioles pass from the carotid artery to the optic nerve. The basilar artery (B.A.) is medial to the third and fourth (IV) cranial nerves. The trigeminal nerve (V) is below the superior cerebellar artery (S.C.A.). (From Saiki N, Rhoton AL Jr: Microsurgical anatomy of the upper basilar artery and the posterior circle of Willis. J Neu-rosurg 46:563-577, 1977.)
make an important part of the limbic system (Figs. 3.7 and 3.9). The most inferior strip is formed by the subiculum, the rounded medial edge of the parahippocampal gyrus; the middle strip is formed by the dentate gyrus, a serrated or beaded strip of gray matter located on the medial surface of the hippocampal formation; and the superior strip is formed by the fimbria of the fornix, a white band formed by the fibers emanating from the hippocampal formation that are directed posteriorly into the crus of the fornix. The choroidal fissure, the cleft along which the choroid plexus of the temporal horn is attached, is situated at the junction of the roof and lateral wall of the middle incisural space between the fimbria below and the pulvinar above. The inferior choroidal point, at the inferior end of the choroidal fissure, is located immediately lateral to the lateral geniculate body. The supratentorial part of the middle incisural space contains the crural and ambient cisterns. The crural cistern, located between the cerebral peduncle and the uncus, is a posterolateral extension of the interpeduncular cistern. The crural cistern opens posteriorly into the ambient cistern, a narrow communicating channel demarcated medially by the midbrain, above by the pulvinar, and laterally by the subiculum, dentate gyrus, and fimbria of the fornix. The ambient cistern is continuous posteriorly with the quadrigeminal cistern. The trochlear nerve is related predominatly to the middle incisural space and is the cranial nerve most intimately related to the free edge. The trochlear nerve arises below the inferior colliculus in the posterior incisural space and passes forward through the middle incisural space between the posterior cerebral and superior cerebellar arteries. Its initial course around the midbrain is medial to the free edge in the space between the tectum and cerebellum. It reaches the lower margin of the free edge at the posterior edge of the cerebral peduncle. It pierces the anterior part of the free edge before entering the cavernous sinus. Posterior Incisural Space The posterior incisural space lies posterior to the midbrain and corresponds to the pineal region (Figs. 3.6 to 3.8). It has a roof, floor, and anterior and lateral walls and extends backward to the level of the apex of the tentorium. The quadrigeminal plate is located at the center of the anterior wall. The anterior wall rostral to
the colliculi is formed by the pineal body and the habenular trigones and commissure. The habenular trigones overlie the habenular nuclei on the posteromedial surface of the pulvinar and are connected across the midline by the habenular commissure. The habenular trigones are continuous anteriorly with the striae medullaris thai-ami, which course along the medial surface of the thalamus. The habenular commissure forms the upper half and the posterior commissure forms the lower half of the attachment of the pineal body to the posterior part of the third ventricle. The part of the anterior wall below the colliculi is formed in the midline by the lingula of the vermis and laterally by the superior cere-bellar peduncles as they ascend beside the lingula. The roof of the posterior incisural space is formed by the lower surface of the splenium, the terminal part of the crura of the fornix, and the hippocampal commissure. The floor of the posterior incisural space is formed by the anterior superior part of the cerebellum. This space extends inferiorly into the cerebellomesencephalic fissure, the cleft opening inferiorly between the anterior superior part of the vermis and the colliculi. The anterior wall of the fissure is formed by the lingula and superior cerebellar peduncles, and the posterior wall is formed by the cerebellum. Each lateral wall is formed by the pulvinar, crus of the fornix, and medial surface of the cerebral hemisphere. The anterior part of the lateral wall is formed by the part of the pulvinar located just lateral to the pineal body. The lateral wall, posterior to the pulvinar, is formed by the segment of the crus of the fornix that wraps around the posterior margin of the pulvinar. The posterior part of the lateral walls is formed by the cortical areas located below the splenium on the medial surface of the hemisphere. These areas include the posterior part of the parahippocampal and dentate gyri and several small nodular collections of gray matter, the fasciolar gyri and Retzius' gyri, which are located at the posterior end of the dentate gyrus (Fig. 3.7). The quadrigeminal cistern, situated posterior to the quadrigeminal plate, is the major cistern in the posterior incisural space. It has also been referred to as the cisterna venae magnae cerebri because the vein of Galen passes through it. The quadrigeminal cistern communicates above with the posterior pericallosal cistern, which extends around the splenium between the cerebral hemispheres; inferiorly into the cerebellomesencephalic cistern, also called the precentral cerebel-
lar cistern, which extends into the cerebellomesencephalic fissure; inferolaterally into the posterior part of the ambient cistern located between the midbrain and the parahippocampal gyrus; and laterally into the retrothalamic cistern, which curves around the posterior margin of the pulvinar medial to the crus of the fornix. The quadrigeminal cistern may communicate with the velum interpositum. Another cavity, the cavum vergae, is located just above the velum interpositum between the hippocampal commissure and the splenium: its roof is formed by the lower surface of the splenium, its floor is formed by the hippocampal commissure, and the lateral walls are formed by the crura. Substances instilled in the subarachnoid space only infrequently enter a cavum vergae. Arterial Relationships Each wall of the third ventricle has surgically important arterial relationships: the posterior part of the circle of Willis and the apex of the basilar artery are below the floor; the anterior part of the circle of Willis and the anterior cerebral and anterior communicating arteries are intimately related to the anterior wall; the posterior cerebral, pericallosal, superior cerebellar, and posterior choroidal arteries pass adjacent to the posterior wall; both the anterior and posterior cerebral arteries send branches into the roof; and the internal carotid, anterior choroidal, anterior and posterior cerebral, and anterior and posterior communicating arteries give rise to perforating branches that reach the walls of the third ventricle. All arterial components of the circle of Willis and the adjacent carotid and basilar artery and their perforating branches may become stretched around third ventricular and selar tumors. The arterial relationships in the anterior incisural space are among the most complex in the brain because it contains all the components of the circle of Willis (Figs. 3.11 to 3.13). Internal Carotid Artery The internal carotid artery exits the cavernous sinus along the medial surface of the anterior clinoid process to reach the anterior incisural space. After entering this space it courses posterior, superior, and lateral to reach the site of its bifurcation below the anterior perforated substance. It is first below and then lateral to the optic nerve and chiasm (Figs. 3.11, 3.13, and 3.14). It sends perforating branches to the optic nerve, chiasm, and tract and to the floor of the third ventricle. These branches pass across the interval between the internal carotid artery and
Figure 3.12. Arterial relationships of the third ventricle. A and С are inferior views of the floor of the third ventricle, and В and D are midsagittal sections through the third ventricle. A and В show the relationship of the main trunks and perforating branches of the following arteries to the third ventricle: internal carotid (C.A.), anterior choroidal (A.Ch.A.}, basilar apex (B.A.), posterior cerebral (P.C.A.), medial posterior choroidal (Med.Post.Ch.A.), lateral posterior choroidal (Lat.Post.Ch.A.), thalamoperforating (Thal. Perf.A.), and thalamogeniculate (Thal.Gen.A.) arteries. С and D show the relationships of the main trunks and perforating branches of the following arteries to the third ventricle: anterior cerebral (A.C.A.), anterior communicating (A..Co.A.), and posterior communicating arteries. The olfactory (Olf.N.) and optic (O.N.) nerves are anterior to the floor of the third ventricle. The structures in the floor are the optic chiasm (O.Ch.), optic tracts (O.Tr.), infundibulum (Infund.), tuber ci-nereum (Tuber Cin.), and mamillary bodies (Mam. В.). The midbrain and cerebral peduncles (Ped.) are inferior to the posterior half of the floor. The anterior perforated substance (Ant. Perf. Subst.) is lateral to the optic tracts. The lateral geniculate (Lat. Gen. B.) and medial geniculate (Med. Gen. B.) bodies are attached to the lower margin of the thalamus near the pulvinar (Pulv.}, lateral to the midbrain. The structures in the anterior wall of the third ventricle are the anterior commissure (Ant. Comm.), lamina terminalis (Lam. Теr.), and optic chiasm. The corpus callosum (Corp. Call.) and septum pellucidum (Sept. Pel.) are above the roof of the third ventricle. The roof is formed of the two layers of tela choroidea (Tela), the fornix, and a vascular layer composed of the internal cerebral veins and the medial posterior choroidal arteries. The oculomotor nerve (HI) exits from the midbrain. (From Yamamoto I, Rhoton AL Jr, Peace DA: Microsurgery of the third ventricle: Part 1. Microsurgical anatomy. Neurosurgery 8:334-356, 1981.)
the optic nerve and may serve as an obstacle to the operative approaches directed through the triangular space between the internal carotid artery, the optic nerve, and the anterior cerebral artery. The internal carotid artery also gives off the superior hypophyseal artery, which runs medially below the floor of the third ventricle to reach the tuber cinereum and joins its mate of the opposite side to form a ring around the infundibulum (10). The supraclinoid (C4) portion of the internal carotid artery is divided into three segments based on the origin of its major branches: the ophthalmic segment extends from the origin of the ophthalmic artery to the origin of the posterior communicating artery; the communicating segment extends from the origin of the posterior communicating artery to the origin of the anterior choroidal artery; and the choroidal segment extends from the origin of the anterior choroidal artery to the bifurcation (Figs. 3.15 to 3.17) (10). Each segment gives off a series of perforating branches with a relatively constant site of termination. The branches arising from the ophthalmic segment pass to the optic nerve and chiasm, infundibulum, and the floor of the third ventricle. The branches arising from the communicating segment pass to the optic tract and the floor of the third ventricle. The branches arising from the choroidal segment pass upward and enter the brain through the anterior perforated substance. Ophthalmic Artery The ophthalmic artery is the first branch of the internal carotid artery above the cavernous sinus (Figs. 3.14 and 3.16). Its origin and proximal segment may be visible below the optic nerve without retracting the nerve, although elevation of the optic nerve away from the carotid artery is usually required to see the segment proximal to the optic foramen. The artery arises from the supraclinoid segment at the carotid artery in most cases, but it may also arise within the cavernous sinus or be absent in a few cases (14, 36). The ophthalmic artery origin has a variable relationship to the tip of the anterior clinoid process. Its origin may vary from as far as 5 mm anterior to 7 mm posterior to the clinoid tip and 2 to 10 mm medial to the clinoid process (14). Most arise anterior and approximately 5 mm medial to the tip. In our studies, 14% of ophthalmic arteries exited from the carotid artery and immediately entered the optic canal; in the remaining 86% the maximal preforaminal length was 7 mm and the mean was 3 mm (14).
Posterior Communicating Artery The posterior communicating artery arises from the posterior wall of the internal carotid artery and courses posteromedially below the optic tracts and the floor of the third ventricle to join the posterior cerebral artery (Figs. 3.11 to 3.19). Its branches penetrate the floor between the optic chiasm and the cerebral peduncle and reach the thalamus, hypothalamus, subthala-mus, and internal capsule. Its posterior course varies depending on whether the artery provides the major supply to the distal posterior cerebral artery. If it is normal, with the posterior cerebral artery arising predominantly from the basilar artery, it is directed posteromedially above the oculomotor nerve toward the interpeduncular fossa. If the posterior cerebral artery has a fetal type configuration in which it arises from the carotid artery, the posterior communicating artery is directed posterolaterally below the optic tract. The oculomotor nerve pierces the dura mater of the roof of the cavernous sinus 2 to 7 mm (average, 5 mm) posterior to the initial supracli-noid segment of the carotid artery (14, 42). Anterior Choroidal Artery The anterior choroidal artery arises from the posterior surface of the internal carotid artery 0.1 to 3.0 mm above the origin of the posterior communicating artery (Figs. 3.11 to 3.20). It is directed posterolaterally below the optic tract between the uncus and cerebral peduncle. It passes through the choroidal fissure near the inferior choroidal point to supply the choroid plexus in the temporal horn. It sends branches into the optic tract and posterior part of the floor that reach the optic radiations, globus pallidus, internal capsule, midbrain, and thalamus (8, 38). Anterior Cerebral and Anterior Communicating Arteries The anterior cerebral artery arises from the internal carotid artery below the anterior perforated substance and courses anteromedially above the optic nerve and chiasm to reach the interhemispheric fissure, where it is joined to the opposite anterior cerebral artery by the anterior communicating artery (Figs. 3.11 to 3.20) (30, 31, 37). The junction of the anterior communicating artery with the right and left A1 segments is usually above the chiasm rather than above the optic nerves. In our studies 70% were above the chiasm and 30% were in a prefixed position above the optic nerves (Fig. 3.21). The shorter A1
segments are stretched tightly over the chiasm, and the longer ones pass anteriorly over the nerves. The arteries with a more forward course are often tortuous and elongated, and some may course forward and rest on the tuberculum sellae or planum sphenoidale (Fig. 3.21). The anterior cerebral artery ascends in front of the lamina terminalis and the anterior wall of the third ventricle and passes around the corpus callosum. It gives off the orbitofrontal artery before reaching the corpus callosum. Often the anterior cerebral artery on one side sends branches across the interhemispheric fissure to supply the medial part of the opposite cerebral hemisphere (31). The distal part of the anterior cerebral artery may be exposed not only above, but also below the corpus callosum because the terminal branch of the pericallosal artery may pass around the splenium and course forward in the roof of the third ventricle, reaching as far anterior as the foramen of Monro (31). The anterior cerebral and anterior communicating arteries give rise to perforating branches that terminate in the whole anterior wall of the third ventricle and reach the adjacent parts of the hypothalamus, fornix, septum pellucidum, and striatum (Figs. 3.22 and 3.23). A precallosal artery may originate from the anterior cerebral or the anterior communicating artery, run upward across the lamina terminalis, and send branches into the anterior wall (Fig. 3.22D). The pericallosal arteries send short and long callosal arteries into the corpus callosum. The short callosal arteries pass directly into the corpus callosum. The long callosal arteries course parallel to the pericallosal arteries, between them and the corpus callosum. The callosal arteries send branches into the corpus callosum that reach the septum pellucidum and the fornix. The recurrent branch of the anterior cerebral artery, which is referred to as the recurrent artery of Heubner, is encountered frequently in approaches to the anterior part of the third ventricle (Figs. 3.11, 3.22, and 3.23). It arises from the anterior cerebral artery in the region of the anterior communicating artery, courses laterally above the bifurcation of the internal carotid artery, and enters the anterior perforated substance (30). The recurrent artery courses anterior to the A1 segment of the anterior cerebral artery and would have been seen when elevating the frontal lobe before visualizing the A1 segment in about two-thirds of cases. Of the remaining one-third, most coursed superior to A1 Some of its branches reach the anterior limb and genu of the internal capsule.
Middle Cerebral Artery The middle cerebral artery courses laterally from its origin below the anterior perforated substance (Figs. 3.11, 3.13, 3.15, 3.18, and 3.19) (9). The major bifurcation of the middle cerebral artery is usually located in the lateral part of the anterior incisural space. The middle cerebral artery has been divided into four segments, M1 through M4 (9). The M1 segment begins at the origin of the middle cerebral artery lateral to the optic chiasm, at the medial end of the sylvian fissure, and extends laterally below the anterior perforated substance, toward the insula. The M1 segment terminates at the limen insulae, the site of a 90° turn, the genu, where the artery curves sharply posterosuperior to form the M2 segment, composed of the trunks coursing over the insula. The M1 segment is subdivided into a prebifurcation and a postbifurcation part. The major division of the artery is a bifurcation in 86% of hemispheres and a trifurcation in the remaining 14% of hemispheres (9). The bifurcation occurs proximal to the genu in most hemispheres.
The middle cerebral artery branches to the anterior perforated substance are called the len-ticulostriate arteries (44). The lenticulostriate arteries are divided into medial, intermediate, and lateral groups, each of which has a unique origin, composition, morphology, and characteristic distribution in the anterior perforated substance. The medial group of lenticulostriate arteries arises on the medial part of the M1 segment near the carotid bifurcation and pursues a relatively direct course to enter the anterior perforated substance. The intermediate lenticulostriate arteries form a complex array of branches resembling a candelabra before entering the anterior perforated substance between the medial and lateral lenticulostriate arteries. The lateral lenticulostriate arteries originate predominantly on the lateral part of the M1 portion, pursue an S-shaped course, and enter the lateral part of the anterior perforated substance. They arise from the parent trunks, travel medially with the parent trunks, then loop sharply posterior, lateral, and superior, and finally turn posteromedial just before penetrating the anterior perforated substance.
Figure 3.13. A and B. Superior view of the arteries in the suprasellar area. A. The posterior part of the optic chiasm is split at the junction with the optic tracts (O. Tr.) to give this view. The proximal portions of the anterior cerebral arteries (A-1) arise from the carotid arteries (C.A.) and pass above the optic nerves (O. N.) and chiasm; the anterior third ventricle lies above the mamillary bodies (Mam. В.). The thalamoperforating arteries (Th.Pe.A.) terminate in the retromamillary area and the interpeduncular fossa. The right posterior choroidal artery (P.Ch.A.) originates from the posterior cerebral artery (P.C.A.). Small branches arise from the posterior cerebral artery and terminate in the peduncle. The left third nerve (III) courses medial to the posterior communicating artery (P.Co.A.). The right posterior communicating artery joins the posterior cerebral artery medial to the third nerve. The left anterior choroidal artery (A.Ch.A.) passes to the peduncle and optic tract. Trigeminal nerves (V) are lateral to the carotid artery. M.C.A., middle cerebral artery. B. Inferior view of the circle of Willis. Fetal type of left posterior cerebral artery. The left third nerve courses under the posterior cerebral artery distal to its junction with the posterior communicating artery and the right nerve courses under the part proximal to the communicating artery. The left thalamoperforating and the posterior choroidal (P.Ch.A.) arteries originate from the proximal part of the posterior cerebral artery. The right premamillary artery (Premam. A.) emerges from the anterior and the left from the middle third of the posterior communicating artery. Small posterior communicating branches course superiorly and medially, terminating in the premamillary area. The left thalamoperforating and right premamillary arteries are well developed in spite of the small trunk of origin. The superior cerebellar artery (S. C. A.) arises below the posterior cerebral artery. No branches arise on the anterior surface of the basilar artery (B.A.) A-2, distal part of the anterior cerebral artery. C. Superior view of the suprasellar area. Arterial branches stretch around the superior extension of a pituitary tumor. Anterior cerebral arteries send branches to the superior surface of the optic nerves and chiasm. The posterior communicating, internal carotid, and posterior cerebral arteries send branches into the area below and behind the chiasm. Recurrent arteries (Rec. A.) arise just distal to the anterior communicating artery. IV, trochlear nerve. (From Saeki N, Rhoton AL Jr: Microsurgical anatomy of the upper basilar artery and the posterior circle of Willis. J Neurosurg 46:563-577, 1977.)
Figure 3.14. Inferior views of the perforating branches of the supraclinoid portion of the internal carotid artery (ICA). A. The supraclinoid portion of the ICA gives rise to the ophthalmic (Ophth. A.), posterior communicating (P.Co.A.), and anterior choroidal (A.Ch.A.) arteries. The supraclinoid portion of the artery is divided into three segments based on the site of origin of these three branches: the ophthalmic segment (C4-Op.) extends from the origin of the ophthalmic artery to the Origin of the PCoA; the communicating segment (C4-Co.) extends from the origin of the PCoA to the origin of the AChA; and the choroidal segment (C4-Ch.) extends from the origin of the AChA to the bifurcation of the carotid artery into the anterior and middle cerebral arteries. The optic nerves (O.N.) are above the ophthalmic arteries. The optic chiasm and optic tracts (O.Tr.) are above the anterior (Ant.Lobe) and posterior (Post. Lobe) lobes of the pituitary gland. The tuber cinereum (Tuber Cin.) is anterior to the apex of the basilar artery (B.A.). The posterior cerebral arteries (P.C.A.) pass around the cerebral peduncles (Cer. Ped.) above the oculomotor nerves (III). The perforating branches arising from the ophthalmic segment pass superior to the anterior lobe to the optic nerve and chiasm and to the anterior part of the tuber cinereum. A single perforating branch arises from the communicating segment on each side and passes upward to the optic tract and the floor of the third ventricle. B. The anterior lobe of the pituitary gland has been reflected backward to show the superior hypophyseal arteries (Sup. Hyp. A.) passing from the ophthalmic segment to the infundibulum (Infund.). The anterior cerebral (A.C.A.) and the anterior communicating (A.Co.A.) arteries are seen above the optic chiasm (O.Ch.). (From Gibo H, Lenkey C, Rhoton AL
Posterior Cerebral Artery The bifurcation of the basilar artery into the posterior cerebral arteries is located in the posterior part of the anterior incisural space below the posterior half of the floor of the third ventricle (Figs. 3.11 to 3.20) (45, 57). A high basilar bifurcation may indent the floor. The posterior cerebral artery encircles the mesencephalon to reach the posterior incisural space and the quad-rigeminal cistern. It courses laterally around the cerebral peduncle, above the oculomotor nerve, and exits the anterior and enters the middle incisural space by coursing between the uncus and the cerebral peduncle. It enters the posterior incisural space from anteriorly and courses through the lateral part of the posterior incisural space. It bifurcates into the parietooccipital and calcarine arteries as it crosses above the free edge of the tentorium. It sends cortical branches to the temporal, occipital, and parietal lobes and callosal branches to the splenium of the corpus callosum. Its callosal branch, the posterior peri-callosal artery, may anastomose with the callosal branches of the anterior cerebral artery. Its branches reach the floor, roof, and posterior and lateral walls of the third ventricle. The thalamogeniculate and the thalamoperfor-ating arteries are two of the larger perforating branches (Figs. 3.13 and 3.19). The thalamoper-forating arteries arise from the proximal part of the posterior cerebral arteries and the posterior part of the posterior communicating arteries, enter the brain through the posterior part of the floor and the lateral walls. The thalamogeniculate arteries arise from the posterior cerebral arteries in the ambient cisterns below the pulvi-nar and course superiorly through the geniculate bodies to reach the adjacent parts of the thala-mus and internal capsule. The medial posterior choroidal arteries arise from the proximal portions of the posterior cerebral arteries in the anterior incisural space and course around the midbrain in the middle incisural space to reach the quadrigeminal cistern in the posterior incisural space (Figs. 3.12 and 3.18 to 3.20) (8, 38). They turn forward at the side of the pineal body and pass above the habenular trigone, course between the layers of the tela choroidea in the velum interpositum, and supply the choroid plexus in the roof of the third ventricle and the body of the lateral ventricle. The
Jr: Microsurgical anatomy of the supraclinoid portion of the internal carotid artery. J Neurosurg 55:560-574, 1981.)
lateral posterior choroidal arteries arise from each posterior cerebral artery in the area lateral to the midbrain and pass superolaterally through the choroidal fissure and around the pulvinar to supply the choroid plexus in the temporal horn and atrium. The posterior choroidal arteries send small branches to the surface of the thalamus along their course. The perforating branches of the posterior cerebral and superior cerebellar arteries and the medial posterior choroidal arteries supply the walls of the posterior incisural space. The posterior cerebral arteries supply the structures above the level of the lower margin of the superior colliculi, and the superior cerebellar arteries supply the structures below the upper margin of the inferior colliculus. There are numerous anastomoses between the two arteries over the surface of the colliculi. The pineal gland and the habenula are supplied by the branches of the medial posterior choroidal arteries.
Figure 3.15. Inferior views of the perforating branches of the supraclinoid portion of the internal carotid artery (ICA). A. The supraclinoid portion of the artery gives rise to the posterior communicating (P.Co.A.), anterior choroidal (A.Ch.A.), middle cerebral (M.C.A.), and anterior cerebral arteries (A.C.A.). The supraclinoid portion of the artery is divided into three segments based on the site of origin of these branches. An ophthalmic segment (C4-Op.) that extends from the origin of the ophthalmic artery (not shown because the ICA was divided above the level of origin of the ophthalmic artery) to the origin of the PCoA; a communicating segment (C4Co.) that extends from the origin of the PCoA to the origin of the AChA; and a choroidal segment (C4-Ch.) that extends from the origin of the AChA to the level of the bifurcation of the ICA into the anterior cerebral and middle cerebral arteries. The ophthalmic segment sends perforating branches to the optic nerves (O.N.), optic chiasm (O.Ch.), and tuber cinereum (Tuber Cin.). The superior hypophyseal arteries (Sup. Hyp. A.) pass to the infundibulum of the hypophysis (Infund.). The communicating segment sends one perforating branch on each side to the optic tracts (O.Tr.) and the region around the mamillary bodies (Mam. Body). The choroidal segment sends its perforating branches into the anterior perforated substance (Anr. Perf. Subst.). The posterior cerebral arteries (P.C.A.) arise from the basilar artery (B.A.) and pass laterally around the cerebral peduncles (Cer. Ped.). The temporal lobe (Temp. Lobe) is lateral to the carotid artery. The middle cerebral arteries pass laterally into the sylvian fissure below the anterior perforated substance. The frontal lobes (Fr. Lobe), gyrus rectus (Gyr. Rectus), and olfactory nerves (Olf. N.) are above the optic nerves. The thalamoperforating
Superior Cerebellar Artery This artery arises from the basilar artery in the posterior part of the anterior incisural space, encircles the midbrain below the posterior cerebral artery, and passes through the quadri-geminal cistern to reach the superior surface of the cerebellum (Figs. 3.18 and 3.19) (13). It dips below the tentorium to reach the superior surface of the cerebellum at the junction of the anterior and middle incisural spaces. It bifurcates into a rostral trunk that supplies the vermis and a caudal trunk that supplies the cerebellar hemisphere as it passes around the lateral margin of the cerebral peduncle to enter the middle incisural spaces. The segment of the artery in the quadrigeminal cistern is exposed in the supra- and infratento-rial operative approaches to the posterior part of the third ventricle, and its cortical branches are exposed in the infratentorial approaches. Perforating branches arising in the quadrigeminal region supply the inferior colliculi.
arteries (Thal. Perf. A.) pass posteriorly between the oculomotor nerves (Ш). B. Inferior view of another specimen. The ophthalmic segment gives rise to perforating branches that pass to the optic nerves, chiasm, and tracts and to the infundibulum and tuber cinereum. No perforating branches arise from the communicating segment on either side in this specimen. The perforating branches from the choroidal segment pass into the anterior perforated substance. (From Gibo H, Lenkey C, Rhoton AL Jr: Microsurgical anatomy of the supraclinoid portion of the internal carotid artery. J Neurosurg 55:560-574, 1981.)
Figure 3.16. Anterior and anteroinferior views of the supraclinoid portion of the internal carotid artery (ICA). A. Anterior view. The optic nerves (O.N.) enter the optic canals medial to the anterior clinoid processes (Ant Clinoid). The infundi-bulum (Infand.) passes inferiorly below the optic chiasm (O.Ch.) to the pituitary gland. The carotid arteries (C.A.) are posterior to the optic nerves. The planum sphenoidale (Planum) is anterior to the chiasmatic sulcus (Ch. Sulc.) and the tuberculum sellae (Tuberculum). The perforating branches of the carotid artery pass medially in the subchiasmatic space. The superior hypophyseal arteries (Sup. Hyp. A.) arise from the carotid artery and pass to the infundibulum. The falciform process (Falc. Process) is a fold of dura mater that passes above the optic nerve proximal to the optic foramen. B. The right optic nerve has been divided at the optic foramen and elevated to show the perforating branches of the supraclinoid portion of the carotid arteries. The right anterior cerebral artery (A.C.A.) was divided at its origin so that the optic nerve and chiasm could be elevated. The carotid artery gives rise to multiple perforating branches as well as the ophthalmic (Ophth. A.), posterior communicating (P.Co.A.), anterior choroidal (A.Ch.A.), and middle cerebral arteries (M.C.A.). The supraclinoid portion of the ICA is divided into three segments based on the origin of its major branches: the ophthalmic segment (C4-Op.) extends from the origin of the ophthalmic artery to the origin of the PCoA, the communicating segment (C4-Co.) extends from the origin of the PCoA to the origin of the AChA, and the choroidal segment (C4-Ch.) extends from the origin of the AChA to the bifurcation of the carotid artery. The perforating branches arising from the ophthalmic segment pass to the optic nerve, chiasm, infundibulum, and floor of the third ventricle. The perforating branches arising from the communicating segment pass to the optic tract and the floor of the third ventricle. The perforating branches arising from the choroidal segment pass upward and enter the brain through the anterior perforated substance (Ant. Perf. Subst.) The diaphragma sellae (Diaph.) surrounds the infundibulum above the pituitary gland. The temporal lobe (Temp. Lobe) is below the middle cerebral artery. C. The left optic nerve has been divided at the optic foramen and the anterior cerebral artery has been divided near its origin so that both optic nerves and the chiasm and tract could be elevated to show the perforating branches of the carotid artery. The Liliequist membrane (Lilieq. Memb.) is posterior to the infundibulum and hides the basilar artery but not the posterior cerebral artery (P.C.A.). The perforating branches of the ophthalmic segment pass upward to the infundibulum and the optic nerve, chiasm, and tract. D. Both optic nerves and both anterior cerebral arteries and the infundibulum have been divided to permit the optic nerves and chiasm to be elevated with a forceps for this view under the optic chiasm and across the diaphragma sellae and dorsum to the upper part of the basilar artery (B.A.) and the oculomotor nerves (III). The oculomotor nerves pass forward below the posterior cerebral arteries. The perforating branches of the supraclinoid segment of the carotid artery pass upward to supply the infundibulum, the optic chiasm and tracts, and the floor of the third ventricle; some enter the brain through the anterior perforated substance. The right AChA is very large. (From Gibo H, Lenkey C, Rhoton AL Jr: Microsurgical anatomy of the supraclinoid portion of the internal carotid artery. J Neurosurg 55:560-574, 1981.1
Figure 3.17. Posterior views of the perforating branches of the supraclinoid portion of the internal carotid artery (ICA). A. The basilar artery (B.A.) and brain stem have been divided and the floor of the third ventricle has been elevated to provide this posterior view of the supraclinoid portion of the ICA. The optic nerves (O.N.) can be followed back to the optic chiasm and the optic tracts (O.Tr.). The infundibulum of the pituitary gland (Infund.) arises from the tuber ciner-eum (Tuber Cin.). The mamillary bodies (Mam. Bodies) lie posterior to the tuber cinereum. The supraclinoid portion of the ICA gives rise to the posterior communicating (P.CO.A.), anterior choroidal (A.Ch.A.), middle cerebral (M.C.A.), and anterior cerebral (A.C.A.) arteries. The PCoA can be followed backward to its junction with its posterior cerebral arteries (P.C.A.). The supraclinoid portion of the ICA is divided into three segments on the basis of the origin of its major branches: the ophthalmic segment (C4-Op.) extends from the origin of the ophthalmic artery to the origin of the PCoA, the communicating segment (Cr-Co.) extends from the origin of the PCoA to the origin of the AChA, and the choroidal segment (C4-Ch.) extends from the origin of the AChA to the bifurcation into the anterior and middle cerebral arteries. The perforating branches arising from the ophthalmic segment in this specimen enter the optic tract, the floor of the third ventricle, and the infundibulum. The superior hypoph-yseal arteries (Sup. Hyp. A.) pass to the infundibulum. A single perforating artery arises from the communicating segment on each side. The posterior wall of the choroidal segment is the source of numerous perforating branches. The pituitary stalk passes through the diaphragma sellae (Diaph.) anterior to the dorsum sel-lae (Dorsum). B. The right half of the dorsum sellae and the right posterior clinoid process (Post. Clinoid) have been removed to expose the pituitary gland and its anterior (Ant. Lobe) and posterior (Post. Lobe) lobes. The infraclinoid segment of the carotid artery (C.A.) is exposed to the right of the pituitary gland. The apex of the basilar artery and the proximal parts of the posterior cerebral and the superior cerebellar (S.C.A.) arteries have been elevated to expose the pituitary stalk and floor of the third ventricle. The intracavernous portion of the carotid artery gives rise to the inferior hypophyseal (Inf. Hyp. A.) and the tentorial (Tent. A.) arteries. The posterior half of the right PCoA is duplicated. The ophthalmic segments give rise to the superior hypophyseal arteries. The left carotid artery has been divided just distal to the origin of the PCoA. C. Posterior views of the anterior part of the circle of Willis. The optic chiasm has been divided posterior to its junction with the optic nerves and anterior to where the infundibulum arises from the floor of the third ventricle. The superior hypophyseal arteries pass medially from the carotid artery to the infundibulum and also give rise to branches coursing to the lower surface of the optic chiasm. The communicating segment gives rise to one perforating branch on the right and two on the left. The choroidal segment gives rise to more perforating branches than the other segments, but they have been divided near their origin. The anterior cerebral artery gives rise to perforating branches coursing to the upper surface of the optic chiasm and nerves. The anterior communicating artery (A.Co.A.) is seen above the optic chiasm. (From Gibo H, Lenkey C, Rhoton AL Jr: Microsurgical anatomy of the supraclinoid portion of the internal carotid artery. J Neuro-surg 55:560574, 1981.)
