Leading Edge
A human characteristic is our willingness to travel vast distances to gather in one place, and for any number of reasons, including warfare, religion, politics, sport, mourning, and entertainment. Modern transport has allowed such gathering to increase in their number of participants and in the speed at which people arrive and disperse to all parts of the globe. Such mass gatherings bring with them the risk of spread of infectious diseases—both at the time of the event and when participants return to their homes—and other health hazards such as stampedes or crush injuries, burns, heat exhaustion, dehydration, trauma, and the potential for environmental and public health hazards. Government bodies that host such mass gatherings must provide food, water, sanitation, shelter, and health security for potentially millions of visitors. Perhaps the largest single mass gathering is the Hindu festival of Kumbh Mela in India. The 2010 event in Haridwar attracted around 50 million pilgrims between January and April, with an estimated 16 million present at the climax of the festival on April 14. Outbreaks of infectious diseases at the Kumbh Mela have been known since cholera was reported in the late 19th century, and deaths following stampedes have happened several times, including an incident that killed five people in 2010. Other notable mass gatherings in 2010 include the World Expo in Shanghai, China, which will attract an estimated 70 million visitors by the time it ends in November, and the FIFA World Cup in South Africa in June and July, during which more than 1 million foreign visitors entered the country (although 2 million people are reported to have gathered in Madrid, Spain, for a parade celebrating the victory of the Spanish team). The largest annual mass gathering is for the Islamic Hajj pilgrimage to Mecca, Saudi Arabia. The number of people attending this event has doubled in the past decade, reaching 2·5 million in 2009, of whom at least 1·6 million were foreign visitors. Ensuring the safety of such huge numbers of people has not always proved easy. Stampedes have caused thousands of deaths in the past, but an investment of US$1·1 billion by the government of Saudi Arabia has eliminated the problem of stampedes since 2007. The first international outbreaks of meningitis related to the Hajj occurred during 1987, and there have been similar incidents www.thelancet.com/infection Vol 10 October 2010
since. However, visitors entering Saudi Arabia for the Hajj are now required to have been vaccinated against meningococcal meningitis (serogroups A, C, Y, and W135 for children over 2 years and adults, and serogroup A for children aged 3 months to 2 years), a policy that has eliminated outbreaks. Saudi Arabian authorities probably, therefore, have more experience in managing mass gatherings than any other national government. It is appropriate then that a Lancet conference on mass gatherings medicine is being held in collaboration with the Kingdom of Saudi Arabia Ministry of Health in Jeddah, Saudi Arabia, on Oct 23–25, 2010. The chairs of the meeting are the Editor of this journal, and Ziad A Memish, the Assistant Deputy Minister of Health for Preventive Medicine. We thank Shahul Ebrahim, Robert Steffen, Maurizio Barbeschi, and Jaouad Mahjour for their invaluable input into arranging the programme and identifying speakers. This conference will be—to our knowledge—the largest such international convention of experts on the health and biosecurity issues related to mass gatherings medicine. Mass gatherings pose complex health-care challenges related to communicable diseases surveillance, vaccination, travel medicine, environmental health, emergency preparedness, crowd management, and national and international security. The meeting will address the issues arising from these challenges through discussion of health-care delivery, the public health risks to countries whose citizens attend mass gatherings, the surveillance needs in both host and home countries, immunisation and other preventive health needs for participants in mass gatherings, the implications of mass gatherings for implementation of the International Health Regulations, and the research opportunities that such gatherings present. One of the outcomes of the meeting will be a series of papers in this journal reviewing the health-care issues related to mass gatherings. The next Hajj will be in mid-November 2010, shortly after the Lancet conference. Our hope is that the conference will serve to disseminate the experience of those professionals accomplished in managing events such as the Hajj, and as a starting point for more internationally coordinated research into mass gatherings medicine. ■ The Lancet Infectious Diseases
Das Fotoarchiv/Still Pictures
Mass gatherings medicine
See Cross-talk page 662 For more on The Lancet Conference on Mass Gatherings Medicine see http://conferences.thelancet. com/massgatherings/
653
Reflection and Reaction
Maximising the public health benefit of antimalarials Published Online September 9, 2010 DOI:10.1016/S14733099(10)70192-4 See Articles page 673
654
Artemisinin combination therapies (ACTs) are integral to current malaria treatment strategies. They effect rapid and complete clearance of multidrug-resistant strains of Plasmodium spp, decrease the transmission potential of the parasite, and limit the emergence of de novo resistance. In The Lancet Infectious Diseases today, Smithuis and colleagues1 report the results of a factorial, open-label, randomised controlled trial at three sites in Myanmar (Burma) in which 808 patients with uncomplicated Plasmodium falciparum malaria were first allocated to one of four fixed-dose ACTs (artesunate–mefloquine, artesunate–amodiaquine,dihydroartemisinin–piperaquine, or artemether–lumefantrine) or a loose combination of artesunate and mefloquine. The study shows that all of the combinations are associated with rapid parasite clearance and, with the exception of artesunate–amodiaquine, low rates of recrudescence. How, therefore, should policymakers decide between one highly efficacious regimen and another? The primary goal of antimalarial treatment should always be to safely eradicate asexual parasites from the host. After this, the most important consideration is to maximise the benefit to public health as gauged by a regimen’s ability to limit the emergence and spread of antimalarial resistance and decrease transmission to the mosquito vector. However, the empirical assessment of these properties is not straightforward. Compared with loose formulations, fixed-dose ACTs have the potential to improve adherence to curative treatment and reduce the risk of parasite exposure to unpartnered artemisinin derivatives, therefore prolonging the useful lifetime of these important drugs. Of the fixed-dose combinations, those with longacting partner drugs such as mefloquine or piperaquine also provide an extended period of post-exposure prophylaxis, reducing the overall frequency of recurrent infections and allowing greater time for haematological recovery.2 However, this benefit might be countered by a long tail of subtherapeutic drug concentrations promoting greater selective pressure for the spread of resistant parasites.3 In individuals who are not immune, mature infectious P falciparum gametocytes are usually detected after the onset of symptoms.4 Variations in the gametocytocidal
activity of antimalarial regimens might therefore translate to tangible differences in the probability that a given infection is transmitted. Artemisinin derivatives have potent activity against stage I–III gametocytes but are ineffective against mature, infectious forms. In the study by Smithuis and colleagues, gametocyte carriage was almost three-times higher after dihydroartemisinin– piperaquine compared with fixed-dose artesunate– mefloquine,1 a finding that has now been replicated in several randomised trials.5–7 The relative transmissibility of these gametocytes is unknown. Primaquine has activity against all gametocyte stages and can substantially reduce the risk of P falciparum gametocyte carriage.8 In a second round of randomisation, the authors allocated patients to schizonticidal treatment with or without a single 0·75mg base/kg dose of primaquine. Since gametocytes are non-pathogenic and primaquine is not active against asexual stages of P falciparum, this treatment had no potential to directly benefit the individuals in the study. Impressively, Smithuis and colleagues show that this dose of primaquine almost completely eliminates the risk of subsequent gametocyte carriage, therefore negating the differences in the gametocytocidal activity of the ACTs. Concerns remain about the potential for a one-off dose of primaquine to induce haemolysis, especially in glucose-6-phosphate dehydrogenase (G6PD) deficient individuals.9 Other than a mild truncation of haematological improvement after treatment with primaquine, the investigators did not detect any significant adverse effects, although this was not the primary endpoint and the prevalence of G6PD deficiency in the study population was unknown.1 A single dose of primaquine has no effect against the hypnozoites of Plasmodium vivax and indeed heterologous recurrence with vivax malaria was common in this study.1 Previous work suggests that this finding is likely to be due to a substantial, but underappreciated, burden of concomitant blood-stage P vivax infection in patients with falciparum malaria in coendemic regions.10,11 Although Smithuis and colleagues show that the long-acting ACTs, artesunate– mefloquine and dihydroartemisinin–piperaquine, are associated with lower rates of heterologous recurrence,1 this suggests suppression of the first of several potential www.thelancet.com/infection Vol 10 October 2010
Reflection and Reaction
liver-stage relapses and thus post-treatment prophylaxis might offer only temporary relief.12 Such high rates of co-infection provide a strong rationale for a unified treatment strategy in which all G6PD-replete patients with microscopically confirmed malaria in regions where P falciparum and P vivax are both endemic receive a full course of primaquine to sterilise circulating gametocytes and rid the liver of P vivax hypnozoites.
3 4
5
6
7
*Ric N Price, Nicholas M Douglas Global Health Division, Menzies School of Health Research, Charles Darwin University, Darwin, NT, Australia; and Centre for Clinical Vaccinology and Tropical Medicine, Nuffield Department of Clinical Medicine, University of Oxford, Oxford, UK
[email protected]
8
We declare that we have no conflicts of interest. RNP is a Wellcome Trust Senior Research Fellow in Clinical Science (091625). NMD is funded by the Rhodes Trust.
10
1
11
2
Smithuis F, Kyaw MK, Phe O, et al. Effectiveness of five artemisinin combination regimens with or without primaquine in uncomplicated falciparum malaria: an open-label randomised trial. Lancet Infect Dis 2010; published online Sept 9. DOI:10.1016/S1473-3099(10)70187-0. Ratcliff A, Siswantoro H, Kenangalem E, et al. Two fixed-dose artemisinin combinations for drug-resistant falciparum and vivax malaria in Papua, Indonesia: an open-label randomised comparison. Lancet 2007; 369: 757–65.
9
12
Hastings IM, Watkins WM. Tolerance is the key to understanding antimalarial drug resistance. Trends Parasitol 2006; 22: 71–77. Stepniewska K, Price RN, Sutherland CJ, et al. Plasmodium falciparum gametocyte dynamics in areas of different malaria endemicity. Malar J 2008; 7: 249. Smithuis F, Kyaw MK, Phe O, et al. Efficacy and effectiveness of dihydroartemisinin-piperaquine versus artesunate-mefloquine in falciparum malaria: an open-label randomised comparison. Lancet 2006; 367: 2075–85. Bassat Q, Mulenga M, Tinto H, et al. Dihydroartemisinin-piperaquine and artemether-lumefantrine for treating uncomplicated malaria in African children: a randomised, non-inferiority trial. PLoS One 2009; 4: e7871. Valecha N, Phyo AP, Mayxay M, et al. An open-label, randomised study of dihydroartemisinin-piperaquine versus artesunate-mefloquine for falciparum malaria in Asia. PLoS One 2010; 5: e11880. Shekalaghe S, Drakeley C, Gosling R, et al. Primaquine clears submicroscopic Plasmodium falciparum gametocytes that persist after treatment with sulphadoxine-pyrimethamine and artesunate. PLoS One 2007; 2: e1023. Shekalaghe SA, ter Braak R, Daou M, et al. In Tanzania, hemolysis after a single dose of primaquine coadministered with an artemisinin is not restricted to glucose-6-phosphate dehydrogenase-deficient (G6PD A–) individuals. Antimicrob Agents Chemother 2010; 54: 1762–68. Looareesuwan S, White NJ, Chittamas S, Bunnag D, Harinasuta T. High rate of Plasmodium vivax relapse following treatment of falciparum malaria in Thailand. Lancet 1987; 2: 1052–55. Mason DP, Krudsood S, Wilairatana P, et al. Can treatment of P vivax lead to a unexpected appearance of falciparum malaria? Southeast Asian J Trop Med Public Health 2001; 32: 57–63. Douglas NM, Anstey NM, Angus BJ, Nosten F, Price RN. Artemisinin combination therapy for vivax malaria. Lancet Infect Dis 2010; 10: 405–16.
Reduction of HIV incidence in men who have sex with men Despite great advances in the collective understanding of the HIV epidemic, men who have sex with men (MSM) continue to be disproportionately affected. In the USA, the Centers for Disease Control and Prevention estimated that more than 55000 new infections occur every year and more than half of these infections are in MSM.1 Similarly, in Canada, MSM are proportionally most affected.2,3 Alarmingly, in western Europe the number of new HIV diagnoses in MSM nearly doubled from 2538 in 1999 to 5016 in 2006,4 and, after infections from immigrants from other countries are accounted for, MSM had the highest number of new infections of any transmission group in western Europe in 2006.5 Elsewhere, significant increases in new HIV infections in MSM have been reported in Asian countries, such as China where the prevalence of HIV infection in MSM in some major cities seems to be rapidly increasing.6 HIV prevalence for MSM is high throughout Latin America, Russia, and some countries in sub-Saharan Africa.7 Data for other regions are often difficult to find, but the number of new infections in MSM is very high in most regions of the world. www.thelancet.com/infection Vol 10 October 2010
In The Lancet Infectious Diseases today, Le Vu and colleagues8 provide further evidence that the number of new infections in MSM is unacceptably high. Although the overall rate of new HIV infections decreased in France from 8930 in 2003 to 6940 in 2008, the number of new infections in MSM was stable. In 2008, 3320 (48%) of new infections in France were attributed to MSM. Incidence of HIV infections was low and stable in 2008 in IDUs, accounting for 1% of new infections. The pattern of the HIV epidemic reported in France is very similar to that in British Columbia, Canada.9 After introduction of highly active antiretroviral therapy (HAART) in 1996, the number of new positive HIV tests in the province fell by about 50%.9 Much of the early reduction in new HIV cases was in MSM. However, the number of new HIV positive tests in MSM has remained largely unchanged in recent years, especially since 2003, with about 150–180 new infections reported every year.10 By contrast, a reduction in the number of new HIV infections was observed in British Columbia in IDUs after use of HAART increased from 2004.10
Published Online September 9, 2010 DOI:10.1016/S14733099(10)70200-0 See Articles page 682
655
Reflection and Reaction
liver-stage relapses and thus post-treatment prophylaxis might offer only temporary relief.12 Such high rates of co-infection provide a strong rationale for a unified treatment strategy in which all G6PD-replete patients with microscopically confirmed malaria in regions where P falciparum and P vivax are both endemic receive a full course of primaquine to sterilise circulating gametocytes and rid the liver of P vivax hypnozoites.
3 4
5
6
7
*Ric N Price, Nicholas M Douglas Global Health Division, Menzies School of Health Research, Charles Darwin University, Darwin, NT, Australia; and Centre for Clinical Vaccinology and Tropical Medicine, Nuffield Department of Clinical Medicine, University of Oxford, Oxford, UK
[email protected]
8
We declare that we have no conflicts of interest. RNP is a Wellcome Trust Senior Research Fellow in Clinical Science (091625). NMD is funded by the Rhodes Trust.
10
1
11
2
Smithuis F, Kyaw MK, Phe O, et al. Effectiveness of five artemisinin combination regimens with or without primaquine in uncomplicated falciparum malaria: an open-label randomised trial. Lancet Infect Dis 2010; published online Sept 9. DOI:10.1016/S1473-3099(10)70187-0. Ratcliff A, Siswantoro H, Kenangalem E, et al. Two fixed-dose artemisinin combinations for drug-resistant falciparum and vivax malaria in Papua, Indonesia: an open-label randomised comparison. Lancet 2007; 369: 757–65.
9
12
Hastings IM, Watkins WM. Tolerance is the key to understanding antimalarial drug resistance. Trends Parasitol 2006; 22: 71–77. Stepniewska K, Price RN, Sutherland CJ, et al. Plasmodium falciparum gametocyte dynamics in areas of different malaria endemicity. Malar J 2008; 7: 249. Smithuis F, Kyaw MK, Phe O, et al. Efficacy and effectiveness of dihydroartemisinin-piperaquine versus artesunate-mefloquine in falciparum malaria: an open-label randomised comparison. Lancet 2006; 367: 2075–85. Bassat Q, Mulenga M, Tinto H, et al. Dihydroartemisinin-piperaquine and artemether-lumefantrine for treating uncomplicated malaria in African children: a randomised, non-inferiority trial. PLoS One 2009; 4: e7871. Valecha N, Phyo AP, Mayxay M, et al. An open-label, randomised study of dihydroartemisinin-piperaquine versus artesunate-mefloquine for falciparum malaria in Asia. PLoS One 2010; 5: e11880. Shekalaghe S, Drakeley C, Gosling R, et al. Primaquine clears submicroscopic Plasmodium falciparum gametocytes that persist after treatment with sulphadoxine-pyrimethamine and artesunate. PLoS One 2007; 2: e1023. Shekalaghe SA, ter Braak R, Daou M, et al. In Tanzania, hemolysis after a single dose of primaquine coadministered with an artemisinin is not restricted to glucose-6-phosphate dehydrogenase-deficient (G6PD A–) individuals. Antimicrob Agents Chemother 2010; 54: 1762–68. Looareesuwan S, White NJ, Chittamas S, Bunnag D, Harinasuta T. High rate of Plasmodium vivax relapse following treatment of falciparum malaria in Thailand. Lancet 1987; 2: 1052–55. Mason DP, Krudsood S, Wilairatana P, et al. Can treatment of P vivax lead to a unexpected appearance of falciparum malaria? Southeast Asian J Trop Med Public Health 2001; 32: 57–63. Douglas NM, Anstey NM, Angus BJ, Nosten F, Price RN. Artemisinin combination therapy for vivax malaria. Lancet Infect Dis 2010; 10: 405–16.
Reduction of HIV incidence in men who have sex with men Despite great advances in the collective understanding of the HIV epidemic, men who have sex with men (MSM) continue to be disproportionately affected. In the USA, the Centers for Disease Control and Prevention estimated that more than 55000 new infections occur every year and more than half of these infections are in MSM.1 Similarly, in Canada, MSM are proportionally most affected.2,3 Alarmingly, in western Europe the number of new HIV diagnoses in MSM nearly doubled from 2538 in 1999 to 5016 in 2006,4 and, after infections from immigrants from other countries are accounted for, MSM had the highest number of new infections of any transmission group in western Europe in 2006.5 Elsewhere, significant increases in new HIV infections in MSM have been reported in Asian countries, such as China where the prevalence of HIV infection in MSM in some major cities seems to be rapidly increasing.6 HIV prevalence for MSM is high throughout Latin America, Russia, and some countries in sub-Saharan Africa.7 Data for other regions are often difficult to find, but the number of new infections in MSM is very high in most regions of the world. www.thelancet.com/infection Vol 10 October 2010
In The Lancet Infectious Diseases today, Le Vu and colleagues8 provide further evidence that the number of new infections in MSM is unacceptably high. Although the overall rate of new HIV infections decreased in France from 8930 in 2003 to 6940 in 2008, the number of new infections in MSM was stable. In 2008, 3320 (48%) of new infections in France were attributed to MSM. Incidence of HIV infections was low and stable in 2008 in IDUs, accounting for 1% of new infections. The pattern of the HIV epidemic reported in France is very similar to that in British Columbia, Canada.9 After introduction of highly active antiretroviral therapy (HAART) in 1996, the number of new positive HIV tests in the province fell by about 50%.9 Much of the early reduction in new HIV cases was in MSM. However, the number of new HIV positive tests in MSM has remained largely unchanged in recent years, especially since 2003, with about 150–180 new infections reported every year.10 By contrast, a reduction in the number of new HIV infections was observed in British Columbia in IDUs after use of HAART increased from 2004.10
Published Online September 9, 2010 DOI:10.1016/S14733099(10)70200-0 See Articles page 682
655
Reflection and Reaction
How can the number of new HIV infections in MSM living in France and elsewhere be reduced? A combined prevention approach, as proposed by Michel Sidibé (the Executive Director of UNAIDS), is needed.11 This approach should include targeted structural interventions directed at specific populations, behavioural interventions directed at individuals, and new biomedical interventions, including expanded coverage of antiretroviral therapy to all HIV infected individuals who meet eligibility criteria for treatment.12 Despite a large body of evidence for the secondary preventive value of antiretroviral therapy, the HIV/AIDS community has failed to fully capitalise on the synergy between treatment and prevention. The aim should not be improved treatment or improved prevention, but rather optimisation of both, because treatment is prevention. Recent therapeutic guidelines13 fully recognise that late initiation of antiretroviral therapy is associated with worse outcomes both at the individual and societal scales. From a public health perspective, expansion of the number of individuals with HIV infection who are eligible for therapy is a unique opportunity to curb the growth of the epidemic. This expansion should decrease HIV/AIDS-related morbidity and mortality in those infected and help to reduce community viral load (and consequently the risk of new infections).14,15 The status quo cannot remain. *Robert S Hogg, David M Moore, Warren D Michelow, Julio S G Montaner British Colombia Centre for Excellence in HIV/AIDS, Vancouver, BC, Canada (RSH, DMM, WDM, JSGM); Faculty of Health Sciences, Simon Fraser University, Vancouver, BC, Canada (RSH); and The School of Population and Public Health (DMM, WDM) and Department of Medicine (JSGM), Faculty of Medicine, University of British Columbia, Vancouver, BC, Canada
[email protected]
RSH has received grant funding from the US National Institutes of Health, UK Medical Research Council, Canadian Foundation for AIDS Research, Canadian Institutes of Health Research, Michael Smith Foundation for Health Research, and the Social Sciences and Humanities Research Council and funding from GlaxoSmithKline and Merck Frosst Laboratories for research and participation in continued medical education programmes. JSGM has received grants from, served as an ad hoc advisor to, and spoken at various events sponsored by Abbott, Argos Therapeutics, Bioject, Boehringer Ingelheim, BMS, Gilead Sciences, GlaxoSmithKline, Hoffmann-La Roche, Janssen-Ortho, Merck Frosst, Pfizer, Schering, Serono, TheraTechnologies, Tibotec, and Trimeris and has received grant funding from the Canadian Institutes of Health Research and National Institutes of Health. All other authors declare that they have no conflicts of interest. 1 2 3
4 5
6
7 8
9
10 11
12
13
14 15
Hall HI, Song R, Rhodes P, et al. Estimation of HIV incidence in the United States. JAMA 2008; 300: 520–29. Public Health Agency of Canada. HIV/AIDS epi updates, November 2007. Ottawa: Public Health Agency of Canada, 2007. Public Health Agency of Canada. Estimates of HIV prevalence and incidence in Canada, 2008. Ottawa: Public Health Agency of Canada, 2009. EuroHIV. HIV/AIDS Surveillance in Europe: end-year report 2006, no 75. Saint-Maurice: Institut de veille sanitaire, 2007. Herida M, Alix J, Devaux I, et al. HIV/AIDS in Europe: epidemiological situation in 2006 and a new framework for surveillance. Euro Surveill 2007; 12: 3312. Ministry of Health of the People’s Republic of China. China 2010 UNGASS country progress report (2008–2009). Beijing: Ministry of Health of the People’s Republic of China, 2010. UNAIDS. Report on the global HIV/AIDS epidemic 2008. Geneva: Joint United Nations Programme on HIV/AIDS, 2008. Le Vu S, Le Strat Y, Barin F, et al. Population-based HIV-1 incidence in France, 2003–08: a modelling analysis. Lancet Infect Dis 2010; published online Sept 9. DOI:10.1016/S1473-3099(10)70167-5. Montaner JSG, Lima VD, Barrios R, et al. Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a population-based study. Lancet 2010; 376: 532–39. British Columbia Centre for Disease Control. STI/HIV annual report. Vancouver: British Columbia Centre for Disease Control, 2008. Sidibé M. Challenges in combating the HIV epidemic, combination prevention and the need for new prevention tools. Joining forces in the devlopment of new prevention technologies meeting; UK Department for International Development, London, UK; Nov 4, 2008. Joint United Nations Programme on HIV/AIDS. Outlook report 2010. http://www.unaids.org/en/KnowledgeCentre/Resources/FeatureStories/ archive/2010/20100713_Outlook_launch.asp (accessed Aug 27, 2010). Thompson MA, Aberg JA, Cahn P, et al. Antiretroviral treatment of adult HIV infection: 2010 recommendations of the international AIDS society—USA panel. JAMA 2010; 304: 321–33. De Cock KM, Gilks CF, Lo Y-R, Guerma T. Can antiretroviral therapy eliminate HIV transmission? Lancet 2009; 373: 7–9. Anema A, Wood E, Montaner J. The use of highly active retroviral therapy to reduce HIV incidence at the population level. CMAJ 2008; 179: 13–14.
Scale-up of diagnostics for multidrug resistant tuberculosis Published Online September 1, 2010 DOI:10.1016/S14733099(10)70188-2 See Articles page 688
656
More than half a century after the discovery of effective drugs, tuberculosis remains an intractable foe. In 2008, 1·8 million people died from the disease, half a million of whom also had HIV.1 Each year an estimated half a million multidrug resistant (MDR) tuberculosis cases develop, of which around only 7% are diagnosed.2 In the 27 high-burden countries for MDR tuberculosis, because
of a lack of laboratory capacity only 1% of new patients received testing for drug susceptibility in 2008.2 There is an urgent need for rapid diagnostics for use in highburden and resource-limited settings. In The Lancet Infectious Diseases today, Jessica Minion and colleagues3 report a meta-analysis of two non-commercial tests for MDR tuberculosis: microscopic-observation drug www.thelancet.com/infection Vol 10 October 2010
Reflection and Reaction
How can the number of new HIV infections in MSM living in France and elsewhere be reduced? A combined prevention approach, as proposed by Michel Sidibé (the Executive Director of UNAIDS), is needed.11 This approach should include targeted structural interventions directed at specific populations, behavioural interventions directed at individuals, and new biomedical interventions, including expanded coverage of antiretroviral therapy to all HIV infected individuals who meet eligibility criteria for treatment.12 Despite a large body of evidence for the secondary preventive value of antiretroviral therapy, the HIV/AIDS community has failed to fully capitalise on the synergy between treatment and prevention. The aim should not be improved treatment or improved prevention, but rather optimisation of both, because treatment is prevention. Recent therapeutic guidelines13 fully recognise that late initiation of antiretroviral therapy is associated with worse outcomes both at the individual and societal scales. From a public health perspective, expansion of the number of individuals with HIV infection who are eligible for therapy is a unique opportunity to curb the growth of the epidemic. This expansion should decrease HIV/AIDS-related morbidity and mortality in those infected and help to reduce community viral load (and consequently the risk of new infections).14,15 The status quo cannot remain. *Robert S Hogg, David M Moore, Warren D Michelow, Julio S G Montaner British Colombia Centre for Excellence in HIV/AIDS, Vancouver, BC, Canada (RSH, DMM, WDM, JSGM); Faculty of Health Sciences, Simon Fraser University, Vancouver, BC, Canada (RSH); and The School of Population and Public Health (DMM, WDM) and Department of Medicine (JSGM), Faculty of Medicine, University of British Columbia, Vancouver, BC, Canada
[email protected]
RSH has received grant funding from the US National Institutes of Health, UK Medical Research Council, Canadian Foundation for AIDS Research, Canadian Institutes of Health Research, Michael Smith Foundation for Health Research, and the Social Sciences and Humanities Research Council and funding from GlaxoSmithKline and Merck Frosst Laboratories for research and participation in continued medical education programmes. JSGM has received grants from, served as an ad hoc advisor to, and spoken at various events sponsored by Abbott, Argos Therapeutics, Bioject, Boehringer Ingelheim, BMS, Gilead Sciences, GlaxoSmithKline, Hoffmann-La Roche, Janssen-Ortho, Merck Frosst, Pfizer, Schering, Serono, TheraTechnologies, Tibotec, and Trimeris and has received grant funding from the Canadian Institutes of Health Research and National Institutes of Health. All other authors declare that they have no conflicts of interest. 1 2 3
4 5
6
7 8
9
10 11
12
13
14 15
Hall HI, Song R, Rhodes P, et al. Estimation of HIV incidence in the United States. JAMA 2008; 300: 520–29. Public Health Agency of Canada. HIV/AIDS epi updates, November 2007. Ottawa: Public Health Agency of Canada, 2007. Public Health Agency of Canada. Estimates of HIV prevalence and incidence in Canada, 2008. Ottawa: Public Health Agency of Canada, 2009. EuroHIV. HIV/AIDS Surveillance in Europe: end-year report 2006, no 75. Saint-Maurice: Institut de veille sanitaire, 2007. Herida M, Alix J, Devaux I, et al. HIV/AIDS in Europe: epidemiological situation in 2006 and a new framework for surveillance. Euro Surveill 2007; 12: 3312. Ministry of Health of the People’s Republic of China. China 2010 UNGASS country progress report (2008–2009). Beijing: Ministry of Health of the People’s Republic of China, 2010. UNAIDS. Report on the global HIV/AIDS epidemic 2008. Geneva: Joint United Nations Programme on HIV/AIDS, 2008. Le Vu S, Le Strat Y, Barin F, et al. Population-based HIV-1 incidence in France, 2003–08: a modelling analysis. Lancet Infect Dis 2010; published online Sept 9. DOI:10.1016/S1473-3099(10)70167-5. Montaner JSG, Lima VD, Barrios R, et al. Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a population-based study. Lancet 2010; 376: 532–39. British Columbia Centre for Disease Control. STI/HIV annual report. Vancouver: British Columbia Centre for Disease Control, 2008. Sidibé M. Challenges in combating the HIV epidemic, combination prevention and the need for new prevention tools. Joining forces in the devlopment of new prevention technologies meeting; UK Department for International Development, London, UK; Nov 4, 2008. Joint United Nations Programme on HIV/AIDS. Outlook report 2010. http://www.unaids.org/en/KnowledgeCentre/Resources/FeatureStories/ archive/2010/20100713_Outlook_launch.asp (accessed Aug 27, 2010). Thompson MA, Aberg JA, Cahn P, et al. Antiretroviral treatment of adult HIV infection: 2010 recommendations of the international AIDS society—USA panel. JAMA 2010; 304: 321–33. De Cock KM, Gilks CF, Lo Y-R, Guerma T. Can antiretroviral therapy eliminate HIV transmission? Lancet 2009; 373: 7–9. Anema A, Wood E, Montaner J. The use of highly active retroviral therapy to reduce HIV incidence at the population level. CMAJ 2008; 179: 13–14.