Figure 3.18. Arterial relationships. A. Right lateral view. The temporal lobe (Temp. Lobe) has been retracted to expose the upper surface of the tentorium (Tent.) and the tentorial edge (Tent. Edge). The uncus, when not retracted, hangs over the free edge antero-lateral to the cerebral peduncle (Ped.). The collateral (Coll. Sulc.) and rhinal sulci (Rhinal Sulc.) separate the parahippocampal (Parahipp. Gyr.) and occipito-temporal gyri (Occip. Temp. Gyr.). The carotid artery (Car. A.) bifurcates below the anterior perforated substance (Ant. Perf. Subst.). The middle cerebral artery (M. C. A.) courses in the sylvian fissure (Sylvian Fiss.) between the temporal and frontal lobes (Front Lobe) and anterior cerebral artery (A.C.A.), which passes between the optic nerve (Optic N.) and olfactory tract (Olf. Tr.). The posterior communicating artery (Post Comm. A.) lies above the oculomotor nerve (III). The anterior choroidal artery (Ant Chor. A.) courses below the optic tract (Optic Tr.). The basilar artery (Bas. A.) gives off the superior cerebellar (S.C.A.) and posterior cerebral arteries (P.C.A.) in the anterior incisural space. The PCA gives off long (Long Circ. A.) and short circumflex (Short Circ. A.) arteries, anterior (Ant Temp. A.), middle (Mid. Temp. A.), and posterior temporal (Ant Temp. A.) middle (Mid. Temp. A.), and posterior temporal (Post Temp. A.) arteries, and medial posterior choroidal (Med. Post. Chor. A.) and cal-carine (Calc. A.) arteries. The SCA bifurcates into rostral (Ro. Tr.) and caudal trunks (Ca. Tr.). The superior (Sup. Coll.) and interior (Inf. Coll.) colliculi are supplied by branches of the SCA and PCA. The lateral mesencephalic sulcus (Lat Mes. Sulc.) separates the peduncular and tegmental parts of the midbrain. The pontomesencephalic sulcus (Pon. Mes. Sulc.) separates the midbrain and pons. B. Enlarged view. The carotid and basilar bifurcations are in the anterior incisural space. The anterior clinoid process (Ant Cli-noid) is lateral to the carotid artery. The lenticulos-triate branches (Lent. Str. A.) arise from the proximal part of the MCA. The anterior choroidal artery passes posteriorly below the optic tract to enter the middle incisural space. The oculomotor nerve enters the dura lateral to the posterior clinoid process (Post Clinoid). C. The temporal lobe has been removed to expose the choroid plexus (Chor. Plex.) within the temporal horn. The mamillary bodies (Mam. Body) lie above the basilar bifurcation. The recurrent artery (Rec. A.) arises from the АСА. (From Ono M, Ono M, Rhoton AL Jr, Barry M: Microsurgical anatomy of the region of the tentorial incisura. J Neurosurg 60:365-99, 1984.)
Figure 3.19. Tentorial incisura: arterial relationships. A. Anterior view. The anterior part of the left frontal lobe (Front. Lobe) has been removed to expose the anterior incisural space, which extends upward around the optic nerves (Optic N.) and chiasm and laterally along the sylvian fissure (Sylvian Fiss.). The internal capsule (Int. Cap.) is located above the anterior incisural space between the caudate nucleus (Caudate Nucl.) and the putamen. The rostrum, genu, and body of the corpus callosum (Corp. Call.) surround the anterior part of the lateral ventricle (Lat. Vent). The fornix extends along the lower margin of the septum
pellucidum (Sept. Pell.) and forms the anterior and superior margin of the foramen of Monro (For. Monro). The middle cerebral artery (M.C.A.) passes laterally in the sylvian fissure above the temporal lobe (Temp. Lobe) and gives rise to the lenticulostriate arteries (Lent. Str. A.). The anterior cerebral artery (A.C.A.) gives rise to the frontopolar (Front. Pol. A.), pericallosal (Pericall. A.), and recurrent (Rec. A.) arteries. Perforating branches (Perf. A.) from the АСА reach the lamina terminalis (Lam. Term.). The olfactory tract (Olf. Tr.) has been divided. The oculomotor nerve (Ш) is exposed behind the carotid artery (Car.A.). Choroid plexus (Chor.Plex.) is attached along the choroidal fissue (Chor.Fiss.) situated between the fornix and thalamus. B. Lateral view. The inferior part of the temporal lobe has been removed. The choroid plexus in the temporal horn (Temp. Horn) has been preserved. The branches of the MCA pass over the insula and loop around the frontal operculum (Front. Operc.). The posterior communicating artery (Post. Comm. A.) passes lateral to the infundibulum (Infiind.) and joins the proximal part of the posterior cerebral artery (P.C.A.). The anterior choroidal artery (Ant. Chor. A.) enters the choroid plexus in the temporal horn. The PCA arises at the bifurcation of the basilar artery (Bas. A.), passes posteriorly around the cerebral peduncle (Ped.), and gives rise to long (Long Circ. A.) and short circumflex (Short Circ. A.) arteries, lateral (Lat. Post. Chor. A.) and medial posterior choroidal (Med. Post. Chor. A.) arteries, and posterior pericallosal (Post. Pericall. A.), posterior temporal (Post. Temp. A.), calcarine (Calc. A.), and parietoocciptal (Par. Occip. A.) arteries. The superior cerebellar artery (S.C.A.) arises from the basilar artery and dips below the tentorial edge (Tent. Edge). The vein of Galen passes above the superior (Sup. Coll.) and inferior colliculi (Inf. Coll.). The lateral mesencephalic sulcus (Lat. Mes. Sulc.) is dorsal to the cerebral peduncle. C. Superolateral view. All of the tentorium except the free edge and the attachment along the petrous ridge has been removed. The basilar artery bifurcates in the anterior incisural space. The trochlear nerve (IV) encircles the brain stem between the PCA and SCA. The abducent nerve (VI) passes above the anterior inferior cerebellar artery (A.I.C.A.). The SCA bifurcates into a rostral (Ro. Tr.) and a caudal trunk (Ca. Tr.) above the trigeminal nerve (V). A bridging vein (Bridg. V.) enters the lateral wall of the cavernous sinus. The oculomotor nerve enters the dura lateral to the posterior clinoid process (Post. Clinoid). D. Another specimen with the temporal horn exposed, lateral view. The anterior choroidal artery gives rise to peduncular perforating branches (Ped. Perf. A.) before entering the temporal horn. The lateral posterior choroidal arteries pass through the choroidal f issue between the f imbria and pulvinar to enter the choroid plexus. The thalamogeniculate arteries (Thal. Gen. A.) enter the medial (Med. Gen. Body) and lateral (Lat. Gen. Body) geniculate bodies and the thalamus. The hippocampal formation on the medial surface of the temporal lobe includes the subiculum of the parahippocampal gyrus (Para-hipp. Gyr.), the dentate gyrus (Dentate Gyr.), and the fimbria. The amygdaloid nucleus (Amygd. Nucl.) is in the anteromedial margin of the temporal horn. E. Posterior incisural space, posterior-superior view. The occipital lobe, splenium, and the posterior part of the fornix have been removed to expose the SCA and PCA and their branches in the posterior incisural space. The SCA gives rise to vermian arteries (Ve. A.), which pass over the culmen just below the tentorial apex. The lateral posterior choroidal arteries pass around the pulvinar to reach the temporal horn and atrium. The medial posterior choroidal arteries encircle the brain stem to enter the roof of the third ventricle. The choroid plexus in the lateral ventricle is attached along the choroidal fissure. The striae terminalis (Str. Term.) courses along the groove between the caudate nucleus and the thalamus. The posterior part of the hippocampal formation (Hippocamp.) is exposed in the floor of the atrium. The tela choroidea (Tela), forming the upper wall of the velum interpositum (Vel. Interpos.), has been removed, but the lower layer, which is attached to the medial surface of the thalamus along the striae medullaris thalami, has been preserved. The suprapineal recess is located between the pineal body and the lower layer of the tela. F. The tentorium, except for the free edge and the strip along the straight sinus (Str. Sinus), and the tela choroidea in the roof of the third ventricle (3 Vent.) have been removed. The lateral posterior choroidal arteries ascend to reach the glomus of the choroid plexus. The medial posterior choroidal arteries pass forward in the roof of the third ventricle above the habenula and the massa intermedia (Massa Inter.). The SCA gives rise to the vermian and hemispheric arteries (He. A.) in the posterior incisural space. The quadrigeminal branches (Quad. A.) of the PCA and SCA pass to the colliculi. (From Ono M, Ono M, Rhoton AL Jr, Barry M: Microsurgical anatomy of the region of the tentorial incisura. J Neurosurg 60:365-399, 1984.)
Figure 3.21. Superior views of the sellar region. A. Carotid arteries are lateral to the optic nerves, anterior cerebral arteries are stretched across the superior surface of the optic chiasm, and the right optic nerve is shorter than the left. B. Long, tortuous anterior cerebral arteries are above the optic chiasm. The A1 segment of the right anterior cerebral artery loops in a complete circle between its origin from the carotid artery and its junction with its mate from the opposite side at the anterior communicating artery. A tortuous left anterior cerebral artery rests against the tubercu-lum sellae covering the prechiasmatic space. (From Renn WH, Rhoton AL Jr: Microsurgical anatomy of the sellar region. J Neurosurg 43:288-298, 1975.)
Figure 3.20. Inferior views with part of the temporal lobes removed to show the choroidal arteries supplying the choroid plexus of the inferior horn and atrium. A. The anterior choroidal arteries (A.Ch.A.) arise from the carotid arteries (C.A.), course posteriorly below the optic tracts (O.Tr.), and pass through the choroidal fissure to enter the choroid plexus (Ch. PL) of the temporal horn (Temp. Horn). The medial posterior choroidal arteries (M.P.Ch.A.) arise from the proximal or P1 segment (P-1) of the posterior cerebral artery (PCA), encircle the brain stem, and pass forward beside the pineal body and below the splenium of the corpus callosum to enter the choroid plexus in the roof of the third ventricle. The lateral posterior choroidal arteries (L.P.Ch.A.) arise from the P2 segment of the PCA (P-2) and course below the pulvinar of the thala-mus (Pulv.) and lateral geniculate bodies (L. Gen. Bo.) to enter the choroid plexus in the temporal horn and atrium. The carotid arteries are lateral to the optic nerves (O.N.) and optic chiasm (O.Ch.). The superior hypophyseal arteries (S. Ну. A.) pass medially from the carotid artery below the optic chiasm to the pituitary stalk (Pit. Stalk). The premamillary artery (Pre-mam. A.) arises from the posterior communicating artery (P.Co.A.) and enters the hypothalamus anterior to the mamillary bodies (Мат. В.). The calcarine (Cal.
A.), parietooccipital (P-O.A.), and posterior temporal arteries (P.T.A.) arise from the PCAs. The cut end of the basilar artery (B. A.) is between the cerebral peduncles (Ped.). B. Inferior view of temporal horn and atrium of right lateral ventricle. Enlarged view of the AChA, LPChA, and MPChA, showing their relationship to the optic tract, cerebral peduncles, brain stem, and pulvinar. The glomus of the choroid plexus projects into the atrium of the ventricle. The lenticulos-triate arteries (Len. Str. A.) arise from the middle cerebral artery (M.C.A.), and the anterior cerebral arteries (A.C.A.) pass above the optic chiasm. (From Fujii K, Lenkey C, Rhoton AL Jr: Microsurgical anatomy of the choroidal arteries: Lateral and third ventricles. J Neurosurg 52:165-188, 1980.)
Figure 3.22. Anterosuperior view of the perforating arteries entering the anterior wall of the third ventricle. A. The optic nerves (O.N.), chiasm (O.Ch.), and tracts (O.Tr.) are medial to the carotid arteries (C.A.) and below the anterior cerebral (A.C.A.) and anterior communicating (A.Co.A.) arteries. The frontal lobes (Fr. Lobe) and anterior perforated substance (Ant. Perf. Subst.) are above the anterior cerebral arteries. The anterior cerebral arteries give rise to a group of perforating arteries that enter the suprachiasmatic area and the upper surface of the optic chiasm. The recurrent artery (Rec. A.) arises from the anterior cerebral artery near the level of the anterior communicating artery. The olfactory nerves (Olf.N.) are above the anterior cerebral arteries. B. The anterior communicating artery gives rise to a series of perforating arteries (Perf. A.) that enter the area around the lamina terminalis (Lam. Теr.). С. A probe elevates the anterior communicating artery to expose two perforating arteries that pass through the lamina terminalis to enter the walls of the third ventricle (3V). The left recurrent artery arises from the anterior communicating artery in a common trunk with a branch to the frontal lobe (Fr. Br.). D. A precallosal artery (Pre.Cal.A.) arises from the anterior communicating artery and passes upward in front of the lamina terminalis to supply the rostrum (Rostrum) of the corpus callosum. (From Yamamoto I, Rhoton AL Jr, Peace DA: Microsurgery of the third ventricle: Part 1. Microsurgical anatomy. Neurosurgery 8:334-356, 1981.)
Internal Cerebral Vein The paired internal cerebral veins originate just behind the foramen of Monro and course posteriorly within the velum interpositum (Figs. 3.24, 3.26, and 3.27). Initially, they follow the gentle convex upward curve of the striae medul-laris thalami and, further distally, as they course along the superolateral surface of the pineal body, they follow the concave upward curve of the inferior surface of the splenium. The union of the paired veins to form the great vein may be located above or posterior to the pineal body and inferior or posterior to the splenium. The length of the internal cerebral vein varies from 19 to 35 mm (average, 30.2) (26).
Figure 3.23. Anterosuperior view of the lamina terminalis. A. The optic nerves (O.N.), chiasm (O.Ch.), and tracts (O.Tr.) are below the lamina terminalis (Lam. Теr.). The carotid arteries (C.A.) pass laterally below the optic nerves. The anterior cerebral (A.C.A.) and anterior communicating (A.Co.A.) arteries are above the optic chiasm. The recurrent arteries (Rec. A.) arise at the level of the anterior communicating artery. The olfactory nerves (Olf. N.) have been divided near their junction with the brain. B. The lamina terminalis has been opened along the dotted line shown in A to expose the cavity of the third ventricle (3V). The optic recess (O. Recess) extends downward between the lamina terminalis and the superior surface of the optic chiasm. The right anterior cerebral artery is hypoplastic, and the left anterior cerebral artery gives rise to both distal segments of the anterior cerebral artery. (From Yamamoto I, Rhoton AL Jr, Peace DA: Microsurgery of the third ventricle: Part 1. Microsurgical anatomy. Neurosurgery 8:334-356, 1981.) Venous Relationships The deep cerebral venous system is intimately related to the walls of the third ventricle (Figs. 3.24 to 3.26) (25, 26). It represents a formidable obstacle to the operative approaches to the third ventricle, especially in the region of the pineal gland, where the internal cerebral vein and the basal vein on each side converge on the great vein.
Great Vein and Straight Sinus The great vein, after being formed by the union of the internal cerebral veins, curves posteriorly and superiorly around the splenium to join the straight sinus at the anterior end of the junction of the falx and tentorium (Figs. 3.24, 3.26, and 3.28). The straight sinus courses posteroinferiorly along the falcotentorial junction to the tor-cular. The junction of the great vein with the straight sinus varies from being nearly flat if the tentorial apex is located below the splenium to forming a sharp angle if the apex is located above the splenium, so that the great vein must turn sharply upward to reach the straight sinus at the apex. The average length of the great vein is 12 mm (range, 8 to 25 mm) (26). The major tributaries of the great vein are the internal cerebral and basal veins, but it also receives blood from numerous other veins in the region. The convergence of multiple veins at the great vein results in a high density of veins in the area. Basal Vein The basal vein is formed below the anterior perforated substance in the anterior incisural space by the union of veins draining the walls of this space. It proceeds posteriorly between the midbrain and the temporal lobe to drain the walls of the middle incisural space before emptying into the internal cerebral or great vein (Figs. 3.24, 3.26, 3.28, and 3.29). The basal vein is divided into anterior, middle, and posterior segments that correspond to the parts of the vein coursing within the anterior, middle, and posterior incisural regions. The anterior and middle incisural regions are drained, almost totally, by tributaries of the basal vein. The veins in the posterior incisural region join the internal cerebral and great veins, as well as the basal vein.
Tributaries of the Deep Veins The tributaries of the internal cerebral, basal, and great veins drain the walls of the lateral and third ventricles, the periventricular white and gray matter, the corpus callosum, the thalamus, the septum pellucidum, the upper midbrain, the choroid plexus of the lateral and third ventricles, and the superior part of the cerebellum.
Figure 3.24. Schematic drawing of the ventricular veins. Lateral (top), anterior (middle), and superior (lower) views. The ventricular veins are divided into a medial and a lateral group. The ventricular veins drain into the internal cerebral (Int. Cer. V.}, basal (Basal V.), and great (V. Galen) veins. The lateral group consists of the anterior caudate vein (Ant. Caud. V.) in the frontal horn; the thalamostriate (Thal. Str. V.), posterior caudate (Post. Caud. V.), and thalamocaudate (Thal. Caud. V.) veins in the body; the lateral atrial vein (Lat. Atr. V.) in the atrium; and the inferior ventricular (Inf. Vent. V.) and amygdalar veins (Amygd. V.) in the temporal horn. The medial group is formed by the anterior septal vein (Ant. Sept. V.) in the frontal horn, the posterior septal veins (Post. Sept. V.) in the body, the medial atrial vein (Med. Atr. V.) in the atrium, and the transverse hippocampal veins (Trans. Hippo. V.) in the temporal horn. The transverse hippocampal veins drain into the anterior (Ant. Long. Hippo. V.) and posterior longitudinal hippocampal (Post. Long. Hippo. V.) veins. The superior cho-roidal veins (Sup. Chor. V.) drain into the thalamostriate and internal cerebral veins, and the inferior choroidal vein (Inf. Chor. V.) drains into the inferior
Ventricular Drainage The veins draining the deep gray and white matter of the cerebrum converge on the large veins coursing in and near the walls of the third ventricle (Figs. 3.24 to 3.29) (26). The veins from the frontal horn, the body of the lateral ventricle, and the surrounding gray and white matter drain into the internal cerebral vein; the veins from the temporal horn and the adjacent periventricular structures drain into the basal veins; and those draining the atrium and adjacent parts of the brain drain into the basal, internal cerebral, and great veins. The veins collecting blood from the periventricular white and gray matter join to form subepedymal channels in the walls of the lateral ventricles. These subependymal channels in the walls of the ventricle are divided into a medial and a lateral group. Both the lateral and the medial group pass through or adjacent to the choroidal fissure to reach the internal cerebral, basal, or great vein. The medial group consists of those veins that traverse the medial wall and roof of the frontal horn and body, the medial wall of the atrium, and the hippocompal surface of the floor of the temporal horn. The lateral group consists of those veins that traverse the lateral wall and floor of the frontal horn and body, the lateral wall of the atrium, and the roof of the temporal horn. In the temporal horn, the roof corresponds to the lateral group and the floor corresponds to the medial group. In the frontal horn, the medial group is formed by the anterior septal veins, which cross the septum pellucidum and fornix to join the internal cerebral vein, and the lateral group is formed by the anterior caudate veins, which cross the ventricular surface of the caudate nucleus and ter-
ventricular vein. The vein of Galen drains into the straight sinus (Str. Sinus). The anterior (Ant. Cer. V.) and deep middle cerebral (Deep. Mid. Cer. V.) veins join to form the basal vein. (From Ono M, Rhoton AL Jr, Peace D, Rodriguez R: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984.)
Figure 3.25. A and B. Ventricular veins. A. Anterior view (along the arrow in the insert) into the frontal horn (Front. Horn) and body of the lateral ventricle (Body Lat. Vent). The frontal horn is located anterior to the foramen of Monro (For. Monro) and has the septum pellucidum (Sept Pell.) in the medial wall, the genu and body of the corpus callosum (Corp. Call.) in the roof, the caudate nucleus (Caudate Nucl.) in the lateral wall, the genu of the corpus callosum in the anterior wall, and the rostrum of the corpus callosum in the floor. The body of the lateral ventricle has the thalamus in its floor, the caudate nucleus in the lateral wall, the body of the fornix and septum pellucidum in the medial wall, and the corpus
minate in the thalamostriate or internal cerebral vein. In the body, the medial group is formed by the posterior septal veins, and the lateral group is formed by the thalamostriate, thalamocau-date, and posterior caudate veins. The thalamostriate vein is the largest tributary of the internal cerebral vein. It courses anteriorly and medially in the groove between the caudate nucleus and the thalamus beneath the striae terminalis. At the level of the foramen of Monro, it curves around the anterior tubercle of the thalamus and joins the internal cerebral vein. The thalamocau-date originates at the lateral angle of the body of the lateral ventricle and runs across the caudate nucleus and thalamus to join the internal cerebral vein more posteriorly than does the thalamostriate vein. The posterior caudate veins cross the lateral wall of the body of the ventricle and empty into the thalamostriate or thalamocaudate vein. The posterior septal veins cross the posterior part of the septum pellucidum and pass around the fornix to enter the internal cerebral vein. In the atrium, the medial group is formed by the medial atrial veins, and the lateral group is formed by the lateral atrial veins. The atrial veins pass through the choroidal fissure to enter the basal, internal cerebral, or great vein. In the temporal horn, the veins draining the roof correspond to the lateral group, and the veins draining the floor correspond to the medial group. The medial group is formed from the transverse hippocampal veins that run in the floor on the sur-
face of the hippocampus and drain into the basal vein. The lateral group is formed by the inferior ventricular vein. It runs in the roof of the temporal horn and passes through the choroidal fissure to join the basal vein. Incisural and Cisternal Veins The main venous trunk related to the anterior incisural space is the basal vein (Figs. 3.28 to 3.30) (25, 26). The veins joining below the anterior perforated substance to form the basal vein include the olfactory vein, which runs posteriorly in the olfactory sulcus; the frontoorbital vein, which courses along the orbital surface of the frontal lobe; the deep middle cerebral vein, which receives the veins from the insula and passes medially across the limen insulae; the uncal veins, which course medially from the uncus; and the anterior cerebral vein, which descends on the lamina terminalis and crosses the optic chiasm to reach the basal vein. The paired anterior cerebral veins are joined across the midline above the optic chiasm by the anterior communicating vein and receive the paraterminal vein from the paraterminal and parolfactory gyri and the anterior pericallosal veins from the rostrum and genu of the corpus callosum. The veins on the surface of the brain stem that form the posterior wall of the anterior incisural space are divided into transversely or vertically oriented groups (21, 22, 25, 26). The transverse veins are the peduncular vein, which passes hor-
callosum in the roof. The choroid plexus (Chor. Plex.) is attached along the choroidal fissure (Chor. Fiss.), the cleft between the fornix and thalamus. The anterior septal veins (Ant. Sept. V.) cross the roof and medial wall of the frontal horn and pass posteriorly toward the foramen of Monro, where they join the anterior end of the internal cerebral veins. The anterior caudate veins (Ant. Caud. V.) cross the lateral wall of the frontal horn and join the thalamostriate vein (Thal. Str. V.), which passes through the foramen of Monro. The superior choroidal vein (Sup. Chor. V.) courses on the choroid plexus in the body. The posterior septal veins (Post. Sept. V.) cross the roof and medial wall of the body and pass through the margin of the choroidal fissure. The posterior caudate veins cross the lateral wall of the body and join the thalamostriate vein, which courses along the striothalamic sulcus (Str. Thal. Sulc.). Anterior (Ant. Superf.. Thal. V.) and superior (Sup. Superf. Thal. V.) superficial thalamic veins cross the surface of the thalamus. The anterior thalamic vein (Ant. Thal. V.) drains the nuclei in the anterior-superior part of the thalamus. B. Anterior-superior view (along the arrow in the insert) into the body, atrium, and occiptal horn (Occip. Horn) of the lateral ventricle. The calcar avis and bulb of the corpus callosum (Bulb Corp. Call.) form the medial wall of the atrium and occipital horn. The floor of the atrium is formed by the collateral trigone (Coll. Trig.). The roof and posterior part of the lateral walls are formed by the tapetum of the corpus callosum. The caudate nucleus is in the anterior part of the lateral wall of the atrium. The medial (Med. Atr. V.) and lateral (Lat. Atr. V.) atrial veins pass forward on the medial and lateral walls of the atrium toward the choroidal fissure. A thalamocaudate vein (Thal. Caud. V.) crosses the lateral wall posterior to the thalamostriate vein. The superior choroidal vein courses toward the foramen of Monro.
Figure 3.25. С and D. Ventricular veins (continued). C. Posterior view (along the arrow in the insert) into the atrium and temporal horn. The atrium has the tapetum of the corpus callosum in the roof, the bulb of the corpus callosum and the calcar avis in its medial wall, the hippocampal formation and the collateral trigone (Coll. Trig.) in the floor, the caudate nucleus and tapetum in the lateral wall, and the crus of the fornix and the pulvinar and choroid plexus in the anterior wall. The floor of the temporal horn (Temp. Horn) is formed by the hippocampus and collateral eminence (Coll. Eminence). The inferior choroidal vein (Inf. Chor. V.) courses on the choroid plexus in the temporal horn. The lateral atrial veins
izontally around the anterior surface of the cerebral peduncle and terminates in the basal vein at the junction of the anterior and middle inci-sural spaces, and the vein of the pontomesence-phalic sulcus, which courses below the peduncular vein in the pontomesencephalic sulcus. The vertically oriented veins on the posterior wall of the anterior incisural space are the median anterior pontomesencephalic vein, which courses in the midline and connects the peduncular veins above with the pontine veins below, and the lateral anterior pontomesencephalic veins, which course on the anterolateral surface of the cerebral peduncle and the pons and join the basal vein superiorly and the vein of the pontomesencephalic sulcus or a transverse pontine vein inferiorly. The venous relationshps in the middle incisural space are relatively simple (Figs. 3.28 to 3.30). The basal vein courses along the upper part of the cerebral peduncle and below the pul-vinar to reach the posterior incisural space. It may infrequently terminate in a tentorial sinus in the free edge at this level. The veins draining the medial wall, from anterior to posterior, are the lateral pontomesencephalic vein, which runs vertically on the lateral surface of the cerebral peduncle and pons and terminates in the basal vein superiorly and the vein of the pontomesencephalic sulcus inferiorly, and the lateral mesencephalic vein, which runs vertically along the
lateral mesencephalic sulcus and terminates above in the basal vein near the medial genicu-late body and below in a petrosal vein. Numerous veins from the lateral wall of the middle incisural space converge on the basal vein near the inferior choroidal point. These veins are the anterior hippocampal vein, which courses posteriorly from the anterior hippocampal sulcus; the uncal veins, which drain the medial surface of the uncus; and the anterior longitudinal hippocampal vein, which courses anteriorly along the dentate gyrus. The venous relationships in the posterior incisural space are the most complex in the cranium because the internal cerebral and basal veins and many of their tributaries converge on the great vein within this area (Figs. 3.38 to 3.30). The internal cerebral veins exit the velum interpositum and the basal veins exit the ambient cistern to reach the posterior incisural space, where they empty into the great vein. The great vein passes below the splenium to enter the straight sinus at the tentorial apex. Other veins converging on this area include the posterior pericallosal veins, which course posteriorly around the splenium to terminate in the great vein or the internal occipital vein; the lateral atrial veins, which drain the lateral wall and roof of the atrium and course medially through the choroidal fissure to terminate in the basal, internal cerebral, or great vein either directly or after
arise on the lateral wall and cross the tail of the caudate nucleus and the pulvinar to pass through the choroidal fissure. The medial atrial veins pass forward and penetrate the crus of the fornix near the choroidal fissure to reach the quadri-geminal cistern. Some of the medial atrial veins also drain the roof and floor. Transverse hippocampal veins (Transv. Hippo. V.) cross the floor of the atrium and temporal horn. Posterior superficial thalamic veins (Post. Superf. Thal. V.) cross the atrial surface of the thalamus. D. Anterior view (along the arrow in the insert) into the temporal horn. The floor of the temporal horn is formed by the collateral eminence and the hippocampal formation. The roof and lateral wall, from medial to lateral, are formed by the thalamus, the tail of the caudate nucleus, and the tapetum of the corpus callosum. The medial wall is little more than the cleft between the inferior surface of the thalamus and the fimbria. The amygdaloid nucleus (Amygd. Nucl.) bulges into the anteromedial part of the temporal horn. The pes hippocampus (Pes Hipp.), the bulbous digitated anterior end of the hippocampal formation, is in the anterior part of the floor. The fimbria of the fornix arises on the surface of the hippocampal formation and passes posteriorly to become the crus of the fornix. The choroid plexus is attached along the choroidal fissure. The inferior ventricular vein (Inf. Vent. V.) drains the roof of the temporal horn and receives the amygdalar vein (Amygd. V.) from the ventricular surface of the amygdaloid nucleus. The inferior choroidal vein joins the inferior ventricular vein. The transverse hippocampal veins draining the floor of the temporal horn pass medially through the choroidal fissure to enter the basal vein or its tributaries. (From Ono M, Rhoton AL Jr, Peace D, Rodiguez R: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 15:621657, 1984.)
Figure 3.26. Midsagittal sections. A. The brain stem has been sectioned at the level of the midbrain. The parahippocampal gyrus (Parahippo. Gyr.) has been depressed to expose the basal vein (Basal V.) between the uncus and the cerebral peduncle (Ped.) and the choroidal fissure (Chor. Fiss.), the cleft between the fimbria and the pulvinar. The anterior septal vein (Ant. Sept. V.) passes posteriorly on the septum pel-lucidum (Sep. Pell.) toward the foramen of Monro (For. Monro) and joins the internal cerebral vein (Int. Cer. V.) in the roof of the third ventricle (3rd Vent). The internal cerebral vein passes posteriorly above the massa intermedia (Massa Inter.) and the choroid
plexus (Chor. Plex.) in the roof of the third ventricle and exits the velum interpositum (Vel. Interpos.), situated between the layers of the tela choroidea (Tela) in the roof of the third ventricle, by passing below the splenium to join the vein of Galen (V. of Galen). The basal vein passes below the lateral geniculate body (Lat. Gen. Body) and pulvinar. The anterior longitudinal hippocampal vein (Ant. Long. Hippo. V.) joins the anterior end of the basal vein. The posterior longitudinal hippocampal (Post. Long. Hippo. V.), internal occipital (Int. Occip. V.), and lateral atrial (Lat. Atr. V.) veins join the posterior end of the basal vein. The posterior pericallosal vein (Post. Pericall. V.) passes around the splenium to join the great vein. The paraterminal vein (Paraterm. V.) passes in front of the lamina terminalis (Lam. Ter.) and anterior commissure (Ant. Comm.) on the paraterminal gyrus (Paraterm. Gyr.) and courses toward the anterior end of the basal vein. The habenular (Hab. Comm.) and posterior (Post. Comm.) commissures form the stalk of the pineal body. The fornix fuses to the lower surface of the splenium posteriorly and passes superior and anterior to the foramen of Monro anteriorly. The cin-gulate gyrus (Cing. Gyr.) is above the corpus callosum (Corp. Call.). The calcarine sulcus (Calc. Sulc.) divides the medial surface of the occipital lobe above the lingual gyrus (Lingual Gyr.). The infundibular (Infund.) and mamillary (Mam. Body) bodies are in the floor of the third ventricle. B. Enlarged view of the inferior part of the choroidal fissure. The veins exiting the temporal horn pass through the choroidal fissure, the cleft between the fimbria of the fornix and the pulvinar. The inferior end of the choroidal fissure, called the inferior choroidal point (Inf. Chor. Point), is located just behind the uncus. The basal vein passes below the medial (Med. Gen. Body) and lateral geniculate bodies. The anterior longitudinal hippocampal vein passes forward and the posterior longitudinal hippocampal vein passes backward on the dentate gyrus. The hippocampal sulcus (Hippo. Sulc.) separates the dentate and parahippocampal gyri, and the fimbriod-entate sulcus (Fimb. Dentate Sulc.) separates the fimbria and the dentate gyrus. The section of the brain stem crosses the red nucleus (Red. Nucl.). C. Enlarged view. The septum pellucidum and body of the fornix have been removed to expose the right lateral ventricle (Lat. Vent). The thalamostriate vein (Thal. Str. V.) passes forward between the caudate nucleus (Caudate Nucl.) and the thalamus and enters the internal cerebral vein. The anterior septal vein passes through the posterior margin of the foramen of Monro and enters the internal cerebral vein in the posterior part of the roof of the third ventricle. The stria medullaris thalami (Str. Med. Thal.) passes posteriorly in the wall of the third ventricle toward the habenular commissure. A superior choroidal vein (Sup. Chor. V.) courses on the surface of the choroid plexus. (From Ono M, Rhoton AL Jr, Peace D, Rodriguez R: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984.)
forming a common stem with the medial atrial veins; the medial atrial veins, which drain the medial wall and roof of the atrium and occipital horn and course anteromedially through the cho-roidal fissure before terminating in the internal cerebral vein; the posterior longitudinal hippocampal vein, which courses posteriorly along the dentate gyrus and terminates in the internal cerebral, basal, or great vein or one of the atrial veins; and the internal occipital veins, which originate in the calcarine and parietooccipital sulci and course anteromedially along the calcarine sulcus to terminate in the great vein. The largest vein from the infratentorial part of the posterior incisural space is the vein of the cerebellomesencephalic fissure, which is also called the precentral cerebellar vein (Figs. 3.28 and 3.30). It originates from the union of the paired veins of the superior cerebellular peduncle and courses anterosuperiorly through the cerebellomesencephalic fissure to terminate with the superior vermain vein in the great vein (21, 26). The veins of the superior cerebellular peduncle arise below the inferior colliculi on the external surface of the superior cerebellar peduncles. The superior vermian and the superior hemispheric veins from the anterosuperior part of the cerebellum arise from the infratentorial part of the posterior incisural space and pass forward under the free edge of the tentorium to empty into the great vein either directly or after joining with the vein of the cerebellomesencephalic fissure. Smaller thalamic, epithalamic, and tectal veins emerge from the walls of the posterior incisural space and terminate in the internal cerebral, basal, or great veins or the vein of the cerebellomesencephalic fissure. Choroidal Veins Two other veins that course near the walls of the third ventricle are the superior and inferior choroidal veins (26). Both lie within the choroid plexus of the lateral ventricles and are covered by ependyma (Figs. 3.24, 3.25, and 3.27 to 3.29). The superior choroidal vein runs in the choroid plexus of the atrium and body of the lateral ventricle. It terminates in either the thalamostriate
or the internal cerebral vein near the foramen of Monro. The inferior choroidal vein drains the choroid plexus of the temporal horn and atrium. It courses downward and forward in the roof of the temporal horn and joins the inferior ventricular vein near where the latter passes through the choroidal fissure or passes directly to the basal vein. Bridging Veins The veins coursing from the surface of the brain to the venous sinuses in the dura mater that are encountered in retracting the brain to reach the walls of the third ventricle include the superficial superior cerebral veins that cross from the superior margin of the cerebrum to the superior sagittal sinus; the cingulate and callosal veins that drain the cingulate gyrus and upper surface of the corpus callosum and pass into the inferior sagittal sinus; the superficial middle cerebral veins that cross the sylvian fissure to enter the sphenoparietal sinus on the inferior margin of the sphenoid ridge; and the inferior cerebral veins that drain the lower portions of the lateral surface of the temporal lobe and the undersur-faces of the occipital and temporal lobes and empty into the transverse and tentorial sinuses. Of the 3 to 12 (average, 6.8) superficial superior cerebral veins per hemisphere, the number anterior and posterior to the central or rolandic vein are close to the same (55). The veins draining the corresponding part of the medial and lateral surfaces of the superior margin of the cerebral hemisphere usually join before entering the superior sagittal sinus. The vein of Labbe, which extends from the sylvian fissure to the transverse sinus, is the largest inferior cerebral vein. There are usually no veins between the medial aspect of the occipital pole and the transverse and straight sinuses, but there are frequently several bridging veins (range, 2 to 7) between the lateral surface of the occipital pole and the transverse sinus (55). There are also numerous bridging veins crossing the interval between the superior surface of the cerebellum and the sinuses in the tentorium. They are more numerous medially than laterally (21).
Figure 3.27. A-F. Superior view. A. The superior part of the cerebral hemisphere has been removed to expose the lateral ventricles. The frontal horn (Front. Horn) extends into the frontal lobe (Front. Lobe). The body of the lateral ventricle (Body Lat. Vent.) is deep to the frontal and parietal lobes (Par. Lobe). The atrium extends posteriorly toward the occipital lobe (Occip. Lobe). The thalamostriate vein (Thal. Str. V.) courses along the junction of the caudate nucleus (Caudate Nucl.) and the thalamus toward the foramen of Monro (For. Monro). The anterior caudate veins (Ant. Caud. V.) course on the caudate nucleus in the lateral wall of the frontal horn. The superior choroidal veins (Sup. Chor. V.) course on the choroid plexus (Chor. Plex.) and join the thalamostriate vein near the foramen of Monro. Medial atrial veins (Med. Atr. V.) course along the medial wall of the atrium. The superior sagittal sinus (Sup. Sag. Sinus) courses in the edge of the falx. B. Enlarged view. The septum pellucidum (Sept. Pell.) separates the lateral ventricles. The anterior caudate and thalamostriate veins pass toward the foramen of Monro. The small anterior superficial thalamic vein (Ant. Superf. Thal. V.) crosses the thalamus near the anterior thalamic tubercle (Ant. Thal. Tuber.). The choroidal fissure (Chor. Fiss.) is located between the fornix and thalamus. The bulb of the corpus callosum (Bulb Corp. Call.) is in the medial wall of the atrium. The cingulate gyrus (Cing. Gyr.) wraps around the corpus callosum (Corp. Call). C. Enlarged view of the frontal horns. The anterior septal veins (Ant. Sept. V.) pass medially on the posterior surface of the genu of the corpus callosum (Genu Corp. Call.) and turn posteriorly on the septum pellucidum. The anterior caudate veins cross the head of the caudate nucleus. The posterior septal veins (Post. Sept. V.) cross the medial wall of the body of the lateral ventricle. On the left side, the anterior septal, thalamostriate, and superior choroidal veins course toward the foramen of Monro. On the right side, the thalamostriate vein is absent and most of the lateral wall of the body is drained by a large thalamocaudate vein
Figure 3.27. G-I. Superior view (continued). G. Another specimen. The thala-mostriate veins receive the superior choroidal, anterior septal, anterior thalamic, and anterior superficial thalamic veins and pass through the choroidal fissure, just behind the foramen of Monro. The striae medullaris thalamic (Str. Med. Thal.) pass along the superomedial margins of the thalami. H. Another specimen. The tela choroidea (Tela) extends across the roof of the third ventricle above the internal cerebral veins. A large superior superficial thalamic vein on the left side receives several small choroidal veins (Chor. V.) and joins the anterior part of the internal cerebral vein. The anterior caudate, anterior septal, anterior thalamic, and superior choroidal veins converge on the anterior end of the internal cerebral vein. I. Another specimen. The right internal cerebral vein receives duplicate thalamostrlate veins, called the anterior (Ant. Thal. Str. V.) and posterior (Post. Thal Str. V.) thalamostriate veins. The right lateral atrial vein (Lat. Atr. V.) joins the posterior thalamostriate vein. On the left side, a single large thalamostriate vein passes through the foramen of Monro to join the internal cerebral vein. The inferior sagittal sinus (Inf. Sag. Sinus) courses in the lower margin of the falx. (From Ono M, Rhoton AL Jr, Peace D, Rodriguez R: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984.)
(Thal. Caud. V.), which receives the posterior caudate vein (Post. Caud. V.) and passes posteriorly in the striothalamic sulcus (Str. Thal. Sulc.). D. The body of the fornix has been removed at the junction of its attachment to the crura posteriorly and the columns anteriorly. The removal of the tela choroidea in the roof of the third ventricle (3 Vent.) exposes the internal cerebral veins (Int. Сет. V.) above the massa intermedia (Massa Inter.). The choroid plexus on the right side has been removed along its attachment to the tenia choroidea (Tenia Chor.). The anterior caudate, anterior septal, anterior superficial thalamic, and superior choroidal veins converge on the left thalamostriate vein at the foramen of Monro. The medial atrial veins unite to form a common trunk. The posterior longitudinal hippocampal veins (Post. Long. Hippo. V.) course on the dentate gyri (Dentate Gur.) and join the internal cerebral veins. The right posterior longitudinal hippocampal vein unites with the thalamocaudate vein before entering the internal cerebral vein. E. Another specimen. The fornix has been removed. The anterior caudate veins pass along the lateral wall of the frontal horn and join the thalamostriate vein on the left side and a thalamocaudate vein on the right side. The superior choroidal, superior superficial thalamic (Sup. Superf. Thal. V.), and anterior septal veins join the internal cerebral veins near the foramen of Monro. On the right side, the lateral wall of the body is drained by a large thalamocaudate vein. A large anterior caudate vein passes along the superolateral angle of the right lateral ventricle to join the thalamocaudate vein. F. Enlarged view. The tributaries converging on the anterior end of the internal cerebral vein at the foramen of Monro include the anterior septal, anterior caudate, thalamostriate, superior choroidal, anterior thalamic (Ant Thal. V.), and anterior superficial thalamic veins.
Figure 3.28. A and B. Cisternal veins. A. Anterolateral view. The insert shows the direction of view. The frontal (Front, Lobe) and temporal (Temp. Lobe) lobes have been retracted away from the floor of the anterior (Ant. Fossa) and middle cranial (Mid. Fossa) fossae. The veins converging on the anterior end of the basal vein (Basal V.) below the anterior perforated substance (Ant. Perf. Subst.) are the deep middle cerebral veins (Deep Mid. Cer. V.) from the sylvian fissure (Sylvian Fiss.); the olfactory vein (Olf. V.), which drains posteriorly along the olfactory tract (Olf. Tr.) near the gyrus rectus (Gyr. Rectus); the frontoorbital veins (Front. Orb. V.), which drain the orbital gyri (Orb. Gyr.); the inferior striate veins (Inf. Str. V.), which exit the anterior perforated substance; and the anterior cerebral veins (Ant. Cer. V.), which are joined above the optic chiasm by the anterior communicating vein (Ant. Comm. V.). The peduncular vein (Ped. V.} passes around the cerebral peduncle (Ped.) above the oculomotor nerve (III) and joins the median anterior pontomesencephalic vein (Med. Ant. Pon. Mes. V.) in the midline and the basal vein laterally. The infundibulum (Infund.) passes inferiorly behind the anterior clinoid process (Ant. Clinoid), optic nerve (Optic N.), and internal carotid artery (Int. Car. A.). The lateral anterior pontomesencephalic vein (Lat. Ant. Pon. Mes. V.) joins the vein of the pontomesencephalic sulcus (V. of Pon. Mes. Sulc.) below and the basal vein above. The inferior thalamic veins (Inf. Thal. V.) arise behind and the premamillary veins (Preman. V.) arise in front of the mamillary bodies. The inferior ventricular vein (Inf. Vent. V.) exits the temporal horn above the parahippocampal gyrus (Parahippo. Gyr.) and enters the basal vein. An uncal vein (Uncal V.) passes medially from the uncus. The trochlear nerve (IV) courses near the tentorial edge (Tent. Edge). B. Lateral view, right side. The temporal lobe has been elevated, as shown in the insert. The tentorium (Tent.) extends along the side of the brain stem. The basal vein passes around the brain stem and joins the vein of Galen (V. of Galen). The tributaries of the basal veins lateral to the brain stem include the lateral mesen-cephalic vein (Lat. Mes. V.), which courses in the lateral mesencephalic sulcus; the inferior ventricular vein, which drains the roof of the temporal horn; the anterior hippocampal vein (Ant. Hippo. V.), which courses along the sulcus between the uncus and the parahippocampal gyrus; the anterior longitudinal hippocampal vein (Ant. Long. Hippo. V.), which courses along the dentate gyrus; and the medial temporal veins (Med. Temp. V.) from the inferomedial surface of the temporal lobe. In the pineal region, the basal vein receives the lateral atrial vein (Lat. Atr. V.) from the lateral wall of the atrium. The internal cerebral veins (Int. Cer. V.) pass above the pineal body. The superior vermian (Sup. Ve. V.) and superior hemispheric (Sup. He. V.) veins from the cerebellum and the vein of the cerebellomesencephalic fissure (V. of Cer. Mes. Fiss.) from the fissure between the midbrain and cerebellum ascend to join the vein of Galen. Tectal veins (Tectal V.) drain the colliculi. A transverse pontine vein (Trans. Pon. V.) crosses the pons.
Figure 3.28. С and D. Cisternal veins (continued). C. Posterior view. The insert shows the direction of view. The occipital (Occip. Lobe) and parietal (Par. Lobe) lobes have been retracted to expose the termination of the internal cerebral and basal veins in the vein of Galen. The internal occipital (Int. Occip. V.) and posterior pericallosal (Post. Pericall. V.) veins join the internal cerebral vein. The posterior longitudinal hippocampal vein (Post. Long. Hippo. V.) passes along the dentate gyrus (Dentate Gyr.) and joins the medial atrial vein (Med. Atr. V.). The lateral mesencephalic, posterior thalamic (Post. Thal. V.), and inferior ventricular veins join the basal vein. Tectal veins pass from the superior (Sup/. Coll.) and inferior colliculi (Inf. Coll.). The medial (Med. Gen. Body) and lateral (Lat. Gen. Body) geniculate bodies are below the pulvinar. The inferior sagittal sinus (Inf. Sag. Sinus) and the vein of Galen join the straight sinus (Str. Sinus). D. Right anterolateral view with the anterior portion of the right cerebral hemisphere removed to expose the upper brain stem and the third ventricle (3 Vent.) in the midline. The brain stem was sectioned at the level of the cerebral peduncle. The anterior cerebral veins join the deep middle cerebral vein to form the basal vein. The basal vein encircles the brain stem and along its course receives the peduncular, inferior ventricular, anterior hippocampal, anterior longitudinal hippocampal, posterior thalamic, lateral atrial, lateral anterior pontomesencephalic, and lateral mesencephalic veins. The superior vermian vein receives the superior hemispheric and tectal veins and the vein of the cerebellomesencephalic fissure. The paraterminal (Paraterm. V.) and anterior pericallosal (Ant. Pericall. V.) veins join the anterior cerebral vein. The internal cerebral vein courses in the velum interpositum (Vel. Interpos.) in the roof of the third ventricle (3 Vent). The collateral eminence (Coll. Eminence) sits above the collateral sulcus (Coll. Sulc.) in the floor of the temporal horn. Septal veins (Sept. V.) cross the septum pellucidum (Sept. Pell.). The choroid plexus (Chor. Plex.) passes through the foramen of Monro (For. Monro) to reach the roof of the third ventricle. (From Ono M, Rhoton AL Jr, Peace D, Rodriguez R: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984.)
Figure 3.29. A-D. A. Lateral view of the left side. The temporal (Temp. Lobe) and frontal (Front. Lobe) lobes have been elevated, as shown in the insert. The tentorium (Tent.) has been opened above the cerebellum. The basal vein (Basal V.) arises below the anterior perforated substance (Ant. Perf. Subst.), passes around the cerebral peduncle (Ped.), and joins the vein of Galen (V. of Galen; dashed lines). The veins converging on the anterior end of the basal vein are the olfactory veins (Olf. V.), which pass along the olfactory tract (Olf. Tr.); the inferior striate veins (Inf. Str. V.), which descend through the anterior perforated substance; the deep middle cerebral veins (Deep Mid. Cer. V.), which pass medially from the sylvian fissure (Sylvian Fiss.); and the anterior cerebral veins (Ant. Cer. V.), which pass across the optic chiasm. The tributaries of the basal veins lateral to the brain stem include the peduncular vein (Ped. V.), which passes around the cerebral peduncle; the lateral mesencephalic vein (Lat. Mes. V.), which courses in the lateral mesencephalic sulcus; and the inferior ventricular vein (Inf. Vent. V.), which drains the roof of the temporal horn. Tributaries of the inferior ventricular vein are the anterior hippocampal vein (Ant. Hippo. V.), which courses along the sulcus lateral to the uncus; the anterior longitudinal hippocampal vein (Ant. Long. Hippo. V.), which courses along the dentate gyrus; and the medial temporal veins (Med. Temp. V.) from the inferomedial surface of the temporal lobe. In the pineal region, the basal vein receives the lateral atrial vein (Lat. Atr. V.) from the lateral wall of the atrium. The internal occipital vein (Int. Occip. V.) from the medial surface of the occipital lobe joins the internal cerebral vein. The superior vermian (Sup. Ve. V.) and superior hemispheric (Sup. He. V.) veins from the cerebellum and the vein of the cerebellomesencephalic fissure (V. of Cer. Mes. Fiss.) ascend to join the vein of Galen. The optic nerve (Optic N.) passes above the carotid artery (Car. A.). The optic tract (Optic Tr.) terminates in the lateral geniculate body (Lat. Gen. Body). The infundibulum (Infund.) passes inferiorly in front of the mamillary bodies (Mam. Body). The uncus protrudes over the free edge (Tent. Edge) of the tentorium above the oculomotor nerve (III). The parahip-pocampal gyrus (Parahippo. Gyr.) is medial to the collateral (Coll. Sulc.) and rhinal (Rhinal Sulc.) sulci. The medial geniculate body (Med. Gen. Body) is superolateral to the inferior (Inf. Coll.) and superior (Sup. Coll.) colliculi. The pontomesencephalic sulcus (Pon. Mes. Sulc.) separates the pons and midbrain. A transverse pontine vein (Trans. Pon. V.) crosses the pons. B. The tip of the temporal lobe has been displaced posteriorly. Tributaries converging on the basal vein below the anterior perforated substance include the deep middle cerebral, insular (Insular V.), inferior striate, olfactory, and anterior cerebral veins. In its course around the midbrain, the basal vein receives the peduncular, lateral mesencephalic, and inferior ventricular veins. The uncal veins (Uncal. V.) drain the semilunar (Semilunar Gyr.) and ambient (Ambient Gyr.) gyri. The premam-illary vein (Premam. V.) joins the peduncular vein. The anterior communicating vein (Ant. Comm. V.) connects the paired anterior cerebral veins above the optic chiasm. C. Enlarged view of the region of the inferior ventricular vein. It exits the ventricle through the choroidal fissure, passes near the medial and lateral geniculate bodies, and receives the uncal, anterior hippocampal, anterior longitudinal hippocampal, and medial temporal veins. The anterior hippocapmal vein arises in the anterior hippocampal sulcus (Ant. Hippo. Sulc.). The brachium of the inferior colliculus (Brach. Inf. Coll.) is directed from the inferior colliculus toward the medial geniculate body. D. Enlarged view. The veins converging on the region of the great vein are the basal, internal cerebral (Int. Cer. V.), lateral atrial, superior vermian, internal occipital, and tectal (Tectal V.) veins and the vein of the cerebellomesencephalic fissure. The lemniscal trigone (Lemniscal Trig.) is posterior to the cerebral peduncle.
Figure 3.29. E-H. E. The posterior part of the basal and lateral atrial veins has been removed (dashed lines) to expose the veins in the quadrigeminal cistern. The tectal veins arise from the region of the quadrigeminal plate and pass posteriorly to join the basal vein and the vein of the cerebellomesencephalic fissure. The veins on the superior surface of the cerebellum converge above the apex of the vermis to form the superior vermian vein. The internal occipital vein has been divided near its origin. The epithalamic veins (Epithal. V.) arise in the region of the pineal body. The vein of Galen enters the straight sinus at the apex of the tentorium (Apex. Tent.). F. The orientation is shown in the insert. The temporal horn (Temp. Horn) has been opened between the superior (Sup. Temp. Gyr.) and the middle temporal (Mid. Temp. Gyr.) gyri to expose the veins crossing the roof and floor of the temporal horn. The inferior ventricular vein drains the anterior part of the roof and receives the amygdalar vein (Amygd. V.) from the ventricular surface of the amygdaloid nucleus (Amygd. Nucl.). The lateral atrial vein drains the posterior part of the roof. The transverse hippocampal veins (Trans. Hippo. V.) pass medially across the hippocampus in the floor of the temporal horn and through the junction of the fimbria of the fornix and hippocampus to reach the basal vein. The inferior choroidal vein (Inf. Chor. V.) courses on the choroid plexus. The medial atrial vein (Med. Atr. V.) drains the medial wall of the atrium. G. Enlarged view. The tenia fimbria is a small ridge on the edge of the fimbria along which the tela choroidea and choroid plexus are attached. The choroid plexus has been detached and reflected upward. The inferior ventricular and inferior choroidal veins pass through the choroidal fissure near the inferior choroid point (Inf. Chor. Point), which is defined as the inferior end of the choroidal fissure. The transverse hippocampal veins cross the floor of the temporal horn. Я. The neural structures forming the roof of the temporal horn have been removed, the choroid plexus has been reflected upward, and the choroidal fissure has been opened to expose the surface of the subiculum and the dentate gyms (Dentate Gyr.) facing the ambient cistern. A posterior superficial thalamic vein (Post. Superf. Thal. V.) joins the posterior longitudinal hippocampal vein (Post. Long. Hippo. V.) on the dentate gyrus. The calcar avis forms the lower half of the medial wall of the atrium. Both the collateral eminence (Coll. Eminence) in the floor of the temporal horn and the collateral trigone (Coll. Trig.) in the floor of the atrium overlie the collateral sulcus on the inferior surface of the temporal lobe. The pes hippocampus (Pes. Hippo.) is the bulbous, digitated anterior part of the hippocampus. (From Ono M, Rhoton AL Jr, Peace D, Rodriguez R: Microsurg-ical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984.)
Figure 3.30. A. Posterior view. Insert shows the orientation. The occipital lobe (Occip. Lobe) has been retracted, and the tentorium (Tent.) has been opened to expose the termination of the internal cerebral (Int. Cer. V.) and basal (Basal V.) veins in the vein of Galen (V. of Galen.). The internal occipital (Int. Occip. V.), posterior pericallosal (Post. Pericall. V.), and posterior longitudinal hippocampal (Post. Long. Hippo. V.) veins drain into the internal cerebral vein. The lateral mesencephalic (Lat. Mes. V.), posterior thalamic (Post. Thal. V.), and inferior ventricular (Inf. Vent. V.) veins join the basal vein. The lingual gyrus (Lingual Gyr.) is below the calcarine sulcus (Calc. Sulc.). The anterior part of the parahip-pocampal gyrus (Parahippo. Gyr.) is lateral to the uncus. The medial geniculate body (Med. Gen. Body) is below the pulvinar. B. Another specimen. The occiptal lobe has been removed to expose the atrium and the glomus of the choroid plexus (Chor. Plex.). The basal vein courses below the pulvinar. A posterior septal vein (Post. Sep. V.) crosses the roof and medial wall of the body of the lateral ventricle (Body Lat. Vent.). A thalamocaudate vein (Thal. Caud. V.) joins the internal cerebral veins above the pineal body. The lower layer of tela choroidea (Tela) in the roof of the third ventricle is attached to the habenular trigones (Hab. Trig.). The superior choroidal vein (Sup. Chor. V.) drains into the internal cerebral vein, and the inferior choroidal vein (Inf. Chor. V.) drains into the basal vein. The oculomotor nerve (Ш) passes along the medial margin of the cerebral peduncle (Ped.). The lateral atrial vein (Lat. Atr. V.) joins the basal vein above the superior colliculus (Sup. Coll.). The lateral mesencephalic vein ascends anterior to the brachium of the inferior colliculus (Brach. Inf. Coll.). The caudate nucleus (Caudate Nucl.) is lateral to the thalamus. C. Right posterolateral view of another specimen. The superior vermian vein (Sup. Ve. V.) receives the vein of the cerebellomesencephalic fissure (V. of Cer. Mes. Fiss.) and the superior hemispheric veins (Sup. H. V.). The tectal veins (Tectal V.) drain the region of the superior and inferior colliculi (Inf. Coll.). D. Another specimen. The internal cerebral vein has been elevated to show the tectal veins joining the superior vermian vein and the vein of the cerebellomesencephalic fissure. The trochlear nerve (IV) arises below the inferior colliculus. E. Midsagittal section of the pineal region. The internal cerebral vein courses in the velum interpositum (Velum Interpos.) above the massa intermedia (Massa Inter.). The velum interpositum is located between the layers of tela choroidea in the roof of the third ventricle (3 Vent.). The internal cerebral vein receives the medial atrial vein (Med. Atr. V.) from the lateral ventricle (Lat. Vent.) and the epithalamic veins (Inf. Epith. V.) from the region of the pineal body. The tectal veins join the superior vermian vein and the vein of the cerebellomesencephalic fissure. The vein of Galen and the inferior sagittal sinus (Inf. Sag. Sinus) drain into the straight sinus. The fourth ventricle (4 Vent.) is behind the pons. F. Inferior view. The cerebellum has been removed by dividing the superior (Sup. Cer. Ped.), middle (Mid. Cer. Ped.), and inferior cerebellar peduncles just posterior to the floor of the fourth ventricle. The colliculi are above the superior medullary velum (Sup. Med. Vel.). The right basal vein enters a sinus in the tentorium (Tent. Sinus), rather than joining the vein of Galen. The posterior mesencephalic (Post. Mes. V.), internal occipital, posterior pericallosal, and posterior longitudinal hippocampal veins drain into the internal cerebral veins. The epithalamic veins arise in the region of the pineal gland. The vein of the cerebellomesencephalic fissure drains into the superior vermian vein. (From Ono M, Rhoton AL Jr, Peace D, Rodriguez R: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984.)