Scale-up of diagnostics for multidrug resistant tuberculosis Published Online September 1, 2010 DOI:10.1016/S14733099(10)70188-2 See Articles page 688
656
More than half a century after the discovery of effective drugs, tuberculosis remains an intractable foe. In 2008, 1·8 million people died from the disease, half a million of whom also had HIV.1 Each year an estimated half a million multidrug resistant (MDR) tuberculosis cases develop, of which around only 7% are diagnosed.2 In the 27 high-burden countries for MDR tuberculosis, because
of a lack of laboratory capacity only 1% of new patients received testing for drug susceptibility in 2008.2 There is an urgent need for rapid diagnostics for use in highburden and resource-limited settings. In The Lancet Infectious Diseases today, Jessica Minion and colleagues3 report a meta-analysis of two non-commercial tests for MDR tuberculosis: microscopic-observation drug www.thelancet.com/infection Vol 10 October 2010
susceptibility assay (MODS) and thin layer agar (TLA). The Foundation for Innovative Diagnostics4 have used largescale operational demonstration projects to provide an evidence base for the use of commercial tests, and for the endorsement of such tests by WHO. Scale-up of automated rapid liquid culture and nucleic acid amplification lineprobe assays has now been recommended;5,6 however, implementation of these techniques in resource-limited settings is challenging. Biosafety laboratories, costs of testing, and the human-resources needed to implement and maintain these tests all require huge and sustained investment. Furthermore, line-probe assays detect multidrug resistance only in smear-positive tuberculosis cases, which is a serious drawback, especially in countries with high HIV prevalence because of the low sensitivity of smear in HIV co-infected individuals. In response, many tuberculosis research groups have developed non-commercial tests that require less sophisticated approaches. Sadly, the usefulness and potential for scaleup has been hard to prove. Independent, large studies are urgently needed to determine if such tests are viable beyond research settings. Minion and colleagues’ analysis is encouraging and reflects our experience that MODS is accurate, robust, and reliable when adopted by tuberculosis reference laboratories; however, insufficient data exist to draw the same conclusions for TLA. WHO’s Strategic Advisory Group has now recommended that MODS be used as an interim solution while the capacity for automated liquid culture and drug susceptibility are developed.7 However, large-scale adoption of MODS will require substantial investment into the development of training programmes, quality control and assurance, reliable supply lines of consumables, reagents, and maintenance of biosafety. Increased diagnosis of MDR tuberculosis will not be useful without rapid access to quality-assured treatment. Potential advantages of the MODS assay compared with commercial techniques might warrant its implementation as an alternative solution rather than an interim one—MODS is faster than automated liquid culture and, unlike line-probe assays, can be used in smear-positive and smear-negative cases. Early detection of smear-negative MDR tuberculosis cases is likely to have a greater programmatic effect on MDR transmission, especially in settings with a high HIV prevalence. MODS might be easier to adopt than commercial techniques in national programmes that www.thelancet.com/infection Vol 10 October 2010
Maxine Caws
Reflection and Reaction
have a strong resource of experienced microscopists but a limited experience of molecular testing. Many questions remain about the implementation of MODS including the full economic costs of the technique (MODS is substantially more labour intensive than commercial liquid culture tests) and the ability to differentiate Mycobacterium tuberculosis from nontuberculous mycobacteria, particularly in settings with a high HIV prevalence. There is no perfect test for MDR tuberculosis that is cheap, rapid, and accurate, and and does not require biocontainment facilities. Choosing between imperfect solutions is never easy. To resolve these questions we should apply the same rigour to diagnostics as we do to treatment with randomised trials of commercial and non-commercial diagnostics in operational settings, full economic costing, and assessment of the effects on actual patient outcomes and incidence rates. Doing so will ensure that scale-up efforts, which are run by programmes such as WHO’s Global Laboratory Initiative,8 apply the best available solutions. *Maxine Caws, Dang Thi Minh Ha Oxford University Clinical Research Unit, Wellcome Trust Major Overseas Programme, Hospital for Tropical Diseases, 190 Ben Ham Tu, Ho Chi Minh City, Vietnam (MC); and Pham Ngoc Thach Hospital for Tuberculosis and Lung Diseases, 120 Hung Vuong, Ho Chi Minh City, Vietnam (DTMH)
[email protected]
657
Reflection and Reaction
We declare that we have no conflicts of interest. 1
2
3
4
5
WHO. Global tuberculosis control: a short update to the 2009 report. 2010. http://whqlibdoc.who.int/publications/2009/9789241598866_eng.pdf (accessed Aug 25, 2010). WHO. Multidrug and extensively drug-resistant TB (M/XDR-TB): 2010 global report on surveillance and response. 2010. http://whqlibdoc.who. int/publications/2010/9789241599191_eng.pdf (accessed Aug 25, 2010). Minion J, Leung E, Menzies D, Madhukar P. Microsocopic-observation drug susceptibility and thin layer agar assays for the detection of drug resistant tuberculosis: a systematic review and meta-analysis. Lancet Infect Dis 2010; published online Aug 31. DOI:10.1016/S1473-3099(10)70165-1. Foundation for Innovative New Diagnostics.http://www.finddiagnostics. org (accessed Aug 25, 2010).
6
7
8
WHO. Use of liquid TB culture and drug susceptibility testing in low and middle income settings. Summary report of the expert group meeting on the use of liquid culture media. March 26, 2007. http://www.who.int/tb/ dots/laboratory/Use%20of%20Liquid%20TB%20Culture_Summary%20 Report.pdf (accessed Aug 25, 2010). WHO. Policy Statement. Molecular line probe assays for rapid screening of patients at risk of multidrug-resistant tuberculosis (MDR-TB). 27 June 2008. http://www.who.int/tb/dots/laboratory/lpa_policy.pdf (accessed Aug 25, 2010). WHO. Strategic and technical advisory group for tuberculosis (STAG-B): report of the ninth meeting. Nov 9–11, 2009. http://www.who.int/tb/ advisorybodies/stagtbreport2009.pdf (accessed Aug 25, 2010). WHO. Global laboratory initiative. http://www.who.int/tb/dots/laboratory/ gli/en/index.html (accessed Aug 25, 2010).
Global burden and investment for neglected diseases? In The Lancet Infectious Diseases, two interesting papers1,2 were published discussing crucial issues concerning neglected diseases. The January, 2010, Editorial questioned how neglected diseases should be (re)defined, and what the best strategies could be to defeat these diseases. In particular, should research and development investment for neglected diseases be driven by the notion of global burden of disease (GBD), and if yes, then how?1,3 In the April, 2010, issue, Mary Moran2 summarised some of the main findings of the Global Funding of Innovation for Neglected Diseases (G-FINDer) 2009 survey,4,5 which tried to form a consensus definition for neglected diseases and to identify their main sponsors and contributions. Research and i development xk expenditure ($)=xi k=1
∑
The survey has several remarkable merits and contributes to clarifying the concept of neglected diseases;6 for this reason, it can inform the efficient allocation of funds to where they are needed the most. However, the setting of funding priorities is not straightforward. Moran notes that research and development sponsors have allocated more funds to HIV/AIDS, malaria, and tuberculosis, even though other diseases also have high levels of disability-adjusted lifeyears (DALYs) and deaths.3,4 This finding, together with the two questions posed by The Lancet Infectious Diseases Editorial stimulated me to write this paper. I used 2004 data from WHO’s Report on GBD,7 and the 2007 research and development funding of neglected
DALYs=Di (millions)
i
∑D
k
Deaths=di (thousands)
k
k=1
k=1
i ∑k=1 xk
i ∑k=1 Dk
i ∑k=1 dk
2 385 117 816
332·4
11 590
i
∑d
Rheumatic fever
1 670 089
1 670 089
5·1
5·1
280
280
0·0007
0·01
0·02
Trachoma
1 679 711
3 349 800
1·3
6·4
0
280
0·001
0·02
0·02
Leprosy
5 619 475
8 969 275
0·02
Typhoid and paratyphoid fever
9 117 212
18 086 487
Bacterial pneumonia and meningitis
32 517 311
0·2 NA
50 603 798 104·6
6·6
285
0·004
0·02
600
885
0·007
NA
0·08
111·2
4240
5125
0·02
0·33
0·44 0·45
Helminths
51 591 838
102 195 636
123·2
47
5172
0·04
0·37
Dengue
82 013 895
184 209 531
0·7
130·2
18
5190
0·08
0·4
0·45
113 889 118
298 098 649
72·3
202·5
2000
7190
0·12
0·61
0·62
Kinetoplastids
125 122 839
423 221 488
4·1
206·6
110
7300
0·18
0·62
0·63
Tuberculosis
410 428 697
833 650 185
34
240·6
1400
8700
0·35
0·72
0·75
Malaria
468 449 438
1 302 099 623
34
274·6
890
9590
0·55
0·83
0·83
1 083 018 193
2 385 117 816
57·8
332·4
2000
11 590
1
1
1
0·52
0·55
Diarrhoeal diseases
HIV/AIDS Gini index
12
5
NA
From left to right, columns of Table 1 are as follows. Arranged in increasing order, for each disease, the first column contains research and development expenditures (xi) made in 2007; the second column contains the cumulative expenditures up to generic disease (i). For each disease, the third column contains the 2004 level of disabilityadjusted life-years (DALYs) (Di), and the fourth column shows the cumulated level of DALYs. The fifth and sixth columns contain, respectively, the number of deaths (di) and their cumulative level. Finally, the seventh, eighth, and ninth column contain, respectively, the cumulative distribution of research and development expenditures, of DALYs and deaths. NA=not available. Table 1 statistics were used to calculate the Gini index (webappendix).
Table 1: The Gini Index for global burden of disease
658
www.thelancet.com/infection Vol 10 October 2010
Reflection and Reaction
We declare that we have no conflicts of interest. 1
2
3
4
5
WHO. Global tuberculosis control: a short update to the 2009 report. 2010. http://whqlibdoc.who.int/publications/2009/9789241598866_eng.pdf (accessed Aug 25, 2010). WHO. Multidrug and extensively drug-resistant TB (M/XDR-TB): 2010 global report on surveillance and response. 2010. http://whqlibdoc.who. int/publications/2010/9789241599191_eng.pdf (accessed Aug 25, 2010). Minion J, Leung E, Menzies D, Madhukar P. Microsocopic-observation drug susceptibility and thin layer agar assays for the detection of drug resistant tuberculosis: a systematic review and meta-analysis. Lancet Infect Dis 2010; published online Aug 31. DOI:10.1016/S1473-3099(10)70165-1. Foundation for Innovative New Diagnostics.http://www.finddiagnostics. org (accessed Aug 25, 2010).
6
7
8
WHO. Use of liquid TB culture and drug susceptibility testing in low and middle income settings. Summary report of the expert group meeting on the use of liquid culture media. March 26, 2007. http://www.who.int/tb/ dots/laboratory/Use%20of%20Liquid%20TB%20Culture_Summary%20 Report.pdf (accessed Aug 25, 2010). WHO. Policy Statement. Molecular line probe assays for rapid screening of patients at risk of multidrug-resistant tuberculosis (MDR-TB). 27 June 2008. http://www.who.int/tb/dots/laboratory/lpa_policy.pdf (accessed Aug 25, 2010). WHO. Strategic and technical advisory group for tuberculosis (STAG-B): report of the ninth meeting. Nov 9–11, 2009. http://www.who.int/tb/ advisorybodies/stagtbreport2009.pdf (accessed Aug 25, 2010). WHO. Global laboratory initiative. http://www.who.int/tb/dots/laboratory/ gli/en/index.html (accessed Aug 25, 2010).
Global burden and investment for neglected diseases? In The Lancet Infectious Diseases, two interesting papers1,2 were published discussing crucial issues concerning neglected diseases. The January, 2010, Editorial questioned how neglected diseases should be (re)defined, and what the best strategies could be to defeat these diseases. In particular, should research and development investment for neglected diseases be driven by the notion of global burden of disease (GBD), and if yes, then how?1,3 In the April, 2010, issue, Mary Moran2 summarised some of the main findings of the Global Funding of Innovation for Neglected Diseases (G-FINDer) 2009 survey,4,5 which tried to form a consensus definition for neglected diseases and to identify their main sponsors and contributions. Research and i development xk expenditure ($)=xi k=1
∑
The survey has several remarkable merits and contributes to clarifying the concept of neglected diseases;6 for this reason, it can inform the efficient allocation of funds to where they are needed the most. However, the setting of funding priorities is not straightforward. Moran notes that research and development sponsors have allocated more funds to HIV/AIDS, malaria, and tuberculosis, even though other diseases also have high levels of disability-adjusted lifeyears (DALYs) and deaths.3,4 This finding, together with the two questions posed by The Lancet Infectious Diseases Editorial stimulated me to write this paper. I used 2004 data from WHO’s Report on GBD,7 and the 2007 research and development funding of neglected
DALYs=Di (millions)
i
∑D
k
Deaths=di (thousands)
k
k=1
k=1
i ∑k=1 xk
i ∑k=1 Dk
i ∑k=1 dk
2 385 117 816
332·4
11 590
i
∑d
Rheumatic fever
1 670 089
1 670 089
5·1
5·1
280
280
0·0007
0·01
0·02
Trachoma
1 679 711
3 349 800
1·3
6·4
0
280
0·001
0·02
0·02
Leprosy
5 619 475
8 969 275
0·02
Typhoid and paratyphoid fever
9 117 212
18 086 487
Bacterial pneumonia and meningitis
32 517 311
0·2 NA
50 603 798 104·6
6·6
285
0·004
0·02
600
885
0·007
NA
0·08
111·2
4240
5125
0·02
0·33
0·44 0·45
Helminths
51 591 838
102 195 636
123·2
47
5172
0·04
0·37
Dengue
82 013 895
184 209 531
0·7
130·2
18
5190
0·08
0·4
0·45
113 889 118
298 098 649
72·3
202·5
2000
7190
0·12
0·61
0·62
Kinetoplastids
125 122 839
423 221 488
4·1
206·6
110
7300
0·18
0·62
0·63
Tuberculosis
410 428 697
833 650 185
34
240·6
1400
8700
0·35
0·72
0·75
Malaria
468 449 438
1 302 099 623
34
274·6
890
9590
0·55
0·83
0·83
1 083 018 193
2 385 117 816
57·8
332·4
2000
11 590
1
1
1
0·52
0·55
Diarrhoeal diseases
HIV/AIDS Gini index
12
5
NA
From left to right, columns of Table 1 are as follows. Arranged in increasing order, for each disease, the first column contains research and development expenditures (xi) made in 2007; the second column contains the cumulative expenditures up to generic disease (i). For each disease, the third column contains the 2004 level of disabilityadjusted life-years (DALYs) (Di), and the fourth column shows the cumulated level of DALYs. The fifth and sixth columns contain, respectively, the number of deaths (di) and their cumulative level. Finally, the seventh, eighth, and ninth column contain, respectively, the cumulative distribution of research and development expenditures, of DALYs and deaths. NA=not available. Table 1 statistics were used to calculate the Gini index (webappendix).
Table 1: The Gini Index for global burden of disease
658
www.thelancet.com/infection Vol 10 October 2010
Reflection and Reaction
diseases5 to consider how research funds could be guided by GBD.1 First, I computed the Gini index8,9 (webappendix)—a common indicator for measuring expenditure inequality—to assess whether 2007 research and development funds were mainly allocated to HIV/AIDS, malaria, and tuberculosis, regarding the total numbers of DALYs and deaths in table 1. The index ranges from 0, when all diseases receive proportionally the same amount of research and development resources, to 1, when only one disease is funded. The index for DALYs and deaths is above 0·5 suggesting that the 2007 research and development resources were not uniformly distributed for all neglected diseases. This result is because more than 60% of DALYs and deaths of the less-funded diseases received about only 20% of the total resources. In particular, bacterial pneumonia and meningitis and diarrhoeal diseases seemed to be underfunded in 2007, according to the levels of DALYs and deaths. However, the index does not consider all of the relevant dimensions for appropriate allocation of research and development resources. To consider these dimensions further, the first two columns of table 2 show, respectively, how many US dollars are invested in research and development for each DALY and each death. Some findings emerged that were not found in the Gini index analysis, in particular, that dengue fever received the highest funding per DALY and per death, but also, that the ratio of deaths to DALYs could be an important indicator to consider for funds allocation. Table 2 shows that for most diseases, the ratio of deaths to DALYs is close to the overall average of 0·33, and use of this ratio to determine the research and development shares would lead to roughly the same amount of investment for all neglected diseases. However, ratios could be corrected by a factor representing the relative burden of each disease, with respect to the total number of DALYs and deaths. If deaths could be decreased by controlling DALYs, then the relative burden of death might also be used to guide fund allocation: D ( Dd ( 332·4 (11332·4 590
(
(
i
i
(
i
=
di 11 590
because the second term of the above equation is the relative burden of DALYs for each disease and the third term is a normalising factor. The last two columns of table 2 show the discrepancies between the indicator www.thelancet.com/infection Vol 10 October 2010
xi
xi
di
Di
di
xi
Di
di
Di
332·4
11 590
2 385 117 816
Rheumatic fever
0·33
6
Trachoma
1·3
Nc
0·85
1123·9
Leprosy Typhoid and paratyphoid fever
NA
Bacterial pneumonia and meningitis
0·31
Helminths
0·05
0·01
0·02
0·0007
0
0·004
0
0·0007 0·002
0·03
0·0006
0·0004
15·2
NA
NA
0·005
0·004
7·7
0·04
0·31
0·36
0·01 0·02
4·3
1097·7
0·004
0·04
0·04
117·17
4556·3
0·03
0·002
0·001
0·03
1·57
56·9
0·03
0·22
0·17
0·05
Kinetoplastids
30·52
1137·5
0·02
0·01
0·01
0·05
Tuberculosis
12·1
293·2
0·04
0·1
0·12
0·17
Malaria
13·8
526·35
0·03
0·1
0·08
0·2
HIV/AIDS
18·8
541·50
0·03
0·17
0·17
0·45
Dengue Diarrhoeal diseases
NC=non-computable.
Table 2: Possible indicators to guide research and development funds allocation
and between the effective research and development shares. Although important, allocation of funds is just one part of the policies that tackle neglected diseases. I have considered several GBD-based elements that might influence decisions. However, many other elements, such as effort fragmentation, time between state of the art research and final treatments, the numbers of affected populations, and the speed and likelihood of an epidemic, are not part of the standard GBD metric, but they might be important in funding decisions. The relation of GBD to funds allocation seems far from obvious; however, I believe that considerations of global burden are unavoidable, and that more work should be pursued in this direction.
See Online for webappendix
Nicola Dimitri University of Siena, Faculty of Economics, Siena 53100, Italy, and Institutions Markets Technologies Lucca Institute for Advanced Studies, Lucca 55100, Italy
[email protected] I declare that I have no conflicts of interest. 1 2 3
4 5
The Lancet Infectious Diseases. Moving away from neglect. Lancet Infect Dis 2010; 10: 1. Moran M., Funding for Research and development and the Financial Crisis. Lancet Infect Dis 2010; 10: 214–15. Mathers CD, Ezzati M, Lopez AD. Measuring the burden of neglected tropical diseases: the global burden of disease framework. PLoS Negl Trop Dis 2007; 1: e114. Moran M, Guzman J, Ropars A, et al. Neglected disease research and development: how much are we really spending? PLoS Med 2009; 6: e30. Moran M, Guzman J, Henderson K, et al. Neglected disease research and development: new times, new trends. http://www.ghtcoalition.org/files/ gfinder_dec2009.pdf (accessed Feb 15, 2010).
659
Reflection and Reaction
6 7
Engels D, Savioli L. Reconsidering the underestimated burden caused by neglected tropical diseases. Trends Parasitol 2006; 22: 363–66. WHO. The global burden of disease: 2004 update. 2008. http://www.who. int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf (accessed Feb 15, 2010).
8
9
Gini C. In: Pizetti E, Salvemini T (eds). Variabilità e mutabilità. Reprinted in memorie di metodologica statistica. Rome: Libreria Eredi Virgilio Veschi, 1955. Amiel Y, Cowell, FA. Thinking about inequality. Cambridge, UK: Cambridge University Press, 1996.
Reconstructing the past of poliovirus eradication efforts
Shehzad Noorani/Majority World/Still Pictures
The March and June, 2010, issues of The Lancet Infectious Diseases1,2 presented two optimistic views of the polio eradication effort that was led by the Global Polio Eradication Initiative (GPEI); however, it might be wise to review the historical record of this initiative and, without losing optimism, be more cautious in prediction of successes. Since the World Health Assembly adopted the goal of eradicating poliovirus infection, major achievements have been made.3,4 These achievements have been enabled by strong societal, financial, and political commitment,5 and, as a result, the circulation of wildtype poliovirus type 2 was interrupted by the initial goal in 2000.6 Nonetheless, circulation of the other two wild-type poliovirus serotypes—P1 and P3—remained in northern Nigeria, northern India, and the Pakistan– Afghanistan border.4,6 By the end of 2009, cases of poliovirus continued to circulate and to produce human disease in these areas and in new settings.7 Crucial analyses of the failure of the malaria, yellow fever, and yaws eradication programmes in the 20th century provided important lessons for the eradication of
A health worker vaccinating an infant in Torkham, a town on the Pakistan–Afghanistan border
660
smallpox.5 Similarly, reconstruction of the past of poliovirus eradication programmes, and application of the lessons learned thus far, might be a wise approach to steering the remaining part of this initiative. Many experts believe that the GPEI should move from purely technical elements and assess the reasons for failed immunisation policies (ie, the introduction of oral poliovirus monovalent formulations of the vaccine during supplemental vaccination campaigns).7–9 Although evidence suggests that the introduction of wild-poliovirus 1 monovalent oral polio vaccine in Bihar and Uttar Pradesh states in India had some scientific merit, the GPEI leadership paid no heed to local expert advice, and concerns about the ethics of vaccination with monovalent vaccines were not addressed.7–9 Wild-type poliovirus 3 emerged iatrogenically in these areas, which led to hundreds of human cases in subsequent years and produced areas of population resistance to further poliovirus vaccination.9 Alternatively, over-reliance on many rounds of monovalent oral poliovirus type 1 vaccine might have led to an immunological vacuum against wild-type poliovirus 3 in children aged 0–5 years.8 Since infection with wild-type poliovirus 3 is as severe as that caused by type 1 poliovirus infection, many children in northern India were left with paralysis, and a tremendous amount of financial effort was expended.7,8 Routine or supplementation immunisation efforts have biological and operational dimensions.5 Bivalent type 1 and type 3 oral poliovirus vaccine might prove to be the missing link in polio eradication, but as history has already shown, this formulation should not be regarded as the panacea. Historical experiences surrounding the smallpox eradication campaign and the elimination of poliovirus circulation in the Americas might need to be revised from managerial, operational, and technical angles. The world has forgotten that the last case of poliomyelitis recorded in the Americas was more than 15 years ago, and this milestone was reached with the use of eradication-quality surveillance www.thelancet.com/infection Vol 10 October 2010
Reflection and Reaction
6 7
Engels D, Savioli L. Reconsidering the underestimated burden caused by neglected tropical diseases. Trends Parasitol 2006; 22: 363–66. WHO. The global burden of disease: 2004 update. 2008. http://www.who. int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf (accessed Feb 15, 2010).
8
9
Gini C. In: Pizetti E, Salvemini T (eds). Variabilità e mutabilità. Reprinted in memorie di metodologica statistica. Rome: Libreria Eredi Virgilio Veschi, 1955. Amiel Y, Cowell, FA. Thinking about inequality. Cambridge, UK: Cambridge University Press, 1996.
Reconstructing the past of poliovirus eradication efforts
Shehzad Noorani/Majority World/Still Pictures
The March and June, 2010, issues of The Lancet Infectious Diseases1,2 presented two optimistic views of the polio eradication effort that was led by the Global Polio Eradication Initiative (GPEI); however, it might be wise to review the historical record of this initiative and, without losing optimism, be more cautious in prediction of successes. Since the World Health Assembly adopted the goal of eradicating poliovirus infection, major achievements have been made.3,4 These achievements have been enabled by strong societal, financial, and political commitment,5 and, as a result, the circulation of wildtype poliovirus type 2 was interrupted by the initial goal in 2000.6 Nonetheless, circulation of the other two wild-type poliovirus serotypes—P1 and P3—remained in northern Nigeria, northern India, and the Pakistan– Afghanistan border.4,6 By the end of 2009, cases of poliovirus continued to circulate and to produce human disease in these areas and in new settings.7 Crucial analyses of the failure of the malaria, yellow fever, and yaws eradication programmes in the 20th century provided important lessons for the eradication of
A health worker vaccinating an infant in Torkham, a town on the Pakistan–Afghanistan border
660
smallpox.5 Similarly, reconstruction of the past of poliovirus eradication programmes, and application of the lessons learned thus far, might be a wise approach to steering the remaining part of this initiative. Many experts believe that the GPEI should move from purely technical elements and assess the reasons for failed immunisation policies (ie, the introduction of oral poliovirus monovalent formulations of the vaccine during supplemental vaccination campaigns).7–9 Although evidence suggests that the introduction of wild-poliovirus 1 monovalent oral polio vaccine in Bihar and Uttar Pradesh states in India had some scientific merit, the GPEI leadership paid no heed to local expert advice, and concerns about the ethics of vaccination with monovalent vaccines were not addressed.7–9 Wild-type poliovirus 3 emerged iatrogenically in these areas, which led to hundreds of human cases in subsequent years and produced areas of population resistance to further poliovirus vaccination.9 Alternatively, over-reliance on many rounds of monovalent oral poliovirus type 1 vaccine might have led to an immunological vacuum against wild-type poliovirus 3 in children aged 0–5 years.8 Since infection with wild-type poliovirus 3 is as severe as that caused by type 1 poliovirus infection, many children in northern India were left with paralysis, and a tremendous amount of financial effort was expended.7,8 Routine or supplementation immunisation efforts have biological and operational dimensions.5 Bivalent type 1 and type 3 oral poliovirus vaccine might prove to be the missing link in polio eradication, but as history has already shown, this formulation should not be regarded as the panacea. Historical experiences surrounding the smallpox eradication campaign and the elimination of poliovirus circulation in the Americas might need to be revised from managerial, operational, and technical angles. The world has forgotten that the last case of poliomyelitis recorded in the Americas was more than 15 years ago, and this milestone was reached with the use of eradication-quality surveillance www.thelancet.com/infection Vol 10 October 2010
Reflection and Reaction
and routine and supplemental rounds of trivalent oral poliovirus vaccine. Virological considerations and academic discussions held at international forums should not dominate the decision-making process in polio eradication. Instead, social autopsies of affected communities should be done to optimise microplanning to reach more unimmunised children.9 There is an urgent need to listen to the missing voice of the affected communities: to address the concerns of parents who are resistant to vaccination, creating persistent clusters of unimmunised children in the last wild-poliovirus strongholds;9 to overcome the public’s fears and suspicions; and to attend to health-care workers who have become fatigued with the programme.10 These efforts will provide a historical legacy for the strengthening of immunisation systems that will ultimately promote immunisation equity for the most disenfranchised populations worldwide. Humility should replace scientific glamour as the banner of the polio eradication endeavour. There is no room for further failure.
www.thelancet.com/infection Vol 10 October 2010
Carlos Franco-Paredes Emory University School of Medicine, Division of Infectious Diseases, Atlanta, GA 30308, USA; and Hospital Infantil de Mexico, Federico Gomez, Mexico City 06720, Mexico
[email protected] I declare that I have no conflicts of interest. 1 2 3 4 5 6 7 8
9
10
Senior K. Polio eradication within 5 years now a real possibility. Lancet Infect Dis 2010; 10: 148–49. The Lancet Infectious Diseases. Polio–a pathogen on a precipice. Lancet Infect Dis 2010; 10: 367. Pallansch MA, Sandhu HS. The eradication of polio–progress and challenges. N Engl J Med 2006; 355: 2508–11. Ehrenfeld E, Chumakov K. Monovalent oral poliovirus vaccines–a good tool but not a total solution. N Engl J Med 2008; 359: 1726–27. Dowdle WR. The principles of disease elimination and eradication. Bull World Health Organ 1998; 76(suppl 2): 22–25. Minor PD. Polio eradication, cessation of vaccination and re-emergence of disease. Nat Rev Microbiol 2004; 2: 473–82. Agarwal RK. Polio eradication in India: a tale of science, ethics, dogmas and strategy. Indian Pediatr 2008; 349–51. Arora NK, Chaturvedi S, Dasgupta R. Global lessons from India’s poliomyelitis elimination campaign. Bull World Health Organ 2010; 88: 232–34. Chaturvedi S, Dasgupta R, Adhish V, et al. Deconstructing social resistance to pulse polio campaign in two North Indian Districts. Indian Pediatr 2009; 46: 963–74. Caplan AL. Is disease eradication ethical? Lancet 2009; 373: 2192–93.
661
Cross-talk
Mass gathering: mass effect? See Leading Edge page 653
662
Despite crass assertions from some media commentators, last year’s fears about the potentially horrendous predations of pandemic H1N1 2009 influenza virus were not ill founded. Nor did scientists dramatise them inappropriately in the quest for research funding. We have, however, been very lucky—for the moment. Now that we understand more about the reassortment of the H1N1, we know that the organism could become established in pigs. From there, a novel strain might well emerge to pose calamitous threats to our wellbeing in the future. And with regards to influenza viruses in general, most experts believe it is not a matter of if, but when. Among the recent patterns of human behaviour that could have worsened matters substantially, Hajj—the Muslim ritual of pilgrimage from around the world to Mecca in Saudi Arabia—features prominently. An annual gathering of millions of pilgrims in extremely crowded conditions known to facilitate the spread of respiratory infections, the Hajj in 2009 coincided not only with the influenza season but also with the spread of the H1N1 virus. Possible consequences, both during the event and afterwards, when huge numbers of people returned to countries with poor health-care resources, are unimaginable. Kamran Khan of the University of Toronto, and colleagues in Canada, Saudi Arabia, and the USA, have now published a valuable review of the global health implications of the 2009 Hajj. Writing in the Journal of Travel Medicine (2010; 17: 75–81), they highlight the scale of the potential problem in 2009 from travel data on the 2008 Hajj. On that occasion, 2·5 million pilgrims took part, 1·7 million of them from countries outside Saudi Arabia. Of these, 91% travelled via commercial flights, the greatest numbers being from Bangladesh, Afghanistan, and Yemen. “Nearly 200 000 pilgrims that performed the Hajj in 2008 originated from the world’s most resource-limited countries where access to H1N1 vaccine and capacity to detect and respond to H1N1 in returning pilgrims are extremely limited”, the authors write. “International efforts may be needed to assist resource-limited countries that are vulnerable to the impact of H1N1 during the 2009 to 2010 influenza season.” Unfortunately, these comments were published too late to be of any material assistance in the way the authors intended. Nevertheless, one must agree with the cannily expressed conclusion of the paper: “The low numbers of H1N1 cases actually observed during the Hajj suggest
that the local and global health implications of this mass gathering were far more limited than their potential”. In June 2009, Saudi Arabia, WHO, and other agencies held a conference to consider measures to mitigate the effects of influenza during the gathering. Health personnel at Jeddah International Airport used infrared thermography to screen pilgrims on arrival, assessed febrile individuals and referred those with symptoms compatible with influenza for PCR testing of nasopharyngeal secretions. Oseltamivir treatment was initiated (and discontinued 6 h later for those with a negative result). However, only a handful of economically prosperous countries were able to vaccinate their pilgrims in sufficient time for them to develop protective immunity before starting the Hajj. And as Khan and his co-workers point out, resistant H1N1 virus, introduced into Mecca, could have spread to pilgrims from other parts of the world, amplifying its global distribution. Only one country, Tunisia, prohibited citizens from participating in the 2009 Hajj, while Saudi Arabia discouraged individuals at high risk of complications from taking part. How many people around the world decided to forego the event in light of the pandemic is unknown. Anecdotal evidence from travel agents suggests only a modest decline in participation. We should not forget, either, that the H1N1 virus, though milder than expected and primarily affecting younger rather than older people, did put patients into intensive care throughout the world, some of whom died. Most of us have indeed had a lucky escape.