Sellar Region The purpose of this section is to review the anatomy of the cranial base important to the various operative approaches to the region of the sella turcica and third ventricle. It deals with the sella turcica, sphenoid bone, sphenoid sinus, cavernous sinus, and pituitary gland. Sphenoid Bone The sphenoid bone is located in the center of the cranial base (28, 40-42). Some part of it is exposed in transcranial operations through the anterior, middle, and posterior cranial fossae and in subcranial approaches through the nose and orbit (Figs. 3.31 and 3.32). The intimate contact of the body of the sphenoid bone with the nasal cavity below and the pituitary gland above has led to the transsphenoidal route being the operative approach of choice for selected tumors involving the third ventricle. The neural relationships of the sphenoid bone
are among the most complex of any bone: the olfactory tracts, gyrus rectus, and posterior part of the frontal lobe rest against the smooth upper surface of the lesser wing; the pons and mesen-cephalon lie posterior to the clival portion; the optic chiasm lies posterior to the chiasmatic sul-cus; and the second through sixth cranial nerves are intimately related to the sphenoid bone. All exit the skull through the optic canal, superior orbital fissure, foramen rotundum, or foramen ovale, all foramina located in the sphenoid bone. The sphenoid bone has many important arterial and venous relationships: the carotid arteries groove each side of the sphenoid bone and may bulge into the sphenoid sinus; the basilar artery rests against its posterior surface; the circle of Willis is located above its central portion; and the middle cerebral artery courses parallel to the sphenoid ridge of the lesser wing. The cavernous sinuses rest against the sphenoid bone, and in-tercavernous venous connections line the walls of the sella turcica and dorsum sellae.
Figure 3.31. Osseous relationships of the sphenoid bone. The sphenoid bone is outlined in each view. A. Anterior view. B. Superior view. C. Lateral view. D. Inferior view. (From Rhoton AL Jr, Hardy DG, Chambers SM: Microsurgical anatomy and dissection of the sphenoid bone, cavernous sinus and sellar region. SurgNeurol 12:63-104, 1979.)
Figure 3.32. A to D. Sphenoid bone. Anterior views. A. Conchal type sphenoid bone. B. Bone with presellar type sphenoid sinus. C. Bone with sellar type sphenoid sinus and well-defined sphenoid ostia. D. Bone with sellar type sphenoid sinus with poorly defined sphenoid ostia and obliquely oriented sphenoidal septae. In the anterior view of the skull the sphenoid bone forms part of the posterior wall and roof of the orbit. In the lateral view the greater wing of the sphenoid bone projects above the zygoma, and the pterygoid plates project below in the lateral view. In the anterior view the sphenoid bone resembles a bat with wings outstretched (Figs. 3.31 and 3.32). It has a central portion called the body; the lesser wings, which spread outward from the superolateral part of the body; the two greater wings, which spread upward from the lower part of the body; and the superior orbital fissure, which is situated between the greater and lesser wings. The vomer, the pterygoid processes, and the medial and lateral pterygoid plates are directed downward from the body. The body of the sphenoid bone is more or less cubical and contains the sphenoid sinus. The superior orbital fissure, through which the oculomotor, troch-lear, and abducens nerves and the ophthalmic division of the trigeminal nerve pass, is formed
on its inferior and lateral margins by the greater wing and on its superior margin by the lesser wing. The inferior surface of the lesser wing forms the posterior part of the roof of each orbit and the exposed surface of the greater wing forms the posterior wall of the orbit. The optic canals are situated above the superomedial margin of the superior orbital fissure. The sphenoid ostia open into the sinus. In the superior view the pituitary fossa occupies the central part of the body and is bounded anteriorly by the tuberculum sellae and posteriorly by the dorsum sellae. The chiasmatic groove, a shallow depression between the optic foramina, is bounded posteriorly by the tuberculum sellae and anteriorly by the planum sphenoidale. The frontal lobes and the olfactory tracts rest against the smooth upper surface of the lesser wing and the planum sphenoidale. The posterior margin of the lesser wing forms a free edge called the sphenoid ridge, which projects into the lateral cerebral fissure to separate the
Figure 3.32 E-H. Sphenoid bone (continued). E. Posterior view. F. Lateral view. G. Superior view. H. Inferior view. (From Rhoton AL Jr, Hardy DG, Chambers SM: Microsurgical anatomy and dissection of the sphenoid bone, cavernous sinus and sellar region. Surg Neural 12:63-104, 1979.) frontal and temporal lobes. The anterior clinoid processes are located at the medial end of the lesser wings, the middle clinoid processes are lateral to the tuberculum sellae, and the posterior clinoid processes are situated at the superolateral margin of the dorsum sellae. The dorsum sellae is continuous with the clivus. The upper part of the clivus is formed by the sphenoid bone and the lower part by the occipital bone. The carotid sulcus is lateral to the body of the sphenoid bone. The superior aspect of each greater wing is concave upward and is filled by the tip of each temporal lobe. The foramen rotundum, through which the maxillary division of the trigeminal nerve passes, is located at the junction of the body and greater wing. The foramen ovale transmits the mandibular division of the trigeminal nerve and the foramen spinosum transmits the middle meningeal artery. When viewed from in-feriorly, the vomer, a separate bone, frequently remains attached to the anterior half of the body
of the sphenoid, and its most anterior portion separates the sphenoid ostia. The pterion and the "keyhole" are two important anatomical landmarks in the region of the greater wing in the lateral view (Fig. 3.31). The pterion is located over the upper part of the greater wing. The "keyhole" is located just behind the junction of the temporal line and the zygo-matic process of the frontal bone several centimeters anterior to the pterion. A burr hole placed over the pterion will be located at the lateral end of the sphenoid ridge. Bone in the region of the pterion is often removed to improve the operative exposure along the sylvian fissure. A burr hole placed at the keyhole will expose the orbit at its lower margin and dura over the frontal lobe at its upper margin. Sphenoid Sinus The sphenoid sinus is subject to considerable variation in size and shape and to variation in
the degree of pneumatization (Figs. 3.32 to 3.34) (3, 7, 12). It is present as minute cavities at birth, but its main development takes place after puberty. In early life, it extends backward into the presellar area and subsequently expands into the area below and behind the sella turcica, reaching its full size during adolescence. As the sinus enlarges, it may partially encircle the optic canals. When the sinus is exceptionally large, it extends into the roots of the pterygoid processes or greater wing of the sphenoid bone and may even extend into the basilar part of the occipital bone. As age advances, the sinus frequently undergoes further enlargement associated with absorption of its bony walls. Occasionally there are gaps in its bone with the mucous membrane lying directly against the dura mater. There are three types of sphenoid sinus in the adult: conchal, presellar, and sellar types, depending on the extent to which the sphenoid bone is pneumatized. In the conchal type the area below the sella is a solid block of bone without an air cavity. In the presellar type of sphenoid sinus, the air cavity does not penetrate beyond a vertical plane parallel to the anterior sellar wall. The sellar type of sphenoid sinus is the most common, and here the air cavity extends into the body of the sphenoid below the sella and as far posteriorly as the clivus. In our previous study in adult cadavers, this sinus was of a presellar type in 24% and of the sellar type in 75% (36). The conchal type is infrequent in the adult. The septae within the sphenoid sinus vary greatly in size, shape, thickness, location, completeness, and relation to the sellar floor (Fig. 3.35). The cavities within the sinus are seldom symmetrical from side to side and are often subdivided by irregular minor septae. The septae are often located off the midline as they cross the floor of the sella. In our previous study a single major septum separated the sinus into two large cavities in only 68% of specimens, and even in these cases the septae were often located off the midline or were deflected to one side (36). The most common type of sphenoid sinus has multiple small cavities in the large paired sinuses. The smaller cavities are separated by septae oriented in all directions. Computerized tomograms of the sella provide the definition of the relationship of the septae to the floor of the sella needed for transsphenoidal surgery. Major septae may be found as far as 8 mm off the midline (36). The carotid artery frequently produces a ser-pigenous prominence into the sinus wall below the floor and along the anterior margin of the sella (Figs. 3.33 and 3.34) (7, 36). Usually, the
optic canals protrude into the superolateral portion of the sinus, and the second division of the trigeminal nerve protrudes into the inf erolateral part. A diverticulum of the sinus, called the op-ticocarotid recess, often projects laterally between the optic canal and the carotid prominence. Removing the mucosa and bone from the lateral wall of the sinus exposes the dura mater covering the medial surface of the cavernous sinus and optic canals. Opening this dura exposes the carotid arteries and optic and trigeminal nerves within the sinus. The sixth cranial nerve is located between the lateral side of the carotid artery and the medial side of the first trigeminal division. The second and third trigeminal divisions are seen in the lower margin of the opening through the lateral wall of the sphenoid sinus. In half of the cases the optic and trigeminal nerves and the carotid arteries have areas where bone 0.5 mm or less in thickness separates them from the mucosa of the sphenoid sinus, and in a few cases the bone separating these structures from the sinus is absent (8, 36). The absence of such bony protection within the walls of the sinus may explain some of the cases of cranial nerve deficits and carotid artery injury after transsphenoidal operations (18). The bone is often thinner over the carotid arteries than over the anterior margin of the pituitary gland. Diaphragma Sellae The diaphragma sellae forms the roof of the sella turcica. It covers the pituitary gland, except for a small central opening in its center, which transmits the pituitary stalk (Fig. 3.36). The diaphragma is more rectangular than circular, tends to be convex or concave rather than flat, and is thinner around the infundibulum and somewhat thicker at the periphery; the opening in its center is large when compared to the size of the pituitary stalk. It frequently is a thin, tenuous structure that would not be an adequate barrier for protecting the suprasellar structures during transsphenoidal operation. An outpouching of the arachnoid protrudes through the central opening in the diaphragma into the sella turcica in about half of the patients. This outpouching represents a potential source of postoperative cerebrospinal fluid leakage (18). Pituitary Gland When exposed from above by opening the diaphragma, the superior surface of the posterior lobe of the pituitary gland is lighter in color than the anterior lobe, and the anterior lobe wraps
Figure 3.33. Stepwise dissection of the lateral wall of the right half of a sellar type sphenoid sinus and adjacent structures (see text under "Incisural and Cister-nal Veins"). A. The sphenoid sinus and sellar area are divided in the midsagittal plane. The optic nerve (Optic N.) is seen proximal to the optic canal. The optico-carotid recess separates the carotid prominence and optic canal. The septum in the posterior part of the sinus is incomplete. B. The sinus mucosa and thin bone of the lateral sinus wall are removed to expose dura mater covering the carotid artery (Carotid A.), the second trigeminal division (V2) just distal to the trigeminal ganglion, and the optic nerve. C. The dura is opened to expose the carotid artery, the optic nerve in the optic canal, the second trigeminal division below the carotid artery, and the abducens nerve (VI) between the first trigeminal division (V1) and the carotid artery. D. Lateral view of the specimen showing area of the cavernous sinus. The oculomotor (HI) and trochlear (IV) nerves are seen above. The intra-cavernous portion of the carotid artery is seen medial to the trigeminal root (V) and the ophthalmic, maxillary, and mandibular divisions (V3) of the trigeminal nerve. The petrous portion of the carotid artery is seen in cross section below the trigeminal nerve. The opening into the sphenoid sinus is located between the first and second trigeminal divisions. E. Trigeminal nerve is reflected forward to expose the carotid artery, the trigeminal impression, the artery of the inferior cavernous sinus (Art. Inf. Cav. Sinus), and the abducens nerve, which splits into three bundles as it passes around the carotid artery. (From Fujii K, Chambers SM, Rhoton AL Jr: Neurovascular relationships of the sphenoid sinus: A microsurgical study. J Neurosurg 50:31-39, 1978.)
Figure 3.34. A. Anterior views of a sellar type sphenoid sinus. The anterior wall of the sella is removed to expose the pituitary gland. The specimen was split at the midline. The air cavity is wider below than above, as is typical in a well-pneumatized specimen. The optic canals are above. Carotid prominences are lateral to the sella. The trigeminal prominence is below the carotid prominence. B. Anterior views of a sellar type sphenoid sinus. The specimen is opened slightly to provide a better view of the lateral wall and the carotid and trigeminal prominences. C. Anterior view of a sellar type sphenoid sinus. Mucosa, dura, and bone of the lateral wall of the sinus are removed to expose parasphenoidal segments of the carotid artery. Sympathetic nerves (Symp. N.) are anterior to the left carotid artery. The orbital contents appear laterally. D. Anterior views of a sellar type sphenoid sinus. Specimen is spread slightly to show the abducens nerve (VI) and the ophthalmic (V1), maxillary (V2), and mandibular (V3) divisions of the trigeminal nerve (V). (From Fujii K, Chambers SM, Rhoton AL Jr: Neurovascular relationships of the sphenoid sinus: A microsurgical study. J Neurosurg 50:31-39, 1978.) around the lower part of the pituitary stalk to form the pars tuberalis (Fig. 3.37). The posterior lobe is more densely adherent to the sellar wall than the anterior lobe. The gland's width is equal to or greater than either its depth or its length in most patients. Its inferior surface usually conforms to the shape of the sellar floor, but its lateral and superior margins vary in shape because these walls are composed of soft tissue rather than bone. If there is a large opening in the diaphragma, the glands tends to be concave superiorly in the area around the stalk. The superior surface may become triangular as a result of being compressed laterally and posteriorly by the carotid arteries (Fig. 3.36). As the anterior lobe is separated from the posterior lobe, there is
a tendency for the pars tuberalis to be retained with the posterior lobe. Intermediate lobe cysts are frequently encountered during separation of the anterior and posterior lobes. Pituitary Gland and Carotid Artery The distance separating the medial margin of the carotid artery and the lateral surface of the pituitary gland usually varies from 1 to 3 mm; however, in some cases the artery will protrude through the medial wall of the cavernous sinus to indent the gland (Fig. 3.37) (3, 14, 36). Heavy arterial bleeding during transsphenoidal hy-pophysectomy has been reported to be caused by carotid artery injury, but may also be caused by a tear in an arterial branch of the carotid artery
Figure 3.35. Septa in the sphenoid sinus. The broken line on the central diagram shows the plane of the section of each specimen from which the drawings were taken, and the large arrow shows the direction of view of the specimens. The planum is above, the dorsum and clivus are below, and the sella is in an intermediate position on each diagram. The heavy dark lines on the drawings show the location of the septae in the sphenoid sinus. A wide variety of septae separate the sinus into cavities that vary in size and shape, seldom being symmetrical from side to side. (From Renn WH, Rhoton AL Jr: Microsurgical anatomy of the sellar region. J Neurosurg 43:288-298, 1975.)
(e.g., the inferior hypophyseal artery) or by avulsion of a small capsular branch from the carotid artery (18). If the carotid arteries indent the lateral surfaces of the gland, the gland does lose its rounded shape and conforms to the wall of the artery, often developing protrusions above or below the artery. Intrasellar tumors are subjected to the same forces, which prevent them from being spherical, and the increased pressure within the tumor increases the degree to which the tumor insinuates into surrounding crevices and tissue planes. Separation of these extensions from the main mass of gland or tumor may explain cases in which the tumor and elevated pituitary hormone levels recur although the hormone levels fell to zero immediately after adenoma removal. Intercavernous Venous Connections Venous sinuses may be found in the margins of the diaphragma and around the gland (36). The intercavernous connections (Fig. 3.38) within the sella are named on the basis of their relationship to the pituitary gland; the anterior
intercavernous sinuses pass anterior to the hypophysis, and the posterior intercavernous sinuses pass behind the gland. Actually, these intercavernous connections can occur at any site along the anterior, inferior, or posterior surface of the gland. The anterior sinus is usually larger than the posterior sinus, but either or both may be absent. If the anterior and posterior connections coexist, the whole structure constitutes the "circular sinus." Entering an anterior intercavernous connection that extends downward in front of the gland during transsphenoidal operation may produce brisk bleeding. However, this usually stops with temporary compression of the channel or with light monopolar diathermy, which serves to glue the walls of the channel together. A large intercavernous venous connection called the basilar sinus often passes posterior to the dorsum sellae and upper clivus. The basilar sinus connects the posterior aspect of both cavernous sinuses and is usually the largest and most constant intercavernous connection across the midline. The superior and inferior petrosal sinuses joint the basilar sinus. The abducens nerve often enters the posterior part of the cavernous sinus by passing through the basilar sinus. Cavernous Sinus The cavernous sinus surrounds the horizontal portion of the carotid artery and a segment of the abducens nerve (Figs. 3.33, 3.36, and 3.39). The oculomotor, trochlear, and ophthalamic divisions of the trigeminal nerve are found in the roof and lateral wall of the sinus (11, 14, 29, 36). The lateral wall of the cavernous sinus extends from the superior orbital fissure in front to the apex of the petrous portion of the temporal bone behind. The oculomotor nerve enters the roof of the sinus lateral to the dorsum sellae. The trochlear nerve enters the roof of the sinus posterolateral to the third nerve, and both nerves enter the dura mater immediately below and medial to the free edge of the tentorium. The ophthalamic division of the trigeminal nerve enters the low part of the lateral wall of the sinus and runs obliquely upward to exit through the superior orbital fissure. The abducens nerve enters the posterior wall of the sinus by passing through the dura lining the upper clivus and courses forward between the intracavernous carotid artery medially and the ophthalmic division laterally. It frequently splits into multiple rootlets in its course lateral to the carotid artery.
Figure 3.36. A. Superior view of the cavernous sinus, anterior and posterior lobe of the pituitary, intracavernous portion of the carotid artery, and meningo-hypophyseal trunk with its three branches: the inferior hypophyseal, tentorial, and dorsal meningeal arteries. The ophthalmic artery arises from the anterior surface of the carotid artery above the clinoid and enters the optic canal under the optic nerve. The dorsum was removed to expose the posterior lobe of the pituitary. The sixth cranial nerve (CN VI) receives a branch from the dorsal meningeal artery and courses laterally around the carotid artery. The dural ostium of the third cranial nerve (CN III) (left) is in the roof of the cavernous sinus, and the right third cranial nerve enters its dural ostium. The carotid artery is exposed in the medial part of the foramen lacerum (left). B. Superior view of the sellar region showing optic nerves, carotid arteries, and the third cranial nerve (CN Ш). Carotid arteries bulge into the pituitary fossa. C. Superior view of the sellar region and both cavernous sinuses. The carotid artery indents the lateral margin of the pituitary gland, and a tongue of pituitary extends over the top of the artery. The sixth cranial nerve (CN VI) passes through the duramater and around the lateral margin of the carotid artery. The third cranial nerve enters the roof of the cavernous sinus lateral to the dorsum sellae. The third through sixth cranial nerves are seen in the lateral wall of the cavernous sinus. D. Superior view of the sellar region. The optical chiasm is reflected forward. A congenitally absent diaphragma exposes the superior surface of the gland. The third cranial nerve is posterior to the carotid arteries. The right ophthalmic artery arises below the optic nerve. (A and С are from Harris FS, Rhoton AL Jr: Anatomy of the cavernous sinus: A microsurgical study. J Neurosurg 45:169-180, 1976. В and С are from Renn WH, Rhoton AL Jr: Microsurgical anatomy of the sellar region. J Neurosurg 43:288-298, 1975.) The branches of the intracavernous portion of the carotid artery are the meningohypophyseal trunk, the artery of the inferior cavernous sinus, and McConnell's capsular arteries (Figs. 3.32, 3.36, and 3.39). The ophthalmic artery may also take origin from the carotid artery within the sinus in few cases (36). The most proximal branch of the intracavernous carotid artery, the
meningohypophyseal trunk, usually arises below the level of the dorsum sellae near the apex of the curve between the petrous and intracavernous segments of the artery. The three branches of the meningohypophyseal artery are the tentorial artery (of Bernasconi-Cassinari), which courses toward the tentorium; the inferior hypophyseal artery, which courses medially to sup-
Figure 3.38. Six sagittal sections of the sellar region showing variations in the intercavernous venous connections within the dura. The variations shown include combinations of anterior, posterior, and inferior intercavernous connections and the frequent presence of a basilar sinus posterior to the dorsum. Either the anterior (lower center) or posterior (lower left) intercavernous connection or both (top center) may be absent. The anterior intercavernous sinus may extend along the whole anterior margin of the gland (lower left). The basilar sinus may be absent (lower right). (From Renn WH, Rhoton AL Jr: Microsurgical anatomy of the sellar region. J. Neurosurg 43:288-298, 1975.)
Figure 3.37. A. Pituitary gland: superolateral view. The posterior lobe is a lighter color and has a different consistency than the anterior lobe. The pars tuberalis partially encircles the stalk. The gland is concave around the stalk. B. Pituitary gland: inferior view. Note the relationship of the anterior and posterior lobes. C. Pituitary gland: anterior and posterior lobe separated. The pars tuberalis partially encircles the stalk. (From Rhoton AL Jr, Hardy DG, Chambers SM: Microsurgical anatomy and dissection of the sphenoid bone, cavernous sinus and sellar region. Surg Neurol 12:63-104, 1979.) ply the posterior part of the capsule of the pituitary gland; and the dorsal meningeal artery, which perforates the dura of the posterior wall of the sinus to supply the region of the clivus and the sixth nerve.
The artery of the inferior cavernous sinus originates from the lateral side of the horizontal segment of the carotid artery distal to the origin of the meningohypophyseal trunk (14). It passes above the abducens nerve and downward medially to the first trigeminal division to supply the dura of the lateral wall of the sinus. In a few cases it arises from the meningohypophyseal trunk. McConnell's capsular arteries, if present, arise from the medial side of the carotid artery and pass to the capsule of the gland, distal to the point of origin of the artery of the inferior cavernous sinus (14). Parkinson's Triangle and the Intracavernous Portion of the Carotid Artery Parkinson described a triangle within the lateral wall of the cavernous sinus through which the intracavernous portion of the carotid artery and its branches might be exposed for the surgical treatment of carotid-cavernous fistulas
(29). The boundaries of the triangle are the lower border of the fourth nerve superiorly, the upper margin of the fifth nerve inferiorly, and the slope along the dorsum sellae and clivus posteriorly (Fig. 3.39). Parkinson opened into the cavernous sinus by incising the dura within Parkinson's triangle starting approximately 6 mm below the dural entrance of the third nerve and extending anteriorly for 2 cm parallel to the slope of the third and fourth nerves and retractng the leaves of the dura. The intracavernous course of the extra-ocular nerves can be located through this opening. Retracting the ophthalmic division of the trigeminal nerve inferolaterally exposes the ab-ducens nerve between the trigeminal nerve and the carotid artery. We found in previous studies that it may be difficult to expose all branches of the intracavernous portion of the carotid artery through this opening (14). Discussion The surgical approaches to the third ventricle may require transcortical or transcallosal incisions; division or displacement of the fornix; opening of the lamina terminalis, septum pellu-cidum, and floor of the third ventricle; and dissection and separation of the tumor from the quadrigeminal plate, the optic nerves, chiasm, and tracts, the pituitary gland and its stalk, the cerebral peduncles, and all of the walls of the third ventricle. Injury of some structures exposed in approaching the third ventricle, such as the olfactory and oculomotor nerves, the optic pathways, and the quadrigeminal plate, causes deficits that are well defined and that correspond to the area injured. The sacrifice of other neural structures has produced variable results: in some cases there was no deficit, and in others the deficit was transient or permanent or resulted in the loss of life. Struc-
Figure 3.39. Superolateral view of the pituitary gland, right cavernous sinus, and intracavernous structures including the carotid artery and the third (CN III), fourth (CN IV), and sixth (CN VI) cranial nerves. A. The lateral dural wall of the cavernous sinus is removed. The tortuous carotid artery bulges superiorly, pushing the interclinoid ligament and cavernous sinus roof upward and indenting the lateral margin of the pituitary gland. The inferior hypophy-seal artery passes to the pituitary gland. The third and fourth cranial nerves enter the roof of the cavernous sinus by passing through the interclinoid ligament. The sixth cranial nerve on the left side enters the dura posterior to the dorsum and passes laterally around the left carotid artery. The sixth cranial nerve is above the superior margin of the trigeminal sensory (CN Vs) and motor (CN VM) roots, and only the first division (CN V1) is exposed distal to the ganglion. B. Further dural removal exposes the trigeminal root and its second (CN V2) and third (CN V3) divisions. The foramen lacerum is exposed lateral to the gasserian ganglion.
The trigeminal root is reflected laterally to show a second branch of the sixth cranial nerve as it passes lateral to the carotid artery. C. The trigeminal root is reflected forward, exposing the carotid artery in the foramen lacerum. A sympathetic nerve bundle is on the surface of the carotid artery in the foramen lacerum. Three rootlets of the sixth cranial nerve are seen passing around the carotid artery. The outline of the carotid artery is marked with a broken line in the areas where it is out of view in the petrous bone and in the cavernous sinus. (From Harris FS, Rhoton AL Jr: Anatomy of the cavernous sinus: A microsurgical study. J Neurosurg 45:169-180, 1976.)
tures sacrificed with variable results include the anterior and posterior parts of the corpus cal-losum and one or both of the fornices. Callosal incisions have resulted in disorders of the inter-hemispheric transfer of information, visuospatial transfer, the learning of bimanual motor tasks, and memory and also in deficits including alexia, apraxia, and astereognosis (20, 34, 46, 52, 56). Division of the fornix on both sides may cause a memory loss (16, 19, 49). The cerebral retraction needed for the anterior and posterior interhemispheric approaches and the cortical incisions for the transventricular surgical approaches have caused convulsions, hemiplegia, mutism, impairment of consciousness, and visual field loss (16, 23, 33). Manipulation of the walls of the third ventricle may cause the following disorders: hypothalamic dysfunction as manifested by disturbances of temperature control, respiration, consciousness, and hypophyseal secretion; visual loss due to damage of the optic chiasm and tracts; and memory loss due to injury to the columns of the fornix in the walls of the third ventricle (23, 24). Dissection in the area of the quadrigeminal plate may cause disorders of eye movement, edematous closure of the aqueduct of Sylvius, blindness from edema in the colliculi or geniculate bodies, and extraocular palsies due to edema of the nuclei of the nerves or the central pathways in the brain stem (6). Splitting the superior part of the cerebellar ver-mis may cause cerebellar edema and brain stem injury (27). Sacrifice of branches of the superficial and deep venous systems has produced inconstant deficits. Before sacrificing these veins, one should try to work around them, displace them out of the operative route, or place them under moderate or even severe stretch (accepting the fact that they may be torn) if it will yield some possibility of their being saved. Another option is to divide only a few of their small branches, which may allow the displacement of the main trunk out of the operative field. Dandy noted that not infrequently one internal cerebral vein had been sacrificed without effect, and on a few occasions both small veins and even the great vein of Galen had been ligated with recovery without any apparent disturbance of function (4-6). On the other hand, it seems that injury to this complicated venous network may cause diencephalic edema, mental symptoms, coma, hyperpyrexia, tachycardia, tachypnea, miosis, rigidity of limbs, and exaggeration of deep tendon reflexes (2, 17, 47, 48). Occlusion of the thalamostriate and other veins at the foramen of Monro may cause
drowsiness, hemiplegia, mutism, and hemor-rhagic infarction of the basal ganglia (16). Obliteration of veins coursing between the cerebrum and the superior sagittal sinus anterior or posterior to the rolandic vein or occlusion of the anterior part of the sagittal sinus, although usually not causing a deficit, may be accompanied by hemiplegia (46). Sacrificing the internal occipital vein or the bridging veins from the occipital pole to the superior sagittal or transverse sinuses may cause hemianopsia (47, 54). Cerebellar swelling after transection of the bridging vein between the cerebellum and tentorium has been reported (27). Numerous arteries including the internal carotid and basilar arteries and the circle of Willis and its branches are frequently exposed during the removal of tumors of the third ventricle, but only infrequently are any of these sacrificed. Symon reported that perforating branches of the posterior communicating artery might be sacrificed in the subtemporal approach to cranio-pharyngioma with a resulting infarct in the basal ganglia, but without obvious deficit (50). Infarction in the distribution of the thalamoperforating branches of the posterior cerebral artery may cause coma and death after the removal of a suprasellar tumor (55). The choroidal arteries are frequently exposed in the approaches through the roof of the third ventricle, but at that point they have given off most of their neural branches. Arterial lesions at the anterior part of the circle of Willis are more likely to result in disturbances in memory and personality, and those at the posterior part of the circle are more likely to result in disorders of the level of consciousness and are frequently combined with disorders of extraocular motion (32, 45, 49). Injuries to the superior cerebellar artery in approaches to the posterior part of the third ventricle may cause a cerebellar deficit. Lesions in the anterior incisural space and in the anterior part of the third ventricle may be approached by the unilateral subfrontal, bifron-tal, frontal-interhemispheric, frontotemporal, subtemporal, transsphenoidal, anterior trans-ventricular, and anterior transcallosal routes. Tumors located anterior to Liliequist's membrane between the optic chiasm and the dia-phragma sella are commonly operated on by the transsphenoidal or subfrontal approach. The transsphenoidal approach is preferred if the tumor extends upward out of an enlarged sella turcica and is located above a pneumatized sphenoid sinus. The subfrontal intracranial approach is reserved for those tumors in the chias-
matic cistern that are not accessible by the transsphenoidal route because they are located entirely above the diaphragma sellae, extend upward out of a normal or small sella, or are located above a nonpneumatized (conchal) type of sphenoid sinus. The subfrontal approach permits exposure of the tumor within the anterior incisural space by four routes: (a) the subchiasmatic approach between the optic nerves and below the optic chiasm; (b) the opticocarotid route directed between the optic nerve and carotid artery; (c) the lamina terminalis approach directed above the optic chiasm through a thined lamina terminalis; and (d) the transfrontal-transsphenoidal approach obtained by entering the sphenoid sinus and sella through the transfrontal craniotomy (43). The subchiasmatic approach is used if the subchiasmatic opening is enlarged by the tumor. The opticocarotid route is selected if parasellar extension of the tumor widens the space between the carotid artery and the optic nerve and the tumor cannot be reached by the subchiasmatic approach. The lamina terminalis approach is selected if the tumor has pushed the chiasm into a prefixed position and extends into the third ventricle to stretch the lamina terminalis so that the tumor is visible through it. The transfrontal-transsphenoidal approach is selected if the tumor grows upward out of the sella, the sphenoid sinus is pneumatized and the tumor does not stretch the lamina terminalis or widen the opticocarotid space, and a prefixed chiasm blocks the subchiasmatic exposure. A bifrontal craniotomy may be used if the tumor extends forward in both anterior cranial fossae and cannot be removed by a unilateral subfrontal exposure. The frontotemporal (pterional) approach is used for a tumor arising from the sphenoid ridge or anterior clinoid process, one arising above the diaphragma and extending along the sphenoid ridge or into the middle cranial fossa, or a lesion accessible through the spaces between the optic nerve and carotid artery or between the carotid artery and the oculomotor nerve. The frontotemporal approach may be combined with a temporal lobectomy to expose lesions extending into the middle incisural space. The anterior transcallosal approach is suitable for lesions located in the anterosuperior part of the third ventricle or extending out of the superior part of the third ventricle into one or both lateral ventricles near the foramen of Monro. The transcallosal approach is easier to perform than the transventricular approach if the ventricles are of a normal size or are minimally enlarged. The anterior transventricular approach is suita-
ble for approaching tumors in the anterosuperior part of the third ventricle, especially if the tumor has a major extension into the anterior part of the lateral ventricle on the side of the approach. It is more difficult to expose the anterior part of the lateral ventricle on the side opposite the craniotomy through the transventricular than through the transcallosal approach. The trans-ventricular approach enters the lateral ventricle in a favorable location for exposure of the superior half of the anterior part of the third ventricle by opening the choroidal fissure or for biopsy or removal of the tumor through an enlarged foramen of Monro, and the transcallosal approach enters the lateral ventricle in a satisfactory position for exposing the superior part of the third ventricle by separating the halves of the body of the fornix in the midline. The ventricular veins provide valuable landmarks in directing one to the foramen of Monro and the choroidal fissure during operations directed through the lateral ventricles. This is especially true if hydrocephalus, a common result of third ventricular tumors, is present because the borders between the neural structures in the ventricular walls become less distinct as the ventricles dilate. The thalamostriate vein is helpful in delimiting the junction of the caudate nucleus and the thalamus because it usually courses along the sulcus separating these structures. The fact that the ventricular veins converge on the choroidal fissure assists in identifying this fissure, which is situated on the periphery of the thalamus. Opening through the fissure in the body of the ventricle will expose the velum inter-positum and the roof of the third ventricle, opening through it in the atrium will expose the quadrigeminal cistern and the pineal region, and opening through it in the temporal horn will expose the crural and ambient cisterns. In the transcortical operative approach through the middle frontal gyrus or the transcallosal approach through the anterior part of the corpus callosum, the veins in the frontal horn are seen to drain posteriorly toward the foramen of Monro because the choroidal fissure does not extend into this area (26). The anterior caudate, anterior septal, superior choroidal, and thalamostriate veins usually join the internal cerebral veins at or near the foramen of Monro. However, these veins may pass through the choroidal fissure behind the foramen of Monro to enter the velum interpositum and course adjacent to the internal cerebral vein for a considerable distance before joining the internal cerebral vein. The junction of the thalamostriate vein with the in-
ternal cerebral vein as seen on the lateral angio-gram usually forms an acute angle at the posterior margin of the foramen of Monro; however, the thalamostriate vein may pass through the choroidal fissure and join the internal cerebral vein posterior to the foramen of Monro. The internal cerebral vein is not seen upon opening into the frontal horn because it courses in the roof of the third ventricle below the body of the fornix. The anterior part of the internal cerebral vein can only be exposed by opening through or displacing the structures forming the roof of the third ventricle. One method of increasing the exposure of the roof of the third ventricle is to section a column of the fornix anterosuperior to the foramen on one side, but this will permit the exposure of no more than a short anterior segment of the internal cerebral vein. To prevent the complications associated with sectioning the fornix, Hirsch et al. sectioned the thalamostriate vein at the posterior margin of the foramen of Monro, rather than damaging the fornix to enlarge the opening in the roof of the third ventricle (16). They stressed that interruption of this vein was harmless; however, some of their patients developed drowsiness, hemiplegia, and mutism, and occlusion of the veins at the foramen of Monro has caused hemorrhagic infarction of the basal ganglia. Other routes to the anterior part of the third ventricle are by the interforniceal approach, in which the body of the fornix is split in the midline and the tela choroidea below the fornix is opened to expose the internal cerebral veins, or the sub-choroidal approach, in which the choroidal fissure is opened between the fornix and thalamus, thus allowing the fornix to be pushed to the opposite side to expose the structures in the roof of the third ventricle (1, 53). The subchoroidal and interforniceal approaches have the advantage of giving access to the central portion of the third ventricle by displacing, rather than dividing, the fibers in the fornix. They also provide a more favorable angle for visualization of the third ventricle than the transforaminal approach. These interforniceal and subchoroidal approaches are uniquely suited to lesions in the superior half of the third ventricle below the roof and behind the foramen of Monro. In both approaches, the tela choroidea below the fornix must be opened to expose the internal cerebral vein and the third ventricle. In the subchoroidal approach, the right half of the fornix and the choroid plexus are displaced to the left side and the third ventricle is entered through the resulting longitudinal opening. A wider opening of the
third ventricle may be provided by displacing the thalamostriate vein and extending the opening into the foramen of Monro. However, occlusion of this vein at the foramen of Monro has caused hemorrhagic infarction of the basal ganglia (16). The maximal limits of the opening are determined by the limits of displacement of these vessels and their branches. The close relationship of the genu of the internal capsule to the foramen of Monro should be kept in mind when operating in this area. The genu of the internal capsule nearly touches the lateral wall of the ventricle in the area lateral to the foramen of Monro near the anterior pole of the thalamus. Approaches to the middle incisural space include the posterior frontotemporal, subtemporal, temporaltransventricular, and lateral suboccip-ital routes. The subtemporal approach with elevation of the temporal lobe is commonly used to expose lesions in the cisterns around the inci-sura. Hemorrhage, venous infarction, and edema after retraction of the temporal lobe during this approach are minimized by placing the lower margin of the craniotomy and the dural exposure at the cranial base so as to reduce the need for retraction and by avoiding occlusion of the bridging veins, especially the vein of Labbe. The ten-torium is frequently divided to increase the exposure or to decompress the brain stem when mass lesions are impacted in the incisura. Resection of part of the parahippocampal gyrus facilitates exposure of the upper part of the middle incisural space. A transventricular approach using a cortical incision in the nondominant inferior temporal gyrus may be used if the lesion involves the temporal horn, choroidal fissure, hippocampal formation, or upper part of the middle incisural space (15). A cortical incision in the occipitotemporal gyrus on the inferior surface of the temporal lobe has been used to minimize visual and speech deficits in exposing the temporal horn of the dominant hemisphere (15). After entering the temporal horn, the surgeon opens the choroidal fissure to expose the middle incisural space. The trochlear nerve is the cranial nerve most frequently injured in the middle incisural space. It can be injured when the free edge is divided and is so thin and friable that it may rupture from gentle retraction on the leaves formed by dividing the tentorium. Lesions in the posterior part of the third ventricle and the posterior incisural space may be approached from above the tentorium along the medial surface of the occipital lobe using an occipital transtentorial approach, through the posterior part of the lateral ventricle using a poste-
rior transventricular approach, and through the corpus callosum using a posterior interhemi-spheric transcallosal approach, or from below the tentorium through the supracerebellar space using an infratentorial supracerebellar approach. The infratentorial supracerebellar and occipital transtentorial approaches, which are most commonly selected for pineal tumors, may be combined with incision of the tentorium approximately 1 cm lateral to the straight sinus. The infratentorial supracerebellar approach is preferred for most lesions in the pineal region because the deep venous system that caps the dorsal aspect of pineal tumors does not obstruct access to the tumor. The approach is best suited to tumors in the midline that grow into the lower half of the posterior incisural space, displacing the quadrigeminal plate and the anterior lobe of the cerebellum. The occipital transtentorial approach is preferred for lesions centered at or above the tentorial edge if there is not a major extension to the opposite side or into the posterior fossa, and for those lesions located above the vein of Galen. The posterior transcallosal approach, in which the splenium is divided, would be used only if the lesion seems to arise in the splenium above the vein of Galen and extends into the posterior incisural space (6). The posterior transventricular approach provides adequate exposure of the atrium and posterior portion of the body of the lateral ventricle and would be the preferred approach to a tumor involving the posterior incisural space if the tumor extends into the pulvinar or involves the atrium or the glomus of the choroid plexus. The preferable approach to the ventricle is through the superior parietal lobule, although Van Wagenen (51) approached the pineal region through the atrium of the lateral ventricle using a cortical incision in the superior temporal gyrus.
6. 7. 8. 9. 10.
1 1
12. 13. 14. 15.
16.
17. 18. 19.
References 1 . Apuzzo ML, Chikovani OK, Gott PS, Teng EL, Zee C, Giannotta SL, Weiss MH: Transcallosal, interfornicial approaches for lesions affecting the third ventricle: Surgical considerations and conse quences. Neurosurgery 10:547-554, 1982. 2. Bailey P: Peculiarities of the intracranial venous system and their clinical significance. Arch Neurol Psychiatry 32:1105, 1934. 3. Bergland RM, Ray BS, Torack RM: Anatomical variations in the pituitary gland and adjacent structures in 225 human autopsy cases. J Neurosurg 28:93-99, 1968. 4. Dandy WE: An operation for the removal of pineal tumors. Surg Gynecol Obstet 33:113-119, 1921. 5. Dandy WE: Benign Tumors in the Third Ventri-
20. 21. 22. 23. 24.
25. 26. 27.
cle of the Brain: Diagnosis and Treatment. Springfield, IL, Charles С Thomas, 1933. Dandy WE: Operative experience in cases of pin eal tumor. Arch Surg 33:19-46, 1936. Fujii K, Chambers SM, Rhoton AL Jr: Neurovascular relationships of the sphenoid sinus: A microsurgical study. J Neurosurg 50:31-39, 1978. Fujii K, Lenkey C, Rhoton AL Jr: Microsurgical anatomy of the choroidal arteries: Lateral and third ventricles. J Neurosurg 52:165-188, 1980. Gibo H, Carver CC, Rhoton AL Jr, et al: Microsurgical anatomy of the middle cerebral artery. J Neurosurg 54:151-169, 1981. Gibo H, Lenkey C, Rhoton AL Jr: Microsurgical anatomy of the supraclinoid portion of the inter nal carotid artery. J Neurosurg 55:560-574, 1981. . Gudmundsson K, Rhoton AL Jr, Rushton JG: De tailed anatomy of the intracranial portion of the trigeminal nerve. J Neurosurg 35:592-600, 1971. Hardy J: Transsphenoidal hypophysectomy. J Neurosurg 34:582-594, 1971. Hardy DG, Peace DA, Rhoton AL Jr: Microsurgical anatomy of the superior cerebellar artery. Neuro surgery 6:10-28, 1980. Harris FS, Rhoton AL Jr: Anatomy of the cavern ous sinus: A microsurgical study. J Neurosurg 45:169-180, 1976. Heros RC: Arteriovenous malformations of the medial temporal lobe: Surgical approach and neuroradiological characterization. J Neurosurg 56:44-52, 1982. Hirsch JF, Zouaoui A, Renier D, Pierre-Kahn A: A new surgical approach to the third ventricle with interruption of the striothalamic vein. Acta Neurochir(Wien) 47:135-147, 1979. Kunicki, A: Operative experiences in eight cases of pineal tumor. J Neurosurg 17:815-823, 1960. Laws ER, Kern EB: Complications of transsphenoidal surgery. Clin Neurosurg 23:401-416, 1976. Little JR, MacCarty CS: Colloid cysts of the third ventricle. J Neurosurg 40:230-235, 1974. Long DM, Chou SN: Transcallosal removal of craniopharyngiomas within the third ventricle. J Neurosurg 39:563-567, 1973. Matsushima T, Rhoton AL Jr, de Oliveria E, et al: Microsurgical anatomy of the veins of the poste rior fossa. J Neurosurg 59:63-105, 1983. Matsushima T, Rhoton AL Jr, Lenkey C: Micro surgery of the fourth ventricle: Part 1. Neurosur gery 11:631-667, 1982. McKissock W. The surgical treatment of colloid cyst of the third ventricle. Brain 74:1-19, 1951. Olivecrona H: Colloidal cysts of the third ventricle, in Olivecrona H, Tonnix W (eds): Handbuch der Neurochirurgie. Berin, Springer-Verlag, 1967, vol 4, pp 101-109. Ono M, Ono M, Rhoton AL Jr, Barry M: Microsurgical anatomy of the region of the tentorial incisura. J Neurosurg 60:365-399, 1984. Ono M, Rhoton AL Jr, Peace DA, Rodriguez R: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 15:621-657, 1984. Page LK: The infratentorial-supracerebellar ex posure of tumors in the pineal area. Neurosurgery
1:36-40, 1977. 28. Pait TG, Zeal A, Harris FS, Paullus WS, Rhoton AL Jr: Microsurgical anatomy and dissection of the temporal bone. Surg Neural 8:363-39-1, 1977. 29. Parkinson D: Transcavernous repair of carotidcavernous fistula: A case report. J Neurosurg 26:420-424, 1967. 30. Perlmutter D, Rhoton AL Jr: Microsurgical anat omy of the anterior cerebral-anterior communi cating-recurrent artery complex. J Neurosurg 45:259-271, 1976. 31. Perlmutter D, Rhoton AL Jr: Microsurgical anat omy of the distal anterior cerebral artery. J Neu rosurg 49:204-228, 1978. 32. Plum F, Posner JB: The Diagnosis of Stupor and Coma. Philadelphia, FA Davis, 1972, pp 64-118. 33. Poppen JL, Reyes V, Horrax G: Colloid cysts of the third ventricle. J Neurosurg 10:242-263, 1953. 34. Preilowski BFB: Possible contribution of the an terior forebrain commissures to bilateral motor coordination. Neuropsychologia 10:267-277, 1972. 35. Rand RW: Transfrontal transsphenoidal craniotomy in pituitary and related tumors using microneurosurgical techniques. In Rand RW (ed): Microneurosurgery. Mosby, St Louis, 1969, pp 74-86. 36. Renn WH, Rhoton AL Jr: Microsurgical anatomy of the sellar region. J Neurosurg 43:288-298, 1975. 37. Rhoton AL Jr: Anatomy of saccular aneurysms. Surg Neurol 14:59-66, 1980. 38. Rhoton AL Jr, Fujii K, Fradd B: Microsurgical anatomy of the anterior choroidal artery. Surg Neurol 12:171-187, 1979. 39. Rhoton AL Jr, Fujii K, Saeki N, et al: Microsurgical anatomy of aneurysms: Part 1. In Hopkins LN, Long DM (eds): Clinical Management oflntracranial Aneurysms. New York, Raven Press, 1982, pp 201-243. 40. Rhoton AL Jr, Harris FS, Renn WH: Microsurgical anatomy of the sellar region and cavernous sinus. Clin Neurosurg 24:54-85, 1977. 41. Rhoton AL Jr, Maniscalco J: Microsurgery of the sellar region. Neuroophthalmology 9:106-127, 1977. 42. Rhoton AL Jr, Hardy DG, Chambers SM: Microsurgical anatomy and dissection of the sphenoid
49. Sweet WH, Talland GA, Ervin FR: Loss of recent bone, cavernous sinus and sellar region. Surg Neurol 12:63-104, 1979. 43. Rhoton AL Jr, Yamamoto I, Peace DA: Microsur gery of the third ventricle: Part 2. Operative ap proaches. Neurosurgery 8:357-373, 1981. 44. Rosner SS, Rhoton AL Jr, Ono M, Barry M: Microsurgical anatomy of the anterior perforating ar teries. J Neurosurg 61:468-485, 1984. 45. Saeki N, Rhoton AL Jr: Microsurgical anatomy of the upper basilar artery and the posterior circle of Willis. J Neurosurg 46:563-577, 1977. 46. Shucart WA, Stein BM: Transcallosal approach to the anterior ventricular system. Neurosurgery 3:339-343, 1978. 47. Stern WE, Batzdorf U, Rich JR: Challenges of surgical excision of tumors in the pineal region. Bull Los Angeles Neurol Soc 36:105-118,1971. 48. Suzuki J, Iwabuchi T: Surgical removal of pineal tumors (pinealomas and teratomas): Experience in a series of 19 cases. J Neurosurg 23:565-571, 1965. memory following section of fornix. Trans Am Neurol Assoc 84:76-82, 1959. 50. Symon L: The intracranial approach to tumours in the area of the sella turcica, in Symon L (ed): Operative Surgery: Neurosurgery. London, Butterworth, 1979, pp 181-186. 51. Van Wagenen WP: A surgical approach for the removal of certain pineal tumors. Surg Gynecol Obstet 53:216-220, 1931. 52. Van Wagenen WP, Herren RY: Surgical division of commissural pathways in the corpus callosum. Arch Neurol Psychiatry 44:740-759, 1940. 53. Viale GL, Turtas S: The subchoroid approach to the third ventricle. Surg Neurol 14:71-76, 1980. 54. Yamamoto I, Kageyama N: Microsurgical anatomy of the pineal region. J Neurosurg 53:205-221, 1980. 55. Yamamoto I, Rhoton AL Jr, Peace DA: Microsur gery of the third ventricle: Part 1. Microsurgical anatomy. Neurosurgery 8:334-356, 1981. 56. Zaidel D, Sperry RW: Memory impairment after commissurotomy in man. Brain 97:263-272, 1974. 57. Zeal AA, Rhoton AL Jr: Microsurgical anatomy of the posterior cerebral artery. J Neurosurg 48:534-551, 1978.
4 Surgery of the Third Ventricle: Regional Embryology IvanS.Ciric,M.D.
The success of a neurosurgical procedure depends among other factors on the operative approach chosen and the surgical technique used. The development of modern neurosurgical techniques (including microsurgery, laser vaporization techniques, and more recently, stereotacticendoscopic procedures) has made easier and safer the complete removal of tumors in sites previously deemed inaccessible or associated with a high incidence of morbidity and mortality. As for the surgical approaches chosen, they have been modified over the years in accordance with the neurosurgeon's ever increasing knowledge of anatomy and physiology of the area under consideration. Understanding of the embryological planes of development should also enter into consideration in choosing the surgical approach to an intracranial tumor, especially in the region of the third ventricle.
cerebral vesicle (32). Posterior to Rathke's pouch and sac, the cerebral vesicle evaginates to form the pituitary stalk and posterior lobe. Rathke's sac eventually gives rise to the anterior pituitary, which remains separated from the surrounding cerebral vesicle, the stalk, and the posterior lobe by a layer of mesoderm that eventually differentiates into the structures of the diaphragma sellae and the pituitary gland capsule (9). It is not surprising, therefore, that a variety of lesions of different embryological layers can occur in and around the third ventricle. The peculiarities of the embryological development of various lesions will reflect in how they relate not only to the
Third Ventricle and the Pituitary Region The third ventricle develops from the dience-phalic vesicle at the rostral end of the neural tube during the third week of gestation. Soon thereafter, the third ventricle becomes surrounded by the rapidly developing cerebral vesicles except along its roof, which begins to invag-inate, trapping the surrounding mesoderm (Fig. 4.1). Pia mater that eventually invests the floor of the third ventricle forms from the layer of mesoderm adjacent to the ventral surface of the diencephalic vesicle (3). Rathke's pouch and sac form from an invagination of the distal end of the stomodeum into the overlying mesoderm and Figure 4.1. Development of the third ventricle (see text).
third ventricle, but also to the arachnoid membrane. A text on surgery of lesions in and around the third ventricle should include a discussion of the regional embryology of both, lesions abutting against the outer boundaries of the third ventricle as well as those occupying its interior. This chapter includes discussion of the developmental anatomy of pituitary adenomas, craniopharyn-giomas, and neuroepithelial cysts of the third ventricle and of pineal region tumors. In addition, in each instance an attempt is made to explain the neuroanatomy of the lesion with the developmental anatomy of the area under consideration and to correlate these with the choice of surgical approach to be used. Pituitary Adenomas Pituitary adenomas develop as intraparenchy-mal lesions of the anterior lobe (Fig. 4.2) (7). Initially, they consist of a microscopic conglomerate of tumor cells arranged in sheets of uniform, polygonal cells with indistinct cell boundaries. These microscopic adenomas can be found in approximately 20% of routine, postmortem examinations of the pituitary (13). Immunoper-oxidase staining will show that these microad-enomas frequently consist of a variety of hyper-secreting cells with one cell type predominant (27). The initial impetus for the development of a microadenoma is not known. It is conceivable that prolonged use of estrogen-containing contraceptives will stimulate prolactin secretion in a previously hormonally inactive microadenoma (37). Approximately 10% of functioning prolac-tin-secreting microadenomas will show substantive growth over a period of 5 years (45). As they
Figure 4.2. Intraparenchymal stage of pituitary tumor development.
grow, microadenomas compress the surrounding normal anterior pituitary tissue into a relatively tough pseudocapsular layer that blends imperceptibly with the periphery of the usually softer and, in the central part, more necrotic, and at times hemorrhagic adenoma (7). To assure a complete removal of a microadenoma, one should also resect this pseudocapsule along a cleavage plane between it and the surrounding normal, yellow-orange anterior pituitary tissue. A surgical cure is therefore possible in this first, intra-parenchymal stage of pituitary (micro)adenoma development. Depending on the direction of its growth, a pituitary microadenoma will eventually reach the pituitary gland capsule along its superior, lateral, or inferior surface (Fig. 4.3). The pituitary gland capsule will become distended and fragmented. Thus, it is common to find, after opening the dura mater, a pituitary macroadenoma still enveloped by an attenuated and partially fragmented pituitary gland capsule. It is only after this thin layer is incised that a spontaneous extrusion of tumor tissue occurs. During this second or sellar stage of their development, pituitary adenomas can still be removed completely and selectively along the pseudocapsule cleavage plane between the tumor and the surrounding, normal anterior pituitary. In about 10% of cases, the tumor will be locally invasive into the surrounding dural structures (11). As pituitary adenomas continue to grow, they do so in the direction of least resistance, namely, toward the diaphragma sellae, which is covered by the arachnoid membrane. As they extend
Figure 4.3. Sellar stage of pituitary tumor development.
through the diaphragma sellae, pituitary adenomas displace, rather than penetrate, the arachnoid membrane in a superior direction (11). Thus, they preserve their original extraarach-noid location in relationship to all of the important intraarachnoid neurovascular structures. Indeed, even pituitary adenomas that invaginate the floor of the third ventricle and reach as high as the foramen of Monro are in fact separated from these structures by a layer of arachnoid membrane (Fig. 4.4). It is primarily for this reason that an extraarachnoid, transsphenoidal approach to pituitary adenomas is preferable in the majority of patients (Fig. 4.5) (20). A transcranial approach, on the other hand, requires transgression of the subarachnoid space with all of its important neurovascular structures (Fig. 4.5). When the adenoma has reached the suprasellar space it is classified as a pituitary macroade-noma with suprasellar extension. Utilizing the transsphenoidal approach, a gross total removal of a macroadenoma, as evident from the absence of any demonstrable tumor tissue on the postoperative computerized tomographic (CT) scan, can be achieved in about 40% of cases. On the other hand, a complete removal of a macroadenoma, as evident from the endocrine cures achieved, will be possible in only onequarter of such cases (11). A gross total removal of a macroadenoma will depend on a number of factors such as the direction of its growth, the height of the suprasellar extension, the width of the diaphragma sellae, the adenoma consistency, and the presence or absence of any dural invasion. Utilizing the transsphenoidal approach, a mac-
Figure 4.5. The transsphenoidal approach (large arrow) to pituitary tumors is confined to the extraarach-noidal space. The transcranial approach (small arrows) requires transgression of a double layer of arachnoid and of the intervening subarachnoid space. roadenoma with a straight suprasellar extension of less than 2 cm is still amenable to a gross total tumor removal. In contrast, an anterior, posterior, or lateral tumor growth and pituitary adenomas with suprasellar extension of more than 2 cm can rarely be removed in a gross total fashion. A constricted diaphragma sellae usually constitutes a contraindication for a transsphenoidal approach. Soft, pulpy, and especially hem-orrhagic and cystic lesions are easier to remove in a gross total fashion as compared to hard, septate lesions. Finally, dural invasion precludes a gross total tumor removal. Craniopharyngiomas During the development of Rathke's sac and its subsequent anterior-superior rotation, columnar epithelium from the hypophyseal duct will be brought into contact with the infundibular area (Fig. 4.6) (43). When this occurs before the development of the pia mater from the mesoderm that separates Rathke's sac from the diencephalic vesicle, these cell nests will be embedded into the neuroepithelium of the diencephalic floor. As the pia mater develops about the fifth week of gestation, these cell nests will be incorporated into the subpial space (Fig. 4.7) (10) where they may undergo squamous metaplasia (17). Intraventricular Craniopharyngiomas develop from an endophytic growth of the squamous epithelium directed toward the third ventricular cavity (10). Proliferation of the squamous
Figure 4.4. Suprasellar stage of pituitary tumor development. Note displacement rather than penetration of the arachnoid membrane by the tumor.
tuitary, they will be excluded from the subpial space by the formation of the pia mater (Fig. 4.9). Depending on their initial position in relationship to the stalk, these extrapial cell nests may give rise to either a completely intraarachnoid (Fig. 4.10) or a partially intra- and partially ex-traarachnoid craniopharyngioma (10). When the lesion is completely extrapial, a gross total sur-
Figure 4.6. Anterior-superior rotation of the anterior pituitary with implanted epithelial cell nests from the hypophyseal duct.
Figure 4.8. Partial intraparenchymal, subpial craniopharyngioma.
Figure 4.7. Epithelial cell nests from a hypophyseal duct implanted into the neuroepithelium and covered by the developing pia mater. epithelium, however, usually proceeds in an ex-ophytic fashion toward the subarachnoid space. In that case, the lesion will be partially intrapar-enchymal, subpial, and adherent to the neural structures of the hypothalamic floor (Fig. 4.8) (10, 18). The partially intraparenchymal craniopharyngiomas cannot be removed completely. These patients have a normal sella turcica, con-traindicating the transsphenoidal approach. When the cell nests from the hypophyseal duct remain adherent to Rathke's sac or are suspended in the mesenchyma of the infundibular area, separated from the developing anterior pi-
Figure 4.9. Epithelial cell nests from the hypophyseal duct suspended in the mesenchyma surrounding the neuroepithelium.
Figure 4.10. Extrapial, subarachnoid craniopharyngioma. gical removal should be possible utilizing micro-surgical and laser vaporization techniques. In patients with extrapial intraarachnoid cranio-pharyngiomas, the sella turcica is also normal, precluding the transsphenoidal approach. A partially intraarachnoid and partially extraarachnoid craniopharyngioma, a rather frequent occurrence, is usually associated with a large sella, making a transsphenoidal approach possible. Because of their partial intraarachnoid location, a removal of these lesions will inevitably result in a breach of the arachnoid membrane and a per-ioperative cerebrospinal fluid fistula requiring repair. When the hypophyseal duct cell nests remain attached to the upper surface of the developing anterior pituitary or are brought into contact with the lowermost part of the stalk, a craniopharyngioma arising in this location will grow within the sella and thus behave as a pituitary adenoma in that it will continue to grow outside the arachnoid membrane (Fig. 4.11) (10). These tumors are eminently resectable utilizing the transsphenoidal approach (21). Because of their high cholesterol content, intrasellar craniophar-yngiomas are easily distinguishable from pituitary adenomas on a magnetic resonance imaging (MRI) study. It is obvious, therefore, that craniopharyn-
Figure 4.11. Extraarachnoidal craniopharyngioma. giomas have a different relationship to the arachnoid membrane from pituitary adenomas, which remain outside the arachnoid membrane regardless of their size. Craniopharyngiomas can be completely within the arachnoid membrane, partially intra- and partially extraarachnoid, or completely extraarachnoid. Neuroepithelial (Colloid) Cyst of the Third Ventricle By the fourth to fifth week of gestation, the rostral end of the neural tube has enlarged into the diencephalic and the two hemispheric vesicles (3). The initial slitlike communication between the diencephalic and the cerebral vesicles is the early choroidal fissure. It enlarges anteriorly to form the foramen of Monro. While the lateral diencephalic wall and the opposing walls of the hemispheric vesicles undergo rapid proliferation to form the thalamus and hypothalamus, the basal ganglia and the lobes of the cerebral hemispheres, the roof of the diencephalon, and the immediately adjacent dorsal wall of the hemispheric vesicles fail to proliferate. Instead, they undergo a U-shaped infolding resulting in the duplication of the overlying mesoderm, which thus gives rise to the velum interpositum. Further infolding of the fibrovascular stroma of the
early velum interpositum along the medial wall of the lateral ventricles gives rise to the choroid plexus (Fig. 4.1) (12). A similar process of invag-ination of the fibrovascular stroma from the velum interpositum into the diencephalic vesicle leads to the formation of the choroid plexus in the roof of the third ventricle. The infolded fibrovascular stroma of the choroid plexus remains covered, however, by a thin layer of neuroepi-thelium that merges imperceptibly with the surrounding ependyma at the base of the choroid plexus. Thus, the mesenchymal stroma of the choroid plexus is in a strict anatomical sense extraventricular. The corpus callosum develops as a connection between the two cerebral vesicles. The fornix develops as a bundle of fibers between the temporal lobe and the hypothala-mus. The segment of the medial wall of the cerebral vesicles between the corpus callosum and the fornix fails to proliferate and persists as the septum pellucidum (12). Neuroepithelial (colloid) cysts develop when the neuroepithelium of the diencephalic roof begins to invaginate into the third ventricle with a subsequent partial sequestration of a cystlike structure. In the majority of cases, the cyst remains firmly attached to the choroid plexus (Fig. 4.12). According to this concept of their origin, the so-called colloid cysts of the third ventricle are best described as neuroepithelial cysts (4, 12, 38). The surgical approaches to lesions in the anterior and middle sections of the third ventricle should follow the planes of embryological development of the diencephalic area (5, 33). This is best executed either along the midline through
Figure 4.12. Development of neuroepithelial (colloid) cyst.
the corpus callosum (14, 29, 39) or from a lateral direction through the lateral ventricle (28). The third ventricle can then be reached through a distended foramen of Monro (14, 28); through the septum pellucidum and the velum interpositum, reaching into the third ventricle between the two internal cerebral veins (2, 8, 31); or through the choroid fissure itself (44). Although sectioning of the fornix anterior to the foramen of Monro, division of the thalamus striate vein (23), and removal of the anterior tubercle of the thalamus have been advocated to enlarge the foramen of Monro, these maneuvers are usually not necessary and should probably be avoided altogether.