Bernard Dixon
www.thelancet.com/infection Vol 10 October 2010
Newsdesk
a public health issue, will there not be those deserving of punishment who escape justice? CPS guidelines point out that the consequences for the “victim” of contracting a sexually transmitted disease fit the criteria for grievous bodily harm. And what of those cases in which there is deliberate deception—an individual who explicitly claims to be negative and goes on to infect their partner? However, there is a risk that prosecutions will encourage the impression that precautions are the prerogative of the infected, eroding efforts to foment a culture of shared responsibility. “One unhelpful message is that HIV is spread only by people who know they have it, and that most people with diagnosed HIV will routinely tell you they have it”, says Azad. Most infections are attributable to the undiagnosed, and to suggest that a handful of criminally liable
people are to blame for the HIV/AIDS epidemic sidelines the responsibilities of the wider population. There are questions of justice. Is not the decision to have unprotected sex an implicit acceptance of the associated risks? What about cases where an HIV-positive defendant used a condom but it split? Or their antiretroviral regimen rendered their viral load undetectable? Weait points out that complaints have tended to be brought by aggrieved ex-partners, but many couples stay together after one infects the other, and no complaint is made. The number of potential criminals in the UK is likely to be several thousand. “There’s been a significant move towards criminalisation both in the UK and other countries”, Azad concludes, “but public health voices are beginning to be heard.” Prosecutorial guidelines demanding a high threshold of evidence have
reduced cases brought to court in England and Wales. Crucial to this is the burgeoning judicial awareness of the limitations of phylogenetic analysis, which cannot reliably establish the direction or source of HIV infection. New police guidelines, agreed with the National AIDS Trust, should prevent a recurrence of the circumstances surrounding the Porter case, and decrease the number and length of investigations. Still, it is hard to shake the idea that prosecutions reinforce stigma and create an impression of those living with HIV as being of especial threat to the general public. Azad believes this is counterproductive. “Stigma is the main reason people find it hard to declare their HIV status; we would do far better address this than to criminalise people.”
Talha Burki
Infectious disease surveillance update West Nile virus in Europe West Nile virus has been reported in several European countries since the initial outbreak in Greece reported on Aug 7. The Greek outbreak is largely limited to the central Macedonia region, with a few cases in the neighbouring district of Larissa. As of Sept 3, 173 cases (all laboratory confirmed) had been reported and there had been 15 deaths. In Romania, health authorities have reported 13 confirmed cases and two probable cases in the past 2 months. Two people with the infection, both older than 75 years, have died. The median age of the people infected was 50 years (range 18–79) and they were from 11 different districts. The Hungarian health authorities have confirmed three cases of infection: two people living in central Hungary and one close to the Romanian border. One case of infection has also been confirmed in the Veneto region of Italy www.thelancet.com/infection Vol 10 October 2010
and one probable case was identified in Portugal. Outside of the European Union, from July to August, 2010, 231 cases of West Nile fever were reported in Russia, including six deaths. On Aug 10, 24 confirmed cases were reported in Israel, with six further presumed cases. A more recent case was imported into Israel from the Netherlands. Most cases in Israel have been in the Central and Tel Aviv region, but there have also been isolated cases reported in the Haifa region.
Encephalitis in India On Sept 8, health officials in India confirmed that a further five people had died due to encephalitis, bringing the total number of deaths in Uttar Pradesh’s eastern region to 254 this year. The outbreak of encephalitis this year began unusually early, by mid-April 210 patients had already been treated in this region. In many
cases the cause has not been specified. Some of the cases are due to Japanese encephlitis virus (JEV) and others have been described as acute encephalitis of undefined aetiology but not due to JEV infection. In 2009, more than 500 cases of encephlitis were reported in the region. Again, only some of these cases were attributed to JEV.
Legionnaires disease in the UK As of Sept 12, 19 cases of legionnaires disease have been identified in Wales in the UK. All patients have needed treatment in hospital. A further four people are being investigated to establish if they are linked to the outbreak. Two of these four people have died: a 70-year-old man who died on Sept 8 and a 64-year-old woman on Sept 6. The source of the outbreak has not yet been identified.
For more on WNV in Europe see http://www.ecdc.europa.eu/en/ activities/sciadvice/Lists/ ECDC%20Reviews/ECDC_ DispForm.aspx?List=512ff74f77d4-4ad8-b6d6-bf0f23083f30 &ID=938&RootFolder=%2Fen%2 Factivities%2Fsciadvice%2FLists %2FECDC%20Reviews For ECDC’s WNV threat assessment see http://www. ecdc.europa.eu/en/healthtopics/ Documents/1009_Threat%20 assessment_West_Nile_Virus.pdf For more on legionnaires disease in Wales see http:// www.wales.nhs.uk/ sitesplus/888/news/17027
Onisillos Sekkides 667
Newsdesk
Preparing for the next pandemic—the lessons of H1N1 See Editorial Lancet 2010; 376: 565
Kin Cheung/AP/Press Association Images
See Leading Edge Lancet Infect Dis 2010; 10: 505
As the influenza A H1N1 2009–10 pandemic is declared over, the process of review and reflection is already underway to establish what lessons can be learned for any future pandemic. The immediate concern is continued vigilance, since H1N1 is now circulating with seasonal influenza variants, and on occasions H5N1. WHO Director General Margaret Chan has encouraged countries to now review their pandemic preparedness, while the independent review of the worldwide response to the pandemic, including the performance of WHO and the functioning of the International Health Regulations, is expected to be completed within 6 months. At the heart of the lessons learned is communication, say experts. Worldwide, “we are now moving into the post-pandemic period”, Chan told a press conference on Aug 10. “Based on experience with past pandemics, we expect the H1N1 virus to take on the behaviour of a seasonal influenza virus and continue to circulate for some years to come”, she predicts. Key objectives for the post-pandemic period are prevention and control of outbreaks, together with monitoring and surveillance. WHO recommends vaccination against H1N1 in either monovalent (H1N1 strain) or trivalent vaccine formulations (which includes
Margaret Chan speaking after the announcement of the post-pandemic phase
664
other seasonal strains: influenza H3 and influenza B), on the basis of what is available. Depending on national advice, vaccination needs to be deemed to prevent severe disease in the younger age groups affected, including pregnant women. Recommended monitoring includes early detection of signal events, monitoring for severe respiratory disease, monitoring groups at highest risk of infection, and virologic surveillance for mutation and antiviral resistance. Chan commended health authorities in India and New Zealand, whose actions “provide a model of how other countries may need to respond in the immediate postpandemic period”. New Zealand and India are a good model, Chan told reporters, because they maintain vigilance at a time when many countries have become “quite complacent”. New Zealand director of public health, Mark Jacobs, told TLID that New Zealand “doesn’t consider itself to be ‘post-pandemic’ just yet”, due to a current second wave of H1N1. Health authorities have essentially continued measures put in place after containment failed in June, 2009, says Jacobs. He cites measures to monitor the progress of the second wave and the effect on health services, ready access to antiviral medication, and regular community healthpromotion information, including encouragement to take up seasonal influenza vaccine. Jacobs says that lessons learned in New Zealand surround the need to reinforce the importance of successful measures, such as detailed, multisectoral planning, high-level ongoing surveillance, regular and open communication, and that basic public health and hygiene measures can make a difference. In terms of the WHO pandemic preparedness guidelines, Jacobs said “I think the guidelines have held up pretty well, but a couple of areas could
be strengthened: recognition that pandemics can differ in their severity, and criteria for deciding when we move to a post-pandemic phase”. He also called for better recognition of circumstances when it might make sense for individual countries to follow containment strategies to slow down the entry and spread of a pandemic virus. “It won’t always be practical”, he added, “but where it is then our experience is that it can be useful in buying time to learn more about the virus, and for health and other key systems to become even better prepared.” Representations by member states to the first meeting of the independent review committee in April, heard these concerns from member states’ representatives. First, said chair Harvey Fineberg, president of the US Institute of Medicine, “it was very important to recognise the challenge of containment as a strategy early in the epidemic. Countries also raised questions about the characterisation of the pandemic phases, and specifically, how the question of severity relates to the definition of phases.” A third issue was equity, he said, and the response in richer versus poorer countries. Although some countries had stockpiles of vaccine, in others, such as Bangladesh, vaccine doses were delivered by WHO just before the pandemic was declared over. Worldwide agreement exists among experts of the need to review the pandemic phases, including the severity. The 2009 definition of the pandemic focuses on extent instead of severity, which is difficult to measure and combine with extent, says Fineberg. Chan proposed that one lesson learned was the need for more flexibility in future pandemic planning. Rather than “planning for the worst, hoping for the best”, she envisioned planning for a best-case scenario, an intermediate scenario,
www.thelancet.com/infection Vol 10 October 2010
and a worst-case scenario. But she denied the suggestion that WHO had hyped this pandemic. “Did we overreact? The answer was no, we did not”, she responded to reporters. Only when precise fatality figures are available in several years, will “history and records show whether or not the estimate by the world’s experts is on the mark”, she said. “The main lesson to be learned here, which is widely underestimated, is that influenza is the most unpredictable disease”, Giuseppe Cornaglia, president of the European Society for Clinical Microbiology and Infectious Diseases, told the journal. This explains how models could fail, and he said it is “too early to say that the forecasts were too pessimistic and it remains a possibility that many forecasts were correct but the timing is wrong”. Since H1N1 persists and coexists with seasonal influenza variants and with avian influenza H5N1, “the possibility of reassortment and further evolution is still there”, he explains, predicting that “we have not seen the last page of the story”. Moreover, he adds, “the whole pandemic was not so dramatic but in countries with an impaired public health system and in immuneimpaired patients, the impact was greater”. WHO stressed the effect on otherwise healthy young adults and recent reports suggest greater severity in preterm babies, pregnant women, and other vulnerable groups. “Overall”, Cornaglia concludes, “the effect of H1N1 was not that dramatic but it was not normal seasonal influenza.” The independent committee is now reviewing documents and having interviews with a view to publication at the World Health Assembly in May, 2011. An unusual aspect of the committee’s work is that representatives of member states and intergovernmental and nongovernmental organisations might attend meetings. Having interviewed key experts, the committee is expected to start talking to responsible WHO officials as the journal went to press. www.thelancet.com/infection Vol 10 October 2010
“We want to offer recommendations that enable the world, WHO, and the nations to do better the next time”, Fineberg has stated. Chan told reporters that the review will provide very valuable lessons that can feed back into the next pandemic planning. Number one, for WHO and for many countries, she said, is communication. “We need to adapt the way we do communication and learn how to work with social media and social network.” A separate subcommittee will review broad communication issues, although WHO is doing an internal review. Better communication is vital between all parties, says Cornaglia, including experts, society, governments, and lay people. “We need to look at communication in the coordination of the response”, and also, he says, “it has not been highlighted that the vaccine has been used safely in millions.” The most essential communication needed is to doctors, he says, “who did not themselves believe in the vaccine”. Refusal of vaccination was 80% in the UK, and higher in some other countries, so health-care workers need education on vaccination, he advocates. A major communication concern was the failure of WHO to release the names and interests of the emergency committee until after the end of the pandemic, leading to accusations of undue influence from the pharmaceutical industry. The review committee has now heard evidence from Paul Flynn, former special rapporteur on a critical Council of Europe report, published in June. Flynn welcomed the publication on the emergency committee members. “A year ago”, he said, “these 15 experts told us the world was facing a grave health emergency. On this occasion, regrettably, they got it wrong: their advice led WHO, the European Union, and national governments to vastly overrate the seriousness of the H1N1 epidemic,
AP/Press Association Images
Newsdesk
The risk of reassortment between H1N1 and H5N1 remains
wasting large sums of public money and scaring the world unnecessarily.” Utmost transparency is vital, says Cornaglia, since it is difficult to find experts that are completely unrelated to industry. Greater transparency, says Flynn, “would help to restore WHO’s credibility—and ensure that its advice is taken seriously when the next pandemic strikes”. From the start, Fineberg urged review committee members to expose “predispositions” that everyone has, due to experience, knowledge, or positions held. Additionally, the committee will recuse experts from specific parts of the process, where necessary. Two experts on the original membership list of the review committee have now been identified as WHO emergency committee members. For the future, “the key is very regular and very open communication, including on the things we don’t know”, says Jacobs. “Of course”, he notes, “there will always be differing views on how best to manage a public health threat, particularly with the benefit of hindsight.” Fineberg has said that the review committee intends to avoid applying “a mindless retrospective view” and instead judge decisions made from a “prospective posture, understanding what was known, and ... what was knowable”.
Kelly Morris 665
Newsdesk
More harm than good? For the Herald-Mail article see http://www.herald-mail. com/?cmd=displaystory&story_i d=249796&format=html&autor eload=true For examples of newspaper coverage of English cases see http://www.thesun.co.uk/sol/ homepage/news/article153799. ece and http://www.dailymail. co.uk/news/article-201850/ Asylum-seeker-HIV-infectioncharge.html For the CPS Code for Crown Prosecutors see http://www.cps. gov.uk/publications/docs/ code2010english.pdf
Boris Roessler/DPA/Press Association Images
For CPS Legal Guidance on cases involving intentional or reckless transmission of infection see http://www.cps. gov.uk/legal/h_to_k/ intentional_or_reckless_sexual_ transmission_of_infection_ guidance/index.html
On Aug 26, a German court convicted HIV-positive pop singer Nadja Benaissa of one count of causing grievous bodily harm (GBH) and two counts of attempted bodily harm. The charges related to her engaging in unprotected sex with three men over 4 years, one of whom contracted HIV as a consequence. She was issued with a 2-year suspended sentence along with 300 h community service. The judge commented that Benaissa—who had faced the possibility of 10 years imprisonment—“had learned to be responsible and deal with her illness”. Several countries maintain HIVspecific laws. Others use existing laws to criminalise certain types of behaviour. Most commonly, cases involving HIV concern reckless—as opposed to intentional—transmission of the virus. Last year in Canada, for example, Johnson Aziga was convicted of two counts of first-degree murder, ten counts of aggravated sexual assault, and one count of attempted aggravated assault after having sex with 13 women without informing them of his status. Seven of these women tested positive for HIV, and two of them later died. Under Canadian law, those who are unaware that a partner has HIV are not deemed to have given consent to sex, hence the assault charges. In England and Wales, cases involving viral transmission are
Nadja Benaissa facing the verdict at the Regional Court in Darmstadt, Germany
666
charged under the GBH section of the Offences Against the Person Act (1861). No crime is committed unless there is infection; exposing someone to the chance of infection is not unlawful (in Scotland, “reckless endangerment” legislation covers such instances). All 12 individuals in England and Wales convicted for transmitting HIV have been convicted for reckless transmission. Under these charges, the prosecution has to prove that the defendant infected the complainant, the defendant knew they were taking an unjustifiable risk (recklessness), and the complainant did not give informed consent. Sarah Jane Porter was reported to the authorities by an ex-boyfriend— himself negative for HIV—who had discovered her status from a third party. The police examined her medical records, and contacted expartners advising them to undergo testing. “It was a fishing expedition”, explains Yusef Azad (National AIDS Trust, London, UK), “and it put the sexual history of anyone with HIV as fair game for the police.” One of Porter’s ex-boyfriends found he was HIV-positive and pressed a complaint. She pleaded guilty to GBH in 2006, and was sentenced to 32 months imprisonment. One newspaper dubbed her the “AIDS Avenger”. Many prosecutions for HIV-related crimes are accompanied by such inflammatory media coverage, which commonly misrepresents the realities of the virus. In an article on the conviction in July of an HIV-positive man from Maryland, USA, for seconddegree assault after he unintentionally spat on a police officer, the HeraldMail did not mention that HIV cannot be transmitted this way. Matthew Weait (Birkbeck University, London, UK) cites the first few cases for reckless transmission in England, where African migrants faced court proceedings to the backdrop of media outrage over immigration from the
region. “Individual cases are framed as moral narratives. It’s the easiest story to tell in a newspaper”, Weait told TLID. England and Wales’ Crown Prosecution Service (CPS) states that, assuming there is sufficient evidence to raise a case against an individual, “a prosecution will usually take place unless the prosecutor is sure that there are public interest factors tending against prosecution which outweigh those tending in favour”. In other words, the CPS decides to prosecute only when they believe this will serve the public good. The CPS prosecutes 1 million cases every year, yet in the past 7 years only 17 people have been brought to court in England and Wales for reckless transmission of a virus (in addition to HIV, there has been one prosecution each for hepatitis B and hepatitis C). There is the possibility that fear of such prosecution might deter others from risky behaviour. And notifying the public of those who have a history of reckless sexual behaviour might not be such a bad thing. Admittedly, there are questions over the rehabilitative value of prison for such individuals and whether education might be a more effective remedy. UNAIDS emphasises that “prevention of HIV must be the primary objective of the policy of criminalisation”. There is little evidence that criminal law is effective in altering sexual practices. Moreover, it is conceivable that individuals might be deterred from visiting counselling and medical services for all sexually transmitted diseases, fearful that their records might subsequently be pored over by the police. “You can treat this like any kind of violent crime”, says Weait, “or you can think of HIV as a public health issue.” There are widespread fears that criminalisation works against the broader interests of public health. But if HIV is to be treated purely as
www.thelancet.com/infection Vol 10 October 2010
Newsdesk
a public health issue, will there not be those deserving of punishment who escape justice? CPS guidelines point out that the consequences for the “victim” of contracting a sexually transmitted disease fit the criteria for grievous bodily harm. And what of those cases in which there is deliberate deception—an individual who explicitly claims to be negative and goes on to infect their partner? However, there is a risk that prosecutions will encourage the impression that precautions are the prerogative of the infected, eroding efforts to foment a culture of shared responsibility. “One unhelpful message is that HIV is spread only by people who know they have it, and that most people with diagnosed HIV will routinely tell you they have it”, says Azad. Most infections are attributable to the undiagnosed, and to suggest that a handful of criminally liable
people are to blame for the HIV/AIDS epidemic sidelines the responsibilities of the wider population. There are questions of justice. Is not the decision to have unprotected sex an implicit acceptance of the associated risks? What about cases where an HIV-positive defendant used a condom but it split? Or their antiretroviral regimen rendered their viral load undetectable? Weait points out that complaints have tended to be brought by aggrieved ex-partners, but many couples stay together after one infects the other, and no complaint is made. The number of potential criminals in the UK is likely to be several thousand. “There’s been a significant move towards criminalisation both in the UK and other countries”, Azad concludes, “but public health voices are beginning to be heard.” Prosecutorial guidelines demanding a high threshold of evidence have
reduced cases brought to court in England and Wales. Crucial to this is the burgeoning judicial awareness of the limitations of phylogenetic analysis, which cannot reliably establish the direction or source of HIV infection. New police guidelines, agreed with the National AIDS Trust, should prevent a recurrence of the circumstances surrounding the Porter case, and decrease the number and length of investigations. Still, it is hard to shake the idea that prosecutions reinforce stigma and create an impression of those living with HIV as being of especial threat to the general public. Azad believes this is counterproductive. “Stigma is the main reason people find it hard to declare their HIV status; we would do far better address this than to criminalise people.”
Talha Burki
Infectious disease surveillance update West Nile virus in Europe West Nile virus has been reported in several European countries since the initial outbreak in Greece reported on Aug 7. The Greek outbreak is largely limited to the central Macedonia region, with a few cases in the neighbouring district of Larissa. As of Sept 3, 173 cases (all laboratory confirmed) had been reported and there had been 15 deaths. In Romania, health authorities have reported 13 confirmed cases and two probable cases in the past 2 months. Two people with the infection, both older than 75 years, have died. The median age of the people infected was 50 years (range 18–79) and they were from 11 different districts. The Hungarian health authorities have confirmed three cases of infection: two people living in central Hungary and one close to the Romanian border. One case of infection has also been confirmed in the Veneto region of Italy www.thelancet.com/infection Vol 10 October 2010
and one probable case was identified in Portugal. Outside of the European Union, from July to August, 2010, 231 cases of West Nile fever were reported in Russia, including six deaths. On Aug 10, 24 confirmed cases were reported in Israel, with six further presumed cases. A more recent case was imported into Israel from the Netherlands. Most cases in Israel have been in the Central and Tel Aviv region, but there have also been isolated cases reported in the Haifa region.
Encephalitis in India On Sept 8, health officials in India confirmed that a further five people had died due to encephalitis, bringing the total number of deaths in Uttar Pradesh’s eastern region to 254 this year. The outbreak of encephalitis this year began unusually early, by mid-April 210 patients had already been treated in this region. In many
cases the cause has not been specified. Some of the cases are due to Japanese encephlitis virus (JEV) and others have been described as acute encephalitis of undefined aetiology but not due to JEV infection. In 2009, more than 500 cases of encephlitis were reported in the region. Again, only some of these cases were attributed to JEV.
Legionnaires disease in the UK As of Sept 12, 19 cases of legionnaires disease have been identified in Wales in the UK. All patients have needed treatment in hospital. A further four people are being investigated to establish if they are linked to the outbreak. Two of these four people have died: a 70-year-old man who died on Sept 8 and a 64-year-old woman on Sept 6. The source of the outbreak has not yet been identified.
For more on WNV in Europe see http://www.ecdc.europa.eu/en/ activities/sciadvice/Lists/ ECDC%20Reviews/ECDC_ DispForm.aspx?List=512ff74f77d4-4ad8-b6d6-bf0f23083f30 &ID=938&RootFolder=%2Fen%2 Factivities%2Fsciadvice%2FLists %2FECDC%20Reviews For ECDC’s WNV threat assessment see http://www. ecdc.europa.eu/en/healthtopics/ Documents/1009_Threat%20 assessment_West_Nile_Virus.pdf For more on legionnaires disease in Wales see http:// www.wales.nhs.uk/ sitesplus/888/news/17027
Onisillos Sekkides 667
Newsdesk
On Aug 10, 2010, WHO declared the end of the influenza pandemic, just over a year since it began. Here is a reminder of how events unfurled. Early 2009—The earliest confirmed case of infection with what WHO subsequently dubbed pandemic H1N1 2009 virus was a 6-month-old girl from San Luis Potosi state, Mexico with symptom onset on Feb 24. Almost 2 months later, on April 19, Mexico declared a national alert. March and April—Mexico experienced outbreaks of respiratory illness and reports of patients with influenzalike illness came from throughout the country. On April 23, several cases of severe respiratory illness were confirmed as infection with what was then referred to as swine-origin influenza A H1N1 virus. Sequence analysis revealed that this was the same strain that had infected two children in California, USA—in a 10-year-old child identified on April 15, and 2 days later in an 8-year-old child living 130 miles away from the first. There was no known connection between these two children. From March 1 to April 30, 1918 suspected cases and 84 deaths were reported in Mexico. Most case reports were of patients admitted to hospital because surveillance efforts were focused there. This potentially skewed initial reports of mortality. April 25—WHO Director General, Margaret Chan, declared the outbreak a health emergency of international concern. By April the US federal government had purchased 50 million courses of oseltamivir and zanamivir for the strategic national stockpile. April 27—Chan raised the pandemic alert from phase 3 (sporadic cases or small clusters insufficient to support a community-level outbreak) to phase 4 (verified human-to-human transmission able to cause a community-level outbreak). By April 28, seven countries had reported confirmed cases of H1N1 infection: Canada, Israel, Mexico, New Zealand, Spain, the UK, and the USA. www.thelancet.com/infection Vol 10 October 2010
April 29—the alert was raised to phase 5 (human-to-human spread of the virus into at least two countries in one WHO region). At this point, WHO requested all countries immediately activate their pandemic preparedness plans. The first death attributed to H1N1 infection outside Mexico occurred in the USA. April 30—Austria, the Netherlands, and Switzerland joined the list of countries with confirmed cases. And between May 2 and May 6, China, Colombia, Costa Rica, Denmark, El Salvador, France, Germany, Guatemala, Italy, Ireland, Portugal, Korea, and Sweden were added. June 3—a case of H1N1 influenza was confirmed in Egypt, the first in Africa. Since Egypt also had the greatest number of cases of avian influenza, fears were expressed that the two viruses might recombine to form what some commentators dubbed the “Armageddon virus”. 16 days later pandemic H1N1 was confirmed in South Africa, marking its spread into sub-Saharan Africa. 25 days later there was the first death in sub-Saharan Africa attributed to the virus. June 11—WHO raised the pandemic alert to phase 6 (community-level outbreaks in at least one other country in a different WHO region in addition to the criteria defined in phase 5) signalling the start of the pandemic. And 5 days later the organisation abandoned the issuing of tables showing the confirmed cases for all countries, stating that the increasing number of cases “is making it extremely difficult, if not impossible, for countries to try and confirm them through laboratory testing”. July 2—Chan described the pandemic as “unstoppable” in a meeting in Cancun, Mexico. Early in the same month, strains of the virus resistant to oseltamivir were found in patients in Denmark and Japan that had been treated with the drug. This did not raise concern since resistant viruses that arise after treatment often lack the ability
Anja Niedringhaus/AP/Press Association Images
Pandemic influenza—a timeline
Margaret Chan at WHO headquarters announcing entry into pandemic phase 6
to infect other people. However, a third resistant virus was also identified in a patient in Hong Kong who had returned from California and who had not been given oseltamivir, suggesting that a resistant strain of the virus was circulating in the USA. Late 2009—By October the pandemic had passed its peak, and, although overall numbers of cases were declining, transmission was sustained in many regions of the world. Feb 18, 2010—WHO published recommendations for the composition of influenza virus vaccines for the upcoming northern hemisphere influenza season. The trivalent vaccine will include a pandemic H1N1-like component. 2010—Over the course of the year, incidence continued to decline, although the virus has persisted in some regions with transmission most intense in parts of India and the temperate southern hemisphere, particularly New Zealand and more recently Australia. With WHO’s announcement that the pandemic has ended we have now entered the so-called post-pandemic phase where virus activity is deemed to have returned to seasonal levels. We are left to ponder whether the responses to the pandemic will be sufficient should a more lethal strain of influenza emerge in the coming years.
For more on the H1N1 pandemic see http://www.who. int/csr/disease/swineflu/en/ index.html For the US CDC’s summary see http://www.cdc.gov/h1n1flu/ cdcresponse.htm For more on avian influenza see http://www.who.int/csr/disease/ avian_influenza/en/
Onisillos Sekkides 663
Newsdesk
Research brief For more on HIV-1 adaption to humoral immunity see Nat Med 2010; 16: 995–97; DOI:10.1038/nm.2203 For more on a possible viral cause for chronic fatigue syndrome see Proc Natl Acad Sci USA 2010; 107: 15874–79; DOI:10.1073/ pnas.1006901107 For more on antisense therapy for filoviruses see Nat Med 2010; 16: 991–94; DOI:10.1038/ nm.2202 For more on altruistic bacteria see Nature 2010; 467: 82–85; DOI:10.1038/nature09354 For more on an antibacterial peptide treatment for A baumannii infections see J Antimicrob Chemother 2010; published online Sept 1; DOI:10.1093/jac/dkq337 For more on a potential new antimalarial drug class see Science 2010; 329: 1175–80; DOI:10.1126/science.1193225 For more on why mosquitoes don’t have malaria see Science 2010; 329: 1353–55; DOI:10.1126/science.1190689
Adapt to survive
Filovirus antisense therapy
Most individuals infected with HIV-1 develop neutralising antibodies against the gp120 and gp41 subunits of the viral envelope. Now, a comparison of antibody neutralisation of HIV-1 variants obtained from people infected between 1985 and 1989 and between 2003 and 2006 suggests that the virus has evolved towards a more antibody neutralisation-resistant phenotype. This enhanced resistance, which coincides with poorer elicitation of neutralising antibody responses, is accompanied by an increase in the length of the variable loops of gp120 that could mask vulnerable epitopes in gp120 from antibody recognition.
Currently there is no treatment for Ebola and Marburg virus infections. However, in a new study, treatment with positively charged phosphorodiamidate morpholino oligomers (PMOplus) directed against Zaire Ebola virus transcripts protected more than 60% of rhesus monkeys when given 30–60 min after infection with this lethal filovirus. Similarly, PMOplus directed against Lake Victoria Marburg virus protected 100% of cynomolgus monkeys against Marburg virus infection. This type of antisense therapy might, therefore, be a useful treatment for human infection with these highly pathogenic viruses.
Always tired? For years, researchers have been trying to discover whether an infection triggers chronic fatigue syndrome (CFS). In a recent attempt to find a cause for CFS, researchers found mouse leukaemia virus (MLV)-related gag gene sequences in 32 (86·5%) of 37 peripheral blood mononuclear cell DNA samples taken in the 1990s from patients with CFS but in only three (6·8%) of 44 healthy blood donors. Seven of eight gag-positive patients tested positive again nearly 15 years later but, the researchers stress, these results do not prove that MLVrelated viruses are involved in the development of CFS.
Charity starts at home Bacteria have a charitable streak that spreads antibiotic resistance, claim researchers. In continuous cultures of Escherichia coli exposed to increasing antibiotic concentrations, they report, most isolates are less antibiotic resistant than the population as a whole. However, a few highly resistant mutants improve the survival of the population by producing indole, a signalling molecule that turns on drug efflux pumps and oxidative stress mechanisms throughout the culture. Thus, a few spontaneous drugresistant mutants boost the culture’s overall fitness by incurring a personal fitness cost. Further studies of this altruistic behaviour should aid the design of more effective antibiotics.
NIBSC/Science Photo Library
Peptide burns off infection
A few drug-resistant individuals can improve the survival of E coli populations
668
Acinetobacter baumannii is a common cause of systemic infection in soldiers returning from Iraq and Afghanistan with burn or blast wounds. Currently these infections are treated with imipenem or colistin, but resistance to these drugs is increasing and colistin is very toxic. Now, an animal study suggests that a proline-rich antibacterial peptide called A3-APO
could provide an effective treatment. Intramuscular injection of A3-APO reduced bacterial counts in the blood and wounds of mice that had had burn wounds inoculated with A baumannii isolated from an injured soldier better and with less toxicity than imipenem or colistin.