Pineal Region Tumors Pineal gland contains a variety of cells of different origin. The predominant cell type is the pineocytoma, a modified neuroepithelial cell that relates closely to interlobular vessels and contains melanin (22). Other component cell types include glial cells, rarely true ganglion cells, neu-roblast-like cells, fibroblasts, lymphocytes, and related cells, connective tissue, and vessels (22). The development of the pineal gland continues through the first decade of life when connective tissue septi containing blood vessels penetrate the parenchyma of the pineal gland and separate it into anastomosing cords and, indeed, into the first half of the second decade, when laminated calcified bodies, so-called calcospherites, develop in the pineal in their adult form. The classification of pineal neoplasms has undergone an evolutionary change since del Rio-Hortega described in 1932 two main cell types in the human pineal gland (16). According to Russell and Rubinstein (35), pineal region tumors can be classified into four main categories. The most common type are tumors of germ cell origin. The germ cells derive from the yolk sac endoderm and can migrate through the embryo. Thus, they can give rise to tumors in diverse anatomical locations, although usually in such midline structures as the brain, mediastinum, retroperi-toneum, and gonads. The most common germ cell tumor is germinoma (dysgerminoma, atypical teratoma). Less common are teratomas, embryonal carcinomas, endodermal sinus tumors, and choriocarcinomas (listed in order of increasing malignancy). Tumors of the pineal parenchymal cells and pineocytomas constitute approximately 20 to 25% (15, 30). Pineal cells have the capability to differentiate into tumors with divergent histolog-ical makeups. Thus, these tumors may assume
the appearance of ganglionic tumors, astrocyto-mas, or retinoblastomas (34). Indeed, all three components have been found in human pineal neoplasms (40). This is in keeping with the phy-logenetic relationship of the pineal cell with both the neurons and the photoreceptor cells of fish and amphibians (22). Glial tumors and tumors of other cell types, especially meningiomas and benign cysts, are of particular interest to neurosur-geons and should always be included in a differential diagnosis. From this brief embryological discussion, it is obvious that surgical approaches to pineal tumors depend not only on their location, but also on their histological type. A detailed discussion of the various surgical approaches to the pineal region is beyond the scope of this chapter. Although some controversy exists as to whether "pineal tumors" should be operated upon or only radiated (1, 25), it seems more appropriate to consider this issue in light of the histological nature of the lesion. Thus, a correct pretreatment diagnosis arrived at by CT scanning (meningiomas), MRI (cholesterol, fat-containing lesions such as dermoid tumors, teratomas), or stereo-tactic biopsy (2) (germinomas, gliomas) is a prerequisite for choosing the optimal treatment. Although hormonal markers may give a hint as to the histological nature of the lesion, they still fall short of providing an accurate histological diagnosis. In conclusion, it could probably be said that treatment of pineal region tumors should be tailored not only to the location of the lesion and its relationship to the surrounding neurovascular structures, but also to the presumed histological nature of the lesion. Thus, it seems reasonable to suggest that such benign lesions as some teratomas, dermoid cysts, meningiomas, and cystic formations can be cured with surgical excision (41). In contrast, malignant gliomas are best treated with stereotactic biopsy and radiation therapy. Whether tumors of germ cell origin should be treated with radiation therapy and shunting (19, 24, 25, 42), with adjuvant chemotherapy (25) in some instances (germinomas, choriocarcinomas), or with surgical excision (6, 26, 36) followed by radiation therapy and chemotherapy has not been settled. The trend has been toward a more aggressive approach that favors partial resection followed by radiation therapy (25). References 1. Abay EO III, Laws ER Jr, Grado GL, et al: Pineal tumors in children and adolescents: Treatment by
2.
3. 4. 5. 6. 7.
8. 9. 10. 11
12. 13. 14. 15. 16.
17. 18. 19. 20. 21. 22.
23.
CSF shunting and radiotherapy. J Neurosurg 55:889895, 1981. Apuzzo MLJ, Chikovani OK, Gott PS, et al: Transcallosal, interforniceal approaches for lesions af fecting the third ventricle: Surgical considerations and consequences. Neurosurgery 10:547-554, 1982. Arey LB: Developmental Anatomy. A Textbook and Laboratory Manual of Embryology. Phila delphia, W В Saunders, 1965, ed 7. Atunes JL, Louise KM, Ganti SR: Colloid cysts of the third ventricle. Neurosurgery 7:450-455, 1980. Bosch DA, Rahn T, Backlund EO: Treatment of colloid cysts of the third ventricle by stereotactic aspiration. Surg Neurol 9:15-18, 1978. Chapman PH, Lindggood RM: The management of pineal area tumors: A recent reappraisal. Cancer 46:1253-1257, 1980. Ciric I: Microsurgical anatomy of pituitary adeno mas in various stages of their development. Pre sented at the Meeting of the Congress of Neurolog ical Surgeons, Washington, DC, October 1978. Ciric I: Neuroepithelial cysts. J Neurosurg 44:134,1976. Ciric I: On the origin and nature of the pituitary gland capsule. J Neurosurg 46:596-600, 1977. Ciric IS, Cozzens JW: Craniopharyngiomas: Transsphenoidal method of approach—for the virtuoso only? Clin Neurosurg 27:169-187, 1980. . Ciric I, Mikhael M, Stafford T, et al: Transsphe noidal microsurgery of pituitary macroadenomas with long-term follow-up results. J Neurosurg 59:395-401, 1983. Ciric IS, Zivin I: Neuroepithelial (colloid) cysts of the septum pellucidum. J Neurosurg 43:69-73, 1975. Costello RT: Subclinical adenoma of the pituitary gland. Am J Pathol 12:205, 1936. Dandy WE: Benign Tumors in the Third Ventri cle of the Brain: Diagnosis and Treatment. Springfield, Charles С Thomas, 1933. DeGirolami U: Pathology of tumors of the pineal region. In Schmidek HH (ed): Pineal Tumors. New York, Masson Publishing, 1977, pp 1-19. del Rio-Hortega P: Pineal gland. In Penf ield W (ed): Cytology and Cellular Pathology of the Nervous System. New York, Paul В Hoeber, 1932, vol 2, pp 635-703. Erdheim J: Uber Hypophysengangsgeschwulste und Hirmcholesteatome. Sitzungsb Kais Akad Wissen Math Naturw Klin 113:537-726, 1904. Grover WD, Rorke LB: Invasive craniopharyngioma. J Neurol Neurosurg Psychiatry 31:580582, 1968. Handa H, Yamashista J: Summary: Current treat ment of pineal tumors. Surg Neurol 16:279, 1981. Hardy J: Transsphenoidal microsurgery of the normal and pathological pituitary. Clin Neuro surg 16:185-217, 1969. Hardy J, Vezina JL: Transsphenoidal neurosurgery of intracranial neoplasm. Adv Neurol 15:261-274, 1976. Herrick MK: Pathology of pineal tumors. In Neuwelt EA (ed): Diagnosis and Treatment of Pineal Region Tumors. Baltimore, Williams & Wilkins, 1984, pp31-60. Hirsch JF, Zouaoui A, Renier D, et al: A new
24. 25. 26. 27.
28. 29. 30. 31.
32. 33.
surgical approach to the third ventricle with interruption of the striothalamic vein. Acta Neurochir(Wien) 47: 135-147, 1979. Jenkins RDT, Simpson WJK, Keen CW: Pineal and suprasellar geminomas: Results of radiation treatment. J Neurosurg 48:99-107, 1978. Jennings S, Gelman R, Hochberg F: Intracranial germcell tumors: Natural history and pathogenesis. J Neurosurg 63:155-167, 1985. Jooma R, Kendall BE: Diagnosis and management of pineal tumors. J Neurosurg 58:654-665, 1983. Kovacs K, Horvath E: Morphology adrenohypophyseal cells and pituitary adenomas. In Imura H (ed): The Pituitary Gland. New York, Raven Press, 1985, pp 25-56. Little JR, MacCarty CS: Colloid cysts of the third ventricle. J Neurosurg 39:230-235, 1974. Long DM, Chou SN: Transcallosal removal of craniopharyngiomas within the third ventricle. J Neurosurg 39:563-567, 1973. McGovern VJ: Tumors of the epiphysis cerebri. J Pathol Bacteriol 61:1-9, 1949. Pecker J, Guy G, Scarabin JM: Third ventricle tumours: Including tumours of the septum pellicidum, colloid cysts and subependymal glomerate astrocytoma. In Vincken PJ, Bruyn GW (eds): Handbook of Clinical Neurology. Amsterdam, North Holland Publishing Co, 1974, vol 17, pp 440-489. Rathke H: Ueber die entstchung der glandula pituitaria. Arch Anat Physiol. Wissen Med 5:482486, 1838. Rhoton AL Jr, Yamamoto I, Peace DA: Microsur gery of the third ventricle: Part 2. Operative ap proaches. Neurosurgery 8:357-373, 1981.
34. Rubinstein LJ: Tumors of the central nervous system. In Atlas of Tumor Pathology, 2nd series, fascicle 6. Washington, DC, Armed Forces Insti tute of Pathology, 1982. 35. Russell DS, Rubinstein LJ: Pathology of Tumors of the Nervous System. Baltimore, Williams & Wilkins, 1977, ed 4, pp 287-290. 36. Sano K, Matsutani M: Pinealoma (germinoma) treated by direct surgery and postoperative irra diation: A long-term follow-up. Childs Brain 8:81-97, 1981. 37. Sherman BM, Harris CE, Schlechte J, et al: Pathogenesis of prolactin-secreting pituitary adeno mas. Lancet 2:1019, 1978. 38. Shuangshoti S, Robert MP, Netsky MG: Neuroepithelial (colloid) cysts: Pathogenesis and relation to choroid plexus ependyma. Arch Pathol Lab Med 80:214-224, 1965. 39. Shucart WA, Stein BM: Transcallosal approach to the anterior ventricular system. Neurosurgery 3:339-343, 1978. 40. Sobel RA, Trice JE, Neilsen SL, et al: Pineoblastoma with ganglionic and glial differentiation. Acta Neuropathol 55:243-246, 1981. 41. Stein BM: Third ventricular tumors. Clin Neuro surg 27:315-331, 1980. 42. Takeuchi J, Handa H, Nagata I; Suprasellar geminoma. J Neurosurg 49:41-48, 1978. 43. von Mihalkovics V: Wirbesaite und Hirnanhang. Arch Mikrosc Anat 11:389-439, 1875. 44. Viale GL, Turtas S: The subchoroid approach to the third ventricle. Surg Neurol 14:71-76, 1980. 45. Weiss MH, Teal J, Gott P, et al: Natural history of microadenomas: Six year follow-up. Neurosur gery 12:180-183, 1983.
5 Physiological Consequences of Complete or Partial Commissural Section Joseph E. Bogen, M.D.
Complete section of the corpus callosum, which has come to be used increasingly as a treatment for medically intractable epilepsy (88), is regularly followed by a wide variety of neurological and neuropsychological deficits. The deficits that are most apparent immediately after operation make up the acute disconnection syndrome (110). Those deficits that persist make up the stabilized syndrome of hemisphere disconnection (12, 14, 15, 100). Partial section of the corpus callosum to facilitate the operative approach to deep midline lesions may produce a fraction of the full syndrome, depending upon the part of the corpus callosum that has been cut as well as the amount and nature of any associated extracallosal damage. In general, most of the disconnection deficits, both acute and chronic, are not present as long as the splenium is spared (4, 52). But certain partial sections do entail a few inevitable deficits, as well as some common complications, whose recognition is important for the surgeon utilizing the transcallosal approaches. In this chapter I present first the syndromes, acute and chronic, consequent to a complete cerebral commissurot-omy (including the corpus callosum and the underlying dorsal hippocampal commissure, the ventral hippocampal commissure between the fornices, the massa intermedia, and the anterior commissure). This is followed by comments on those particular aspects of the full-blown syndrome that are apt to appear with different partial sections. One should keep in mind that cal-losal lesions, surgical as well as naturally occur-
ring, are often accompanied by damage to or pressure upon neighboring structures, resulting in several distinct types of signs: (a) signs of hemisphere disconnection, (b) neighborhood signs, and (c) nonlocalizing signs such as men-ingismus or signs of increased intracranial pressure. In this chapter, attention is given mainly to signs of the first type. A recent history of studies of the corpus callosum explains how the disconnection signs have come to be understood and emphasized (12). Other extensive reviews are available (8, 16, 27, 28, 37, 42, 70, 80, 81, 106). Historically Salient Signs Before callosal syndromes are considered in detail, it is useful to have an overall view of those signs of hemispheric disconnection that have been emphasized historically and whose reality has been confirmed by more recent investigations. After hemisphere disconnection in the human, unilateral tactile anomia, left hemialexia, and unilateral apraxia are typical. That is, a righthander with complete cerebral commissurotomy cannot name aloud objects correctly manipulated (hence, recognized) with the left hand, cannot read aloud written material presented solely to the left halffield of vision, and cannot execute with the left hand actions verbally named or described by the examiner, although these actions are readily imitated when demonstrated. The apraxia usually recedes within a few months, whereas the hemialexia and unilateral anomia can persist for years.
The relationship of apraxia to the corpus cal-losum was pointed out by Liepmann and Mass (71). They considered the corpus callosum instrumental in left hand responses to verbal command: the verbal instruction was comprehended only by the left hemisphere and the left hand followed instructions delivered by a route involving callosal interhemispheric transfer from left to right and then by what we now call "contra-lateral control," that is, by right hemisphere control of the left hand. Necessarily then, callosal interruption would result in an inability to follow verbal commands with the left hand although there would be no loss of comprehension (as expected from a left hemisphere lesion) and there would be no weakness or incoordination of the left hand (as expected from a right hemisphere lesion). We now recognize the notion that spatial or pictorial instructions understood by the right hemisphere may require callosally mediated interhemispheric communication for correct right hand execution. This right-to-left aspect of callosal function was not part of Liepmann's original callosal concept although, in retrospect, it seems a natural corollary. Liepmann himself considered the left hemisphere to be the organizer of complex (particularly learned) motor behavior. Whether and in what way the left hemisphere is dominant for skilled movements generally (and not just those linguistically related) are currently matters of active controversy (56, 58, 63, 66, 115). Acute Disconnection Syndrome (As Observed in RightHanders Operated from the Right Side) During the first few days after complete cerebral commissurotomy, the patients commonly respond reasonably well, with their right limbs, to simple commands. But they are easily confused by three- or even two-part commands, each part of which is obviously understood. The patients often lie quietly and may seem mildly "akinetic," although cooperating when stimulated. There is sometimes an "imperviousness" resembling that often seen with naturally occurring genu lesions (2, 12). The patients are often mute even when willing to write short (usually one-word) answers. This mutism is of special interest and is discussed in more detail under "Postcallosotomy Mutism." The left-sided apraxia to verbal command is usually severe and can be mistaken for hemiple-gia. Similarly, a disregard for the left half-field of vision can be mistaken for a hemianopsia. Left side weakness in the first week or so due to
retraction edema in the right hemisphere sometimes confounds the picture. The neurological status is difficult to evaluate not only because the seemingly flaccid and unresponsive left extremities may exhibit coordinated movements, but also because, as the patient improves, there may be competitive movements between the left and right hands. Moreover, some patients have focal motor seizures, manifested by clonic contractions on alternating sides of the body and without loss of consciousness, occasionally followed by transient unresponsiveness of whichever limbs were involved. The patients commonly have bilateral Babinski signs as well as bilaterally absent superficial abdominal reflexes. Well-coordinated but repetitive reaching, groping, and grasping with the left hand sometimes resembles a grasp reflex; grasp reflexes may actually be present bilaterally for a day or two. When forced grasping cannot be elicited (by inserting two fingers into the patient's palm from the ulnar side, with some distal stroking), it is nevertheless possible in most cases to demonstrate a proximal traction response (PTR) (i.e. the patient is unable to relax the hand grip when the examiner pulls so as to exert traction on the elbow and shoulder flexors). The left arm hypo-tonia, the left arm PTR, the responses bilaterally to plantar stimulation, and the mutism were regularly observed in our cases; two representative graphic summaries are shown in Figures 5.1 and 5.2. Figure 5.1 indicates the intensity and duration of mutism and other features of the acute disconnection syndrome in our first patient, W.J. (21-23). His commissural section was done without entering the third ventricle; two exposures, before and behind the rolandic veins, made it possible to confirm under direct vision that his callosal section was complete. The patient's condition worsened at the end of the first week, associated with an alkalotic hyponatremia present on Days 5 to 7 (the patient had had diabetes insipidus ever since his previous head injury, and vas-opressin was used throughout his hospital stay). When the hyponatremia was corrected, the patient rapidly improved with respect to alertness and left side coordination, with subsidence of the hypotonia and PTR. But some degree of mutism persisted for another month. This patient's mutism as well as that of several others has been described in detail elsewhere (11). As discussed below, the duration of mutism seems to be related to the extent of extracallosal damage; in this case there was distinct atrophy of the right frontal lobe quite evident at the time of operation, as
Figure 5.1. Features of the acute disconnection syndrome as observed postop-eratively in Patient W.J., a man aged 48 at the time of operation on Feb. 2, 1962. The capital letter N means "none" (i.e., normal). A solid block in black indicates a marked deficit, whereas interrupted black bars mean that the deficit was present but mild, the observations having been made by the author (J.E.B.). Where the solid block or interrupted bars are gray, the observation was made by someone else, usually a resident.
Figure 5.2. Features of the acute disconnection syndrome as observed postop-eratively in Patient N.G., a woman aged 30 at the time of operation on Sept. 5, 1963. The symbols are the same as for Figure 5.1. well as a left parietotemporal electroencephalogram (EEG) focus and bitemporal EEG foci on sphenoid recordings (92). Figure 5.2 shows the daily postoperative observations of key features of the acute disconnection syndrome in Patient N.G. (13, 18, 23). Her anterior commissure and massa intermedia were sectioned under direct vision after opening into the third ventricle between the fornices. Recent magnetic resonance imaging studies have con-
firmed that her callosotomy was complete and that there was a paucity of extracommissural damage. That her extracommissural damage was slight is probably important with respect to the relatively brief duration of her mutism. Her postoperative course was overall quite smooth (small amounts of vasopressin were used on Days 3 to 6). Even so, there was a marked left apraxia to verbal command for many weeks. There was also, as in every one of our 12 right-handers
tested after complete commissurotomy, a persistent anomia in the left hand. The left hand anomia has been such a regularly appearing, easily tested, and persistently present phenomenon that a separate section, "Unilateral (Left) Tactile Anomia," is devoted to its further description. N.G.'s postoperative course was summarized at the end of a month by the author: Two hours after operation the patient did not react to pain and an hour later there was bilateral flexion to sternal rub. An hour later there was a good grip on the right to request, the right toe sign was positive and the entire left side was flaccid. In succeeding hours the left side remained flaccid to passive motion and to verbal request but the patient used both hands together to pull the covers over her chest. Hyperreflexia and positive toe sign on the right side subsided after several days; a tonic grasp and proximal traction response on the right rapidly decreased but could be intermittently elicited as late as a week postoperatively. On the first postoperative day there was occasional mumbling including "yes"and "no," on the second day there was intelligible speech and on the third day the patient was well oriented for place and person although not for time and did not recognize her husband. Disorientation for time and date and intermittent inability to recognize the doctors persisted for nearly two weeks. Complete apraxia, tonic grasp and PTR remained pronounced on the left side for two weeks. By the end of the third week, the left grasping had largely subsided; no verbal instructions were followed with the left hand but poses and gestures were copied excellently. The patient graduated from wheelchair to walker after two weeks but continued to have difficulty controlling the left leg. By the fourth week the patient was walking very well although still apraxic with the left hand. A positive Babinski sign has been present not only contralateral to the retracted hemisphere but also, for at least a day or two, on the ipsilat-eral side. An ipsilateral (as well as contralateral) toe sign was also seen in our one patient operated from the left, Patient C.C. (9, 23). The best explanation for this ipsilateral Babinski sign is probably that the diaschistic shock effect on the unexposed hemisphere after an extensive cal-losal section is sufficient to depress for a few days the corticofugal inhibition of the primitive extensor response (10, 107). In any event, the Babinski signs rapidly subside bilaterally and are not present over the long term (24). An ipsilateral Babinski sign is usually not observed when the callosal section is partial. It is not possible to include here all of the daily progress notes on this patient, who had a rather smooth, almost ideal postoperative course after a complete section, but the following provides some idea of the patient's status at the end of 2l/2 weeks:
Progress Note of Sept. 23, 1963 The patient has progressed rapidly in the past week and is now up walking; she eats by herself without difficulty and discusses things at some length although after a lengthy conversation she tends to become a little bit confused and says, "you people must have taken out my brains." When the patient walks she tends to have a slight shuffle but there is no asymmetry. However, if one asks her to walk on her heels she does this readily with the right foot but does not do it with the left foot. If one asks her to walk standing up on her tiptoes she again does this only with the right foot; she could not rise up on either the heel or the toe of the left foot. The hand is progressing and whereas some days ago there was a PTR on the left the patient now has no difficulty letting go with the left hand when asked to do so, even if considerable stretch is applied to the flexor joints. However, apraxia of the left hand continues in that the patient has considerable difficulty carrying out any command with the left hand and often, while laughing, reaches over with the right hand to make the left hand behave in a more appropriate or precise fashion. When the patient was asked to put her left hand on her head she could not do this, but when she was asked to imitate the examiner when he put his left hand on his head she did it without any difficulty. She was then asked, "now put your left hand on your head" and she did this right away. She was then asked, "put your left hand on your left ear" and the left hand went on top of the head. This could easily be duplicated with any combination of movements; that is, the patient would be shown how to do something and do it excellently with the left hand and then she would be asked to do something else and the hand would do what she had previously just done. . . . Without warning, I put my hand in front of my forehead so that a finger was placed in each supra-orbital notch with the palm of my hand up over my forehead and the little finger sticking out to the side. No sooner had I done this without saying anything, when the patient's left arm went up in complete imitation, to the exact positioning of the fingers. After her arm was lowered, the patient was then asked, "just lift your arm up in the air." The left arm went up with palm on the forehead, in an imitation of the previous peculiar position. Reflex testing showed no asymmetry; the knee jerks were quite hyperactive bilaterally but the reflexes at the ankle, biceps and triceps were normal. Plantar stimulation produced some withdrawal on each side with no suggestion of dorsiflexion. Testing with the pinwheel disclosed that on the left the patient felt the pinwheel to be somewhat less sharp than the corresponding parts of the right side; this hypalgesia is similar to what she had before surgery and is quite distinct from the "shock-like" sensation which W. J. has on the left side of his body. Vibratory sense is apparently quite normal on the right side but the patient is not cooperative enough to give reliable results over the left side. As the severity of the acute disconnection syndrome subsides in a week or so, a phenomenon variously called "intermanual conflict" or "the alien hand" appears. Almost all complete com-
missurotomy patients manifest some degree of intermanual conflict during the early postoperative period. For example, a few weeks after operation, Patient R.Y. (23) was seen buttoning his shirt with one hand while the other hand was following along just behind undoing the buttons. Intermanual conflict was observed after commissurotomy by Wilson et al. (110) and by Ake-laitis (1), who called it "diagnostic dyspraxia." The phenomenon has been described in many case reports of callosal infarcts or tumors. Even our youngest patient (L.B.) with no appreciable apraxia to verbal command in the long term, manifested this "alienation" 3 weeks after operation. While doing the block design test uniman-ually with his right hand, his left hand came up from beneath the table and was reaching for the blocks when he slapped it with his right hand and said, "That will keep it quiet for a while." Such behavior progressively subsides after callosotomy, probably because of other integrative mechanisms supplementing or replacing com-missural function. More recently, examples of intermanual conflict have been reported in detail by Rayport et al. (87) and Ferguson (43). In rare instances, the intermanual conflict may reappear even years later. An example is the following from a follow-up examination done on Feb. 26, 1973. (Patient A.M. was operated on Jul. 7, 1964, at age 31). The most interesting finding in the entire examination is the frequent occurrence of well-coordinated movements of the left arm which are at cross-purposes with whatever else is going on. These sometimes seem to occur spontaneously, but on other occasions are clearly in conflict with the behavior of the right arm. For example, when attempting a Jendrassic reinforcement, the patient reached with his right hand to hold his left, but the left hand actually pushed his right hand away. While testing finger-to-nose test (with the patient sitting), his left hand suddenly started slapping his chest like Tar-zan. The patient follows verbal instructions readily with the right hand, but poorly if at all with the left hand, although he can raise and lower the left arm with considerable reliability, the defect being more obvious the more distal the joints involved in the requested act. After a variety of maneuvers (the left hand often following a visual demonstration fairly well) it was possible to test the grip of the left hand while keeping the thumb extended—so it was apparent that there is a negative Wartenberg thumb sign on the left as well as on the right. In most patients, the intermanual conflict progressively subsides in a few months or even weeks. Patient N.G. had only mild intermanual conflicts in the hospital and none were noted after her return home. Her status 5 months post-operatively was as follows:
Progress Note of Feb. 2, 1964 N.G. is doing excellently. There is still some lack of initiative and an occasional defect in memory, but otherwise her husband and she were both delighted with her progress when I visited them in their home today. The surgical wound has healed well and has a good cosmetic appearance except for a small depression in the most anterior burr hole. She has no headaches. The patient is eating well and has gained weight although she has been doing calisthenics each morning to the TV. She sleeps well except for an occasional nocturia (she needs no help with personal care); she recently had a urethral dilatation and she takes mandelamine 4 times a day along with her Dilantin 100 mg and Mebaral 50 mg. She has had no seizures (in particular no convulsive movements) unless one wishes to suspect as seizures an occasional inattention and some numbness of the left side. This absence of seizures is in spite of the fact that the patient has had two menstrual periods since surgery, the most recent a few weeks ago. The numbness of the left side of the body is apparently continuous at a low level, but once or twice a day it is quite bothersome for 5 to 10 minutes at a time. It involves more the leg (thus resembling her aura before surgery) and also the arm but not the face. In spite of the "numbness" the patient has no disability with the left hand, which she uses quite naturally for dressing, cooking, etc. Although she occasionally has trouble fastening her bra behind her back she says that she dresses with ease and demonstrated for me twohanded tying of shoe laces, this being done in a manner indistinguishable from normal. The patient also shuffled a pack of cards and dealt them deftly. She dealt onehanded with either hand (slightly better with the left) and at my request she unbuttoned and rebuttoned the top of her blouse with her right hand while simultaneously dealing cards with the left hand. The patient wrote well with the right hand but could make no legible characters with the left hand on several trials. However, copying of designs was somewhat better with the left. Her deficit with the right hand appeared to be a deficit of conceptualization whereas the deficit in the left hand appeared to be one of muscular coordination. It is of interest that neither the patient nor her husband could recall any incident of intermanual conflict since her arrival home. On blindfold testing today, the patient followed my verbal commands with the left quite well, her unilateral apraxia seemingly overcome. However, the anomia was still present in the left hand to full extent. I repeatedly demonstrated this with various objects while her husband sat watching in open mouthed silence (Fig. 5.2). The patient's personality seems to me essentially unchanged from before surgery; she laughs easily and appropriately. Her husband says she talks less ("she used to talk my ear off") but that she socializes well with her neighbors. She is perhaps a little more interested in sexual activity than before the operation. Her memory occasionally fails in that she may sometimes accompany her husband somewhere and on the next day ask him what they did. The memory loss occasionally takes the form of absent minded-
ness; while helping her husband prepare lunches for the family, "she sometimes puts 2 or 3 slices of lunch meat in a sandwich instead of one." During my visit the patient was quite interested and alert and not only responded appropriately to questions, but spoke spontaneously on a number of subjects and did not appear to have any memory difficulties at this time. Her husband makes each night a list of housework for her to do the next day, and, "she does everything well; but if I forget to make the list, she will just sit around the next day and do nothing." He makes out a menu and she cooks the meals. "I used to have trouble getting things to come out at the same time," she said, "but the meals are better now." (Her husband nodded agreement.) N.G. remained seizure-free for 8 years and then had status epilepticus a few months after her family physician discontinued her anticon-vulsant medication (13). Her seizures were readily controlled when her medication was reinsti-tuted, and her EEG abnormalities cleared during the next 18 months. When N.G. was examined in November 1984, she continued to be seizure-free, to maintain her household efficiently, and to participate in laboratory studies at Cal Tech. Her left-handed anomia was still readily demonstrable 21 years postoperatively. This persistence of unilateral tactile anomia was also true of all of our other patients (17). Chronic, Stabilized Syndrome of Hemisphere Disconnection Overall Effects Within a few months after operation, the symptoms of acute hemisphere disconnection become compensated to a remarkable degree. In personality and social situations the patient appears much as before. However, when input is lateralized (i.e., presented to only one hemisphere) each hemisphere can be shown to operate independently to a large extent. Each hemisphere seems to have its own learning processes and its own separate memories. Split-brain patients soon accept the idea that they have capacities of which they are not conscious, such as lefthand retrieval of objects not nameable. They may quickly rationalize such acts, sometimes in a transparently erroneous way (46). But even many years after operation the patients will occasionally be quite surprised when some well-coordinated or obviously well-informed act has just been carried out by the left hand. This is particularly common under conditions of continuously lateralized input (116, 118). Visual Effects: Double Hemianopia Visual material can be presented selectively to a single hemisphere by having the patient fix his
gaze on a projection screen onto which pictures of objects or symbols are backprojected to the right, left, or both visual half-fields using exposure times of 0.1 second or less. The patients can read and describe material of various kinds in the right half-field essentially as before operation. When stimuli are presented to the left half-field, however, the patients usually report that they see "nothing" or at most "a flash of light." The disconnection (if it includes the splenium) can sometimes be demonstrated with simple confrontation testing. The patient is allowed to have both eyes open but does not speak and is allowed to use only one hand (sitting on the other hand, for example). Using the free hand, the subject indicates the onset of a stimulus, such as the wiggling of the examiner's fingers. With such testing there may seem to be a homonymous hemianopia in the half-field contralateral to the indicating hand. When the patient is tested with the other hand, there seems to be a homonymous hemianopia in the other half-field. This situation must be distinguished from the much more commonly occurring extinction or hemiinattention deficit from a hemispheric lesion, in which the patient tends to indicate only one stimulus when the stimuli are in fact bilateral (109). An observable difference is that the double hemianopia is a symmetrical phenomenon (the deficit occurs on each side) whereas extinction or hemiinattention is typically onesided, more commonly for the left side. In most patients there eventually appears a condition in which no field defect can be demonstrated by a casual confrontation technique. This apparently depends mainly upon the ability of each hemisphere to direct the head and eyes. For example, if the patient is instructed to point with the right hand and if the examiner then wiggles the fingers in the patient's left visual half-field, the patient's head and eyes quickly turn to the left and then the right hand points to the correct target. If turning of the head is prevented, a leftward glance will suffice for the patient's need for a cue. Or the right hand may point to the left visual half-field as soon as it is apparent that there is no suitable stimulus in the right visual half-field. This "cheating" by the left hemisphere can often be detected by providing no stimulus at all. Furthermore, if stimuli are presented simultaneously in both visual halffields, only the stimulus in the right half-field is described by the patient, that is, by the left hemisphere. There is usually no verbal response to the stimulus in the left visual half-field until the left hemisphere realizes
that the patient's left hand is also in action, pointing to the left half-field stimulus. When a patient has been tested repeatedly, so that the occurrence of bilateral stimuli can be anticipated, both stimuli may be identified by a single hand; if using the right hand, the patient will point first to the right and then to the left half-field. In those patients who have been frequently tested, the appearance of a stimulus in the left visual half-field is occasionally recognized in spite of our attempts to circumvent the various cross cueing strategies. However, even in our patient who does the best (L.B.), performance on confrontation testing of visual fields is still distinguishable from normal. Auditory Suppression After cerebral commissurotomy, the patient readily identifies single words (and other sounds) if they are presented to one ear at a time. But if different words are presented to the two ears simultaneously ("dichotic listening") only the words presented to the right ear will be reliably reported (40, 79, 95). This large right ear advantage is usually considered the result of two concurrent circumstances: (a) the ipsilateral pathway (from the left ear to the left hemisphere) is suppressed by the presence of the simultaneous but differing inputs, as it is in intact individuals during dichotic listening (65, 103). (b) The contralateral pathway from the left ear to the right hemisphere conveys information that ordinarily reaches the left (speaking) hemisphere by the callosal pathway and that has now been severed. Although left ear words are rarely reported, their perception by the right hemisphere is occasionally evidenced by appropriate actions of the left hand (50). Contralateral ear suppression commonly appears after hemispherectomy or the creation of another large hemispheric lesion. Because there is usually suppression of the right ear by left hemisphere lesions, suppression of the left ear by a left hemisphere lesion has been called "paradoxical ipsilateral extinction." Further observations have led to the conclusion that, whether the lesion is in the left or the right hemisphere, if it is close to the midline the suppression of left ear stimuli is probably attributable to interruption of interhemispheric pathways (32, 76, 96). Unilateral Apraxia The degree of left hand dyspraxia is subject to large individual differences. Immediately after operation all of our patients showed some left-sided apraxia to verbal commands such as "Wig-
gle your left toes" or "Make a fist with your left hand." Left limb dyspraxia can be attributed to the simultaneous presence of two deficits: poor comprehension by the right hemisphere (which has good control of the left hand) and poor ipsilateral control by the left hemisphere (which understands the commands). Subsidence of the dyspraxia, therefore, can result from two compensatory mechanisms: increased right hemisphere comprehension of words or increased left hemisphere control of the left hand. The extent of ipsilateral motor control can be tested by flashing to right or left visual half-field sketches of thumb and fingers in different postures for the subject to mimic with one or the other hand. Responses are poor with the hand on the side opposite the visual input, simple postures such as a closed fist or an open hand being attainable after further recovery. As recovery proceeds, good ipsilateral control is first attained for responses carried out by the more proximal musculature. After several months, most of the patients can form a variety of hand and finger postures with either hand to verbal instructions, such as, "Make a circle with your thumb and little finger" and the like. Even many years later, however, some degree of apraxia can be demonstrated (115). Somesthetic Effects The lack of interhemispheric transfer after hemisphere disconnection can be demonstrated with respect to somesthesis (including touch, pressure, and proprioception) in a variety of ways. Cross Retrieval of Small Test Objects Unseen objects in the right hand are handled, named, and described in normal fashion. In contrast, attempts to name or describe the same objects held out of sight in the left hand consistently fail. In spite of the patient's inability to name an unseen object in his left hand, identification of the object by the right hemisphere is evident from appropriate manipulation of the item, showing how it is used, or by retrieval of the same object with the left hand from among a collection of other objects screened from sight. What distinguishes the split-brain patient from normal is that excellent same hand retrieval (with either hand) is not accompanied by the ability to retrieve with one hand objects felt with the other. Cross Replication of Hand Postures Specific postures impressed on one (unseen) hand by the examiner cannot be mimicked in the
opposite hand. Also, if a hand posture in outline form is flashed by tachistoscope to one visual half-field, it can be copied easily by the hand on that side but usually not by the other hand. A convenient way to test for lack of interhem-ispheric transfer of proprioceptive information is as follows: the patient extends both hands beneath the opaque screen (or vision is otherwise excluded) and the examiner impresses a particular posture on one hand. For example, one can put the tip of the thumb against the tip of the little finger and have the other three fingers fully extended and separated. The split-brain patient cannot mimic with the other hand a posture being held by the first hand. This procedure should be repeated with various postures and in both directions. Cross Localization of Finger Tips After complete cerebral commissurotomy there is a partial loss of the ability to name exact points stimulated on the left side of the body. This defect is least apparent, if at all, on the face and it is most apparent on the distal parts, especially the finger tips. This deficit is not dependent upon language; it can be shown in a nonverbal (picture identification) fashion, in which case the deficit is present in both directions (right-to-left and vice versa). An easy way to demonstrate the defect is to have the subject's hands extended, palms up (with vision excluded). One touches the tip of one of the four fingers with the point of a pencil, asking the patient to then touch the same point with the tip of the thumb of the same hand. Repeating this maneuver many times produces a numerical score, about 100% in normals for either hand. In the absence of a parietal lesion, identification of any of the four finger tips by putting the thumb tip upon the particular finger can be done at nearly 100% level by the split-brain patient. One then changes the task so that the finger tip is to be indicated, not by touching it with the thumb of the same hand, but by touching the corresponding finger tip of the other hand with the thumb of that (other) hand. Sometimes the procedure should be demonstrated with the patient's hand in full vision until the patient understands what is required. This cross localization cannot be done by the split-brain patient at much better than chance level (25%), whereas most normal adults do better than 90%. An incompetence to cross localize or cross match has been found in young children (44) possibly because their commissures are not yet fully functioning (113).
Verbal Comprehension by the Right Hemisphere Auditory comprehension of words by the disconnected right hemisphere is suggested by the subject's ability to retrieve with the left hand various objects if they are named aloud by the examiner. Visual comprehension of printed words by the right hemisphere is often present; after a printed word is flashed to the left visual half-field, the subject is often able to retrieve with the left hand the designated item from among an array of hidden objects. Control by the left hemisphere is excluded in these tests because incorrect verbal descriptions given immediately after a correct response by the left hand show that only the right hemisphere knew the answer. While the disconnected right hemisphere's receptive vocabulary can increase considerably over the years, this single word comprehension is rarely accompanied by speech. The most extreme cases (to date) of right hemisphere language ability in right-handed (and left hemisphere-speaking) split-brain subjects are two with right hemisphere speech, both with the anterior commissure uncut (75, 93). Right hemisphere language in the split-brain subject has other limitations, with syntactic ability being rudimentary at best. After studying a few cases in great depth for over 10 years, Zaidel concluded: Whereas phonetic and syntactic analysis seem to specialize heavily in the left hemisphere, there is a rich lexical structure in the right hemisphere. The structure of the right hemisphere lexicon appears to be unique in that it has access to a severely limited short term verbal memory, and it has neither phonetic encoding nor grapheme-to-phoneme correspondence rules . . . [this] represents the limited linguistic competence that can be acquired by a nonlin-guistic, more general purpose (or other purpose) cognitive apparatus (117). Right Hemisphere Dominance After commissurotomy, each hemisphere can be tested separately, demonstrating in a positive way those things that each hemisphere can do better than the other, rather than inferring what a hemisphere does from loss of function when it is injured. Representative reviews are included in the References (16, 69, 83, 99, 118). The rapidly growing literature on hemispheric specialization was thoroughly summarized by Brad-shaw and Nettleton (25) and by Trevarthen (105) and more recently was discussed by Gordon (51). Unilateral (Left) Agraphia Right-handers can write legibly, albeit not fluently, with the left hand. This ability is com-
monly lost with callosal lesions, especially those that cause unilateral apraxia. An inability to write to dictation is common with left hemisphere lesions, almost always affecting both hands. The left hand may be dysgraphic if affected by a right hemispheric lesion, such as a frontal lesion causing forced grasping. That the left dysgraphia after callosal sectioning is not simply attributable to an incoordination or paresis can be established if one can demonstrate some other ability in the left hand requiring as much control as would be required for writing. The left hand may spontaneously doodle or it may copy various designs or diagrams. It is not so much the presence of a deficit but rather the contrast between certain deficits and certain retained abilities that is most informative. Simple or even complex geometric figures can often be copied by a left hand that cannot write or even copy writing previously made with the patient's own right hand (9, 19, 67, 115). Unilateral (Left) Tactile Anomia The most useful single sign of hemisphere disconnection is unilateral tactile anomia: this is an inability to name or to describe an object when it is felt by one hand whereas it is readily named (or well described if the name is unknown) when it is placed into the other hand or when it is presented either to vision or to audition. This unilateral tactile anomia was present in every one of our complete commissurotomy patients, in the left hands of 12 righthanders and in the right hand of our 1 left-handed patient. Others have repeatedly confirmed this finding, and it has also been observed when the patient has had a callosotomy sparing the anterior commissure (46, 74). Not only is unilateral anomia quite regular in its appearance, but it is also quite persistent, whenever appropriately tested, for as long as 20 years (17) (and longer, as shown by my more recent testing). In additon to its regularity and its persistence, the demonstration of this sign requires a minimum of equipment and time, and the interpretation of results is usually quite clear. Of the many maneuvers developed in the laboratory to test split-brain patients, this is the principal one to be adopted as part of a routine neurological examination. Although there are many signs of brain bisection (as described), one of the most convincing ways to demonstrate hemisphere disconnection is to ask the patient to feel with one hand and then name various small, common objects such as a button, coin, safety pin, paper clip, pencil
stub, rubber band, or key. Vision must be excluded. A blindfold is notoriously unreliable; it is better to hold the patient's eyelids closed, put a pillowcase over the patient's head, or use an opaque screen. The patient with a hemisphere disconnection is generally unable to name or describe an object in the left hand although he readily names objects in the right hand. Sometimes the patient will give a vague description of the object although unable to name it, but there is a contrast with the ability to name the object readily when it is placed into the right hand. To establish hemisphere disconnection, other causes of unilateral anomia, particularly aster-eognosis (or a gross sensory deficit) must be excluded. The most certain proof that the object has been identified is for the subject to retrieve it correctly from a collection of similar objects. Such a collection is most conveniently placed in a paper plate about 12 to 15 cm in diameter, around which the subject can shuffle the objects with one hand while exploring for the test object. Even without evidence of correct retrieval, aster-eognosis can be reasonably excluded by observing the rapid, facile, and appropriate manipulation of an object despite an inability to name or verbally describe it. In testing for unilateral anomia, the examiner must be aware of strategies for circumventing the defect. For example, the patient may manipulate it to produce a characteristic noise, or the patient may identify a pipe or some other object by a characteristic smell and thus circumvent the inability of the left hemisphere to identify, by palpation alone, an object placed in the left hand. Syndrome after Corpus Callosotomy When the corpus callosum is sectioned in its entirety at one sitting the acute disconnection syndrome appears in almost all respects as it is seen after a complete cerebral commissurotomy including the anterior commissure and massa intermedia. This state after callosotomy includes transient mutism; hence this symptom when it follows a complete section is not reasonably ascribed to molestation of third ventricular structures (Fig. 5.3). Moreover, mutism does not necessarily follow a frontal commissurotomy that does include molestation of the third ventricular structures but spares the splenium. These two observations taken together suggest that post-commissurotomy mutism requires explanation on some other basis than in terms of third ventricle retraction.
Figure 5.3. Degree of mutism (marked, mild, or none) after operation on 15 patients with medically intractable epilepsy. Thirteen patients underwent complete commissurotomy and 2 patients (D.M. and N.F.) underwent frontal commis-surotomy (52). The symbols are the same as for Figure 5.1. Staying out of the third ventricle does avoid some problems, including transient diabetes in-sipidus. And there may be less likelihood of aseptic meningitis or other complications leading to hydrocephalus, a point that was often advanced by Wilson and colleagues (59, 110). Experiments with monkeys have shown that if the anterior commissure is left intact it can compensate for loss of the splenium with respect to interhemispheric transfer of certain kinds of visual information (57). But the anterior commissure cannot compensate completely for splenial loss in the human because hemialexia usually is present after splenial section. Indeed, most of the stabilized syndrome seen after a complete cerebral commissurotomy is also seen (i.e., has not been compensated) after a callosotomy sparing the anterior commissure (46, 74). This is perhaps not surprising because the anterior commissure is only Vioo the size of the corpus cal-losum. On the other hand, we can appreciate how significant it might be when we consider the wealth of information that is conveyed over one
optic nerve—the diameter of which is about the same as that of the anterior commissure. This question is complicated by the fact that the size of the anterior commissure is quite variable; a diameter difference of 3 or 4 times has been reported (112). The discrepancy between monkeys (transfer of learning by the anterior commissure) and humans (inability of the anterior commissure to compensate for callosotomy) may reflect differences between recently acquired memories as opposed to longstanding ones. Current evidence suggests that memory deficits can be expected after commissurotomy that includes the anterior commissure, even when the splenium is spared (30, 78, 114). Syndrome after Frontal Commissurotomy By "frontal commissurotomy" we refer (52) to section of the anterior two-thirds of the corpus callosum together with anterior commissurotomy; section of the anterior commissure was done under direct vision and the third ventricle was entered between the two fornices so the
Figure 5.4. Frontal commissurotomy: Sketched here are the structures sectioned during a frontal commissurotomy. This includes most of the corpus callosum (sparing the splenium), the ventral hippocampal commissure (between the forn-ices), and the anterior commissure. section also included the ventral hippocampal commissure (Fig. 5.4). The same operation was used on a few occasions by Wilson et al. (110). When the splenium is spared, as it is in the case of frontal commissurotomy, very little of either the acute or the chronic disconnection syndrome is seen. In particular, mutism did not occur in our four operations for intractable seizures with bitemporal foci. Because these patients had retraction within the third ventricle to allow section of the anterior commissure under direct vision, the mutism in cases with more extensive section (i.e., complete commissurotomy) was not likely of third ventricular origin. Moreover, all of these four had retraction of the medial aspect of the right frontal lobe, comparable with that in the complete cases; retraction on the supplementary motor cortex does not therefore seem to be a sufficient explanation for the mutism after callosotomy, an issue discussed in more detail under "Postcallosotomy Mutism." In the long term, Preilowski could show some deficits of motor coordination in two of our pa-
tients with frontal commissurotomy (85). But exhaustive testing for the usual disconnection deficits (as described under "Chronic, Stabilized Syndrome of Hemisphere Disconnection"), has shown that, with retention of the splenium, such deficits should not be expected (17, 47, 52, 55). Memory deficits have been observed after complete commissurotomy and also in lesser degree after frontal commissurotomy sparing the splenium. This is probably in part attributable to deficits in combining linguistic representations with visual or spatial images. We need further research using a wide variety of memory tests before and after operation, particularly with prolonged follow-up. Syndrome after Callosotomy Sparing the Splenium Section of the corpus callosum that spares the splenium (as well as the anterior commissure) but is otherwise complete can be readily accomplished via an exposure anterior to the rolandic
Figure 5.5. Anterior callosotomy. This sketch shows how anterior callosotomy differs from frontal commissurotomy, which is illustrated in Figure 5.4. There are two principal differences: the anterior commissure is spared, and third ventricle is not entered by separating the fornices. bridging veins (Fig. 5.5). This procedure, which avoids entry into the third ventricle, has come to be the most popular version of commissural section for epilepsy. After this operation one does not observe most of the acute disconnection syndrome. This smoother postoperative course is not surprising because the acute disconnection syndrome does not usually occur after frontal commissurotomy, which is essentially the same procedure plus section of the anterior commissure and massa intermedia. Absence of significant mutism with callosotomy sparing the splenium has been reported by Rayport et al. (87) as well as in personal communications to myself from M. Rayport, M. Apuzzo, J. Vries, J. Wada, and G. Yasargil. On the other hand, a recent abstract by Ross et al. (91) reported transient (1 - to 10-day) mutism with this procedure. This difference could depend upon the amount of retraction, particularly if there has been retraction of the left hemisphere below the falx as well as the usual right hemi-
sphere retraction. Or it might reflect differences in the criteria of the observers. Perhaps most important are differences in the patients; even a lesser callosal section may be followed by mutism when it is associated with a concurrent thalamic lesion. Midcallosal Section The most common variant of callosal section is an incision through the trunk sparing not only the splenium but also most of the genu as well. Such an incision affords ready access to both foramina of Monro with practically no obligatory physiological cost as presently assessable (3, 4, 52, 72, 84, 94, 111). With some possible exceptions (occasional brief mutism, inconstant auditory effects, tactile transfer deficits) any impairments after callosal trunk section are typically of the neighborhood type; for example, the language deficits allegedly secondary to callosal trunk section have been reported in only one
case, one in which the callosum was approached in a right-hander by retraction of the left hemisphere (36). Mutism in Special Cases Mutism can be expected, at least for a time, if partial callosal section is done in a right-hander to remove or to biopsy a right thalamic lesion; this conclusion is based on three personal cases and two cases of others on which I was consulted. Mutism for several weeks followed callosal trunk incision for the removal of a colloid cyst in a case in which right thalamic injury had occurred during a previous emergency ventriculostomy (personal communication from Dr. Paul LaPrade). But significant mutism did not appear after callosal trunk section in five other personal cases including a small left thalamic arteriovenous malformation (AVM) (not resected) and two small AVMs of the septum (with resection in both cases). As mentioned under "Syndrome after Frontal Commissurotomy," no mutism appeared after frontal commissurotomy in four of our epilepsy patients and no mutism was seen in two personal tumor patients (both still being followed after 10 and 13 years) with quite extensive mid-callosal section for a third ventricle craniopharyngioma and a third ventricle glioma (84). Inconstant Auditory Effects As pointed out under "Auditory Suppression," a naturally occurring lesion near the callosal trunk can result in suppression of one ear when tested dichotically. Hence, one might except similar changes after section of the callosal trunk. A few changes have sometimes been detected (personal communications from E. Zaidel and from E. Teng); this requires further investigation. In any case, the auditory effects are unlikely to be of clinical importance, although one can anticipate certain patients (for example, a symphony conductor) in whom slight auditory alterations might be quite important. Effects on Tactile Interaction? The somesthetic nontransfer described under "Somesthetic Effects" has not been detected in most cases of midcallosal section. But if the task is made more difficult, so that normal individuals commonly make mistakes, some deficits (as compared with normal subjects) have been elicited (7, 64). Further studies along this line will no doubt be forthcoming; they are clearly needed and the growing popularity of this approach will afford frequent opportunities for both pre- and postoperative testing.
Posterior Callosal (Splenial) Section Whereas more anterior section of the corpus callosum entails minimal long term physiological cost, section of the splenium (Fig. 5.6) commonly causes a left hemialexia (33, 62, 73, 104, 108). This inability to read in the left visual half-field may be accompanied by so-called color anomia, an inability to give the names of colors presented to the patient's view although the colors can be matched and the patient can give (speak or write) the color names of objects, for example, "yellow" when asked the color of a banana (49). When the hemialexia and associated deficits are most severe they can be mistaken for a left homonymous hemianopsia. Although use of a tachistoscope is more precise, it is not necessary because the deficit has often been demonstrated without such equipment. In a recent review (101), hemialexia is considered of two types: an inability to match written words with objects and an inability to read aloud written words or letters. Both of these are said to be mimicked by the condition of visual hemineglect. The foregoing hemideficits may not prove totally disabling for most persons, although they can be so in individuals whose occupations involve seeing large groups of symbols more or less simultaneously, such as a symphonic score. Where the splenial disconnection can be quite crippling for almost any literate person is when the callosal lesion is combined with a left occipital lesion or indeed any lesion causing a right hemianopsia. Such individuals typically have al-exia without agraphia; that is, they can write but are unable to read, even what they have just written correctly to dictation. This remarkable dissociation of reading from writing has been known for nearly a century (6, 35). One explanation is as follows: When the patient has a right homonymous hemianopsia resulting from a left occipital lobe lesion, nothing can be seen, much less read, in the right half-field. Hence, visual information can reach the left hemisphere language zone only from the left half-field via the right occipital cortex and transfer via the splenium. If another (pr confluent) splenial lesion has disconnected the right occipital cortex from the left hemisphere, the left hemisphere retains competence to write to dictation but no longer has access to information arriving in the right occipital lobe from the left visual half-field. As its proponents have recognized, there are some difficulties with this explanation of alexia without agraphia. These patients can often name objects or pictures of objects visualized in the left half-field, showing that information can reach the language zone from the left half-field. More-
Figure 5.6. Posterior callosotomy: In this operation, in addition to the splenium, part of the dorsal hippocampal commissure has been divided as it is usually adherent to the undersurface of the corpus callosum. over, alexia without agraphia can occur without an accompanying loss of the right visual half-field (54, 61). And alexia without agraphia can occur in cases with the splenium largely intact. Reading seems to be a multistage process that can be disturbed in a variety of ways (54, 60, 68). One can readily see that if splenial section were accompanied by left occipital damage the patient might be unable to read altogether and not just be hemialexic. Hence, retraction of the left occipital pole is to be avoided in approaching the splenium, although there may be situations, such as an AVM in the medial aspect of the left occipital lobe, where it seems necessary. In one such personal case, the patient was "cured" of her AVM and was still able to read albeit with difficulty, but she lost entirely her ability, said to have been phenomenal, to sight read music for the piano. Approaching the splenium from the left may be preferable in the rare case where a left-hander has been shown to be a right hemisphere speaker, preferably by use of the Wada carotid amobarbital technique.
Postcallosotomy Mutism After complete section of the cerebral commissures, there was in almost every case a postoperative period of mutism, of varying duration, during which speech was absent or extremely sparse, although comprehension and writing were retained. The patient did not talk even when quite cooperative and having some ability to write. This emissive deficit lasted, in milder form at least, in 11 righthanded patients for nearly 3 months, 3 days, 20 years, 4 weeks, 4 months, 2 days, 4 weeks, 2 weeks, 1 week, 4 weeks, and 4 weeks (see Fig. 5.3). In 1 left-hander (P.D.) mutism lasted 8 months. In all cases, there was little if any paraphasia; when ability to talk returned there was no nominal amnesia; in some cases there was a definite lack of bodily spontaneity and motor initiative for a time, but only partially correlated in duration with the loss of speech. As the mutism subsided, there was a stage of partial recovery that usually included hoarseness or whispering, but without para-
phasia, anomia, or novel semantic or syntactic errors, except for 1 righthander (C.C.) operated from the left. After complete cerebral commissurotomy there was almost always at least some degree of mutism. The notable exception in our series (see Fig. 5.3) was Patient M.K., whose convulsive disorder, beginning on the left side, was associated with a right ventricular dilatation as well as a left footdrop present since childhood. Compensation as a result of this right cerebral atrophic change may have contributed to the absence of postoperative mutism. On the other hand, this was our only case in which a very large bone plate (nearly a hemicraniotomy) was left out and replaced later to minimize pressure effects during postoperative swelling. Clearing of the mutism occurred in all cases but one. This patient (A.M.) continues, even years later, to speak very little—and when he does speak, his speech is poorly phonated and badly articulated. He had severe brain damage before the operation and we believe that he was, to an unusual degree, dependent upon a compensatory function of his corpus callosum for artic-ulatory control. In this case, the prolonged speechlessness was almost certainly not attributable to callosal section alone. This case further illustrates the previously emphasized point that postcommissurotomy deficits depend in large part upon the nature and amount of preexisting, extracallosal damage (11, 52, 100). Postcommissurotomy mutism was originally considered a simple neighborhood sign, a partial akinetic mutism from retraction affecting the anterior end of the third ventricle (29, 90) during section of the anterior commissure. However, a number of patients have now been seen with similar retraction, but a spared splenium, who did not become mute, including N.F. and D.M. (shown in Fig. 5.3). In contrast to the cases of complete commissurotomy, in these (and two other patients, D.B. and B.K.) all of the same structures (including the massa intermedia and the anterior commissure) were severed except the splenium and there was no immediate postoperative mutism (52). Such cases afford evidence not only against a third ventricle origin for the mutism, but also against a right supplementary cortex origin. On the other hand, Ross et al. (91) reported mutism after anterior callosal section but not after posterior callosal section. How to reconcile this apparent conflict is one task of this section, as well as for the future. After complete callosotomy in two stages, Ray-
port et al. (87) observed in three of eight cases a marked decrease in spontaneous speech unaccompanied by paraphasia or comprehension deficit or inability to sing. The authors suggested that, in these most affected cases, mixed hand dominance may have been an important consideration. Of particular importance was the absence of any language or speech problem after the first stage (rostrum, genu, and most of the trunk). The mutism appeared only after the second stage (splenium and remainder of the trunk). This result does not totally eliminate retraction on supplementary motor cortex as a partial contributory cause, but edema probably did not play a role because the second stage followed the first by 2 months, 36 months, and 19 days. Rayport and colleagues have suggested that mutism could result from an interhemispheric conflict (see also Ref. 11). But they emphasize more the aspect of mixed dominance, which might indicate a greater role than usual of the corpus callosum in the production of speech, as was probably the situation in Patient A.M. (Fig. 5.3) and possibly the patient reported by Sussman et al. (102). The role of the corpus callosum in speech may be much greater in certain patients. We can probably be reassured with respect to the risk of severe postoperative mutism by a favorable response to carotid amobarbital testing. This test can give ambiguous results, but it can, we believe, be used to exclude a critical dependence upon the commissures. That is, if the patient continues to speak intelligibly when the hemisphere minor for speech is narcotized, then disconnection of this minor hemisphere will probably not deprive the patient of an essential resource with respect to speech production. Although severe, long term mutism may occur only when callosal section is done in certain patients who are peculiarly dependent for speech on these pathways, there remains to be explained the mutism lasting several weeks in callosotomy patients without either widespread damage or anomalous lateralization from other causes. A diaschisis secondary to deafferentation of speech areas accounts for the deficits in terms of left hemisphere dysfunction even when the left hemisphere is unmolested. This interpretation is consistent with the appearance of a right Babinski sign (a sign of left hemisphere malfunction) and it may help to explain why our patients (and those of Rayport) whose spleniums were spared had no mutism. This means that the diaschisis of the left hemisphere (from a complete section) must affect the speech "centers" or "circuits" more than it affects writing "centers"
or "circuits." This implies, in turn, that the writing function is more robust or resistant in some sense than that for speech, in spite of having developed later in both phylogenesis and ontogenesis. A more speculative possibility is that speech usually requires interhemispheric integration for control of the larynx and other midline structures in the following sense: One can suppose that left hemisphere speech ordinarily includes a corollary discharge (97) to the other hemisphere. When the commissures are completely severed a downstream interhemispheric conflict occurs at the level of the motor nuclei for the larynx, which results in dysphonia. This explanation is consistent with most of the evidence. In particular, it fits the case of one patient (M.K.) who had no mutism at all—this was the patient with right hemisphere atrophy from childhood. It also fits the report of Sussman et al. (102) that their patient uttered several complex sentences shortly after operation although he was otherwise speechless for 16 months. Those concerned with the normal physiology of speech will be more interested in the temporary (several weeks to months) mutism after splenial section (either as part of a total calloso-tomy or as a second stage) than the longlasting mutism (many months to years) of those whose anomalous laterality is associated with longstanding cortical lesions. But for the surgeon using a transcallosal approach to the third ventricle, neither of these is apt to be as important as transient mutism (several days to weeks) after section of callosal segments (such as the trunk) well anterior to the splenium. The mutism reported by Ross et al. (91) lasted 1 to 10 days; hence, it is more relevant to the problems of third ventricle access than the severe, prolonged type of mutism. We have already alluded to the notion that the lack of speech could be considered a mutism of the type often associated with akinesia of either cingulate or subfron-tal origin. This suggests that we think of the postcommissurotomy state as a sort of "forme fruste" of akinetic mutism, in which the mutism is much more evident than is the akinesia. The occurrence of mutism in callosotomy cases without entry into the third ventricle could be explained as the result of subf rontal retraction during section of the rostrum. Arguing against this is the paucity of postoperative mutism in four of our patients (D.M., N.F., D.B., and B.K.) whose splenium was spared but who underwent section of both the rostrum and the anterior commissure under direct vision. The cause for transient mutism may be differ-
ent from one case to the next, and in any given case there may be more than one factor operative, (a) The commissural fibers may have been acting in a compensatory fashion because of previous brain injury, (b) Speech may require some absolute number of relevant brain connections. In a marginal case the callosotomy may be sufficient to bring the number of connections below the minimum required, (c) The callosotomy may produce a diaschisis in the speaking hemisphere, from which it recovers only slowly, (d) The cal-losum may carry a corollary discharge for speech output whose sudden loss results in downstream interhemispheric conflict, (e) Temporary circul-tory alterations (particularly of the internal cerebral veins) may transiently derange the more sensitive functions of the basal ganglia. (f) Damage to one or the other fornix may be contributory, (g) Trauma to the anterior third ventricle during division of the anterior commissure or during division of the rostrum from in front may be relevant in some cases, (h) Supplementary motor cortex dysfunction may be contributory in some cases, (i) Decussating fibers (for example, from medial frontal cortex to opposite basal ganglia) may be relevant so that their interruption plus medial frontal cortex edema combines to produce mutism when associated with a thala-mic or striatal lesion. Concluding Remarks The multifactorial causation of postcalloso-tomy mutism affords a measure of our ignorance, representing as it does the difficulties in predicting the likelihood of this usually transient but nonetheless distressing malfunction. On the other hand, the very existence of this riddle affords an opportunity for someone to unravel it. Likewise, the lack of replicable signs or symptoms from section of the anterior two-thirds of the corpus callosum represents a territory still unknown and awaiting exploration. That the transcallosal approach to the third ventricle is largely without obligatory physiological cost is both a boon to the operator and a challenge to the scientist. Neurosurgeons have historically been both therapeutic and investigative, and we can expect that continued exploitation of the transcallosal approach will benefit our patients both at the time of treatment and by augmenting our understanding of what Bremer (26) once called, "the highest and most elaborate activities of the brain." Acknowledgment I am grateful for library assistance from S. Zeind and staff, including P. Logan, C. DeCicco, L. Walden,
and V. Caullay of the Huntington Memorial Hospital, and for word processing by Sally Johnstone. References 1 . Akelaitis AJ: Studies on the corpus callosum: IV. Diagnostic dyspraxia in epileptics following par tial and complete section of the corpus callosum. Am J Psychiatry 101:594-599, 1944-1945. 2. Alpers BJ, Grant FC: The clinical syndromes of the corpus callosum. Arch Neurol Psychiatry 25:67-86, 1931. 3. Antunes JL, Louis KM, Ganti SR: Colloid cysts of the third ventricle. Neurosurgery 7:450-455, 1980. 4. Apuzzo MLJ, Chikovani OK, Gott PS, Teng EL, Zee CS, Giannotta SL, Weiss MH: Transcallosal, interfornicial approaches for lesions affect ing the third ventricle: Surgical considerations and consequences. Neurosurgery 10:547-554, 1982. 5. Benes V: Sequelae of transcallosal surgery. Childs Brain 9:69-72, 1982. 6. Benson DF: Alexia. In Vinken PJ, Bruyn CW (eds): Handbook of Clinical Neurology. Amster dam, Elsevier-Biomedical, 1985, revised ed vol 1. 7. Bentin S, Sahar A, Moscovitch M: Intermanual information transfer in patients with lesions in the trunk of the corpus callosum. Neuropsychology 22:601 -611, 1984. 8. Berlucchi G: Anatomical and physiological as pects of visual functions of corpus callosum. Brain Res 37:371 -392, 1972. 9. Bogen JE: The other side of the brain: I. Dysgraphia and dyscopia following cerebral commissurotomy. Bull Los Angeles Neurol Soc 34:73105, 1969. 10. Bogen JE: Hemispherectomy and the placing re actions in cats. In Kinsbourne M, Smith WL (eds): Hemispheric Disconnection and Cerebral Function. Springfield, IL, Charles С Thomas, 1974. 1 1 . Bogen JE: Linguistic function in the short term following cerebral commissurotomy. In AvakianWhitaker H, Whitaker HA (eds): Studies in Neurolinguistics, vol 2, New York, Academic Press, 1976, pp 193-224. 12. Bogen JE: The callosal syndrome. In Heilman KM, Valenstein E (eds): Clinical Neuropsychology. New York, Oxford University Press, 1978, pp 308-359. 13. Bogen JE: Concluding overview. In Reeves A (ed): Epilepsy and the Corpus Callosum. New York, Plenum, 1985. 14. Bogen JE: Split-brain syndromes. In Handbook of Clinical Neurology, vol 45, ch 9, pp 1-8. Amsterdam, Elsevier, 1985. 15. Bogen JE: The stabilized syndrome of hemi sphere disconnection. In Benson DF, Zaidel E (eds): The Dual Brain: Hemispheric Speciali zation in the Human. New York, Guilford Press, 1985. 16. Bogen JE, Bogen GM: The other side of the brain: III. The corpus callosum and creativity. Bull Los Angeles Neurol Soc 34:191-220, 1969. 17. Bogen JE, Campbell A, Thompson A: Long-term persistence of unilateral anomia following com-
18. 19.