New antimalarials The emergence of artemisininresistant malaria parasites means there is an urgent need for new antimalarial drugs with a different mechanism of action. Spiroindolones (spirotetrahydro-β-carbolines), a class of molecules recently identified in a chemical library screen, could fulfil this need. The optimised spiroindolone, NITD609, which inhibits protein synthesis in Plasmodium falciparum, kills blood-stage P falciparum and Plasmodium vivax clinical isolates and P falciparum strains resistant to other antimalarial drugs at nanomolar concentrations. NITD609 also has pharmacokinetic properties consistent with once-daily oral dosing and is effective in a rodent malaria model.
Mosquitoes don’t get malaria Most Anopheles gambiae mosquitoes, the major malaria vector in Africa, are not infected with malaria parasites even though the female insects are frequently exposed to the parasite through their human blood meals. Researchers investigating innate immunity in mosquitoes have now solved this puzzle. Invasion of the mosquito midgut by malaria ookinetes, they report, disrupts the barriers that normally prevent gut bacteria coming into direct contact with epithelial cells. The bacteria then trigger a long-lived antibacterial response that reduces the survival of malaria parasites on rechallenge, a discovery that could have implications for malaria control.
Jane Bradbury www.thelancet.com/infection Vol 10 October 2010
Media Watch
Film The vaccine war “For the first time, despite repeated reassurances to the contrary; despite convincing scientific evidence confirming the safety of the vaccine programme; despite serious criticisms of the probity of the source of the [MMR] rumours; some sections of the public have decided that they know better, and that they do not need to follow the advice of experts”.
“Polio: gone. Hib: gone. Diphtheria: gone”, enthuses Paul Offit, one of the inventors of the rotavirus vaccine, early on in The Vaccine War. There are 16 diseases on the US childhood and adolescent vaccination roster, adds the US Centres for Disease Control and Prevention’s Melinda Wharton. “It’s a really good thing”, she says. But not everyone sees it that way. As the nurse prepared to give her daughter a hepatitis B inoculation, Jennifer Margulis was outraged. “Why am I supposed to vaccinate my new born baby against a sexually transmitted disease?” she demanded. There is a constituency of American parents who are inclined to agree with Margulis, and if the movement gathers pace, the consequences could be severe. Take measles, a disease more infectious than smallpox or influenza. Following publication of now discredited findings suggesting a link between the measles, mumps, and rubella vaccine in 1998, the UK saw a pronounced fall in vaccine coverage. In Dublin, coverage had fallen to 79% by 2000. The same year, an outbreak of measles saw 110 infants taken to hospital. Three of them died. “And measles is a preventable disease”, sighs Eric Frombonne in The Vaccine War. An interesting and concise film, written, directed, and produced by John Palfreman, much of The Vaccine War centres on Ashland, Oregon—a college town of around 20 000 people located near the border with California. Ashland is known for its yearly Shakespeare Festival and its organic food stores. It also has one of the lowest rates of vaccination in the USA, with an estimated 28% of children entering kindergarten incompletely inoculated. “It’s an outbreak waiting to happen”, frets one mother. Jenny McCarthy, an actress and model, seems to be the symbolic figurehead of the antivaccine movement. The Vaccine War gives her, and other opponents of America’s admittedly hefty vaccination schedule, plenty of room to make their case. The film is largely focused on the entirely unproven link between autism and vaccination: if not MMR itself, then the overall number of inoculations, or the mercury-based compound thimerosal. “It doesn’t take a rocket scientist to draw a possible correlation between the vaccines being given and the brain injuries that our children are suffering”, insists one father. No, it does not. But it does take a scientist. And what seems to be so disturbing about the growing ranks of parents who distrust vaccinations is their reluctance to www.thelancet.com/infection Vol 10 October 2010
Frontline
Quiet Killers. The Fall and Rise of Deadly Diseases (2007). Robert Baker.
Jenny McCarthy—anti-vaccine activist
credit the word of science. “You don’t have any science that can show me that [my son’s] regression wasn’t caused by the six vaccines”, says one participant, oblivious to the absurdity of demanding proof of a negative. Part of this reaction is to be expected, of course, most people are not equipped to apprehend complicated matters of science. One mother politely waits for Ashland’s public health officer to explain the idea of herd immunity, and how her decision not to have her own child vaccinated could cause others to fall sick. “I really don’t believe it”, she responds. Besides, parents with disabled children are under a great deal of stress and desperate to find an explanation for their child’s disability. “It is entirely reasonable to ask whether MMR causes autism”, stresses Offit. Fombonne points out that autism does tend to develop at the same age as children begin receiving vaccinations, for many this is all too tempting a correlation. Furthermore, this is a generation that didn’t see the damage wrought by, say, polio. For them vaccination itself has become the problem. Finally, and perhaps most damagingly, there seems to be a general inability to sift information gathered from the internet, so that the word of respected scientists may not be considered any more valuable than that from of any number of bizarre websites. Nonetheless, a peculiar remark from McCarthy is rather telling. “We keep saying ‘study it’ but they won’t”, she says, brimming with irritation. The Vaccine War lists several studies from around the world that have disproved the various theories on autism that have prompted suspicion of vaccination. “Scientifically the matter is settled”, says Danish researcher Anders Hviid. “But there are a lot of people who don’t listen to science, who don’t understand it, and they are going to be difficult to sway with the scientific arguments.”
The Vaccine War Written, directed, and produced by John Palfreman. FRONTLINE. To view the film online go to http://www.pbs.org/wgbh/ pages/frontline/vaccines/view/
Talha Burki 669
Correspondence
Reducing antibiotic use in Croatia
Vera Vlahović-Palčevski
Benedikt Huttner and colleagues’1 Review discussed the characteristics and outcomes of 22 public campaigns designed to reduce antibiotic use. The campaigns were done at a national or regional level in highincome countries between 1990 and 2007, and aimed to educate the public about appropriate antibiotic use in outpatients. In Croatia, classified as an emerging and developing economy by the International Monetary Fund, but as a high-income economy by the World Bank, a similar regional campaign was done in Primorje-Gorski kotar county (about 305 000 inhabitants), for the first time in February and March 2009, and again in February and March 2010. Croatia is one of several countries with a high antibiotic consumption.2 The decision to launch a campaign was supported by the Croatian Government and at a regional level by the City of Rijeka Government and the Primorje-Gorski kotar County. The approximate cost of a 1 month regional campaign was
€50 000, not including staff salaries. The campaign staff, who included specialists in clinical pharmacology, paediatrics and public health, general practitioners, a psychologist, pharmacists, and medical students, were voluntary. Like most other campaigns, this campaign focused on respiratory tract infections, and targeted general-care, public-care, and primary-care physicians. The key messages of the campaign were that most respiratory tract infections are caused by viruses and cannot be treated with antibiotics, that misuse of antibiotics promotes bacterial resistance, and that physician advice on how and when to take antibiotics should be followed. The interventions used were multifaceted. With permission, visual identity was taken from the European Centre for Disease Prevention and Control Campaign Communication Materials for the European Antibiotic Awareness Day.3 Interventions that targeted the public included press conferences, media advertisements and interviews, adverts on public transport, leaflets, free telephone helplines, a weekend stand in the main pedestrian zone
Figure: A stand in the main pedestrian zone in Rijeka, Croatia, 2010 After asking for advice on antibiotics, the lady was offered chocolate by children dressed in hedgehog masks.
670
with merchandise carrying the message “cold or flu—no antibiotics” (figure), and physicians advising how to take antibiotics responsibly. Furthermore, lectures were given to high-school students, parents of children attending elementary schools, and teachers in day-care centres. Interventions for healthcare professionals were primarily to inform physicians about the public campaign, and to invite them to take an active role in it, and anonymous questionnaires were distributed to all primary-care physicians within the county. A symposium was organised on the Antibiotic Awareness Day, Nov 18, 2009, with lectures focusing on the treatment of most common infections and the distribution of guidelines for the use of antibiotics in primary care—a comprehensive booklet covering various infections. The second year campaign involved the same activities as the first year. When assessing the outcomes of the campaigns, short-term, midterm, and long-term effects should be identified. We were able to assess only short-term outcomes, such as the knowledge and attitudes of patients to antibiotics. Findings of an anonymous questionnaire, which was administered to the general population, noted that 35% of patients before the first campaign, and 25% after it, thought that antibiotics are drugs to treat influenza, therefore, patient knowledge had improved. Midterm outcomes might be measured by the amount of antibiotics used. Although we did not expect a decrease in antibiotic use after the first campaign, the use of prescribed co-amoxiclav fell by 4%. Long-term outcomes and changes in resistance patterns could not yet be seen. If financially supported, we will continue with educational activities to reduce future unnecessary antibiotic use. I would like to encourage other transitional countries with high antibiotic use to perform similar www.thelancet.com/infection Vol 10 October 2010
Correspondence
Vera Vlahović-Palčevski
[email protected] Unit for Clinical Pharmacology, University Hospital Rijeka, University of Rijeka Medical Faculty, Krešimirova 42 51000, Rijeka, Croatia The campaign is part of the project funded by the Croatian Ministry of Science, Education and Sports (grant number 0062–3537), and is financially supported by the Croatian Ministry of Health and Social Welfare, Governments of the City of Rijeka and County of Primorje and Gorski kotar and Jadran Pharmacies, Rijeka, Croatia. I declare that I have no conflicts of interest. 1
2
3
Huttner B, Goossens H, Verheij T, Habarth S, for the CHAMP consortium. Characteristics and outcomes of public campaigns aimed at improving the use of antibiotics in outpatients in high-income countries. Lancet Infect Dis 2010; 10: 17–31. Goossens H, Ferech M, Coenen S, Stephens P, for the European Surveillance of Antimicrobial Consumption Project Group. Comparison of outpatient systemic antibacterial use in 2004 in the United States and 27 European countries. Clin Infect Dis 2007; 44: 1091–95. European Centre for Disease Prevention and Control. European Antibiotic Awareness Day: campaign communication materials— Nov 18, 2008. http://www.ecdc.europa.eu/ en/eaad/Documents/EAAD_Guidelines_for_ use.pdf (accessed March 9, 2010).
Monitoring and adjusting for future needs in response to changing HIV policies We agree with Jose Castillo-Mancilla and Monica Carten1 that immigrants and refugees from developing countries who are infected with HIV have substantially different medical and social characteristics than those from developed countries. The investigators propose a National HIV in Foreign-born Registry to obtain data for immigrants with HIV who receive care in the USA. Over the past 20 years, we have followed the demographic, clinical health-care use, and cost of care for patients receiving HIV therapy in www.thelancet.com/infection Vol 10 October 2010
the Southern Alberta HIV Cohort. In 2001, Canada updated the Immigration and Refugee Protection Act (IRPA)2 to introduce HIV screening, and to clarify guidelines for the acceptance of refugees and sponsored immigrants. Before being accepted, unsponsored immigrants with HIV should show that they are unlikely to impose a substantial burden on health-care resources. In a subsequent clarification of the IRPA, it was declared that HIV did not, as such, pose a serious enough public-health threat for automatic denial to immigration applicants. The result of this new act and of the global HIV pandemic, has been an increase in immigrants and refugees with HIV to Canada, with most being diagnosed at their immigration medical screening examinations.3 Among new patients to our HIV cohort, there are more foreign-born patients with HIV—from 15·8% in 2001 to 43·3% in 2009—31·6% of whom accounted for all patients in active care on December 31, 2009 (figure); 73% of immigrants and refugees were from sub-Saharan Africa. We agree with Castillo-
Mancilla and Carten that collection of this information is useful because of its clinical implications and usefulness in programme planning. A higher proportion of immigrants and refugees than Canadian-born patients often presented at an advanced stage of disease (ie, CD4 count cells per μL), and required more intense medical care because of the presence of more comorbidities such as tuberculosis, toxoplasmosis, and hepatitis B.3 Foreign-born patients often experience social problems because of their immigration history (eg, refugee camps) and their adjustment to new environments, which leads to the involvement of other health-care disciplines such as social workers. Most foreignborn patients (especially those from sub-Saharan Africa) have HIV-1 non-B subtypes, requiring more customised care. However, once in care, health outcomes and costs for new immigrants and refugees was roughly equivalent to those for Canadian-born patients.4 Because of very substantial social concerns, Castillo-Mancilla and Carten’s national registry1 would best be
100
Proportion of new patients starting care (%)
campaigns to improve antibiotic use and preserve their effectiveness for as long as possible.
80
60
40
20
0 1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Year Africa
Europe
North America
Asia
South America
Oceania
Unknown
Canada
Figure: Proportion of new patients starting care at the Southern Alberta Clinic Categorised by birth country, the proportion of new patients starting care at the Southern Alberta Clinic, Calgary, Canada, from 1995 to 2010.
671
Correspondence
Vera Vlahović-Palčevski
[email protected] Unit for Clinical Pharmacology, University Hospital Rijeka, University of Rijeka Medical Faculty, Krešimirova 42 51000, Rijeka, Croatia The campaign is part of the project funded by the Croatian Ministry of Science, Education and Sports (grant number 0062–3537), and is financially supported by the Croatian Ministry of Health and Social Welfare, Governments of the City of Rijeka and County of Primorje and Gorski kotar and Jadran Pharmacies, Rijeka, Croatia. I declare that I have no conflicts of interest. 1
2
3
Huttner B, Goossens H, Verheij T, Habarth S, for the CHAMP consortium. Characteristics and outcomes of public campaigns aimed at improving the use of antibiotics in outpatients in high-income countries. Lancet Infect Dis 2010; 10: 17–31. Goossens H, Ferech M, Coenen S, Stephens P, for the European Surveillance of Antimicrobial Consumption Project Group. Comparison of outpatient systemic antibacterial use in 2004 in the United States and 27 European countries. Clin Infect Dis 2007; 44: 1091–95. European Centre for Disease Prevention and Control. European Antibiotic Awareness Day: campaign communication materials— Nov 18, 2008. http://www.ecdc.europa.eu/ en/eaad/Documents/EAAD_Guidelines_for_ use.pdf (accessed March 9, 2010).
Monitoring and adjusting for future needs in response to changing HIV policies We agree with Jose Castillo-Mancilla and Monica Carten1 that immigrants and refugees from developing countries who are infected with HIV have substantially different medical and social characteristics than those from developed countries. The investigators propose a National HIV in Foreign-born Registry to obtain data for immigrants with HIV who receive care in the USA. Over the past 20 years, we have followed the demographic, clinical health-care use, and cost of care for patients receiving HIV therapy in www.thelancet.com/infection Vol 10 October 2010
the Southern Alberta HIV Cohort. In 2001, Canada updated the Immigration and Refugee Protection Act (IRPA)2 to introduce HIV screening, and to clarify guidelines for the acceptance of refugees and sponsored immigrants. Before being accepted, unsponsored immigrants with HIV should show that they are unlikely to impose a substantial burden on health-care resources. In a subsequent clarification of the IRPA, it was declared that HIV did not, as such, pose a serious enough public-health threat for automatic denial to immigration applicants. The result of this new act and of the global HIV pandemic, has been an increase in immigrants and refugees with HIV to Canada, with most being diagnosed at their immigration medical screening examinations.3 Among new patients to our HIV cohort, there are more foreign-born patients with HIV—from 15·8% in 2001 to 43·3% in 2009—31·6% of whom accounted for all patients in active care on December 31, 2009 (figure); 73% of immigrants and refugees were from sub-Saharan Africa. We agree with Castillo-
Mancilla and Carten that collection of this information is useful because of its clinical implications and usefulness in programme planning. A higher proportion of immigrants and refugees than Canadian-born patients often presented at an advanced stage of disease (ie, CD4 count cells per μL), and required more intense medical care because of the presence of more comorbidities such as tuberculosis, toxoplasmosis, and hepatitis B.3 Foreign-born patients often experience social problems because of their immigration history (eg, refugee camps) and their adjustment to new environments, which leads to the involvement of other health-care disciplines such as social workers. Most foreignborn patients (especially those from sub-Saharan Africa) have HIV-1 non-B subtypes, requiring more customised care. However, once in care, health outcomes and costs for new immigrants and refugees was roughly equivalent to those for Canadian-born patients.4 Because of very substantial social concerns, Castillo-Mancilla and Carten’s national registry1 would best be
100
Proportion of new patients starting care (%)
campaigns to improve antibiotic use and preserve their effectiveness for as long as possible.
80
60
40
20
0 1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Year Africa
Europe
North America
Asia
South America
Oceania
Unknown
Canada
Figure: Proportion of new patients starting care at the Southern Alberta Clinic Categorised by birth country, the proportion of new patients starting care at the Southern Alberta Clinic, Calgary, Canada, from 1995 to 2010.
671
Correspondence
undertaken with the context of care provision. For routine clinical practice, clinical and laboratory adjustments to care such as the recognition of posttraumatic stress, acknowledgement of social pressure for childbearing and breast feeding, routine viral subtype testing for recognition of possible viral load quantification difficulties,5 and underlying resistance polymorphisms6 might need to be introduced. Extra vigilance and effort is needed towards the social needs of immigrants to ensure timely access and use of health-care services that lead to comparable outcomes between foreign-born and local-born patients. The effect of changes to immigration policy for individuals with HIV is very evident in Canada7 and will continue to evolve in
672
the future. Similar changes will probably occur in the USA after the changes to the 2010 travel ban. Early establishment of an appropriate registry or mechanism to recognise care and prevention needs for immigrants and refugees to the USA could have benefits.
4
5
*Harmut B Krentz, John Gill
[email protected]
6
Southern Alberta Clinic, Sheldon M Chumir Health Centre, Calgary, AB T2R OX7, Canada 1
2
3
Castillo-Mancilla JR, Carten M. The need for a US registry of HIV in foreign-born people. Lancet Infect Dis 2010; 10: 370–71 Citizenship and Immigration Canada. Immigrant and Refugee Protection Act. 2001. http://laws.justice.gc.ca/PDF/ Statute/I/I-2.5.pdf (accessed May 25, 2010). Krentz HB, Gill MJ. Clinical, demographic, and health resource utilization characteristics of HIV-infected sub-Saharan immigrants and refugees followed at the Southern Alberta Clinic. Can J Infect Dis Med Microbiol 2006; 16: 255P. http://www.pulsus.com/cahr2006/ abs/255.htm (accessed May 25, 2010).
7
Krentz H, Gill MJ. The five-year impact of an evolving global epidemic, changing migration patterns, and policy changes in a regional Canadian HIV population. Health Policy 2009; 90: 296–302. Church D, Lloyd T, Klein M, Beckthold B, Laupland K, Gill J. Beyond B: HIV-1 viral load multi-assay comparison in a cohort of Canadian patients with diverse HIV subtype infections. 17th Conference on Retroviruses and Opportunistic Infections; San Francisco; Feb 16–19, 2010; 944. http://retroconference. org/2010/PDFs/944.pdf (accessed May 25, 2010). Manosuthi W, Butler DM, Pérez-Santiago J, et al. Protease polymorphisms in HIV-1 subtype CRF01_AE represent selection by antiretroviral therapy and host immune pressure. AIDS 2010; 24: 411–16. Public Health Agency of Canada. Population-Specific HIV/AIDS Status Report: people from countries where HIV is endemic—black people of African and Caribbean descent living in Canada. 2009. http://www.phac-aspc.gc.ca/aids-sida/ publication/ps-pd/africacaribbe/pdf/psspreport-eng.pdf (accessed May 25, 2010).
www.thelancet.com/infection Vol 10 October 2010
Articles
Effectiveness of five artemisinin combination regimens with or without primaquine in uncomplicated falciparum malaria: an open-label randomised trial Frank Smithuis, Moe Kyaw Kyaw, Ohn Phe, Thein Win, Pyay Phyo Aung, Aung Pyay Phyo Oo, Arkar Linn Naing, Mya Yee Nyo, Naing Zaw Htun Myint, Mallika Imwong, Elizabeth Ashley, Sue J Lee, Nicholas J White
Summary Background Artemisinin-combination therapy (ACT) is recommended as first-line treatment of falciparum malaria throughout the world, and fixed-dose combinations are preferred by WHO; whether a single gametocytocidal dose of primaquine should be added is unknown. We aimed to compare effectiveness of four fixed-dose ACTs and a loose tablet combination of artesunate and mefloquine, and assess the addition of a single gametocytocidal dose of primaquine. Methods In an open-label randomised trial in clinics in Rakhine state, Kachin state, and Shan state in Myanmar (Burma) between Dec 30, 2008, and March 20, 2009, we compared the effectiveness of all four WHO-recommended fixed-dose ACTs (artesunate–mefloquine, artesunate–amodiaquine, dihydroartemisinin–piperaquine, artemether– lumefantrine) and loose artesunate–mefloquine in Burmese adults and children. Eligible patients were those who presented to the clinics with acute uncomplicated Plasmodium falciparum malaria or mixed infection, who were older than 6 months, and who weighed more than 5 kg. Treatments were randomised in equal numbers within blocks of 50 and allocation was in sealed envelopes. All patients were also randomly assigned to receive either a single dose of primaquine 0·75 mg base/kg or not. Patients were followed up for 63 days. Treatment groups were compared by analysis of variance and multiple logistic regression. The primary outcome was the 63 day recrudescence rate. This study is registered with clinicaltrials.gov, number NCT00902811. Findings 155 patients received artesunate–amodiaquine, 162 artemether–lumefantrine, 169 artesunate–mefloquine, 161 loose artesunate–mefloquine, and 161 dihydroartemisinin–piperaquine. By day 63 of follow-up, 14 patients (9·4%; 95% CI 5·7–15·3%) on artesunate–amodiaquine had recrudescent P falciparum infections, a rate significantly higher than for artemether–lumefantrine (two patients; 1·4%; 0·3–5·3; p=0·0013), fixed-dose artesunate–mefloquine (0 patients; 0–2·3; p<0·0001), loose artesunate–mefloquine (two patients; 1·3%; 0·3–5·3; p=0·0018), and dihydroartemisinin–piperaquine (two patients 1·3%; 0·3–5·2%; p=0·0012). Hazard ratios for re-infection (95% CI) after artesunate–amodiaquine were 3·2 (1·3–8·0) compared with the two artesunate–mefloquine groups (p=0·01), 2·6 (1·0–6–0) compared with artemether–lumefantrine (p=0·04), and 2·3 (0·9–6·0) compared with dihydroartemisinin– piperaquine (p=0·08). Mixed falciparum and vivax infections were common: 129 (16%) had a mixed infection at presentation and 330 (41%) patients had one or more episodes of Plasmodium vivax infection during follow-up. The addition of a single dose of primaquine (0·75 mg/kg) reduced P falciparum gametocyte carriage substantially: rate ratio 11·9 (95% CI 7·4–20·5). All regimens were well tolerated. Adverse events were reported by 599 patients, most commonly vomiting and dizziness. Other side-effects were less common and were not related to a specific treatment.
Lancet Infect Dis 2010; 10: 673–81 Published Online September 9, 2010 DOI:10.1016/S14733099(10)70187-0 See Reflection and Reaction page 654 Médecins sans Frontières— Holland, Yangon, Myanmar (F Smithuis MD, M Kyaw Kyaw MBBS, O Phe BSc, T Win MBBS, P P Aung MBBS, A P P Oo MBBS, A L Naing MBBS, M Y Nyo MBBS, N Z H Myint MBBS); Medical Action Myanmar, Yangon, Myanmar (F Smithuis, O Phe); Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand (M Imwong PhD, E Ashley MRCP, S J Lee PhD, Prof N J White FRS); and Centre for Tropical Medicine, Nuffield Department of Clinical Medicine, Churchill Hospital, Oxford, UK (E Ashley, S J Lee, Prof N J White) Correspondence to: Prof Nicholas J White, Faculty of Tropical Medicine, Mahidol University, Wellcome Unit, 420/6 Rajvithi Rd, Bangkok, 10400, Thailand
[email protected]
Interpretation Artesunate–amodiaquine should not be used in Myanmar, because the other ACTs are substantially more effective. Artesunate–mefloquine provided the greatest post-treatment suppression of malaria. Adding a single dose of primaquine would substantially reduce transmission potential. Vivax malaria, not recurrent falciparum malaria, is the main complication after treatment of P falciparum infections in this region. Funding Médecins sans Frontières (Holland) and the Wellcome Trust Mahidol University Oxford Tropical Medicine Research Programme.
Introduction Artemisinin-based combination therapy (ACT) is recommended by WHO for the treatment of uncomplicated falciparum malaria.1 The success of this recent policy change will depend on the efficacy of the combination components, high population coverage, low costs, correct dosing, and ensuring good adherence to prescribed treatment. To improve adherence and acceptability, and prevent one drug being taken without its partner, ACTs www.thelancet.com/infection Vol 10 October 2010
are preferably formulated in fixed-dose combinations.2–5 Four fixed-dose ACTs are now available: two new combinations (artesunate–mefloquine and artesunate– amodiaquine) now join artemether–lumefantrine and dihydroartemisinin–piperaquine. Artemisinin and its derivatives reduce gametocyte carriage,6 but they do not prevent transmission from gametocytaemia present at the time of treatment.7 A single gametocytocidal dose of primaquine was widely 673
Articles
recommended in low transmission areas before the introduction of ACTs.8 With the greater effects of artemisinins on gametocyte carriage there has been uncertainty whether primaquine should be added to ACT regimens.9 This question is of increasing importance as countries move from malaria control to elimination, which will require effective, well tolerated medicines and reduction of transmission. Our aim was to compare the efficacy of the four available fixed-dose ACTs and the currently used loose tablet combination of artesunate with mefloquine and to assess the effectiveness of adding a single gametocytocidal dose of primaquine.
Methods Patients Between Dec 30, 2008, and March 20, 2009, we recruited patients into our open-label randomised study at three clinics in Rakhine state in western Myanmar (Burma),2,3,4,10 two clinics in Kachin state in northern Myanmar, and one clinic in Shan state in northeast Myanmar. Patients older than 6 months who weighed more than 5 kg and presented with acute uncomplicated Plasmodium falciparum malaria (parasite density 500–200 000 parasites per μL) or mixed infection were enrolled into the study after fully informed consent was obtained from them or their carer. Patients were excluded if they were pregnant, had severe malaria, had severe acute malnutrition (weight-for-height below 70% of median with or without symmetrical peripheral oedema), had taken antimalarial drugs within the past 48 h, had taken mefloquine during the past 9 weeks, or had known history of hypersensitivity to any of the study drugs.
Randomisation After patients were screened and enrolled into the study, they were stratified prospectively into three age groups (1–4 years, 5–14 years, and older than 14 years). Patients were randomly assigned in equal numbers to receive one of the five different treatments. They were then randomly assigned either a single dose of primaquine 0·75 mg base/kg (Government Pharmaceutical Organisation, Bangkok, Thailand) or not. Treatment allocations were put in sealed envelopes in blocks of 50 for each age-group, and random assignment was achieved by patients drawing an envelope from a box after enrolment. When the box was empty, another 50 envelopes were added.
Procedures For the four fixed-dose combinations the standard dosage instructions of the manufacturer for weight and age ranges were followed. For the loose tablets regimen of artesunate plus mefloquine, the target dose was artesunate 4 mg/kg per day for 3 days (total 12 mg/kg) plus mefloquine 25 mg base/kg on day 0; the number of pills given was rounded off to the nearest quarter of a 674
tablet. For fixed dose artesunate–mefloquine hydrochloride we used 25 mg plus 55 mg or 100 mg plus 220 mg tablets, the target dose was 4·0 mg/kg per day plus 8·8 mg/kg per day for 3 days. For fixed-dose artemether–lumefantrine we used 20 mg plus 120 mg tablets twice daily for 3 days, the target dose was 3·3 mg/kg per day plus 19·8 mg/kg per day; patients were advised to consume some fatty food (or mothers were encouraged to breastfeed treated infants) before each dose. For fixed-dose dihydroartemisinin– piperaquine we used 40 mg plus 320 mg or 20 mg plus 160 mg tablets, the target dose was 2·5 mg/kg per day plus 20 mg/kg per day. For fixed dose artesunate– amodiaquine we used 25 mg plus 67·5 mg, 50 mg plus 135 mg, or 100 mg plus 270 mg tablets, the target dose was 4 mg/kg per day plus 10·8 mg base/kg per day. The first dose was taken under supervision. For children, tablets were crushed and syrup was added. All subsequent doses were self administered. Patients received sealed plastic bags containing the remaining doses and were instructed clearly about their subsequent treatment, emphasising the importance of taking primaquine after food, and taking their medicines even when their symptoms had subsided. Patients were asked to return weekly for 9 weeks for assessment, and at any other time if they became ill. Microscopists examining blood films were unaware of treatment allocation. Haemoglobin was measured on day 63. Patients with recurrent falciparum malaria were treated with artesunate–mefloquine (loose tablets). Patients who had already received this treatment were given dihydroartemisinin–piperaquine. PCR genotyping was used to distinguish recrudescence from re-infection, by assessment of variable blocks within merozoite surface proteins 1 and 2, and glutamate-rich protein.11 Patients who developed intercurrent Plasmodium vivax or Plasmodium malariae infections received chloroquine (25 mg base/kg) and continued follow-up. If the patients were again positive for P vivax 14 days or later, they were retreated with chloroquine and defined as having a new P vivax episode (as a result of treatment failure, relapse, or, less likely, a new infection).
Statistical analysis A sample size of 160 patients (80 with primaquine) per study group (total 800 patients) allowed a cure rate (100–recrudescence rate) of 95% per group to be estimated with 5% precision, and allowed estimation of effectiveness equivalence with a maximum allowable difference of 10% (90% power and 95% confidence) between groups with a follow-up drop-out rate of up to 20%. Treatment groups were compared by two-way factor analysis: ANOVA for continuous variables and multiple logistic regression for categorical data. A test for trend was used for comparisons across age groups. Time-toevent outcomes, including time to first recurrence (vivax or falciparum), time to recrudescence (new infections www.thelancet.com/infection Vol 10 October 2010
Articles
28 664 patients screened 27 853 ineligible 16 436 no malaria 633 parasitaemia outside range 72 gametocytes only 8103 P vivax only 4 pregnant 2605 refused consent 811 randomly assigned to treatment group
3 excluded from the study due to laboratory error
155 AA (71 with PQ, 84 no PQ) 17 <5 years 56 5–14 years 82 >14 years
162 AL (78 with PQ, 84 no PQ) 17 <5 years 60 5–14 years 85 >14 years
5 lost to follow-up
150 analysed on day 63
169 AM-F (86 with PQ, 83 no PQ) 18 <5 years 63 5–14 years 88 >14 years
7 lost to follow-up
9 lost to follow-up
162 analysed on day 63
153 analysed on day 63
161 AM-L (79 with PQ, 82 no PQ) 18 <5 years 57 5–14 years 86 >14 years
161 DP (83 with PQ, 78 no PQ) 17 <5 years 62 5–14 years 82 >14 years
12 lost to follow-up
6 lost to follow-up
155 analysed on day 63
149 analysed on day 63
Figure 1: Trial profile AA=artesunate–amodiaquine. PQ=primaquine. DP=dihydroartemisinin–piperaquine. AL=artemether–lumefantrine. AM-L=loose tablet regimen of artesunate–mefloquine. AM-F=fixed-dose artesunate–mefloquine.
censored), and time to new infection (recrudescent infections censored) were assessed by the Kaplan-Meier method. Patients lost to follow-up were censored at the last time seen. When failure rates were zero, CIs were estimated with the exact binomial method with the effective sample size. All other treatments were compared against the current standard of care, which is artesunate plus mefloquine (loose tablets), and then against fixed-dose artesunate–mefloquine hydrochloride, with the log-rank test, or the Wilcoxon-Breslow test of AA
equality if survival lines crossed. Cox regression was used to quantify risks. Person-gametocyte weeks (PGW) were defined as the number of weeks in which gametocytaemia was patent (excluding at admission) divided by the duration of follow-up, expressed per 1000 person-weeks. The protocol was approved by the Myanmar Department of Health and by the Médecins sans Frontières ethics review board. This study is registered with clinicaltrials. gov, number NCT00902811. AM-F
AM-L
155
162
169
161
161
Primaquine single dose
71
78
86
79
83
Women
55 (36%)
44 (27%)
62 (37%)
52 (32%)
50 (31%)
Number of patients
AL
DP
Age groups (years) 1–4
17 (11%)
17 (11%)
18 (11%)
18 (11%)
17 (11%)
5–14
56 (36%)
60 (37%)
63 (37%)
57 (35%)
62 (39%)
>14
82 (51%)
82 (53%)
85 (53%)
88 (52%)
86 (53%)
Mean haemoglobin g/dL (SD)
11·3 (0·26)
11·4 (0·27)
11·5 (0·25)
11·5 (0·26)
11·3 (0·24)
Anaemic (haemoglobin <10 g/dL)
48 (31%)
54 (33%)
46 (27%)
48 (30%)
50 (31%)
Geometric mean parasite count per μL (range)
7077 (521–163 881)
8709 (557–151 714)
7934 (554–177 000)
7849 (504–139 600)
7879 (546–176 000)
Mixed infections
26 (17%)
25 (15%)
21 (12%)
31 (19%)
26 (16%)
Gametocytaemia on admission
53 (34%)
54 (33%)
50 (30%)
46 (29%)
61 (38%)
AA=artesunate–amodiaquine. AL=artemether–lumefantrine. AM-F=artesunate–mefloquine fixed-dose combination. AM-L=artesunate–mefloquine loose tablets. DP=dihydroartemisinin–piperaquine.
Table 1: Baseline characteristics
www.thelancet.com/infection Vol 10 October 2010
675
Articles
Total target Median (IQR; range) doses (mg/kg) Artesunate–mefloquine (loose)
12/25
12 (11·7– 12·2; 10·7–14·2); 25 (24·5–25·3; 22·3–29·5)
Artesunate–mefloquine (fixed dose)
12/26·4
14·6 (12·5– 16·7; 7·0– 30·0); 32·2 (27·5– 36·7; 15·5–66·0)
Artemether–lumefantrine
9·9/59·4
10·9 (9·8– 12·0; 7·9– 16·0); 65·4 (58·8– 72·0; 47·2–96·0)
Dihydroartemisinin– piperaquine
7·5/60
7·3 (6·5– 8·4; 4·4– 16·0); 58·2 (51·7– 67·4; 35·3–128·0)
Artesunate–amodiaquine
12/32·4
Primaquine
12·3 (10·7–14·3; 8·6–16·7); 33·1 (28·9–38·6; 23·2–45·0)
0·75
0·76 (0·7–0·82; 0·53–1·07)
Table 2: Drug doses given
Number by day 28; proportion (95% CI)
Number by day 63; Median days to proportion (95% CI) event (IQR); range
p value for day 63 comparison with AM-L
p value for day 63 comparison with AM-F
Recurrence of P falciparum AA
16; 10·5 (6·6–16·6)
28; 18·6 (13·2–25·8)
28 (17·5–42); 14–63
0·001
<0·0001
15; 9·7 (6·0–15·6)
42 (21–56); 14–63
0·09
0·01
0; 0 (0–2·2)
5; 3·1 (1·3–7·2)
42 (35–56); 35–63
0·43
··
AM-L
2; 1·3 (0·3–5·2)
7; 4·7 (2·3–9·7)
42 (28–56); 14–56
··
0·43
DP
5; 3·2 (1·3–7·4)
14; 9·0 (5·4–14·7)
0·14
0·02
AL
5; 3·2 (1·3–7·4)
AM-F
38·5 (28–56); 14–63
Recrudescent P falciparum infections* AA
12; 8·0 (4·6–13·6)
14; 9·4 (5·7–15·3)
21 (14–28); 14–49
0·002
0·0001
AL
1; 0·6 (0·1–4·5)
2; 1·4 (0·3–5·3)
42 (21–63); 21–63
0·95
0·14
AM-F
0; 0·0 (0–2·2)
0; 0 (0–2·3)
0·14
··
AM-L
1; 0·6 (0·1–4·5)
2; 1·3 (0·3–5·3)
31·5 (14–49); 14–49
··
0·14
DP
0; 0·0 (0–2·3)
2; 1·3 (0·3–5·2)
38·5 (35–42); 35–42
0·96
0·14
··
Re-infections with P falciparum† AA
2; 1·4 (0·3–5·5)
11; 8·2 (4·6–14·3)
42 (35–63); 21–63
0·02
0·05
AL
3; 1·9 (0·6–5·8)
10; 6·5 (3·6–11·8)
38·5 (28–49); 14–63
0·05
0·13
AM-F
0; 0·0 (0–2·2)
5; 3·1 (1·3–7·2)
42 (35–56); 35–63
0·58
··
AM-L
0; 0·0(0–2·5)
3; 2·1 (0·7–6·4)
56 (42–56); 42–56
··
0·58
DP
3; 1·9 (0·6–5·8)
9; 5·9 (3·1–11·1)
49 (28–56); 21–63
0·09
0·21
AM-F=artesunate–mefloquine fixed-dose combination. AM-L=artesunate–mefloquine loose tablets. AA=artesunate– amodiaquine. AL=artemether–lumefantrine. DP=dihydroartemisinin–piperaquine. *PCR indeterminate and new infections censored. †Recrudescent and indeterminate parasitaemias censored.
Table 3: Recurrence of Plasmodium falciparum after antimalarial treatment
Role of the funding source The sponsor of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results Figure 1 shows the trial profile. Of the patients screened for malaria, 3416 met the inclusion criteria. Most of the patients that did not agree to participate did so because they lived too far away from the study clinic. 713 of the patients who were randomly assigned to a treatment group were from Rakhine state, 66 were from Kachin state, and 32 were from Shan state. Three patients were 676
excluded after random assignment because they were incorrectly reported as positive for P falciparum on day 7 and so were treated again. Of the 808 eligible patients, 397 (49·1%) also received primaquine (figure 1). All patients survived, although 39 (4·9%) did not complete the 63 day follow-up. Baseline characteristics were similar among the five treatment groups (table 1). Table 2 shows the median (IQR) and range of doses received. All patients cleared parasitaemia by day 7. 69 patients had recurrence of P falciparum within 63 days (table 3). Recurrence was less common in patients in the fixeddose artesunate–mefloquine group than in the fixed-dose artesunate–amodiaquine group, fixed-dose artemether– lumefantrine group, and the fixed-dose dihydroartemisinin–piperaquine group but was not significantly different to that in the loose tablet artesunate–mefloquine recipients (table 3, figure 2). Thus recurrence was less common in the two artesunate–mefloquine groups combined than in the other ACT groups (for artesunate– amodiaquine p<0.0001, for artemether–lumefantrine p=0.0097, and for dihydroartemisinin–piperaquine p=0.02). Recurrence of P falciparum parasitaemia was significantly (p=0·011) more common in children younger than 5 years (13 [15%] of 87 children) than in those aged 5–14 years (28 [9%] of 298), or in adults (28 [7%] of 423). 38 new infections and 20 recrudescent parasitaemias occurred by day 63. Results for nine patients were indeterminate (two artesunate–amodiaquine, three artemether–lumefantrine, three dihydroartemisinin– piperaquine, and one loose tablet artesunate–mefloquine), and data for two patients (one artesunate–amodiaquine and one loose artesunate–mefloquine) were missing. If indeterminate or missing results are treated as censored findings, then the treatment failure rate after fixed-dose artesunate–amodiaquine was significantly higher than for the other treatments (for fixed-dose artesunate–mefloquine p<0·0001, for artesunate–mefloquine loose p=0·0018, for artemether–lumefantrine p=0·0013, and for dihydroartemisinin–piperaquine p=0·0012; table 3). If indeterminate or missing results are treated as failures, then failure rates after fixed-dose artesunate–mefloquine (0%; 95% CI 0–2·3) were significantly lower than those for other treatments (for artesunate–amodiaquine [11·3%; 95% CI 7·2–17·5] p<0·0001, for artesunate–mefloquine loose [2·7%; 1·0–7·0] p=0·0355, for artemether– lumefantrine [3·4%; 1·4–7·95] p=0·0191, and for dihydroartemisinin–piperaquine [3·3%; 1·4–7·65] p=0·0207) and failure rates after fixed-dose artesunate–amodiaquine were significantly higher than with the other regimens; fixeddose artemether–lumefantrine (p=0·0042), the loose tablet regimen of artesunate–mefloquine (p=0·0025), fixed-dose artesunate–mefloquine (p<0·0001), and fixed-dose dihydroartemisinin–piperaquine (p=0·004). If chloroquine treatment of intercurrent vivax malaria had any significant antimalarial effect it would have reduced late cases of recrudescence. Chloroquine did not affect cure rates except www.thelancet.com/infection Vol 10 October 2010
Articles
www.thelancet.com/infection Vol 10 October 2010
47·2–63·1) groups. The other three treatment groups had malaria-free rates of less than 50%, and all were significantly worse than fixed-dose artesunate– mefloquine (artesunate–amodiaquine HR 1·86, 95% CI 1·32–2·62; artemether–lumefantrine 2·46, 1·77–3·43; and the loose tablet regimen of artesunate plus mefloquine 1·79, 1·28–2·52). 264 patients (33%) presented with patent gametocytaemia, which was more common in children younger than 5 years (46 [53%] of 87) than in children
Patients with recurrent falciparum (%)
25
AA AL AM-F AM-L DP
20
15
10
5
0 0
7
14
21
28
35
42
49
56
63
132 147 161 145 148
128 145 160 141 146
124 141 159 140 143
122 139 157 137 140
Follow-up (days) Number at risk AA AL AM-F AM-L DP
155 162 169 159 160
155 162 169 159 160
152 161 165 155 159
145 154 165 151 158
138 151 165 148 156
135 149 164 146 152
Figure 2: Comparative efficacy of the five ACT regimens against falciparum malaria in Myanmar Recurrences are recrudescence plus re-infections. AA=artesunate–amodiaquine. AL=artemether–lumefantrine. AM-F=fixed-dose artesunate–mefloquine. AM-L=loose tablet regimen of artesunate–mefloquine. DP=dihydroartemisinin–piperaquine.
60
AA AL AM-F AM-L DP
50 Cumulative vivax (%)
in the fixed-dose artesunate–amodiaquine group: there were two cases of recrudescence in the 59 (4·6%; 1·2–8·0) who did not receive chloroquine versus 12 of 96 (13·2%; 7·7–22·0) who did (p=0·04). Primaquine did not affect recrudescence rates. The risk of new infection was more than three-times higher for patients who received artesunate–amodiaquine compared with the artesunate–mefloquine groups combined (hazard ratio [HR] 3·2; 95% CI 1·3–8·0; p=0·01) and more than twice that with artemether–lumefantrine (2·6; 1·0–6·6; p=0·04) and dihydroartemisinin– piperaquine (2·3; 0·9–6·0; p=0·08). 129 patients (16%) had mixed infections at presentation, and these were more common in children (102 [27%] of 385) than in adults (27 [6%] of 423; risk ratio [RR] 4·2, 95% CI 2·8–6·2; p<0·0001). All patients with mixed infections responded to ACT treatment. 330 patients had P vivax infection during follow-up (figure 3): 259 had one episode (68 had mixed infections initially), 55 had two, and three had three. Thus the 330 patients had 404 episodes of P vivax during follow-up. Of the 679 patients presenting with P falciparum infections only, 235 (35%) had subsequent P vivax malaria compared with 95 (74%) of those with mixed infection initially (p<0·0001). Fewer cases of P vivax were recorded in patients who received fixed-dose artesunate–mefloquine than in patients who received fixed-dose artesunate– amodiaquine, fixed-dose artemether–lumefantrine, or the loose tablet regimen of artesunate plus mefloquine (table 4). The median time to the identification of P vivax was 35 days after fixed-dose artemether–lumefantrine, 42 days after the loose tablet regimen of artesunate plus mefloquine, 49 days after fixed-dose artesunate– amodiaquine and fixed-dose artesunate–mefloquine, and 56 days after fixed-dose dihydroartemisinin–piperaquine (p=0·0001). Thus 12 of the 14 cases of recrudescence of P falciparum infection after artesunate–amodiaquine happened before the median time to P vivax relapse for this group (table 2), which argues against a major effect of chloroquine in suppressing recrudescence. The median time to P vivax recurrence was shorter for children aged 0–14 years (42 days; IQR 35–56) than for adults (56; 42–63; p=0·0001). Children were nearly three times more likely to develop intercurrent vivax malaria than were adults, after adjusting for treatment and mixed infections at admission (HR 2·91; 95% CI 2·28–3·70; p<0·0001). Single-dose primaquine did not affect P vivax rates (data not shown). Pooling all recurrences of falciparum or vivax malaria together provided a combined measure of efficacy and post-treatment prophylaxis, and it also addresses any possible confounding by weak activity of chloroquine against recrudescent P falciparum, because patients were censored at the time of any recurrence of malaria. The day 63 malaria-free rate was highest in the fixeddose artesunate–mefloquine (64·3%; 95% CI 56·4–71·1) and the dihydroartemisinin–piperaquine (55·5%,
40 30 20 10 0 0
7
14
21
28
35
42
49
56
63
124 94 158 122 141
103 79 150 104 132
90 69 131 90 121
77 64 112 77 104
Follow-up (days) Number at risk AA AL AM-F AM-L DP
155 162 169 159 160
155 162 169 159 160
152 161 165 155 159
144 151 165 150 158
137 137 165 145 154
134 116 163 134 149
Figure 3: Cumulative proportion of patients with vivax malaria during follow-up AA=artesunate–amodiaquine. AL=artemether–lumefantrine. AM-F=fixed-dose artesunate–mefloquine. AM-L=loose tablet regimen of artesunate–mefloquine. DP=dihydroartemisinin–piperaquine.
677
Articles
Number of patients
p value for Number of Proportion of P vivax infections by comparison P vivax with AM–L infections day 63 (95% CI)
p value for comparison with AM–F
Artesunate–amodiaquine 26
18
84 (66–96)
0·62
0·05
No P vivax at day 0
129
41
38 (30–48)
0·25
0·15
All
155
59
46 (38–55)
0·10
0·02
With P vivax at day 0
Artemether–lumefantrine 25
23
··*
<0·0001
<0·0001
No P vivax at day 0
137
62
49 (41–58)
0·03
<0·0001
All
162
85
57 (49–65)
0·004
<0·0001
With P vivax at day 0
Artesunate–mefloquine (fixed dose) 21
11
55 (35–77)
0·01
··
No P vivax at day 0
148
44
31 (24–40)
0·01
··
All
169
55
34 (28–42)
<0·001
··
With P vivax at day 0
Artesunate–mefloquine (loose) 31
28
93 (81–99)
··
0·01
No P vivax at day 0
130
47
42 (33–51)
··
0·01
All
161
75
53 (45–61)
··
0·0001
With P vivax at day 0
Dihydroartemisinin–piperaquine 26
15
68 (49–86)
0·01
0·69
No P vivax at day 0
135
41
34 (26–43)
0·02
0·93
All
161
56
39 (32–48)
<0·001
0·67
With P vivax at day 0
AM-F=artesunate–mefloquine fixed-dose combination. AM-L=artesunate–mefloquine loose tablets. *Day 63 rate could not be calculated because all patients either lost to follow-up or relapsed by day 56.
Table 4: Intercurrent Plasmodium vivax infections
After ACT (no primaquine)*
p value for comparison with AM-F
After ACT (with primaquine)*
p value for comparison with AM-F
Artesunate–amodiaquine
94·4
<0·001
7·50
0·43
Artemether–lumefantrine
58·2
0·009
4·65
0·86
Artesunate–mefloquine fixed-dose
29·2
··
4·02
··
Artesunate–mefloquine loose tablets
34·6
0·58
4·56
0·88
112·8
<0·001
7·17
0·45
Dihydroartemisinin–piperaquine Total
65·51†
5·49†
ACT=artemisinin combination therapy. combination. *Person gametocytaemia weeks per 1000 person-weeks of follow-up. †Comparison of all patients receiving primaquine versus patients not receiving primaquine (rate ratio 11·9; 95% CI 7·4–20·5; p≤0·0001).
Table 5: Plasmodium falciparum gametocyte carriage*
aged 5–14 years (121 [41%] of 298) and in adults (97 [23%] of 423; p<0·0001). Independent of age, gametocytaemia was more common among moderately anaemic patients (haemoglobin <8·0 g/dL; 49 [70%] of 70) than among patients with higher values (215 [29%] of 738; RR 2·4, 95% CI 2·0–2·9; p<0·0001). Without primaquine the treatment regimens had widely different gametocyte carriage rates (table 5). With primaquine all the treatment regimens had low gametocyte carriage, which overall was about 12-times lower than without (table 5; figure 4). New gametocytaemia on day 7 was also reduced by primaquine (one of 272 vs ten of 268; 0·10, 0·01–0·76; p=0.006). 678
246 patients (30%) had a haemoglobin concentration of less than 10·0 g/dL (mild anaemia), and 70 (9%) a concentration of less than 8·0 g/dL (moderate anaemia) at presentation. Young children were more likely to be anaemic at presentation (28 [32%] of 87) than older children (26 [9%] of 298) and adults (16 [4%] of 423; RR 5·5; 95% CI 3·6–8·4; p<0·0001). On day 63, eight (1%) of 693 patients were moderately anaemic and 72 (10%) were mildly anaemic. The mean increase of haemoglobin was similar among the five treatment groups, was unaffected by intercurrent P vivax, but was slightly reduced by primaquine (0·75 g/dL vs 1·04 g/dL; p=0·036; mean difference 0·295 g/dL; 95% CI 0·199–0·570). 599 patients (74%) reported adverse events during the study (table 6). 13 patients (1·6%) vomited within the first hour (nine patients within 30 min and four within 30–60 min), four after the loose tablet regimen of artesunate plus mefloquine, three after artesunate–amodiaquine, two after artemether–lumefantrine, two after fixed-dose artesunate–mefloquine, and two after dihydroartemisinin– piperaquine. None vomited again after treatment was repeated. Dizziness was more common after fixed dose artesunate–mefloquine than after artemether–lumefantrine (RR 1·24; 95% CI 1·03–1·48; p=0·03) and dihydroartemisinin–piperaquine (1·22; 1·02–1·45). Other side-effects were less common and were not related to a specific treatment. The only side-effect attributable to primaquine was abdominal pain (table 7). There were no cases of urticaria or other potentially serious skin reactions, convulsions, behaviour disturbances, blackwater fever, or severe anaemia (<50 g/L).
Discussion Control and ultimately elimination of falciparum malaria depends on providing effective and well tolerated medicines and minimising transmission. In Myanmar, our large study shows that artesunate–mefloquine, the first-line treatment in the main study region since 1996, remains very effective. The new fixed-dose artesunate– amodiaquine was well tolerated, but was not effective enough to be recommended as a first-line treatment. The other available ACTs were all well tolerated and highly effective, suggesting good adherence to the prescribed treatments, although only the first dose was observed (ie, the normal context of use). Addition of a single dose of primaquine to ACT regimens has a very large additional effect on gametocytaemia, and therefore on malaria transmission potential, making any residual differences in gametocyte carriage after the primary infection between the five ACT regimens insignificant. From an operational perspective, fixed-dose combinations are simple to give and avoid monotherapy, thereby protecting against resistance. Findings from our previous studies have shown that artesunate– mefloquine, the first-line treatment in Myanmar since 1996, is very effective.2–4,10 The new fixed-dose artesunate– mefloquine has the greatest efficacy of all the ACTs www.thelancet.com/infection Vol 10 October 2010
Articles
www.thelancet.com/infection Vol 10 October 2010
40
Patients with gametocytaemia (%)
assessed.5 Artesunate–mefloquine regimens were also associated with greater post-treatment prophylaxis than the other drugs,12 although differences with dihydroartemisinin–piperaquine were not significant. Although artesunate–mefloquine regimens were associated with slightly more dizziness than were fixeddose artemether–lumefantrine or dihydroartemisinin– piperaquine,2–4,9,10,13,14 high cure rates argue against a significant effect on adherence. Malaria transmission in Myanmar is seasonal and generally low, but access to effective drugs is limited, and the clinical epidemiology is similar to that of higher-transmission settings with a high prevalence of anaemia, high rates of P falciparum gametocytaemia, and substantial discrepancy between cure rates in adults and children.2–4,10 The artemisinin component of ACTs reduces gametocyte carriage,6,7,15 and patients who received artesunate–mefloquine regimens had the lowest rates of gametocyte carriage after treatment, probably because of the higher dose of the artemisinin derivative compared with artemether–lumefantrine and dihydroartemisinin–piperaquine, and the greater antimalarial effect of mefloquine compared with amodiaquine. The use of ACTs in parts of Rakhine state since 19962–4,10 has resulted in a gradual but substantial decrease in falciparum malaria in regions where the drugs are available. The high gametocyte carriage rates probably contribute to differences in the effect of ACTs on malaria transmission and thus incidence in Myanmar compared with that in other areas in the region, such as the western border of Thailand, where gametocyte carriage rates are low, treatment delays are brief, and, as a result of ACT deployment, falciparum malaria incidence has fallen substantially.16 Greater coverage of effective drug treatments with more potent transmission-blocking activity is expected to result in greater effects on malaria incidence. The 8-aminoquinolines have a rapid and powerful sterilising effect on mature gametocytes.17,18 Single-dose primaquine has been widely recommended in the past, and used extensively in eradication campaigns; however, its safety profile has not been well characterised.19 Primaquine causes abdominal discomfort, particularly if taken on an empty stomach, and potentially serious haemolytic anaemia in patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency.20 Single doses have been judged safe, particularly when prevalent genotypes confer mild G6PD deficiency, although a recent study has challenged this assumption.21 In our study there were no severe episodes of haemolysis, although we did note a small but significant adverse effect on recovery from anaemia. We did not test for G6PD deficiency, and more frequent measurements of haemoglobin concentrations were not possible. Mixed P falciparum and P vivax infections were common. 330 patients (41%) had at least one episode of vivax malaria in the 63 day follow-up. The risk was
No primaquine Primaquine
30
20
10
0 0
7
14
21
28
35
42
Follow-up (days)
Figure 4: Effects of a single dose of primaquine on Plasmodium falciparum gametocyte carriage after artemisinin combination treatment All five ACT regimens pooled. No gametocytaemia was detected after day 42.
AA (n=155)
AL (n=162)
AM-F (n=169)
AM-L (n=161)
DP (n=161)
Total (n=808)
p value*
Dizziness
91 (59%)
86 (53%)
111 (66%)
Nausea
27 (17%)
28 (17%)
33 (20%)
104 (65%)
87 (54%)
479 (59%)
0·06
30 (19%)
28 (17%)
146 (18%)
Anorexia
29 (19%)
17 (11%)
25 (15%)
0·98
21 (13%)
22 (14%)
114 (14%)
0·34
Diarrhoea
17 (11%)
12 (7%)
11 (7%)
16 (10%)
20 (12%)
76 (9%)
0·32
Abdominal pain
20 (13%)
23 (14%)
23 (14%)
26 (16%)
19 (12%)
111 (14%)
0·83
Palpitations
35 (23%)
25 (15%)
38 (23%)
41 (26%)
32 (20%)
171 (21%)
0·23
Sleeplessness
22 (14%)
14 (9%)
25 (15%)
24 (15%)
18 (11%)
103 (13%)
0·33
2 (1%)
6 (1%)
0·38†
10 (6%)
48 (6%)
0·18
Headache
2 (1%)
2 (1%)
0
Vomiting in first 24 h
7 (5%)
6 (4%)
9 (5%)
0 16 (10%)
AA=artesunate–amodiaquine. AL=artemether–lumefantrine. AM-F=artesunate–mefloquine fixed-dose combination. AM-L=artesunate–mefloquine loose tablets. DP=dihydroartemisinin–piperaquine. *For comparison of frequency across the five treatment groups, adjusted for primaquine allocation. †p value for differences between treatment groups not adjusted for primaquine allocation.
Table 6: Number of patients reporting side-effects at least once during study
Dizziness
With primaquine (n=397)
No primaquine (n=411)
Total (n=808)
p value*
239 (60%)
240 (58%)
479 (59%)
0·61
Nausea
78 (20%)
68 (17%)
146 (18%)
0·26
Anorexia
48 (12%)
66 (16%)
114 (14%)
0·11
Diarrhoea
31 (8%)
45 (11%)
76 (9%)
0·13
Abdominal pain
64 (16%)
47 (11%)
111 (14%)
0·05
Palpitations
86 (22%)
85 (21%)
171 (21%)
0·73
Sleeplessness
50 (13%)
53 (13%)
103 (13%)
0·90
4 (1%)
2 (1%)
6 (1%)
0·44†
25 (6%)
23 (6%)
48 (6%)
0·69
Headache Vomiting in first 24 h
*For comparison between primaquine allocation, adjusted for treatment groups. †p value not adjusted for differences between treatment groups.
Table 7: Number of patients reporting side-effects after treatment with primaquine
679
Articles
particularly high in young children—56 children (64%) younger than 5 years had vivax malaria after acute falciparum malaria. This compares with P falciparum recrudescence rates of less than 10%. Treatment of vivax relapses with chloroquine might have provided weak additional activity against P falciparum, suppressing some late recrudescence, particularly for the less effective artesunate–amodiaquine. Although because most recrudescence happened within 4 weeks, and most vivax infections were not noted until after 5 weeks, this effect is unlikely to have been large. Thus for the four highly effective ACT regimens the main clinical problem after effective treatment of acute falciparum malaria in Myanmar is recurrent P vivax malaria not P falciparum.22 Recurrence of vivax malaria was over four times more likely than was recurrence of falciparum malaria, emphasising the public health importance of vivax infections after falciparum malaria. If safe radical treatment was available, then these very high rates would argue for radical treatment of all malaria in this region where mixed infections are so common. This is, to our knowledge, the first comparison of all currently available fixed-dose artemisinin combination treatments for falciparum malaria. Our findings suggest that artesunate–mefloquine, artemether– lumefantrine, or dihydroartemisinin–piperaquine are all good treatments of falciparum malaria in Myanmar. The new fixed dose artesunate–mefloquine had the highest cure rates, the lowest rates of gametocyte carriage, and the most effective suppression of P vivax malaria. Adverse effects were generally mild, despite the high doses received by some patients, and did not substantially affect adherence. Although giving the entire 25 mg/kg dose of mefloquine as loose tablets was directly observed, suppression of recurrent vivax malaria was significantly worse than with the fixed combination, suggesting better absorption of mefloquine with the latter.5,23,24 The addition of a single gametocytocidal dose of primaquine was well tolerated and highly effective and did not cause serious adverse effects, although more studies to characterise the safety with different G6PD deficiency genotypes are needed. The addition of a single dose of primaquine to ACTs could have a major effect on malaria transmission from treated patients, and could have a crucial role in elimination programmes.25 Contributors FS, EA, and NJW designed the study; FS, MKK, OP, PPA, ALN, MYN, NZHM, and TW did the investigation; MI did the genotyping studies; and SJL did the analyses. All authors contributed to the data interpretation and writing of the paper. Conflicts of interest FS has been supported by the Drugs for Neglected Diseases initiative to attend scientific conferences. NJW is Chairman of the Drugs for Neglected Diseases Initiative Fixed Dose Artemisinin Combination Treatment and co-chairman of the WHO GMP antimalarial treatment guidelines committee. The other authors declare no conflicts of interest.