20.
21. 22. 23.
24.
25. 26. 27. 28. 29. 30.
31.
32. 33.
34. 35.
plete cerebral commissurotomy. Proc Soc Neu-rosci 7:945, 1981. Bogen JE, Fisher D, Vogel PJ: Cerebral commis surotomy: A second case report. JAMA 194:1328-1329, 1965. Bogen JE, Gazzaniga MS: Cerebral commissur otomy in man: Minor hemisphere dominance for certain visuospatial functions. J Neurosurg 23:394-399, 1965. Bogen JE, Sperry RW, Vogel PJ: Commissural section and the propagation of seizures. In Jas per HH, Ward AA, Pope A (eds): Basic Mecha nisms of the Epilepsies. Boston, Little, Brown, 1969. Bogen JE, Vogel PJ: Cerebral commissurotomy in man. Bull Los Angeles Neurol Soc 27:169172, 1962. Bogen JE, Vogel PJ: Treatment of generalized seizures by cerebral commissurotomy. Surg Forum 14:431-433, 1963. Bogen JE, Vogel PJ: Neurologic status in the long term following cerebral commissurotomy. In Michel F, Schott В (eds): Les syndromes de Dis connexion Calleuse chez l,Нотте. Lyon, Hopital Neurologique, 1975, pp 227-251. Botez MI, Bogen JE: The grasp reflex of the foot and related phenomena in the absence of other reflex abnormalities following cerebral commis surotomy. Acta Neurol Scand 54:453-463, 1976. Bradshaw JL, Nettleton NC: Human Cerebral Asymmetry. Englewood Cliffs, NJ, Prentice-Hall, 1983. Bremer F: Physiology of the corpus callosum. Res Publ Assoc Res Nerv Ment Dis 36:424-428, 1956. Bremer F, Brihaye J, Andre-Balisaux G: Physio logie et pathologie du corps calleux. Arch Suisses Neurol Psychiat 78:31-87, 1956. Brion S, Jedynak CP: Les Troubles du Transfert Interhemispherique. Paris, Masson, 1975. Cairns HR: Disturbances of consciousness with lesions of the brainstem and diencephalon. Brain 75:109-146, 1952. Campbell AL, Bogen JE, Smith A: Disorganiza tion and reorganization of cognitive and sensorimotor functions in cerebral commissurotomy: Compensatory roles of the forebrain commis sures. Brain 104:493-511, 1981. Cobben A, Seron X, Gillet J, Bonnal J: Absence de signe de deconnexion lors d'un examen neuropsychologique differe dans quatre cas de lesions callosales anterieures et medianes partielles d'origine vasculaire et neurochirurgicale. Acta Neurol Belg 78:207-216, 1978. Damasio H, Damasio AR: "Paradoxic" ear extinc tion in dichotic listening: Possible anatomic sig nificance. Neurology (NY) 29:644-653, 1979. Damasio AR, Chui HC, Corbett J, Kassel N: Pos terior callosal section in a non-epileptic patient. J Neurol Neurosurg Psychiatry 43:351-356, 1980. Dandy WE: Operative experience in cases of pin eal tumor. Arch Surg 33:19-46, 1936. Dejerine J: Contribution a l etude anatomo-pathologique et clinique des differentes varietes de cecite verbale. С R Seances Memoires Soc Biol
36.
37. 38.
39.
40. 41.
42. 43.
44.
45. 46. 47.
48. 49. 50. 51.
52.
53. 54.
44 (vol 4 of Ser 9) (Second section-Memoires):6190, 1892. Dimond SJ, Scammell RE, Brouwers EYM, Weeks R: Functions of the centre section (trunk) of the corpus callosum in man. Brain 100:543562, 1977. Doty RW, Negrao N: Forebrain commissures and vision. In Jung R (ed): Handbook of Sensory Physiology VII/3. Berlin, Springer-Verlag, 1972. Doty RW, Overman WH: Mnemonic role of forebrain commissures in macaques. In Harnad S, et al (eds): Lateralization in the Nervous Sys tem. New York, Academic Press, 1977. Downer JLC: Changes in visual gnostic func tions and emotional behavior following unilat eral temporal pole damage in the split-brain monkey. Nature 191:50-51, 1961. Efron R, Bogen JE, Yund EW: Perception of dichotic chords by normal and commissurotomized human subjects. Cortex 13:137-149, 1977. Escourolle R, Hauw JJ, Gray F, Henin D: Aspects neuropathologiques des lesions du corps calleux. In Michel F, Schott В (eds):Les Syndromes de Disconnexion Calleuse Chez I'Homme. Lyon, Hopital Neurologique, 1975. Ethelberg S: Changes in circulation through the anterior cerebral artery. Ada Psychiatr Neurol Suppl 75:3-211, 1951. Ferguson SM: Some neuropsychiatric observa tions on behavioral consequences of corpus cal losum section. In Reeves A (ed): Epilepsy and the Corpus Callosum. New York, Plenum, 1985. Galin D, Johnstone J, Nakell L, Herron J: Devel opment of the capacity for tactile information transfer between hemispheres in normal chil dren. Science 204:1330-1332, 1979. Gazzaniga MS, Freedman H: Observations on visual processes after posterior callosal section. Neurology (NY) 23:1126-1130, 1973. Gazzaniga MS, LeDoux JE: The Integrated Mind. New York, Plenum, 1978. Gazzaniga MS, Risse GL, Springer SP, Clark E, Wilson DH: Psychologic and neurologic conse quences of partial and complete cerebral commissurotomy. Neurology (NY) 25:10-15, 1975. Geschwind, N: Disconnexion syndromes in ani mals and man. Brain 88:237-294, 585-644, 1965. Geschwind N, Fusillo M: Color-naming defects in association with alexia. Arch Neurol 15:137146, 1966. Gordon HW: Verbal and Nonverbal Cerebral Processing in Manjor Audition. Thesis, Califor nia Institute of Technology, 1973. Gordon HW: The neurobiological basis of hemisphericity. In Trevarthen С (ed): Brain Circuits and Theories of Mind. Cambridge, Cambridge University Press, 1986. Gordon HW, Bogen JE, Sperry RW: Absence of deconnexion syndrome in two patients with par tial section of the neocommissures. Brain 94:327-336, 1971. Gott PS, Saul RE: Agenesis of the corpus cal losum: Limits of functional compensation. Neurology(NY) 28:1272-1279, 1978. Greenblatt SH: Neurosurgery and the anatomy of reading: A practical review. Neurosurgery 1:6-15, 1977.
55. Greenblatt SH, Saunders RL, Culver CM, Bogdanowicz W: Normal interhemispheric transfer with incomplete section of the splenium. Arch Neurol 37:567-571, 1980. 56. Haaland KY, Delaney HD: Motor deficits after left or right hemisphere damage due to stroke or tumor. Neuropsychologia 19:17-27, 1981. 57. Hamilton CR: Mechanisms of interocular equiv alence. In Ingle D, Goodale M, Mansfield R (eds): Advances in the Analysis of Visual Behavior. Cambridge, MA, MIT Press, 1982, pp 693-717. 58. Hampson E, Kimura D: Hand movement asym metries during verbal and nonverbal tasks. Can J Psychol 38:102-125, 1984. 59. Harbaugh RE, Wilson DH, Reeves AG, Gazzaniga MS: Forebrain commissurotomy for epilepsy: Re view of 20 consecutive cases. Acta Neurochir (Wien) 68:263-275, 1983. 60. Hecaen H, Kremin H: Neurolinguistic research on reading disorders resulting from left hemi sphere lesions: Aphasic and pure alexias. In Whitaker H, Whitaker HA (eds): Studies in Neu rolinguistic s. New York, Academic Press, 1976, pp 861-872. 61. Henderson VW, Friedman RB, Teng EL, Weiner JM: Left hemisphere pathways in reading: Infer ences from pure alexia without hemianopia. Neurol (NY) 35:962-968, 1985. 62. Iwata M, Sugishita M, Toyokura Y, Yamada R, Yoshioka M: Etude sur le syndrome de discon nexion visuo-linguale apres la transection du splenium du corps calleux. J Neurol Sci 23:421432, 1974. 63. Jason GW: Hemispheric asymmetries in motor function: I. Left-hemisphere specialization for memory but not performance. Neuropsychologie 21:35-45, 1983. 64. Jeeves MA: Some limits to interhemispheric in tegration in cases of callosal agenesis and partial commissurotomy. In Russell IS, von Hof MW, Berlucchi G (eds): Structure and Function of Cerebral Commissures. Baltimore, University Park Press, 1979, ch 37. 65. Kimura D: Functional asymmetry of the brain in dichotic listening. Cortex 3:163-178, 1967. 66. Kimura D, Archibald Y: Motor functions of the left hemisphere. Brain 97:337-350, 1974. 67. Kumar S: Short-term memory for a non-verbal tactual task after cerebral commissurotomy. Cortex 13:55-61, 1977. 68. Landis T, Regard M, Serrat A: Iconic reading in a case of alexia without agraphia caused by brain tumor: A tachistoscopic study. Brain Lang 11:45-53, 1980. 69. Levy J: Cerebral asymmetries as manifested in split-brain man. In Kinsbourne M, Smith WL (eds): Hemispheric Disconnection and Cerebral Function. Springfield, IL, Thomas, 1974. 70. Levy-Valensi J: Le Corps Calleux (Paris Theses 448). Paris, G Steinheil, 1910. 71. Liepmann H, Mass O: Fall von linksseitiger Agraphie und Apraxie bei rechtsseitiger Lahmung. J Psychol Neurol 10:214-227, 1908. 72. Long DM, Leibrock L: The transcallosal ap proach to the anterior ventricular system and its application in the therapy of cran/opharyngioma. Clin Neurosurg 27:160-168, 1980. 73. Maspes PE: Le syndrome experimental chez
74.
75.
76.
77.
78.
79.
80. 81. 82. 83. 84.
85.
86.
87.
88. 89.
90.
1'homme de la section du splenium du corps calleux: Alexie visuelle pure hemianopsique. Rev Neural 80:100-113, 1948. McKeever WF, Sullivan KF, Ferguson SM, Rayport M: Typical cerebral hemisphere disconnec tion deficits following corpus callosum section despite sparing of the anterior commissure. Neu ropsychologia 19:745-755, 1981. McKeever WF, Sullivan KF, Ferguson SM, Rayport M: Right hemisphere speech development in the anterior commissure-spared commissurotomy patient: A second case. Clin Neuropsychol4:l7-22, 1982. Michel F, Peronnet F: Extinction gauche au test dichotique: Lesion hemispherique ou lesion commissurale? In Michel F, Schott В (eds): Les Syn dromes de Disconnexion Calleuse chez I'Homme. Lyon, Hopital Neurologique, 1975. Milner AD, Jeeves MA: A review of behavioral studies of agenesis of the corpus callosum. In Russell IS, van Hof MW, Berlucchi G (eds): Struc ture and Function of Cerebral Commissures. Baltimore, University Park Press, 1979. Milner B: Analysis of memory disorder after cer ebral commissurotomy. In Trevarthen С (ed): Es says in Honour of R. W. Sperry. Cambridge, Cambridge University Press, 1985. Milner B, Taylor L, Sperry RW: Lateralized sup pression of dichotically presented digits after commissural section in man. Science 161:184186, 1968. Mingazzini G: Der Balken. Berlin, Springer, 1922. Myers JJ: Right hemisphere language: Science or fiction? Am Psychol 39:315-320, 1984. Myers RE: Function of the corpus callosum in interocular transfer. Brain 79:358-363. Nebes RD: Hemispheric specialization in commissurotomized man. Psychol Bull 81:1-14, 1974. Ozgur MH, Johnson T, Smith A, Bogen JE: Transcallosal approach to third ventricle tumor: Case report. Bull Los Angeles Neurol Soc 42:5762, 1977. Preilowski BFB: Possible contribution of the an terior forebrain commissures to bilateral motor coordination. Neuropsychologie 10:267-277, 1972. Rayport M, Ferguson SM, Corrie WS: Outcomes and indications of corpus callosum section for intractable seizure control. Appl Neurophysiol 46:47-51, 1983. Rayport M, Corrie WS, Ferguson SM: Results of two-stage corpus callosum section for seizure control in clinically and electroencephalographically defined cases. In Reeves A (ed): Epilepsy and the Corpus Callosum. New York, Plenum, 1985. Reeves A (ed): Epilepsy and the Corpus Cal losum. New York, Plenum, 1985. Risse GL, LeDoux J, Springer SP, Wilson DH, Gazzaniga MS: The anterior commissure in man: Functional variation in a multisensory system. Neuropsychologia 16:23-31, 1978. Ross ED, Stewart RM: Akinetic mutism from hypothalamic damage: Successful treatment with dopamine agonists. Neurology (NY) 31:1435-1439, 1981.
91. Ross MK, Reeves AG, Roberts DW: Postcommissurotomy mutism. Ann Neurol 16:114, 1984. 92. Rovit RL, Gloor P, Rasmussen T: Sphenoidal electrodes in the electrographic study of patients with temporal lobe epilepsy. J Neurosurg 18:151-158, 1961. 93. Sidtis JJ, Volpe ВТ, Wilson DH, Rayport M, Gaz zaniga MS: Variability in right hemisphere lan guage function after callosal section: Evidence for a continuum of generative capacity. J Neurosci 1:323-331, 1981. 94. Shucart WA, Stein BM: Transcallosal approach to the anterior ventricular system. Neurosurgery 3:339-343, 1978. 95. Sparks R, Geschwind N: Dichotic listening in man after section of neocortical commisures. Cortex 4:3-16, 1968 96. Sparks R, Goodglass H, Nickel B: Ipsilateral ver sus contralateral extinction in dichotic listening from hemispheric lesions. Cortex 6:249-260, 1970. 97. Sperry RW: Neural basis of the spontaneous optokinetic response produced by visual inversion. J Comp Physiol Psychol 43:482-489, 1950. 98. Sperry RW: Cerebral organization and behavior. Science 133:1749-1757, 1961. 99. Sperry RW: Lateral specialization in the sur gically separated hemispheres, in Schmitt FO, Worden FG (eds): Neuroscience 3rd Study Prog. Cambridge, MIT Press, 1974. 100. Sperry RW, Gazzaniga MS, Bogen JE Interhemispheric relationships: The neocortical commis sures. Syndromes of hemisphere disconnection. Handbook Clin Neurol 4:273-290, 1969. 101. Sugishita M, Shinohara A, Shimoji T: Does a posterior lesion of the corpus callosum cause hemialexia? In Reeves A (ed): Epilepsy and the Corpus Callosum. New York, Plenum, 1985. 102. Sussman NM, Gur RC, Gur RE, O'Connor MJ: Mutism as a consequence of callosotomy. J Neu rosurg 59:514-519, 1983. 103. Teng EL: Dichotic ear difference is a poor index for the functional asymmetry between the cere bral hemispheres. Neuropsychologia 19:235240, 1981. 104. Trescher HH, Ford FR: Colloid cyst of the third ventricle: Report of a case. Operative removal with section of posterior half of the corpus cal losum. Arch Neurol Psychiatry 37:959-973, 1937. 105. Trevarthen C: Hemispheric specialization. In Handbook of Physiology—The Nervous Sys tem HI. Washington DC, American Society of Physiology, 1984, pp 1129-1190. 106. Unterharnscheidt F, Jalnik D, Gott H: Der balkenmangel. Monograph Ges Neurol Psychiat 128:1-232, 1968. 107. Van Gijn J: The Plantar Reflex. Rotterdam, Krips Repro-Meppel, 1977. 108. Wechsler AF: Transient left hemialexia. Neurol ogy (NY) 22:628-633, 1972. 109. Weinstein EA, Friedland RP: Hemi-Inattention and Hemisphere Specialization. New York, Raven Press, 1977. 110. Wilson DH, Reeves A, Gazzaniga M, Culver C: Cerebral commissurotomy for control of intract able seizures. Neurology (NY) 27:708-715, 1977.
1 1 1 . Winston KR, Cavazzuti V, Arkins T: Absence of neurological and behavioral abnormalities after anterior transcallosal operation for third ven tricular lesions. Neurosurgery 4:386-393, 1979. 112. Yamamoto I, Rhoton AL, Peace DA: Microsurgery of the third ventricle: Part 1. Neurosurgery 8:334-356, 1981. 113. Yakovlev PI, Lecours AR: The myelogenetic cycles of regional maturation of the brain. In Minkowski (ed): Regional Development of the Brain in Early Life. Edinburgh, Blackwell, 1967. 114. Zaidel D, Sperry RW: Memory impairment after commissurotomy in man. Brain 97:263-272, 1974. 115. Zaidel D, Sperry RW: Some long-term motor ef-
fects of cerebral commissurotomy in man. Neuropsychologia 15:193-204, 1977. 116. Zaidel E: Linguistic competence and related functions in the right hemisphere of man follow ing cerebral commissurotomy and hemispherectomy. Ph.D. thesis, California Institute of Technology. Dissertation Abstr Intern 34: 2350B (University Microfilms 73-26,481). 117. Zaidel E: Lexical organization in the right hemi sphere. In Buser P, Rougeul-Buser A (eds): Cer ebral Correlates of Conscious Experience. Am sterdam, Elsevier, 1978. 118. Zaidel E: Disconnection syndrome as a model for laterality effects in the normal brain. In Hellige J (ed): Cerebral Hemisphere Asymmetry: Meth ods, Theory and Application. New York, Praeger, 1983. ~
6 Pathological Correlates of Amnesia and the Anatomical Basis of Memory Antonio R. Damasio, M.D., Ph.D., and Gary W. Van Hoesen, Ph.D.
The early understanding of the structural basis of human memory derived from the anatomical study of two distinct groups of patients. The first comprised patients with Wernicke-Korsakoff's psychosis. The second was made up of patients with temporal lobectomies performed to alleviate seizures refractory to medical treatment. The current views on the anatomy of memory draw from a much larger base of data, including not only the study of such classical cases, but also recently described partial and global amnestic syndromes associated with focal lesions in (a) the basal forebrain, (b) the lateral and polar temporal cortices, and (c) the association cortices of separate sensory modalities. Additionally, the study of senile dementia of the Alzheimer type is contributing important understanding in this area. In this chapter we review these recent developments and indicate why they enlarge our working concept of the relation between brain structure and memory. We also review the anatomy of the system most frequently involved in memory disorders, the limbic system, and note some fundamental aspects of the neuropsychological testing of the amnesias. Amnesia Associated with Medial Temporal and Diencephalic Damage Decades of histopathological study of alcoholic patients with amnesia have revealed damage in diencephalic and mesencephalic regions. The consistent lesions in the hypothalamus, espe-
cially in mamillary bodies, and the dorsomedial nuclei of the thalamus have been related to the memory impairment. Considerable controversy has surrounded the issue of whether the mamillary bodies or the thalamic lesions are crucial in causing the amnesia (41). Our view is that, if the damage is bilateral, both loci can be associated with memory impairment. Quite commonly, this is not even an issue as both loci are involved. The reasonable conclusion is that these structures are part of a system of cortical and sub-cortical neural units necessary for learning, recall, and recognition (11, 86). Future and more detailed analysis of this type of data may well result in the inclusion, as part of this system, of additional structures in the high, paramidline brain stem. Nonetheless, the currently available correlations are quite unequivocal, so long as it is kept in mind that extensive neurochemical changes accompany this disorder and undoubtedly play a crucial role. The work on temporal lobectomies has also led to unequivocal correlations. Patients with extensive, unilateral, temporal excisions develop a partial amnesia that affects predominantly verbal or nonverbal material depending on whether the lobectomy is placed in the dominant or non-dominant hemisphere (45). The inclusion, in the excision, of parahippocampal cortex, hippocampus, and amygdala seems necessary for the appearance of amnesia (93). On the other hand, patients with bilateral damage of the medial temporal lobes encompassing the hippocampal formation, the overlying parahippocampal gyrus,
and part of the amygdala developed a major an-terograde learning defect (65). The best-studied subject with such a lesion, Patient H.M., was operated on in the 1950s and still gives no evidence of learning, at a conscious level, any material that he has come in contact with since then (15, 73). This anterograde amnesia covers material learned in any sensory modality. The exception is the learning of some visuomotor skills, which he has clearly mastered without being aware that learning has taken place (13). H.M. can neither recall nor recognize anterograde material of contextual and generic types. The terms generic or semantic used in association with memory denotes knowledge about the basic properties of a stimulus, such as class membership and function. Contextual or episodic memory denotes knowledge of a specific stimulus, its relationship to other stimuli in a given situation, and its relationship to the perceiver and may contain a temporal element. The concepts of episodic and semantic memory were proposed by Tulving (74). The largely equivalent terms contextual and generic are our own. This anterograde defect covers both verbal and nonverbal information. In contrast to this pervasive anterograde defect, H.M. has only a small retrograde memory impairment. The patient retains normal memory for persons, objects, and experiences up to the age of 16 years, when he started having seizures. That means that his contextual and generic memories are normal up to that age. For the period beyond the age of 16 years, there is a recently uncovered contextual memory impairment, noticeable when the patient is compared with controls (14). The defect is far less marked than the anterograde impairment. Both H.M. and the patients with alcoholic Kor-sakoff's syndrome have normal perception, language, and motor function and preserved intellectual skills that allow them, both in real situations and in laboratory testing, to exhibit a large range of intelligent behaviors (11, 15). Although some cognitive strategies are impaired, the disorder of recall, recognition, and learning dominates the clinical picture. The only other impairment of significance in these amnestic patients has to do with emotion and affect. Most of these patients have some disturbance of affect, emotional display, or motivation. But these disturbances are mild in comparison to the memory disorder (11, 15). Both H.M. and the patients with alcoholic Kor-sakoff's syndrome preserve their knowledge of skills and are able to acquire new visuomotor skills. Their "procedural" memory and "proce-
dural learning" are spared, i.e., their knowledge of "how" to act in some specific circumstances is preserved. This contrasts with their "declarative" memory and "declarative" learning, i.e., their knowledge of "that," which is impaired (12). We might take this qualification a bit further by saying that, in the retrograde compartment of their memory, only the contextual component of declarative knowledge is disturbed, whereas the generic component of declarative knowledge is intact. None of these patients has difficulty recognizing the myriad separate items that compose their universe in their essential, generic nature or the function of those components, i.e., they can recognize any human face as a face, any dog as a dog, any house or airplane as such, the beauty of a landscape as beauty indeed. Their disturbance pertains only to those memories in which a particular person, animal, object, quality, or experience must be placed in the context of the subject who was previously in contact with them, i.e., assigned a specific relation to the beholder and maybe even a temporal position. The hallmark of these amnestic disorders is the contrast between the vast store of intact generic knowledge and intellectual skills and the gross impairment of contextual knowledge (17). Recent Developments Amnesia after Medial and Lateral Temporal Lobe Damage Recent developments in the understanding of the anatomical basis of amnesia include the description of cases in which the profile of amnesia and the anatomical lesions correlated with it were considerably different. One of the most significant is the description of a global amnestic syndrome, similar to that seen in H.M., but with a far more pervasive retrograde amnesia encompassing all previously acquired contextual memory (18). In addition to the predictable bilateral medial temporal lobe lesions, that syndrome is associated with bilateral lesions in both lateral aspects of the temporal lobe including its pole, bilateral insular cortices, and bilateral lesions in the basal forebrain region. The cause is herpes simplex encephalitis. Further investigations are necessary to establish unequivocal correlations between some features of the syndrome, especially the overwhelming retrograde amnesia, and the different sets of paired lesions seen in such cases. But a reasonable working hypothesis is that damage to the polar and lateral temporal cortices and the insular cortex is the best corre-
late of the retrograde amnesia. Those are the areas conspicuously spared in amnestic patients who have a largely preserved retrograde memory, such as H.M. Amnesia after Basal Forebrain Damage Another important development is the description of an amnestic syndrome encompassing an-terograde and retrograde memory, milder than that reported with temporal lobe damage and related exclusively to contextual material, caused by unilateral or bilateral lesions of the basal forebrain area (18, 85). This is the area that includes the substantia innominata (which contains a large part of the nucleus basalis of Mey-nert), the diagonal band nuclei, the nucleus ac-cumbens and septal nuclei, and part of the for-nix. These lesions are caused by vascular damage in the territories supplied by the anterior communicating artery, anterior cerebral artery, and recurrent artery of Heubner. The immediate cause is rupture of aneurysms in one of these arteries. The exact pathological mechanism for this damage is varied— infarction due to ischemia in a vascular territory subject to sudden loss of blood flow, infarction due to embolism from the aneurysmal sac, or direct damage by hemorrhage from ruptured aneurysm. Not uncommonly, there is additional damage in one of the orbitofrontal regions (especially in gyrus rec-tus) and occasionally in the medial frontal lobe, especially in the subcallosal portion of the cin-gulate gyrus. The amnesia is more or less pure depending on how much additional damage of this type is present. A component of the basal forebrain, the nucleus basalis of Meynert, is the site of some changes in its cholinergic neurons in Alzheimer's disease (91) and these changes might be related, in part, to the amnesia seen in Alzheimer's patients. This area is of obvious special importance to neurosurgeons engaged in the treatment of aneurysms or in the management of third ventricle tumors. It is apparent that damage in this region, whether or not it involves the fornix, is likely to cause memory impairment. Because that impairment may be only partial, the potential for missing its presence or minimizing its importance is considerable. Such impairments may cast a major shadow on what might otherwise be an uneventful recovery from a successful operation. Amnesia in Alzheimer's Disease No less important has been the description of specific and preponderant pathological involvement of the entorhinal cortex and subiculum in
Alzheimer's disease (31). The entorhinal cortex is located along the anterior part of the parahip-pocampal gyrus. The subiculum is a component of the hippocampal formation. Both entorhinal and subicular cortices serve as staging areas for input to and output from the hippocampal formation. The net effect of such lesions is to disconnect the hippocampus from the cerebral cortex and from a variety of subcortical structures in the limbic system and diencephalon. This isolation is incompatible with normal function and must be no less devastating than surgical resection or vascular or encephalitic damage. In practice, after such lesions develop, the hippocampus will no longer be able to operate on the basis of the multimodal cortical inputs that normally reach it via the perforant pathway that originates in the superficial layers of the entorhinal cortex (32). As if that were not impairment enough, however, the lesions of Alzheimer's disease also block the exit of outputs at the level of the subiculum. These selective lesions are likely to be a major cause, perhaps the principal cause, of the contextual amnesia that hallmarks the early phase of Alzheimer's disease. Other Developments Other important new findings in the field of human amnesia include: (a) the realization that syndromes such as that of visual agnosia can be understood as disorders of visually triggered memory that impair the recognition, recall, and learning of certain types of visual stimuli (16, 20, 21); (b) the description of defects in learning related to unilateral lesions capable of compromising the processing of information conveyed from one sector of the body alone (i.e., a learning defect of material presented in one hand only (61, 62)); (c) the description of cases of amnesia with lesions of the medial thalamic nuclei caused by stroke or by direct injury (25, 70); and (d) the clarification of a question regarding the role of the temporal stem in the development of amnesia caused by medial temporal lobe damage. The hypothesis that the amnesia associated with medial temporal lesions might be related to section of the temporal stem rather than damage to the hippocampus proper has been posited (30). Recent experimental work in the monkey, specifically designed to test the temporal stem hypothesis, failed to support the claim (94). Furthermore, because the temporal stem carries some of the outputs from the hippocampus, it would be only natural that its damage might lead to amnesia. That certainly does not disqualify the hippocampus from a role in memory.
Anatomical Underpinnings of Memory Reflection on this vast body of classical and recent anatomical and cognitive findings permits the elaboration of a more comprehensive list of neural systems crucial to memory processing. It also permits some