680
Acknowledgments Fixed-dose artesunate–amodiaquine and artesunate–mefloquine were provided by Jean-Rene Kiechel of the Drugs for Neglected Diseases Initiative. We thank the staff of the Vector Borne Disease Control department, Department of Health, Myanmar for their support, and Kasia Stepniewska for her advice on the statistical analysis. We also thank the clinical and laboratory staff of Médecins sans Frontières (Holland) in Sittwe, Maungdaw, Lashio, Myitkyitina, and Bahmo for their hard work. SJL, MI, and NJW are supported by The Wellcome Trust of Great Britain as part of the Wellcome Trust Mahidol University Oxford Tropical Medicine Research Programme. References 1 WHO. Guidelines for the treatment of malaria, 2nd edn. Geneva: World Health Organization, 2010. 2 Smithuis F, Kyaw MK, Phe O, et al. Efficacy and effectiveness of dihydroartemisinin-piperaquine versus artesunate-mefloquine in falciparum malaria: an open-label randomised comparison. Lancet 2006; 367: 2075–85. 3 Smithuis F, van der Broek I, Katterman I, et al. Optimising operational use of artesunate-mefloquine; a randomised comparison of four treatment regimens. Trans R Soc Trop Med Hyg 2004; 98: 182–92. 4 Smithuis F, Shahmanesh M, Kyaw MK, Savran O, Lwin S, White NJ. Comparison of chloroquine, sulfadoxine/pyrimethamine, mefloquine and mefloquine-artesunate for the treatment of falciparum malaria in Kachin State, North Myanmar. Trop Med Int Health 2004; 9: 1184–90. 5 Ashley EA, Stepniewska K, Lindegårdh N, et al. Population pharmacokinetic assessment of a new regimen of mefloquine used in combination treatment of uncomplicated falciparum malaria. Antimicrob Agents Chemother 2006; 50: 2281–85. 6 Price RN, Nosten F, Luxemburger C, et al. The effects of artemisinin derivatives on malaria transmissability. Lancet 1996; 347: 1654–58. 7 Bousema JT, Schneider P, Gouagna LC, et al. Moderate effect of artemisinin-based combination therapy on transmission of Plasmodium falciparum. J Infect Dis 2006; 193: 1151–59. 8 Burgess RW, Bray RS. The effect of a single dose of primaquine on the gametocytes, gametogony and sporogony of Laverania falciparum. Bull World Health Organ 1961; 24: 451–56. 9 Suputtamongkol Y, Chindarat S, Silpasakorn S, et al. The efficacy of combined mefloquine-artesunate versus mefloquine-primaquine on subsequent development of Plasmodium falciparum gametocytemia. Am J Trop Med Hyg 2003; 68: 620–23. 10 Smithuis FM, Monti F, Grundi M, et al. In vivo sensitivity of Plasmodium falciparum to chloroquine, sulphadoxinepyrimethamine, and mefloquine in Rakhine State, western Myanmar. Trans R Soc Trop Med Hyg 1997; 91: 468–72. 11 Brockman A, Paul RE, Anderson TJ, et al. Application of genetic markers to the identification of recrudescent Plasmodium falciparum infections on the northwestern border of Thailand. Am J Trop Med Hyg 1999; 60: 14–21. 12 White NJ. How antimalarial drug resistance affects post-treatment prophylaxis. Malaria J 2008; 11: 7–9. 13 Zwang J, Ashley EA, Karema C, et al. Safety and efficacy of dihydroartemisinin-piperaquine in falciparum malaria: a prospective multi-centre individual patient data analysis. PLoS One 2009; 4: e6358. 14 van Vugt M, Brockman A, Gemperli B, et al. Randomised comparison of artemether-benflumetol and artesunate-mefloquine in the treatment of multi-drug resistant falciparum malaria. Antimicrob Agents Chemother 1998; 42: 135–39. 15 Okell LC, Drakeley CJ, Ghani AC, Bousema T, Sutherland CJ. Reduction of transmission from malaria patients by artemisinin combination therapies: a pooled analysis of six randomized trials. Malar J 2008; 7: 125. 16 Carrara VI, Sirilak S, Thonglairuam J, et al. Deployment of early diagnosis and mefloquine-artesunate treatment of falciparum malaria in Thailand: the Tak Malaria Initiative. PLoS Med 2006; 3: e183. 17 MacKerras MJ, Ercole QN. Observations on the action of quinine, atebrin and plasmoquine on the gametocytes of Plasmodium falciparum. Trans R Soc Trop Med Hyg 1949; 42: 455–63.
www.thelancet.com/infection Vol 10 October 2010
Articles
18
19
20
21
Shekalaghe S, Drakeley C, Gosling R, et al. Primaquine clears submicroscopic Plasmodium falciparum gametocytes that persist after treatment with sulphadoxine-pyrimethamine and artesunate. PLoS One 2007; 2: e1023. Hill DR, Baird JK, Parise ME, Lewis LS, Ryan ET, Magill AJ. Primaquine: report from CDC expert meeting on malaria chemoprophylaxis I. Am J Trop Med Hyg 2006; 75: 402–15. Beutler E, Duparc S; G6PD Deficiency Working Group. Glucose-6-phosphate dehydrogenase deficiency and antimalarial drug development. Am J Trop Med Hyg 2007; 77: 779–89. Shekalaghe SA, ter Braak R, Daou M, et al. In Tanzania, hemolysis after a single dose of primaquine coadministered with an artemisinin is not restricted to glucose-6-phosphate dehydrogenasedeficient (G6PD A–) individuals. Antimicrob Agents Chemother 2010; 54: 1762–68.
www.thelancet.com/infection Vol 10 October 2010
22
23
24
25
Ratcliff A, Siswantoro H, Kenangalem E, et al. Two fixed-dose artemisinin combinations for drug-resistant falciparum and vivax malaria in Papua, Indonesia: an open-label randomised comparison. Lancet 2007; 369: 757–65. Price RN, Simpson JA, Teja-Isavatharm P, et al. Pharmacokinetics of mefloquine combined with artesunate in children with acute falciparum malaria. Antimicrob Agents Chemother 1999; 43: 341–46. Simpson JA, Price RN, ter Kuile FO, et al. Population pharmacokinetics of mefloquine in patients with acute falciparum malaria. Clin Pharmac Ther 1999; 66: 472–84. White NJ. The role of anti-malarial drugs in eliminating malaria. Malar J 2008; 7 (suppl 1): S8.
681
Articles
Population-based HIV-1 incidence in France, 2003–08: a modelling analysis Stéphane Le Vu, Yann Le Strat, Francis Barin, Josiane Pillonel, Françoise Cazein, Vanina Bousquet, Sylvie Brunet, Damien Thierry, Caroline Semaille, Laurence Meyer, Jean-Claude Desenclos
Summary Lancet Infect Dis 2010; 10: 682–87 Published Online Se ptember 9, 2010 DOI:10.1016/S14733099(10)70167-5 See Reflection and Reaction page 655 Institut de Veille Sanitaire, Saint-Maurice, France (S Le Vu PhD, Y Le Strat PhD, J Pillonel MSc, F Cazein PharmD, V Bousquet DVM, C Semaille MD, J-C Desenclos PhD); National Institute of Health and Medical Research (INSERM), Centre for Research in Epidemiology and Population Health, U1018, Le Kremlin-Bicêtre, France (S Le Vu, Prof L Meyer PhD); Department of Medicine, University of Paris-South, Le Kremlin-Bicêtre, France (S Le Vu, L Meyer); Epidemiology and Public Health Service, Bicêtre Hospital, Assistance Publique–Hôpitaux de Paris, Le Kremlin-Bicêtre, France (L Meyer); and INSERM U966, National HIV Reference Centre, University Hospital Bretonneau, François Rabelais University, Tours, France (Prof F Barin PhD, S Brunet BSc, D Thierry BSc) Correspondence to: Dr Stéphane Le Vu, Department of Infectious Diseases, Institut de Veille Sanitaire, 12 rue du Val d’Osne, Saint-Maurice 94415, France
[email protected]
Background Routine national incidence testing with enzyme immunoassay for recent HIV-1 infections (EIA-RI) has been done in France since January, 2003. From the reported number of HIV infections diagnosed as recent, and accounting for testing patterns and under-reporting, we aimed to estimate the incidence of HIV infection in France in 2003–08. Methods We analysed reports from the French National Institute for Public Health Surveillance for patients who were newly diagnosed with HIV between January, 2003, and December, 2008. Missing data were imputed with multiple imputation. Patients were classified with non-recent or recent infection on the basis of an EIA-RI test, which was calibrated with serial measurements from HIV seroconverters from the French ANRS-PRIMO cohort. We used an adapted stratified extrapolation approach to calculate the number of new HIV infections in men who have sex with men (MSM), injecting drug users (IDUs), and heterosexual men and women by nationality. Population sizes were obtained from the national census and national behavioural studies. Findings After accounting for under-reporting, there were 6480 (95% CI 6190–6780) new diagnoses of HIV infection in France in 2008. We estimate that there were 6940 (6200–7690) new HIV infections in 2008, suggesting an HIV incidence of 17 per 100 000 person-years. In 2008, there were 3550 (3040–4050) new infections in heterosexuals (incidence of 9 per 100 000 person-years), 3320 (2830–3810) in MSM (incidence of 1006 per 100 000 person-years), and 70 (0–190) in IDUs (incidence of 86 per 100 000 person-years). Overall HIV incidence decreased between 2003 and 2008 (p<0·0001), but remained comparatively high and stable in MSM. Interpretation In France, HIV transmission disproportionately affects certain risk groups and seems to be out of control in the MSM population. Incidence should be tracked to monitor transmission dynamics in the various population risk groups and to help to target and assess prevention strategies. Funding French National Institute for Public Health Surveillance (InVS) and French National Agency for Research on AIDS and Viral Hepatitis (ANRS).
Introduction The HIV-1 epidemic in France and other European countries has chiefly been monitored by analysis of data for new diagnoses of HIV infection, which are reported by regional or national case surveillance.1 However, because of the long and variable time from infection to diagnosis, case surveillance of new HIV diagnoses does not show present patterns of virus transmission. In the past 15 years, laboratory-based methods have been developed to estimate incidence of HIV with a cross-sectional approach.2,3 This method proved applicable with case-based surveillance data.4 To monitor the dynamic of HIV infection in France, routine incidence testing with an enzyme immunoassay for recent HIV infections (EIA-RI) has been implemented as part of the national HIV case surveillance since its introduction in 2003. We aimed to estimate HIV incidence in France by use of this serological assay for recent infection.
Methods Procedures For calibration of the EIA-RI assay, we defined an assay threshold that discriminated recent HIV infection from 682
longstanding infection, and estimated the distribution of time spent in a recently infected state—called the recentinfection-testing algorithm (RITA) duration5—from a reference population sample. The EIA-RI test had initially been developed to detect recent HIV infection through an algorithm that combined standardised measures of antibody binding to the immunodominant epitope of gp41 and the V3 region of gp120.6 In this first design of the EIA-RI assay, recent infection was defined as being infected for less than 180 days and the biomarkers threshold was estimated for the specific purpose of classification according to time since infection. To estimate incidence of HIV infection, we further defined the properties of the EIA-RI assay as follows. The time for which the EIA-RI assay biomarker remains less than a given threshold from the beginning of the infection (RITA duration) was estimated from a large number of seroconverter samples. Therefore, the recent-infection status is a transient biological state rather than a dichotomised time since infection. Furthermore, because the original development of the test showed that the most discriminatory antigen was immunodominant epitope www.thelancet.com/infection Vol 10 October 2010
Articles
alone or in combination with V3,6 we calibrated the assay with immunodominant epitope alone to improve model parsimony. The reference population sample was 952 serial measurements from 298 seroconverters from the French National Agency for Research on AIDS and Viral Hepatitis (ANRS) PRIMO cohort7 between 1996 and 2006. The proportion (43%) of samples from patients infected with viruses of non-B subtypes was much the same as that in new diagnoses. Date of infection was estimated as previously described.6 Sequential serum samples from these patients were obtained during the first 2 years after diagnosis of HIV primary infection and while patients were untreated (from one to seven samples per patient). These samples were tested as dried serum spot by EIA-RI. The calibration process also involved an estimation of the proportion of patients with long-term HIV infection who would test as recent infection. This proportion was designated the false-recent rate. The estimate was based on 250 chronically infected patients, who were tested more than 2 years after HIV diagnosis, but were clinically without AIDS, and 143 patients tested at the clinical AIDS stage (ANRS SEROCO and HEMOCO cohorts8). Since Jan 1, 2003, newly diagnosed HIV infections in France have been mandatorily reported to the National Institute for Public Health Surveillance. The notification form included demographic data (sex, age, and nationality), clinical stage, transmission category, and history of previous HIV testing.9 Remnant serum from the diagnosis sample was sent as a dried serum spot to the National HIV Reference Centre (Tours, France), where the EIA-RI test was done. We analysed reports from patients newly diagnosed with HIV between January, 2003, and December, 2008. Patients diagnosed with clinical AIDS were classified as having a non-recent infection. We accounted for completeness of reporting to the surveillance system and provided yearly incidence estimates for six subgroups of the population: men who have sex with men (MSM), injecting drug users (IDUs), heterosexual French-national men and women, and heterosexual non-French-national men and women. Men and women whose reported route of transmission was not sex or drug use accounted for less than 1% of diagnoses and were grouped in the heterosexual category. To calculate incidence, we obtained subpopulation sizes from the national census and a French national random probability survey of sexual behaviours (CSF).10 The proportion of men aged 18–69 years in the CSF who reported having had sex with men within the previous 12 months was applied to the overall male population as of 2008. We obtained the number of IDUs from the French Monitoring Centre for Drugs and Drug Addictions.11 We worked out the number of exclusively heterosexual adults aged 18–69 years from the proportion of adults reporting opposite-sex relationships in the CSF and applied this value to the overall population, from which numbers of IDUs and MSM were discounted. Sizes of overall and non-national population in France www.thelancet.com/infection Vol 10 October 2010
were obtained from the National Institute of Statistics and Economic Studies.12
Statistical analysis For calibration of the EIA-RI assay, both linear and nonlinear random-effects mixed models were tested to characterise the growth of immunodominant epitope response (measured as a standardised optical density value) against time since infection. For the linear model, the natural logarithm of both optical density values and time were chosen. Non-linear models used a Gompertz link function and untransformed values. The optical density threshold was chosen to minimise the falserecent rate. Missing data for diagnoses of HIV infection were estimated by use of multiple imputation by chained equations (ICE) with Stata 9.2.13 Because the history of previous HIV testing had to be completed before imputation of the delay between a last negative test (if any) and the positive test was possible, imputation consisted of a two-stage process. First, history of previous HIV testing was jointly imputed with nationality, transmission group, clinical stage, and EIA-RI result, and five datasets were generated. Second, the time delays between tests were estimated conditionally on a previous HIV test, and three datasets were imputed from those obtained at the first stage. We generated 15 datasets and calculated estimates and standard errors by use of Rubin’s rules. We did a non-parametric estimation of the distribution of reporting delays to adjust the number of diagnoses reported.14 The resulting estimated number was compared with the number of confirmed diagnoses obtained by a national postal survey of testing activity in all public and private laboratories in France, to estimate the completeness of reporting to the surveillance system (Le Strat Y, unpublished data). To construct our incidence model, we used a stratified extrapolation approach to estimate the yearly incidence of HIV infections in various transmission groups in France. Our model was based on the method developed by Karon and colleagues.15 HIV diagnoses were stratified into subpopulation groups, and diagnoses observed as recent infection were assumed by the model to be a random sample of the population of HIV infections occurring within 1 year (incident cases). The model assumed that (after imputation) a result for recent infection testing was available for every HIV diagnosis, and calculated the corresponding sampling probability as the probability of being tested within 1 year after infection (p1) multiplied by the probability of being detected as recent infection when diagnosed within a year after infection (pw). p1 was estimated separately for individuals who reported a negative test before diagnosis (termed repeat testers) and for those diagnosed at their first test (new testers). These conditional probabilities were primarily established by testing history (p1 for repeat testers), proportion of HIV infections diagnosed at AIDS 683
Articles
stage (p1 for new testers), and mean RITA duration (for pw). We used a distribution of the AIDS incubation periods according to the European AIDS case definition, corresponding to a median incubation time between infection and AIDS of 10 years.16 The number of incident cases within a subpopulation was then derived from the size of the sample divided by its sampling probability. For N diagnoses, we calculated the true number of HIV diagnoses in recent infection (Nr) as follows: Nr=
Or–N(1–false-recent rate) false-recent rate
With Or, the number of cases observed as recent. Variances and 95% CIs were calculated by use of the delta method and included the variability due to multiple imputation and variance associated with the estimation of completeness of case reporting. Temporal trends were assessed by means of variance-weighted least-square regressions.17 We did a sensitivity analysis by simultaneously simulating different values of the false recency rate from 0% to 5% and proportions of repeat testers in new diagnoses from 20% to 80%. We assessed the robustness of incidence model results by comparison of the range of estimates within these extreme scenarios and 95% CIs. Apart from multiple imputation, analyses were done with SAS software version 9.1.
Role of the funding source This study was supported by the French National Institute for Public Health Surveillance (InVS) and French National Agency for Research on AIDS and Viral Hepatitis (ANRS). The sponsor of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. SLV had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results For calibration of EIA-RI, the chosen optical density threshold led to a mean RITA duration of 179·7 days (95% CI 167·2–192·2) and a false recent rate of 0·8% (0–3·1%) in patients without AIDS and 5·6% (2·7–10·8%) in patients with AIDS. The range of possible RITA
durations was 25–731 days, with 95% of durations lasting less than 358 days, and 99% less than 480 days. 26 760 new diagnoses of HIV infection were reported between Jan 1, 2003, and Dec 31, 2008. With underreporting (estimated average of 37%), we estimated that 42 330 (95% CI 40 030–44 840) people were newly diagnosed with HIV in this time. Data were missing for 29·2% of diagnoses for category of transmission group, 23·5% for EIA-RI testing results, and 19·2% for the history of a previous negative test. Table 1 shows the number and characteristics of new diagnoses by year, after multiple imputation and accounting for underreporting. Overall number of diagnoses decreased by an average of 3·7% per year from 2003 to 2008 (p<0·0001). 25% of new HIV diagnoses were classified as recent by the EIA-RI assay; this proportion did not vary during 2003–08 (table 1). MSM were the most commonly (40%) diagnosed during recent infection, compared with Frenchnational heterosexual women (28%) and men (22%), heterosexual non-French-national women (16%) and men (12%), and IDUs (15%). The proportion of people who previously tested HIV negative before diagnosis increased from 2003 to 2008 (p<0·0001; table 1). Although the baseline proportion varied between different transmission groups, the same increasing trend was observed in all groups (data not shown). In 2003–08, 28% of new diagnoses were in heterosexual men, 37% in heterosexual women, 32% in MSM, and 3% in IDUs. Female nonFrench-nationals, mainly from sub-Saharan Africa, accounted for 25% of all HIV diagnoses. Between 2003 and 2008, the percentage of female non-French-nationals with new diagnosis of HIV decreased, whereas percentage of newly diagnosed MSM increased (figure 1). We estimated that nearly 7000 people were newly infected with HIV in France in 2008, of whom 48% were MSM (table 2). Of new infections attributed to heterosexual transmission, about half were in women. Only 1% of new HIV infections occurred in IDUs. NonFrench-nationals living in France accounted for around 23% of all new infections and 45% of the infections by heterosexual transmission. Overall HIV incidence decreased significantly from 8930 new infections in 2003 to 6940 in 2008 (p=0·002). This decrease was recorded for all heterosexual groups, whereas HIV incidence was
Year of diagnosis 2003
Overall 2004
2005
2006
2007
2008
EIA-RI result Recent
24·1%
23·7%
25·2%
24·8%
26·4%
27·0%
25·2%
Non-recent (non-AIDS)
56·3%
58·7%
58·2%
59·8%
58·6%
59·8%
58·7%
Non-recent at AIDS stage* Ever tested negative for HIV infection HIV diagnoses
19·7%
17·6%
16·6%
15·4%
14·9%
13·2%
16·1%
32·4%
37·4%
43·8%
48·7%
53·1%
58·9%
46·2%
7370 (6880–7920) 7580 (7120–8090) 7480 (7090–7900) 6990 (6620–7400) 6440 (6140–6750) 6480 (6190–6780)
42 330 (40 030–44 840)
Data are (%) or number (95% CI). Missing data were redistributed by multiple imputation. *Diagnoses at AIDS stage were classified as non-recent irrespective of routine incidence testing (EIA-RI) result.
Table 1: New HIV-1 diagnoses in France, 2003–08
684
www.thelancet.com/infection Vol 10 October 2010
Articles
high and stable among MSM and low and stable among IDUs (figure 2). Incidence was 200 times higher in MSM than in the French-national heterosexual population, 18 times higher in IDUs, and nine times higher in nonFrench nationals. Incidences for people from subSaharan Africa were 29 times higher in men and 69 times higher in women than they were for respective Frenchnational heterosexuals (data not shown).
40
35
French-national women French-national men Non-French-national women Non-French-national men Men who have sex with men* Injecting drug users†
30
Overall incidence of HIV infection in France decreased between 2003 and 2008. Our results support previous findings of a disproportionately high number of transmissions among MSM and a very low number of transmissions among IDUs. Because the test for recent infection has been routinely applied since the start of case surveillance in 2003, we were able to calculate trends in incidence of HIV transmission in the most relevant groups of population in France. With 48% of all new infections and a persistently high incidence of 1% per year, the HIV epidemic seems to be out of control in the MSM population. In France and several other industrialised countries, the number of new diagnoses of HIV in MSM has increased in recent years.18–20 In one review,21 incidence of HIV transmission in MSM was stable and high during 1995–2005 in selected samples in several industrialised countries. Incidence measurements were obtained from selected samples of MSM in urban communities, but not for entire countries. This sampling strategy might explain why the reported mean incidence of 2–3% per year was high compared with our estimates of 1% per year in France as a whole. Our estimated incidence for the general population in France in 2008 (17 per 100 000 person-years) is comparable to those reported for the USA in 2006 (23 per 100 000 person-years).4 In both countries, around half of yearly new infections were in MSM. Several factors explain the high rate of HIV transmission between MSM in France. Behavioural studies suggest an increase in unprotected anal sex and number of sexual partners in MSM with and without HIV infection.22 An increase in transmission of primary and secondary syphilis and rectal lymphogranuloma venereum has been reported between MSM.23 These risk factors, combined with a high HIV prevalence in the MSM population, are probably interacting to maintain the high incidence, despite the probable effect of antiretroviral treatment for reduction of transmission at the population scale.24,25 Renewed safer-sex initiatives or new alternative prevention strategies targeting MSM are urgently needed. Conversely, the number of new infections related to drug injection remained very low during 2003–08 (1–2% of new infections every year). This situation is probably attributable to, at least in part, a successful harmreduction policy that was established in France in the mid 1990s to reduce unsafe injecting practices by promotion of access to clean needle and syringes and www.thelancet.com/infection Vol 10 October 2010
New diagnoses (%)
25
Discussion
20
15
10
5
0 2003
2004
2005
2006
2007
2008
Year
Figure 1: New HIV-1 diagnoses in France, 2003–08 *All nationalities. †All nationalities, both sexes. New HIV-1 infections Estimated (95% CI) population size Heterosexual French women French men
3550 (3040–4050)
40 836 530
810 (620–1000)
18 363 590
4 (3–5)
1140 (830–1440)
18 848 440
6 (4–8) 54 (40–68)
Foreign women
940 (700–1180)
1 739 760
Foreign men
660 (460–870)
1 884 740
Men who have sex with men* Injecting drug users† Overall
Incidence per 100 000 person-years (95% CI)
3320 (2830–3810) 70 (0–190) 6940 (6200–7690)
9 (7–10)
35 (24–46)
329 950
1006 (857–1155)
81 000
86 (0–192)
41 247 480
17 (15–19)
Population (aged 18–69 years) size estimates at Jan 1, 2008 from Bajos and colleagues, Costes and colleagues,11 and National Institute of Statistics and Economic Studies.12 *All nationalities. †All nationalities, both sexes. 10
Table 2: Estimated new HIV-1 infections and incidence for France in 2008, by transmission group
opioid-substitution treatments.26 The low number of new infections in IDUs might also be attributed to a secular change in the pattern of drug use and the large number of deaths in drug users during the late 1980s and early 1990s. Incidence of HIV infections in heterosexuals has decreased since 2003. However, without precise data for sexual partners, characterisation of heterosexuals at high risk of HIV infection in France is difficult from a surveillance system perspective. The decline we report in incidence in heterosexuals might be attributed to the effect of potent antiretroviral treatment within a population in which, unlike for MSM, HIV prevalence is low.24 The proportion of patients receiving highly active antiretroviral therapy increased in France from 75% in 685
Articles
4000
Number
3000
2000
1000
20 0 20 3 0 20 4 0 20 5 06 20 0 20 7 08
20 0 20 3 04 20 0 20 5 06 20 0 20 7 08
20 0 20 3 0 20 4 0 20 5 06 20 0 20 7 08
20 0 20 3 0 20 4 0 20 5 06 20 0 20 7 08
20 0 20 3 04 20 0 20 5 06 20 0 20 7 08
20 0 20 3 0 20 4 0 20 5 06 20 0 20 7 08
0
French national heterosexual women
Non-French-national heterosexual women
French national heterosexual men
Non-French-national heterosexual men
MSM
IDUs
Figure 2: Estimated number of new HIV-1 infections by transmission group in France, 2003–08 MSM=men who have sex with men. IDUs=injecting drug users.
2001 to 85% in 2008, and the proportion of treated patients with viral load less than 500 copies per mL increased from 70% in 2002 to 92% in 2008.27 Furthermore, investigators reported18 an increase in the mean age at HIV diagnosis for heterosexuals between 2003 and 2008, which is consistent with the decreasing trend in heterosexual transmission we noted. For heterosexuals, non-French nationals were most affected, especially women from sub-Saharan Africa. As in other European countries, the epidemic in Africa is having an important role in transmission rates in France.1,28 The number of new infections has, however, decreased between 2003 and 2008 in immigrant populations (figure 2). This decrease is unlikely to be explained by a change in migration pattern, because immigration was stable between 2003 and 2007.18 Specific prevention efforts targeting migrants started since 2002 might have been effective.29 Our model and the available data have several limitations. The incidence estimation model is mainly based on three parameters: the number of HIV diagnoses classified as recent infection, the mean RITA duration, and the probability of diagnosis within 1 year after infection. The first two parameters were dependent on the assay calibration process. In a previous study,30 we noted that factors such as viral subtype or geographical origins of patients affected the EIA-RI test results. Therefore, we ensured that the reference sample used for assay calibration was sufficiently diverse for virus subtypes, geographical origin, and time since infection. Because we were able to identify, through surveillance data, individuals diagnosed at AIDS stage and correct the estimated number of new infections to account for false recent infection, we addressed the main concerns about misclassification attributed to all tests for recent infection.3 For the third parameter—probability of diagnosis within 1 year after infection—the model required the 686
assumption of independence between infection and testing time.15 In certain circumstances, this assumption does not hold. In particular, testing could be motivated by seroconversion illness or recent exposure, resulting in overestimation of the number of recent infections and thus incidence.31 The effect of this potential bias needs to be addressed through analysis of questionnaire data on motivation for seeking tests (particularly a question about recent exposure). This item has been added to the questionnaire used in the national HIV case-reporting system. Calculation of incidence necessitates, as a denominator, a precise estimation of the size of the different at-risk subpopulations. Potentially socially stigmatised behaviours such as sex between men or drug use are prone to under-reporting in questionnaire surveys. Therefore, use of national behavioural data to extrapolate these behaviours to the overall population might have led to an underestimate of the size of the at-risk populations, and thus an overestimate of incidences. Furthermore, we could not distinguish from available data whether diagnoses were attributable to infection acquired in France or abroad. Our estimates of incidence for people not of French origin are for the population that can potentially be diagnosed in France, once infected. Reliability of our estimates is dependent on the stability of migration in HIV-infected individuals, which cannot be measured in France. Our results provide a new perspective on the HIV epidemic in France, which could not be garnered from data for HIV-diagnosis reporting alone.18,19 Despite an overall decline in HIV incidence, the high rates estimated for MSM and sub-Saharan Africans living in France warrant renewed prevention strategies. Incidence should be tracked to monitor transmission dynamics in the various population risk groups and to help target and assess prevention strategies. www.thelancet.com/infection Vol 10 October 2010
Articles
Contributors SLV had the idea for the study, analysed the data, and drafted the report. SLV, VB, and YLS designed the statistical analysis. FB, SB, DT did the biological analyses and helped to write the report. CS, JP, FC, and LM contributed to data collection, provided input to the study design, and revised the report. LM and J-CD supervised the experiments and reviewed the report. Conflicts of interest We declare that we have no conflicts of interest. Acknowledgments We thank the French National Agency for Research on AIDS and Viral Hepatitis (ANRS) AC23 working group for funding support and we are especially grateful to Joseph Lellouch, Ahmadou Alioum, Daniel Commenges, Dominique Costagliola, Véronique Doré, and Nadine Job-Spira for their help and valuable comments; Christiane Deveau and Laurent Tran for their help with data collection and monitoring of ANRS cohorts data; Joseph Prejean, Irene Hall, and Ruigang Song from the Division of HIV/AIDS Prevention of US Centers for Disease Control and Prevention; Daniela De Angelis from Medical Research Council; and the WHO Technical HIV Incidence Assay Working Group for useful discussions. ANRS funded the PhD scholarship for VB. References 1 European Centre for Disease Prevention and Control/WHO Regional Office for Europe. HIV/AIDS surveillance in Europe 2008. Stockholm: European Centre for Disease Prevention and Control, 2009. 2 Janssen RS, Satten GA, Stramer SL, et al. New testing strategy to detect early HIV-1 infection for use in incidence estimates and for clinical and prevention purposes. JAMA 1998; 280: 42–48. 3 Le Vu S, Pillonel J, Semaille C, et al. Principles and uses of HIV incidence estimation from recent infection testing—a review. Euro Surveill 2008; 13: 1–6. 4 Hall HI, Song R, Rhodes PH, et al. Estimation of HIV incidence in the United States. JAMA 2008; 300: 520–29. 5 WHO Technical Working Group on Statistical Approaches for Development, Validation and Use of HIV Incidence Assays. Final report. Geneva: World Health Organization, 2009. 6 Barin F, Meyer L, Lancar R, et al. Development and validation of an immunoassay for identification of recent human immunodeficiency virus type 1 infections and its use on dried serum spots. J Clin Microbiol 2005; 43: 4441–47. 7 Goujard C, Chaix M, Lambotte O, et al. Spontaneous control of viral replication during primary HIV infection: when is “HIV controller” status established? Clin Infect Dis 2009; 49: 982–86. 8 Madec Y, Boufassa F, Avettand-Fenoel V, et al. Early control of HIV-1 infection in long-term nonprogressors followed since diagnosis in the ANRS SEROCO/HEMOCO cohort. J Acquir Immune Defic Syndr 2009; 50: 19. 9 Semaille C, Cazein F, Pillonel J, et al. Four years of surveillance of recent HIV infections at country level, France, mid 2003–2006: experience and perspectives. Euro Surveill 2008; 13: 17–22. 10 Bajos N, Bozon M, Beltzer N. Enquête sur la sexualité en France : pratiques, genre et santé. Paris: Editions La Découverte, 2008. 11 Costes J, Vaissade L, Colasante E, et al. Prévalence de l’usage problématique de drogues en France—estimations 2006. SaintDenis: Observatoire Français des Drogues et des Toxicomanies, 2009. 12 National Institute of Statistics and Economic Studies. National population census. http://www.insee.fr/fr/themes/detail.asp?reg_ id=0&ref_id=ir-sd2008&page=irweb/sd2008/dd/sd2008_population. htm (accessed July 26, 2010).
www.thelancet.com/infection Vol 10 October 2010
13
14
15
16
17 18
19
20 21
22
23
24
25
26
27
28
29
30
31
Van Buuren S. Multiple imputation of discrete and continuous data by fully conditional specification. Stat Methods Med Res 2007; 16: 219–42. Brookmeyer R, Liao JG. The analysis of delays in disease reporting: methods and results for the acquired immunodeficiency syndrome. Am J Epidemiol 1990; 132: 355–65. Karon JM, Song R, Brookmeyer R, Kaplan EH, Hall HI. Estimating HIV incidence in the United States from HIV/AIDS surveillance data and biomarker HIV test results. Stat Med 2008; 27: 4617–33. Cooley PC, Myers LE, Hamill DN. A meta-analysis of estimates of the AIDS incubation distribution. Eur J Epidemiol 1996; 12: 229–35. Greene WH. Econometric analysis. New York: Macmillan, 1993. Cazein FC, Pillonel J, Imounga L, et al. Surveillance du dépistage et du diagnostic de l’infection VIH et du sida, France, 2008. BEHWeb 2009; 2: 1–15. Sullivan PS, Hamouda O, Delpech V, et al. Reemergence of the HIV epidemic among men who have sex with men in North America, western Europe, and Australia, 1996–2005. Ann Epidemiol 2009; 19: 423–31. Griensven FV. The global epidemic of HIV infection among men who have sex with men. Curr Opin HIV AIDS 2009; 4: 300–07. Stall R, Duran L, Wisniewski SR, et al. Running in place: implications of HIV incidence estimates among urban men who have sex with men in the United States and other industrialized countries. AIDS Behav 2009; 13: 615–29. Velter A, Bouyssou-Michel A, de Busscher PO, Jauffret-Roustide M, Semaille C. Enquête presse gay 2004. Saint-Maurice: Institut de Veille Sanitaire, 2007. Bouyssou-Michel A, Gallay A, Janier M, et al. Surveillance de la syphilis en France, 2000–2006: recrudescence des diagnostics en 2006. Bull Epidemiol Hebd 2008; 5–6: 39–42. Anema A, Wood E, Montaner JS. The use of highly active retroviral therapy to reduce HIV incidence at the population level. CMAJ 2008; 179: 13–14. Katz MH, Schwarcz SK, Kellogg TA, et al. Impact of highly active antiretroviral treatment on HIV seroincidence among men who have sex with men: San Francisco. Am J Public Health 2002; 92: 388–94. Emmanuelli J, Desenclos J. Harm reduction interventions, behaviours and associated health outcomes in France, 1996–2003. Addiction 2005; 100: 1690–1700. Yéni P. Prise en Charge Médicale des Personnes Infectées par le VIH: Rapport 2010 (version préliminaire). La documentation française. Paris: Ministère de la Santé et des Sports, 2010. Hamers FF, Downs AM. The changing face of the HIV epidemic in western Europe: what are the implications for public health policies? Lancet 2004; 364: 83–94. Lydié N, Beltzer N, Fénies K, Halfen S, Lert F, Le Vu S. Les populations africaines d’Ile-de-France face au VIH/sidaconnaissances, attitudes, croyances et comportements. Saint-Denis: Institut National de Prévention et d’Éducation à la Santé, 2007. Le Vu S, Meyer L, Cazein F, et al. Performance of an immunoassay at detecting recent infection among reported HIV diagnoses. AIDS 2009; 23: 1773–79. Remis RS, Palmer RWH. Testing bias in calculating HIV incidence from the serologic testing algorithm for recent HIV seroconversion. AIDS 2009; 23: 493–503.
687
Articles
Microscopic-observation drug susceptibility and thin layer agar assays for the detection of drug resistant tuberculosis: a systematic review and meta-analysis Jessica Minion, Erika Leung, Dick Menzies, Madhukar Pai
Summary Lancet Infect Dis 2010; 10: 688–98 Published Online September 1, 2010 DOI:10.1016/S14733099(10)70165-1 See Reflection and Reaction page 656 Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, QC, Canada (J Minion MD, E Leung MSc, Prof D Menzies MD, M Pai MD); Respiratory Epidemiology and Clinical Research Unit, Montreal Chest Institute, Montreal, QC, Canada (J Minion, E Leung, Prof D Menzies, M Pai); and Department of Medical Microbiology and Immunology, University of Alberta, Edmonton, AB, Canada (J Minion) Correspondence to: Dr Madhukar Pai, Department of Epidemiology, Biostatistics and Occupational Health, McGill University, 1020 Pine Avenue West, Montreal, QC, Canada H3A 1A2
[email protected]
Background Simple, rapid, and affordable tests are needed to detect drug resistance in Mycobacterium tuberculosis. We did a systematic review and meta-analysis to investigate the accuracy of microscopic-observation drug susceptibility (MODS) and thin layer agar (TLA) assays for rapid screening of patients at risk of drug-resistant tuberculosis. Methods In accordance with protocols and methods recommended by the Cochrane Diagnostic Test Accuracy Working Group, we systematically searched PubMed, Embase, and Biosis for reports published between January, 1990, and February, 2009. We included studies investigating detection of drug resistance in M tuberculosis with the MODS or TLA assay, and in which an accepted reference standard was used. Data extracted from the studies were combined by use of bivariate random-effects regression models and hierarchical summary receiver operating characteristic curves to estimate sensitivity and specificity for detection of resistance to specific drugs. Findings We identified 12 studies, of which nine investigated the MODS assay and three investigated the TLA assay. For the MODS assay of rifampicin resistance, pooled estimates were 98·0% (95% CI 94·5–99·3) for sensitivity and 99·4% (95·7–99·9) for specificity. For the MODS assay of isoniazid resistance with a 0·1 μg/mL cutoff, pooled sensitivity was 97·7% (94·4–99·1) and pooled specificity was 95·8% (88·1–98·6), but with a 0·4 μg/mL cutoff, sensitivity decreased to 90·0% (84·5–93·7) and specificity increased to 98·6% (96·9–99·4). All assessments of rifampicin and isoniazid resistance with the TLA assay yielded 100% accuracy. Mean turnaround time was 9·9 days (95% CI 4·1–15·8) for the MODS assay and 11·1 days (10·1–12·0) for the TLA assay. Interpretation MODS and TLA assays are inexpensive, rapid alternatives to conventional methods for drug susceptibility testing of M tuberculosis. Our data and expert opinion informed WHO’s recommendation for use of selected non-commercial drug susceptibility tests, including MODS, as an interim solution until capacity for genotypic or automated liquid culture drug susceptibility testing is developed. Funding Stop TB Department of WHO.
Introduction The growing problem of drug resistance in Mycobacterium tuberculosis is accompanied by increasing demand for quick, cheap, and easy techniques to detect resistance.1 Patients infected with multidrug resistant (MDR) or extensively drug resistant (XDR) strains of M tuberculosis need treatment regimens that include second-line drugs in addition to extended treatment duration.2 Even when treated appropriately, MDR and XDR tuberculosis have substantially worse outcomes3 and longer infectious periods than does drug-susceptible tuberculosis.4 Therefore, prevention of the spread of MDR and XDR tuberculosis is important, for which timely identification of such cases is the first and most crucial step. Several methods to detect drug resistance are available, but none clearly satisfies the demands of quick, cheap, and easy. Traditional agar-based methods can take months for results. Commercial drug susceptibility testing with liquid culture decreases turnaround times but requires expensive equipment. Molecular detection of gene mutations associated with drug resistance has also been developed, with variable sensitivity reported especially for
688
in-house methods.5 Commercial versions of line-probe assays have high accuracy6,7 and were recently endorsed by WHO for rapid screening of MDR tuberculosis.8 However, the expertise needed for a laboratory to offer molecular diagnostics is too great for these tests to be implemented in many resource-constrained settings with poor laboratory infrastructure. Although liquid cultures and line-probe assays have been endorsed by WHO and phased implementation is underway in many countries, interim measures are needed to meet the needs of low-income settings with high rates of MDR and XDR tuberculosis. Noncommercial techniques to test drug susceptibility with inexpensive and widely available laboratory equipment and supplies—such as microscopic-observation drug susceptibilty (MODS) and thin layer agar (TLA) assays— could be implemented in these settings with minimum cost and training. In both MODS and TLA testing, drugfree and drug-containing media (liquid for MODS, solid for TLA) are inoculated with specimens from patients, and cultures are microscopically examined for early growth or microcolonies.9,10 Growth of M tuberculosis in www.thelancet.com/infection Vol 10 October 2010
Articles
or on drug-free media indicates a positive culture, whereas growth of M tuberculosis in or on both drugfree and drug-containing media indicates resistance. We did a systematic review of published reports and a meta-analysis of studies examining the diagnostic accuracy, performance (including contamination rate and turnaround time), and outcomes important to patients of MODS and TLA assays for the detection of drug resistance in M tuberculosis. We followed a standard protocol for systematic reviews and meta-analyses,11 and used methods recommended by the Cochrane Diagnostic Test Accuracy Working Group.12
including outcomes important to patients, were reviewed and summarised narratively.
Assessment of study quality We used the QUADAS criteria13 for assessment of quality of diagnostic accuracy in studies to assess quality characteristics that were judged to be important for this review: masked interpretation of the test result with reference standard results and vice-versa; complete verification of test results with the reference standard; recruitment of patients or collection of specimens either consecutively or randomly; and cross-sectional study design (not case-control).
Methods Search strategy and study selection
Data extraction
We searched PubMed, Embase, and Biosis for reports published between January, 1990, and February, 2009 (inclusive). The first search used the terms tuberculosis[mesh] OR mycobacter*[ti] OR acid-fast[ti] OR tuberculous[ti] and the second search used the terms MODS[tw] OR “microscopic-observation”[tiab] OR “drugsusceptibility”[ti] OR (drug[ti] AND susceptibility[ti]) OR microcolony[tiab] OR (liquid[ti] AND culture[ti]) OR thinlayer[tiab] OR (thin[tiab] AND layer[tiab] AND agar[tiab]) OR agar[tiab] OR (mycobacterial[ti] AND culture[ti]); results from the first and second searches were combined with the Boolean operator AND so that all studies identified were listed in the results of both the first and second searches. All database searches were done independently by an experienced librarian and JM. We restricted the search to reports published in English, French, or Spanish. Studies were selected independently by two reviewers (JM and EL) and disagreements were resolved by consensus. Titles and abstracts were screened for relevance and any citations identified by either reviewer were selected for full-text review. Articles retrieved for full-text review and full reasons for exclusion are available from the authors. Reference lists from selected studies were hand searched and experts and test developers were contacted to identify additional studies. Predetermined eligibility criteria for studies included in the primary analysis were assessment of the MODS or TLA assay for drug susceptibility of M tuberculosis, and use of an accepted reference standard. Accepted reference standards included indirect proportion methods (eg, a sensitive isolate has <1% resistant population whereas a resistant isolate has >1%), absolute concentration, resistance ratio, commercial liquid systems (BACTEC 460, Becton Dickinson Diagnostic Instrument Systems, Sparks, MD, USA; MGIT 960, Becton Dickinson Diagnostic Instrument Systems; MB/BacT, Organon Teknika, Turnhout, Belgium), or microdilution methods. We included studies in which either the specimens from patients were used for direct inoculation or previously cultured isolates were used for indirect inoculation. Studies in which outcomes other than accuracy were measured,
We created and piloted a data extraction form with a subset of eligible studies and used experience gained from the pilot study to finalise the extraction form. All studies included in the final review were extracted independently by two reviewers (JM and EL) and any disagreements were resolved by consensus. Data were extracted for every drug tested to construct two-by-two tables of true positive, false positive, false negative, and true negative values. We defined true positives as isolates identified as drug resistant by the MODS or TLA assay and the reference method. False positives were isolates identified as resistant by the MODS or TLA assay, but susceptible by the reference method. False negatives were isolates identified as susceptible on MODS or TLA assay, but resistant in the reference method. True negatives were isolates identified as susceptible by the MODS or TLA assay and the reference method.
www.thelancet.com/infection Vol 10 October 2010
Statistical analysis We calculated the sensitivity and specificity for every drug tested. Results are presented separately by index test (MODS or TLA) for every drug tested. Subgroup analysis was done according to the drug concentration used to define the resistance cutoff. For sensitivity and specificity subgroup estimates, pooled with a randomeffects model, we regarded subgroups with overlapping 95% CIs to be homogeneous groups and combined these groups for the primary analysis. Otherwise results are presented separately for different cutoffs. Subgroup analysis was also done with exclusion of studies with substantial protocol deviations or using microdilution reference standards. We also planned a subgroup analysis by the type of specimen used (ie, direct samples from patients or indirect isolates). Other outcomes extracted included the following: turnaround time, defined as the time from specimen receipt or processing in the laboratory to availability of results in the laboratory; contamination rate, defined as the proportion of specimens contaminated with bacterial or fungal growth on first inoculation or appropriate assessment of cross-contamination; cost estimates, which include costs for reagents, supplies, equipment, 689
Articles
labour, overheads, or other justified related costs; and outcomes important to patients. Data were analysed with Stata/IC (version 11.0). Forest plots of sensitivity and specificity estimates and their 2067 records identified 730 from PubMed 610 from Embase 727 from Biosis
5 records identified from other sources
2018 excluded 666 duplicate studies 1352 on the basis of title or abstract
54 full-text articles reviewed
27 excluded 13 not MODS or TLA 11 reviews, letters, and editorials 2 no outcomes of interest 1 inappropriate reference standard
95% CIs were constructed from every study with MetaDiSc software (version 1.4) and by use of exact methods for proportions.14 Sensitivity and specificity estimates tend to be correlated and vary according to thresholds (either explicit or implicit cutoff values determining whether results are positive or negative), so we analysed hierarchical summary receiver operating characteristic (HSROC) curves to explore the effect of the thresholds and produce an overall summary of test accuracy.15–17 Accuracy measures were pooled by use of bivariate random-effects regression models,18 with the user-written program metandi in Stata.19 Heterogeneity of accuracy estimates was assessed by I2 statistic.20 Subgroups with fewer than four studies were combined by use of univariate random-effects models because bivariate random-effects regression models do not converge with small numbers of studies. Contamination rates were combined by use of weighted means and turnaround times were combined by use of simple means.
Role of the funding source 11 assessing the TLA assay*
WHO had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. The authors had full access to all study data and are solely responsible for the decision to submit for publication.
17 assessing the MODS assay*
8 excluded 8 assessing detection only*
3 included in analysis
8 excluded 6 assessing detection only* 2 assessing outcomes other than accuracy (discussed narratively)
Results We identified 2072 citations, of which 1406 unique articles remained after exclusion of duplicate articles, and 54 remained after screening of titles and abstracts (figure 1). These articles were retrieved for full-text review, of which 12 were eligible for inclusion in the primary analysis: nine of MODS testing21–29 and three of the TLA
9 included in analysis
Figure 1: Study selection MODS=microscopic-observation drug susceptibility. TLA=thin layer agar. *One study assed both the MODS and TLA assays.
Samples
Country
Drugs tested
Reference
Inoculation* Smear positive Sputum
HIV positive
Microscopic-observation drug susceptibility assay 207
Peru
Rifampicin, isoniazid, ethambutol, streptomycin
MABA
Direct
NS
NS
NS
Caviedes et al (2000)22
88
Peru
Rifampicin, isoniazid
MABA
Direct
100%
100%
NS
Devasia et al (2009)23
239
USA
Ofloxacin
Proportion method
Indirect
NA
NA
NA
Ethiopia
Rifampicin, isoniazid
MGIT 960
Direct
100%
100%
NS NS
Moore et al (2004)21
Ejigu et al (2008)24
58
Mello et al (2007)25†
180
Mengatto et al (2006)26 Moore et al (2006)27 Park et al (2002)28 Shiferaw et al (2007)29
64 338 53
Brazil
Rifampicin, isoniazid
Proportion method
Direct
NS
100%
Argentina
Rifampicin, isoniazid, ethambutol, streptomycin
Proportion method
Indirect
NA
NA
NA
Peru
Rifampicin, isoniazid, ethambutol, streptomycin
Proportion method
Direct
5·9%
100%
7·9%
USA
Rifampicin, isoniazid, ethambutol, streptomycin
Proportion method
Indirect
NA
NA
NA
247
Ethiopia
Rifampicin, isoniazid
Proportion method
Direct
100%
100%
NS
Thin layer agar assay Martin et al (2009)30
147
Belgium
Rifampicin, ofloxacin, kanamycin
Proportion method
Indirect
NA
NA
NA
Robledo et al (2008)31
95
Columbia
Rifampicin, isoniazid
Proportion method
Direct
100%
100%
NS
Schaberg et al (1995)32
197
Germany
Rifampicin, isoniazid, ethambutol, streptomycin, pyrazinamide
Proportion method
Direct
48%
100%
1%
MABA=microplate alamar blue assay. NS=not specified (studies using direct inocula). NA=not applicable (studies using indirect inocula). *Direct inoculation refers to direct application of a specimen from a patient (processed or unprocessed) to drug-containing and drug-free media; indirect inoculation refers to application of a previously isolated strain of Mycobacterium tuberculosis to drug-containing and drug-free media. †Resistance defined as growth in drug-containing wells on day 14 after growth was detected in drug-free wells, which is opposed to the standard procedure to detect the presence of growth in drug-containing wells on the same day as growth is detected in drug-free wells.
Table 1: Characteristics of studies included
690
www.thelancet.com/infection Vol 10 October 2010
Articles
assay30–32 (table 1; see webappendix for primary accuracy data from included studies). We identified two additional studies that did not measure accuracy, and so were not included in the main meta-analysis, but reported on other outcomes of interest.33,34 All but one study presented results for resistance to rifampicin and all but two presented results for resistance to isoniazid. Data were also available from some studies for resistance to ethambutol, streptomycin, ofloxacin, kanamycin, and pyrazinamide. Three MODS studies and one TLA study used previously cultured isolates (indirect inoculates) and the remainder used clinical specimens (direct inoculates) for resistance testing. Four studies (three MODS and one TLA) used only smear-positive specimens, and seven studies (five MODS and two TLA) used only sputum specimens. Only two studies, one of each assay type, provided information on the HIV status of patients providing clinical specimens for testing. All studies verified their complete study sample with a reference standard and eight of 12 had a cross-sectional design Microscopic-observation drug susceptibility assay (n=9)
Thin layer agar assay (n=3)
Recruitment of specimens or patients Prospective Retrospective Unclear
4 2 3
2 1 0
6 3
2 1
5 3 1
1 0 2
Study design Cross-sectional Unclear Sampling Consecutive or random Convenience Unclear
Verification with accepted reference Complete
9
3
Masked interpretation of index and use of reference test Yes Unclear
5 4
2 1
Table 2: Study quality
(table 2). Half the studies used either consecutive or random samples, and half stated that test interpretation was masked with respect to the reference standard results. Six studies were done prospectively, three were done retrospectively, and three did not clearly report type of recruitment. Sensitivity and specificity estimates for assessment of rifampicin resistance are shown for the MODS assay (figure 2) and TLA assay (figure 3). The accuracy of studies that used a 1 μg/mL concentration cutoff did not differ from those that used 2 μg/mL (data not shown) so results for all rifampicin assessments were combined. Estimates of sensitivity and specificity are also presented for resistance to isoniazid with the MODS assay (figure 4) and TLA assay (figure 5). For the MODS assay, data were stratified by drug concentrations of 0·1 μg/mL and 0·4 μg/mL; data differed significantly between these cutoff values so all results for the MODS assay of isoniazid resistance were stratified accordingly. Several studies tested both drug concentrations and so have two sets of results. For the TLA assay, studies used drug concentrations of 0·2 μg/mL and 0·25 μg/mL isoniazid. Estimates from the MODS assay had less variability with rifampicin than with isoniazid for sensitivity (range 92–100% vs 82–100%) and specificity (83–100% vs 78–100%). Sensitivity and specificity estimates for resistance to rifampicin and isoniazid with the TLA assay were all 100%, but few studies were included (three for rifampicin, two for isoniazid). For drugs tested in only one study for each type of assay, both the MODS23 and TLA30 assays reported resistance to ofloxacin with 100% accuracy (table 3). Resistance to each of ethambutol, streptomycin, pyrazinamide, and kanamycin was also assessed with the TLA assay which seemed reliable but few data were available (table 3). For drugs tested in more than one study, we pooled estimates for sensitivity and specificity (table 4). For the MODS assay, sensitivity and specificity estimates were highest for resistance to rifampicin. Use of the 0·1 μg/mL isoniazid cutoff was associated with significantly higher sensitivity, but lower specificity (not significant), than was the
Sensitivity (95% CI)
See Online for webappendix
Specificity (95% CI)
Moore et al (2004)21
1·00 (0·40–1·00)
0·99 (0·96–1·00)
Caviedes et al (2000)22
1·00 (0·79–1·00)
0·97 (0·90–1·00)
Ejigu et al (2008)24
0·95 (0·75–1·00)
1·00 (0·91–1·00)
Mello et al (2007)25
0·96 (0·89–0·99)
0·83 (0·74–0·90)
Moore et al (2006)27
1·00 (0·90–1·00)
1·00 (0·99–1·00)
Shiferaw et al (2007)29
0·92 (0·75–0·99)
1·00 (0·98–1·00)
Mengatto et al (2006)26
1·00 (0·86–1·00)
1·00 (0·91–1·00)
Park et al (2002)28
1·00 (0·92–1·00)
1·00 (0·72–1·00)
0
0·2
0·4
0·6
0·8
1·0
0
0·2
0·4
0·6
0·8
1·0
Figure 2: Forest plot of accuracy of the microscopic-observation drug susceptibility assay for rifampicin resistance Open squares represent studies using direct inoculation with specimens from patients; red squares represent studies using indirect inoculation with isolates. Size of the square is proportionate to the size of the study.
www.thelancet.com/infection Vol 10 October 2010
691
Articles
Sensitivity (95% CI)
Specificity (95% CI)
Robledo et al (2008)31
1·00 (0·69–1·00)
1·00 (0·96–1·00)
Schaberg et al (1995)32
1·00 (0·63–1·00)
1·00 (0·98–1·00)
Martin et al (2009)30
1·00 (0·97–1·00)
1·00 (0·86–1·00)
0
0·2
0·4
0·6
0·8
1·0
0
0·2
0·4
0·6
0·8
1·0
Figure 3: Forest plot of accuracy of the thin layer agar assay for detection of rifampicin resistance Open squares represent studies using direct inoculation with specimens from patients; red squares represent studies using indirect inoculation with isolates. Size of the square is proportionate to the size of the study.
Sensitivity (95% CI)
Specificity (95% CI)
0·1 ug/mL concentration cutoff Caviedes et al (2000)22
1·00 (0·85–1·00)
0·98 (0·92–1·00)
Ejigu et al (2008)24
0·97 (0·84–1·00)
0·92 (0·74–0·99)
Mello et al (2007)25
0·97 (0·91–0·99)
0·78 (0·68–0·86)
Shiferaw et al (2007)29
0·94 (0·84–0·99)
0·97 (0·93–0·99)
Mengatto et al (2006)26
1·00 (0·87–1·00)
0·97 (0·86–1·00)
Park et al (2002)28
1·00 (0·92–1·00)
1·00 (0·69–1·00)
Moore et al (2004)21
0·82 (0·66–0·92)
0·97 (0·93–0·99)
Caviedes et al (2000)22
1·00 (0·85–1·00)
0·98 (0·92–1·00)
Ejigu et al (2008)24
0·94 (0·79–0·99)
0·96 (0·80–1·00)
Moore et al (2006)27
0·85 (0·74–0·92)
1·00 (0·98–1·00)
Shiferaw et al (2007)29
0·90 (0·77–0·97)
0·98 (0·96–1·00)
Mengatto et al (2006)26
0·88 (0·70–0·98)
1·00 (0·91–1·00)
Park et al (2002)28
0·95 (0·85–0·99)
1·00 (0·69–1·00)
0·4 ug/mL concentration cutoff
0
0·2
0·4
0·6
0·8
1·0
0
0·2
0·4
0·6
0·8
1·0
Figure 4: Forest plot of accuracy of the microscopic-observation drug susceptibility assay for detection of isoniazid resistance Open squares represent studies using direct inoculation with specimens from patients; red squares represent studies using indirect inoculation with isolates. Size of the square is proportionate to the size of the study.
Sensitivity (95% CI) 31
Specificity (95% CI)
Robledo et al (2008)
1·00 (0·69–1·00)
1·00 (0·96–1·00)
Schaberg et al (1995)32
1·00 (0·63–1·00)
1·00 (0·98–1·00)
0
0·2
0·4
0·6
0·8
1·0
0
0·2
0·4
0·6
0·8
1·0
Figure 5: Forest plot of accuracy of the thin layer agar assay for detection of isoniazid resistance Both studies used direct inoculation with specimens from patients. Size of the square is proportionate to the size of the study.
Study
Specimen
True positive
False positive
False negative
True negative
Sensitivity (95% CI)
Specificity (95% CI)
233
100% (54·1–100)
100% (98·4–100)
Microscopic-observation drug susceptibility assay Ofloxacin
Devasia et al (2009)23
Indirect
6
0
0
Thin layer agar assay Ethambutol
Shaberg et al (1995)32
Direct
5
1
0
191
100% (47·8–100)
99·5% (97·1–100)
Streptomycin
Shaberg et al (1995)32
Direct
21
0
0
176
100% (83·9–100)
100% (97·9–100)
Pyrazinamide
Schaberg et al (1995)32
Direct
6
0
0
189
100% (54·1–100)
100% (98·1–100)
Ofloxacin
Martin et al (2009)30
Indirect
39
0
0
95
100% (91·0–100)
100% (96·2–100)
Kanamycin
Martin et al (2009)30
Indirect
67
1
0
77
100% (94·6–100)
98·7% (93·1–100)
Table 3: Drugs assessed by single studies
692
www.thelancet.com/infection Vol 10 October 2010
Articles
Number of studies
Pooled sensitivity (95% CI)
Pooled specificity (95% CI)
I2* (p value)
I2* (p value)
All included studies Microscopic-observation drug susceptibility assay 1 or 2 μg/mL rifampicin
8
98·0% (94·5–99·3)
0·0% (0·86)
99·4% (95·7–99·9)
68·5% (0·002)
0·1 μg/mL isoniazid
6
97·7% (94·4–99·1)
22·4% (0·27)
95·8% (88·1–98·6)
84·4% (<0·0001)
0·4 μg/mL isoniazid
7
90·0% (84·5–93·7)
45·9% (0·09)
98·6% (96·9–99·4)
18·8% (0·29)
2·5 μg/mL ethambutol
4
60·0% (47·3–71·6)
46·6% (0·13)
95·4% (83·1–98·9)
80·6% (0·001)
2 μg/mL streptomycin†
3
70% (62–78)
58·0% (0·09)
99% (97–100)
69·3% (0·54)
6 μg/mL streptomycin†
3
44% (32–57)
69·3% (0·04)
99% (97–100)
0% (0·45)
Thin layer agar assay 3
100% (97–100)
0% (1·0)
100% (99–100)
0% (1·0)
0·2 or 0·25 μg/mL isoniazid† 2
1 μg/mL rifampicin†
100% (91–100)
0% (1·0)
100% (99–100)
0% (1·0)
Studies with stringent exclusion criteria applied‡ Microscopic-observation drug susceptibility assay 1 or 2 μg/mL rifampicin
5
98·7% (89·4–99·9)
45·4% (0·12)
99·9% (95·8–100)
0·1 μg/mL isoniazid
4
97·6% (90·9–99·4)
41·6% (0·16)
96·6% (93·4–98·3)
0% (0·73) 0% (0·57)
0·4 μg/mL isoniazid
5
90·2% (84·4–93·9)
5·9% (0·37)
99·0% (97·3–99·6)
2·4% (0·39)
2·5 μg/mL ethambutol†
3
67% (55–77)
0% (0·81)
97% (95–99)
6 μg/mL streptomycin†
2
45% (32–58)
84·5% (0·01)
100% (94–100)
91·2% (<0·0001) 0% (1·0)
*I2 statistic measures the percentage of total variation across studies due to heterogeneity. †Pooled estimates calculated with the univariate random-effects model because too few studies were available for the bivariate random-effects model to converge. ‡Only studies in which estimates were affected by stringent exclusion criteria are listed.
Table 4: Pooled accuracy estimates, stratified by test and drug
0·4 μg/mL cutoff. Subgroup analysis of MODS studies also showed a significant difference between cutoff values of streptomycin (6 μg/mL vs 2 μg/mL) for detection of resistance, so results were stratified. For the MODS assay of ethambutol and streptomycin resistance, sensitivity for detection of resistant strains was poor, but specificity for classification of susceptible strains was fairly accurate. In two studies of the MODS assay, a microdilution method was used as the reference standard,21,22 and a third study used a substantially different protocol to read resistance from the MODS assay.25 Such study characteristics could have been grounds for exclusion, so we repeated our analysis without these three studies, but none of the estimates changed significantly (table 4). We did subgroup analyses of two important characteristics of study quality, blinding and sample selection, to explore the effects of these characteristics on pooled estimates of accuracy for detection of rifampicin resistance. Comparison of five studies that clearly reported masked interpretation of the index and use of reference tests with three studies that did not report blinding showed no significant differences in pooled sensitivity or specificity estimates (data not shown). Similarly, differences in pooled estimates of accuracy did not differ between the four studies in which random or consecutive sampling was used to select specimens and the four studies that used convenience sampling or did not clearly report the method of sample selection (data not shown). Too few studies were available to compare the effects of direct versus indirect inoculation on the performance of the TLA assay. For the MODS assay, sensitivity and www.thelancet.com/infection Vol 10 October 2010
Number Pooled sensitivity of studies (95% CI)
I2* (p value)
Pooled specificity (95% CI)
I2* (p value)
1 or 2 μg/mL rifampicin Direct specimen
6
Indirect specimen†
2
96·8% (92·4–98·7) 100% (95–100)
3·3% (0·40) 0% (1·0)
99·0% (94·3–99·8) 92·3% (<0·0001) 100% (93–100)
0% (1·0)
0·1 μg/mL isoniazid Direct specimen
4
Indirect specimen†
2
96·4% (92·4–98·4) 100% (95–100)
0% (0·53)
94·2% (83·6–98·1)
0% (1·0)
98% (89–100)
89·7% (<0·0001) 0% (0·49)
0·4 μg/mL isoniazid Direct specimen
7
88·6% (82·7–92·7)
54·1% (0·07)
98·5% (96·8–99·3) 33·0% (0·20)
Indirect specimen†
2
93% (84–98)
13·5% (0·28) 100% (93–100)
0% (1·0)
*I2 statistic measures the percentage of total variation across studies due to heterogeneity. †Pooled estimates calculated with the univariate random-effects model because too few studies were available for the bivariate random-effects model to converge.
Table 5: Accuracy of the microscopic-observation drug susceptibility assay when used on direct versus indirect specimens
specificity for detection of resistance to rifampicin or isoniazid (at either drug concentration) did not differ significantly between the types of inoculate (table 5). However, accuracy estimates with direct specimens were lower in all cases than with indirect specimens (range –3·2% to –4·4% for sensitivity, –1·0% to –3·8% for specificity). No direct head-to-head comparisons were identified for MODS versus TLA drug susceptibility 693
Articles
1·0
0·8
Sensitivity
0·6
0·4
0·2
0 1·0
0·8
0·6
0·4
0·2
0
Specificity
Figure 6: Hierarchical summary receiver operating characteristic plot of the microscopic-observation drug susceptibility assay for detection of rifampicin resistance Open squares represent individual studies, with the size of the square proportionate to the size of the study. The summary point is a closed circle, representing sensitivity and specificity estimates pooled with a bivariate random-effects model. The hierarchical summary receiver operating characteristic curve is truncated outside of the area for which data exist.
For more on MODS testing procedures see http://www.modsperu.org
694
testing on the same specimens, or for direct versus indirect specimen inoculation. We plotted sensitivity (or true positives rate) and one minus specificity (or false positives rate) in an HSROC curve for the MODS assay of rifampicin resistance (figure 6) and isoniazid resistance (figure 7). Sensitivity seemed generally more variable than did specificity, and this effect was more pronounced for isoniazid than for rifampicin MODS testing. Median or mean turnaround times were provided in six MODS studies and two TLA studies, and were stratified by type of inoculate (table 6). Mean turnaround times with the MODS assay were slightly shorter than that with the TLA assay, and, for the MODS assay, use of direct specimens was associated with longer turnaround times than was use of indirect isolates, but 95% CIs overlapped for both comparisons, suggesting that neither difference was significant. All studies reporting bacterial or fungal contamination rates for the MODS or TLA assay used direct specimens for inoculation (table 7). The contamination rate for the TLA assay was significantly lower than that for the MODS assay, with non-overlapping 95% CIs, but data were only available from two TLA studies and three MODS studies.
None of the studies assessing the MODS or TLA assay included a thorough investigation for cross-contamination between cultures. However, in a separate study by Moore and colleagues’,34 a combination of reculturing, molecular typing, and clinical epidemiology was used to identify cross-contaminated cultures from the MODS assay, MB/BacT (Biomerieux, Durham, NC, USA), and Löwenstein-Jensen culture. Of 362 positive samples (11% positivity in 3416 samples), 17 false positive cultures were identified in 14 specimens: 12 with MODS testing, four with MB/BacT, and one with Löwenstein-Jensen culture. Most false positive samples from MODS testing showed evidence of growth in only one drug-free well, and if the definition of a positive culture was revised to demand growth in two drug-free wells, the number of crosscontaminated cultures reduced to three. This amended definition has been incorporated into the standard operating procedure for MODS testing for tuberculosis.9 Four studies reported estimates for the cost of drug susceptibility testing with the MODS or TLA assay.22,23,26,27 Caviedes and colleagues (2000)22 estimated the reagent and supplies costs for testing against two drugs (rifampicin and isoniazid) to be US$1·72 per sample for the MODS assay and $1·60 per sample for the TLA assay. Equivalent estimates for testing against four drugs (rifampicin, isoniazid, ethambutol, and streptomycin) were $1·80 per sample for MODS testing and $2·92 per sample for TLA testing. Similar costs for MODS testing were subsequently reported in 2006 by the same research group at $2 per sample,27 and by Mengatto and colleagues (2006)26 at $1·57 per sample for testing against two drugs and $2·17 per sample for testing against four drugs. Devasia and colleagues (2009)23 reported costs of $1·38 per sample for MODS testing against ofloxacin (including reagents and supplies) after purchase of an incubator (estimated cost $8000) and an inverted microscope (estimated cost $4000). None of the studies considered the costs of labour, capital costs, or overhead costs associated with MODS or TLA testing. All prices were reported for the year of publication. Although outcomes important to patients were not directly assessed, Nic Fhogartaigh and colleagues33 reported a retrospective database analysis of patients from Lima, Peru, whose specimens were selected by their physicians to undergo MODS testing. Of 209 patients with positive culture samples, more than 80% had culture confirmation or availability of the test result from the MODS assay before any standard method. In 41·4% of patients with positive culture samples, the results from MODS testing should have prompted a modification in management of patients.
Discussion From pooled estimates of nine studies, MODS testing had high accuracy for detection of rifampicin resistance, but showed slightly lower sensitivity for detection of isoniazid resistance. We identified only three studies www.thelancet.com/infection Vol 10 October 2010
Articles
www.thelancet.com/infection Vol 10 October 2010
1·0
0·8
0·6 Sensitivity
assessing the TLA assay; however, in all studies, the assay had 100% concordance with the reference standards for detection of rifampicin or isoniazid resistance. For detection of isoniazid restistance with the MODS assay, studies in which a cutoff of 0·1 μg/mL was used had significantly higher sensitivity, but somewhat lower specificity, than did studies in which a cutoff of 0·4 μg/mL was used. In view of pharmacokinetic and pharmacodynamic evidence supporting further lowering of isoniazid cutoffs to define resistance, we recommend use of the conservative 0·1 μg/mL cutoff for cases in which only one concentration can be used.35 MODS testing of ethambutol and streptomycin showed variable specificity but generally low sensitivity. Difficulty with standardisation of susceptibility testing of M tuberculosis against many drugs is not unique to the MODS assay and variable concordance between other methods has also been reported.36 For other antituberculous drugs tested in only one study, the TLA assay had 100% sensitivity and more than 98% specificity. Although differences in accuracy between studies using direct specimens for inoculation and those using indirect isolates were not significant, direct inoculation was associated with lower sensitivity and specificity. Turnaround times for both the MODS and TLA assays were much faster than is conventional proportion method drug susceptibility testing and similar to commercial liquid systems.37 For the TLA assay, all turnaround time estimates came from studies using direct specimens, but stratification by specimen type for the MODS assay showed a longer turnaround time for direct inoculation than for indirect inoculation. MODS testing seemed to have a higher contamination rate than did the TLA assay. This review had several strengths, including a broad and inclusive search of published reports together with efforts to identify unpublished studies. Study selection and data extraction was done independently by two reviewers. Additionally, we used rigorous statistical methods—bivariate random-effects models where possible and HSROC curves for assessment of diagnostic accuracy—which have been recommended by the Cochrane Diagnostic Test Accuracy Working Group as the methods of choice for diagnostic meta-analyses.12 However, the review also had several limitations. First, few studies were available for some assessments, especially for the TLA assay and for drugs other than rifampicin and isoniazid, which meant that bivariate models would not converge and some subgroup analyses had to be done with less sophisticated statistical methods. The small number of studies also meant that important study characteristics, including type of reference standard used, method of decontamination, and schedule of microscopic examination, were not assessed. Second, although we specifically reviewed drug susceptibility testing by MODS and TLA and their agreement with reference standards, when these assays are used on direct specimens they are implicitly being used for detection
0·4
0·2
0 1·0
0·8
0·6
0·4
0·2
0
Specificity
Figure 7: Hierarchical summary receiver operating characteristic plot of the microscopic-observation drug susceptibility assay for detection of isoniazid resistance Open squares represent individual studies, with the size of the square proportionate to the size of the study. The summary point is a closed circle, representing sensitivity and specificity estimates pooled with a bivariate random-effects model. The hierarchical summary receiver operating characteristic curve is truncated outside of the area for which data exist.
Microscopic-observation drug susceptibility assay
Thin layer agar assay
Number of studies
Number of studies
Time (days)
Time (days)
Overall
6
9·9 (6·0–21·0; 4·1–15·8)
2
11·1 (11·0–11·2; 10·1–12·0)
Direct only
4
11·6 (6·0–21·0; 1·5–21·7)
2
11·1 (11·0–11·2; 10·1–12·0)
Indirect only
2
6·5 (6·0–7·0; 0·2–12·9)
0
··
Data are means of values in original studies (range; 95% CI). Turnaround times are defined as the number of days between specimen receipt or processing in the laboratory to the availability of results within the laboratory.
Table 6: Turnaround times
and identification of M tuberculosis as well. Our estimates of test accuracy do not account for the variability between studies due to differential isolation or inappropriate speciation from direct specimens; however, a thorough review and meta-analysis of the performance of MODS and TLA assays for diagnosis of active tuberculosis has been done to specifically address this issue.38 Third, no thorough evaluations of cost-effectiveness were available, and no studies focused specifically on outcomes in patients. The translation of accurate, timely, and useful results into outcomes important to patients is difficult to show and is dependent on many programmatic factors. 695
Articles
Contamination rate*
Comparison† Solid culture
Liquid culture
Microscopic-observation drug susceptibility assay
7·4% (0·4–8·1; 6·6–8·2; n=3)
12·9% (1·0–14·2; 11·9–13·9; n=3)
4·2% (4·0–4·4; 3·6–4·9; n=2)
Thin layer agar assay
1·4% (0–4·1; 0·4–3·5; n=2)
10·5% (··; 5·2–18·5; n=1)
2·1%(··; 0·3–7·4; n=1)
Data are means of values in original studies, weighted by the total number of specimens in each study (range; 95% CI; number of studies). *Defined as the proportion of specimens contaminated on first inoculation. †Contamination rate of the reference standard method of isolation done on the same specimens by the same laboratory.
Table 7: Contamination rates of the microscopic-observation drug susceptibility or thin layer agar assay with comparison to the reference standard method
Nevertheless, Nic Fhogartaigh and colleagues’ study33 provides good evidence that MODS culture could improve care, but this finding needs confirmation in studies with large population samples. Fourth, publication bias is a concern with all systematic reviews. Statistical and graphical approaches for publication bias (eg, funnel plots and regression asymmetry tests) are not recommended for diagnostic meta-analyses.12 In this context, we are unable to exclude the possibility of bias against publication of studies with poor performance of the MODS or TLA assay. Microcolony detection methods such as MODS and TLA assays are inexpensive, rapid alternatives for drug susceptibility testing for M tuberculosis. Along with the nitrate reductase assay, these non-commercial methods to detect resistance could fulfil a pressing need for tuberculosis diagnostics. Although extensive and stringent studies have been done to prove the accuracy of commercial liquid culture systems and rapid molecular detection of drug resistance both in research settings and under programmatic conditions, such large-scale assessments are unlikely to be funded for non-commercial laboratory techniques such as the MODS or TLA assay. Thus, policy makers and decision makers have to weigh the potential risks and benefits of implementation of diagnostic techniques that are less standardised than are commercial methods and have few data available on their performance in varied real-world settings. Despite these concerns, the need for rapid, simple, and affordable methods to detect drug-resistant tuberculosis is becoming increasingly urgent as rates of MDR and XDR tuberculosis rise steadily.39 The goal to strengthen laboratories and build the infrastructure and capacity needed to implement rapid, direct molecular detection and liquid culture technologies is important, but such advances will take time to implement. However, efforts are already underway to scale up the use of rapid line-probe assays in several high-burden countries. Tuberculosis control programmes considering implementation of a non-commercial technique to test drug susceptibility need to carefully assess the capabilities of their laboratories, timelines for anticipated growth, improvements in infrastructure and human resources, and the overall capacity and projected budgets of their programme to establish whether implementation of a non-commercial diagnostic technique, such as the MODS or TLA assay, is feasible, and assess whether a 696
more standardised diagnostic technique, such as commercial liquid culture systems and line-probe assays, could instead be implemented. In addition to line-probe assays that are already approved by WHO, a highly sensitive automated molecular assay (Xpert MTB/RIF, Cepheid, CA, USA) is now an option for detection of both M tuberculosis and rifampicin resistance within 2 h with minimal hands-on time.40 This assay promises to decentralise molecular diagnosis since it can potentially be used at the point of treatment in a microscopy centre or in a tuberculosis or HIV clinic.41 Several features of MODS and TLA assays need further research.42 First, direct inoculation with specimens from patients substantially reduces turnaround time but introduces the greatest risk of variability, threatening the principles of traditional proportion method drug susceptibility testing. Head-to-head studies of direct inoculation with specimens from patients versus indirect inoculation with isolates would help to quantify this variability. Direct inoculation might have to be reserved for smear-positive specimens (similar to molecular detection technologies), but no published data are available to support this recommendation. Second, MODS and TLA assays both use microcolony detection to identify positive cultures. Qualitative assessment of the specificity of microcolony detection is reported briefly by Caviedes and colleagues;22 however, rigorous assessment of the ability of technologists to differentiate M tuberculosis from non-tuberculous mycobacteria is lacking. The need for an additional species identification test would not only affect the complexity and turnaround time of these diagnostics, but also the biosafety level needed for laboratories using them. If species identification by microcolony detection is indeed highly accurate, sealed cultures prepared for the TLA or MODS assay would not need to be opened after initial inoculation with specimens from patients. Without the need to handle cultured isolates, drug susceptibility testing could be done with low biohazard risk. Until data are available to precisely define the specificity of microcolony cording as definitive identification of M tuberculosis, one of the two drug-free wells in the MODS assay could include para-nitrobenzoic acid, a specific inhibitor of M tuberculosis, to help distinguish growth of M tuberculosis from growth of nontuberculous mycobacteria (Moore D, London School of Hygiene and Tropical medicine, London, UK, personal communication). Additionally, research is underway in www.thelancet.com/infection Vol 10 October 2010
Articles
collaboration with the Foundation for Innovative New Diagnostics to develop a diagnostic test that is similar to the TLA assay but incorporates a simple colorimetric indicator to easily identify growth of M tuberculosis. Third, we have presented cost estimates from the reviewed reports, but these estimates are not a true representation of the costs of non-commercial techniques. The estimates are simple totals of supplies and reagents needed per specimen and therefore are very low, but the additional expenses associated with implementation of the assays could substantially increase costs. Costs associated with labour, equipment, infrastructure, training, and quality control and assurance are likely to differ greatly between microcolony methods and conventional techniques for isolation and drug susceptibility testing, and would vary with the setting of implementation and volume of specimens processed. The increased labour requirements of microcolony methods would make MODS or TLA assays less cost effective in developed countries with high labour costs (JM, unpublished data). Finally, for research on non-commercial techniques to be comparable across laboratories and settings, common standardised operating procedures should be followed whenever possible. Standardised operating procedures are available from the groups developing MODS and TLA assays and the developers supporting the implementation of the MODS assay have recommended procedures for quality assurance and laboratory accreditation. Researchers wanting to contribute to the development of these techniques are encouraged to refer to these documents and to report on their experiences in diverse and pragmatic settings to build a strong evidence base on which policy makers can base their decisions. Only by use of such globally collaborative strategies will non-commercial techniques gain the level of rigorous evidence provided by profit-driven commercial products. The data in this report were presented to a WHO expert group in September, 2009, to help guide policy decisions and recommendations on the use of diagnostic tests for rapid detection of drug resistance in M tuberculosis. WHO now recommends that selected non-commercial methods for drug susceptibility testing be used as an interim solution in resource-constrained settings, under clearly defined programmatic and operational conditions, in reference laboratories or those with sufficient culture capacity until capacity for genotypic or automated liquid culture drug susceptibility testing is developed.43 Specifically, WHO endorsed the MODS assay, as a direct or indirect test, for rapid screening of patients with suspected MDR tuberculosis. The WHO expert group agreed that evidence was insufficient to recommend the use of the TLA assay for rapid screening of these patients, but this assay is a promising diagnostic technique and further research is encouraged. WHO’s policy emphasises that time to detection of MDR tuberculosis might not be faster with www.thelancet.com/infection Vol 10 October 2010
indirect inoculation of these tests than with conventional methods of drug susceptibility testing, and in their present forms these non-commercial assays are unable to detect XDR tuberculosis.43 These limitations underscore the need for continuing development and evaluation of novel tuberculosis diagnostics. Contributors JM developed the study protocol, searched for published reports, selected the studies, did data extraction, analysis, and interpretation, and wrote the report. EL selected the studies and extracted the data, and contributed to data interpretation and writing of the report. DM and MP contributed to development of the study protocol, data interpretation, and writing of the report. Conflicts of interest MP was a consultant for the non-profit organisation Foundation for Innovative New Diagnostics at the time that this report was submitted for publication, and has become a consultant for the non-profit organisation the Bill & Melinda Gates Foundation since this report was submitted for publication; neither of these organisations had any involvement in this report. All other authors declare that they have no conflicts of interest. Acknowledgements This study was funded by the Stop TB Department of WHO as part of their assessment of evidence regarding non-commercial culture methods for rapid screening of patients at risk of drug-resistant tuberculosis. JM is supported by a Quebec Respiratory Health Training Program fellowship, DM is supported by a salary award from Fonds de la Recherche en Santé du Québec, and MP is supported by a New Investigator Award from the Canadian Institutes of Health Research; none of these agencies had any involvement in this report. We thank Karin Weyer (Stop TB Department, WHO, Geneva, Switzerland) for helpful input and feedback; and David Moore, Anandi Martin, Jaime Robledo, Susan Dorman, and Juan C Palomino for provision of unpublished or additional data. References 1 Wright A, Zignol M, Van Deun A, et al, for the Global Project on Anti-Tuberculosis Drug Resistance Surveillance. Epidemiology of antituberculosis drug resistance 2002–07: an updated analysis of the Global Project on Anti-Tuberculosis Drug Resistance Surveillance. Lancet 2009; 373: 1861–73. 2 WHO. Treatment of tuberculosis: guidelines for national programmes, 3rd edn. WHO/CDS/TB/2003.313. Geneva: World Health Organization, 2003. 3 Orenstein EW, Basu S, Shah NS, et al. Treatment outcomes among patients with multidrug-resistant tuberculosis: systematic review and meta-analysis. Lancet Infect Dis 2009; 9: 153–61. 4 Wells CD, Cegielski JP, Nelson LJ, et al. HIV infection and multidrug-resistant tuberculosis: the perfect storm. J Infect Dis 2007; 196 (suppl 1): S86–107. 5 Pai M, Ramsay A, O’Brien R. Evidence-based tuberculosis diagnosis. PLoS Med 2008; 5: e156. 6 Ling DI, Zwerling AA, Pai M. GenoType MTBDR assays for the diagnosis of multidrug-resistant tuberculosis: a meta-analysis. Eur Respir J 2008; 32: 1165–74. 7 Morgan M, Kalantri S, Flores L, Pai M. A commercial line probe assay for the rapid detection of rifampicin resistance in Mycobacterium tuberculosis: a systematic review and meta-analysis. BMC Infect Dis 2005; 5: 62. 8 WHO. Molecular line probe assays for rapid screening of patients at risk of multidrug resistant tuberculosis (MDR-TB). June 27, 2008. http://www.who.int/tb/features_archive/policy_statement.pdf (accessed Aug 21, 2010). 9 Coronel J, Roper H, Caviedes L, Moore D. MODS: a user guide (v12.1). Microscopic observation drug susceptibility assay. Lima, Peru: Universidad Peruana Cayetano Heredia, 2008. http://www. modsperu.org/MODS_user_guide.pdf (accessed Aug 21, 2010). 10 Martin A, Palomino JC. Procedure manual: thin layer agar (TLA) microcolony detection, version 3. Antwerp, Belgium: Institute of Tropical Medicine, Mycobacteriology Unit, 2009.
For standardised operating procedures for the MODS assay see http://www.modsperu.org and the TLA assay see http:// www.tbevidence.org/documents/ rescentre/sop/TLA.pdf, and for more on evidence-based tuberculosis diagnosis see http://www.tbevidence.org
697
Articles
11
12 13
14
15
16
17
18
19
20 21
22
23
24
25
26
27
698
Pai M, McCulloch M, Gorman JD, et al. Systematic reviews and meta-analyses: an illustrated, step-by-step guide. Natl Med J India 2004; 17: 86–95. Leeflang MM, Deeks JJ, Gatsonis C, Bossuyt PM. Systematic reviews of diagnostic test accuracy. Ann Intern Med 2008; 149: 889–97. Whiting P, Rutjes A, Reitsma J, Bossuyt P, Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003; 3: 25. Zamora J, Abraira V, Muriel A, Khan K, Coomarasamy A. Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol 2006; 6: 31. Pai M, McCulloch M, Enanoria W, Colford JM Jr. Systematic reviews of diagnostic test evaluations: what’s behind the scenes? ACP J Club 2004; 141: A11–13. Littenberg B, Moses LE. Estimating diagnostic accuracy from multiple conflicting reports: a new meta-analytic method. Med Decis Making 1993; 13: 313–21. Toft N, Nielsen SS. Summary receiver operating characteristics (SROC) and hierarchical SROC models for analysis of diagnostic test evaluations of antibody ELISAs for paratuberculosis. Prev Vet Med 2009; 92: 249–55. Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005; 58: 982–90. Harbord RM, Whiting P. metandi: meta-analysis of diagnostic accuracy using hierarchical logistic regression. In: Sterne JAC, ed. Meta-analysis in Stata: an updated collection from the Stata Journal. College Station, TX, USA: Stata Press, 2009: 181–99. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003; 327: 557–60. Moore DAJ, Mendoza D, Gilman RH, et al. Microscopic observation drug susceptibility assay, a rapid, reliable diagnostic test for multidrug-resistant tuberculosis suitable for use in resource-poor settings. J Clin Microbiol 2004; 42: 4432–37. Caviedes L, Lee TS, Gilman RH, et al. Rapid, efficient detection and drug susceptibility testing of Mycobacterium tuberculosis in sputum by microscopic observation of broth cultures. The Tuberculosis Working Group in Peru. J Clin Microbiol 2000; 38: 1203–08. Devasia RA, Blackman A, May C, et al. Fluoroquinolone resistance in Mycobacterium tuberculosis: an assessment of MGIT 960, MODS and nitrate reductase assay and fluoroquinolone cross-resistance. J Antimicrob Chemother 2009; 63: 1173–78. Ejigu GS, Woldeamanuel Y, Shah NS, Gebyehu M, Selassie A, Lemma E. Microscopic-observation drug susceptibility assay provides rapid and reliable identification of MDR-TB. Int J Tuberc Lung Dis 2008; 12: 332–37. Mello FC, Arias MS, Rosales S, et al. Clinical evaluation of the microscopic observation drug susceptibility assay for detection of Mycobacterium tuberculosis resistance to isoniazid or rifampin. J Clin Microbiol 2007; 45: 3387–89. Mengatto L, Chiani Y, Imaz MS. Evaluation of rapid alternative methods for drug susceptibility testing in clinical isolates of Mycobacterium tuberculosis. Mem Inst Oswaldo Cruz 2006; 101: 535–42. Moore DA, Evans CA, Gilman RH, et al. Microscopic-observation drug-susceptibility assay for the diagnosis of TB. N Engl J Med 2006; 355: 1539–50.
28
29
30
31
32
33
34
35
36 37
38
39
40
41
42
43
Park WG, Bishai WR, Chaisson RE, Dorman SE. Performance of the microscopic observation drug susceptibility assay in drug susceptibility testing for Mycobacterium tuberculosis. J Clin Microbiol 2002; 40: 4750–52. Shiferaw G, Woldeamanuel Y, Gebeyehu M, Girmachew F, Demessie D, Lemma E. Evaluation of microscopic observation drug susceptibility assay for detection of multidrug-resistant Mycobacterium tuberculosis. J Clin Microbiol 2007; 45: 1093–97. Martin A, Paasch F, Von Groll A, et al. Thin layer agar for detection of resistance to rifampicin, ofloxacin and kanamycin in Mycobacterium tuberculosis isolates. Int J Tuberc Lung Dis 2009; 13: 1–4. Robledo J, Mejia GI, Paniagua L, Martin A, Guzman A. Rapid detection of rifampicin and isoniazid resistance in Mycobacterium tuberculosis by the direct thin-layer agar method. Int J Tuberc Lung Dis 2008; 12: 1482–84. Schaberg T, Reichert B, Schulin T, Lode H, Mauch H. Rapid drug susceptibility testing of Mycobacterium tuberculosis using conventional solid media. Eur Respir J 1995; 8: 1688–93. Nic Fhogartaigh CJ, Vargas-Prada S, Huancare V, Lopez S, Rodriguez J, Moore DA. Physician-initiated courtesy MODS testing for TB and MDR-TB diagnosis and patient management. Int J Tuberc Lung Dis 2008; 12: 555–60. Moore DA, Caviedes L, Gilman RH, et al. Infrequent MODS TB culture cross-contamination in a high-burden resource-poor setting. Diagn Microbiol Infect Dis 2006; 56: 35–43. Gumbo T. New susceptibility breakpoints for first-line antituberculosis drugs based on antimicrobial pharmacokinetic/ pharmacodynamic science and population pharmacokinetic variability. Antimicrob Agents Chemother 2010; 54: 1484–91. Kim SJ. Drug-susceptibility testing in tuberculosis: methods and reliability of results. Eur Respir J 2005; 25: 564–69. Cruciani M, Scarparo C, Malena M, Bosco O, Serpelloni G, Mengoli C. Meta-analysis of BACTEC MGIT 960 and BACTEC 460 TB, with or without solid media, for detection of mycobacteria. J Clin Microbiol 2004; 42: 2321–25. Leung E, Minion J, Pai M, Menzies D. Using MODS and/or TLA techniques for active tuberculosis diagnosis: a systematic review and meta-analysis. American Thoracic Society International Conference; New Orleans, LA, USA; May 14–19, 2010. Abstr A1768. WHO. Global tuberculosis control 2009: epidemiology, strategy, financing. http://www.who.int/tb/publications/global_ report/2009/pdf/full_report.pdf (accessed Aug 21, 2010). Boehme CC, Nabeta P, Hillemann D, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med 2010; published online Sept 1. DOI:10.1056/NEJMoa0907847. Small PM, Pai M. Tuberculosis diagnosis—time for a game change. N Engl J Med 2010; published online Sept 1. DOI:10.1056/ NEJMe1008496. Minion J, Pai M. Expanding the role of the microscopic observation drug susceptibility assay in tuberculosis and HIV management. Clin Infect Dis 2010; 50: 997–99. WHO. Non-commercial culture and drug-susceptibility testing methods for screening of patients at risk of multi-drug resistant tuberculosis: policy statement (July, 2010). http://www.who.int/tb/ dots/laboratory/whopolicy_noncommercialculture_and_ dstmethods_july10.pdf (accessed Aug 24, 2010).
www.thelancet.com/infection Vol 10 October 2010
Review
Needle-free influenza vaccination Jean-Pierre Amorij, Wouter L J Hinrichs, Henderik W Frijlink, Jan C Wilschut, Anke Huckriede
Vaccination is the cornerstone of influenza control in epidemic and pandemic situations. Influenza vaccines are typically given by intramuscular injection. However, needle-free vaccinations could offer several distinct advantages over intramuscular injections: they are pain-free, easier to distribute, and easier to give to patients, and their use could reduce vaccination costs. Moreover, vaccine delivery via the respiratory tract, alimentary tract, or skin might elicit mucosal immune responses at the site of virus entry and better cellular immunity, thus improving effectiveness. Although various needle-free vaccination methods for influenza have shown preclinical promise, few have progressed to clinical trials—only live attenuated intranasal vaccines have received approval, and only in some countries. Further clinical investigation is needed to help realise the potential of needle-free vaccination for influenza.
Introduction Influenza is a major cause of morbidity, mortality, and economic loss that affects about 5–10% of the world’s population each year.1 Epidemic influenza strains can cause life-threatening disease in immunocompromised individuals. Pandemic influenza viruses have caused substantial morbidity and mortality worldwide, and are a constant public health threat. Vaccination is the main method of influenza prevention and is the most costeffective measure for control of the disease. Vaccination against epidemic influenza strains, usually with split or subunit vaccine formulations, is recommended in the European Union for people with chronic diseases, people older than 65 years of age, immunocompromised patients, residents of nursing homes, and health-care workers.2 Vaccination is also the main method of prophylaxis to protect against pandemic influenza.3 Prepandemic mockup vaccines based on influenza A H5N1 virus strains, formulated as adjuvanted split or whole inactivated virus (WIV) vaccines, received approval in Europe at the beginning of 2009.4 During the 2009 H1N1 influenza pandemic, these newly approved vaccine formulations and conventional H1N1 vaccines were used in accordance with national immunisation plans worldwide. Conventional needle-based intramuscular influenza vaccines stimulate the production of serum antibodies, which by transudation to the lungs protect the lower respiratory tract against influenza infection.5 However, protection against initial infection of the upper respiratory tract is poor because of an absence of antibody induction in the nasal mucosa. Other disadvantages of needle-based intramuscular vaccines are safety problems (ie, the risk of needle stick injuries and the risk of infection by reuse of needles), low acceptance among patients with needle phobia, and logistic challenges in mass vaccination programmes. Several of these problems can be avoided with the use of needle-free vaccination systems.6–12 Systems in development can be classified according to the route by which they are given, either via mucosal tissues (nasal, pulmonary, oral cavity, or gastrointestinal) or via the skin (intradermal or transcutaneous). Each of these routes has its own advantages and disadvantages related to ease of vaccination and the type of immune response elicited. www.thelancet.com/infection Vol 10 October 2010
In this Review, we discuss the development of needlefree influenza vaccines and give an overview of approaches that have entered clinical assessment, such as live-attenuated influenza virus (LAIV), WIV, and splitvirus and subunit vaccine formulations.
Development of non-parenteral influenza vaccines Need for improved vaccines WHO urges countries to introduce and broaden influenza vaccination programmes to achieve high vaccination coverage in at-risk populations.13 Major issues such as vaccine acceptance and cost have to be addressed before this can be achieved. Needle-free vaccination might reduce costs because, unlike needlebased vaccination, it does not require trained health-care personnel. Furthermore, people with a phobia of needles will be more likely to accept needle-free vaccination than they would injection-based vaccination. In a study by Sendi and colleagues,14 1552 (97%) of 1600 participants chose an intranasal vaccine spray when given the choice between the spray and an intramuscular injection. Vaccine acceptance would also benefit from improved effectiveness. Intramuscular vaccines, which induce only systemic IgG responses, protect from severe complications of influenza but not necessarily from early disease symptoms. In addition to serum IgG, vaccines given via mucosal tissues can also induce local IgA responses in the upper respiratory tract, which can neutralise the virus at the point of entry. IgA is more cross-reactive than IgG and can provide some protection against drift and even shift virus variants.15–19 Needle-free vaccination, especially if there is no requirement for trained health-care personnel, would be ideal for mass vaccination campaigns. Logistic problems associated with supply and disposal of syringes and needles and safety risks associated with injections would be reduced.20 Compared with injection, most methods of non-parenteral vaccination could improve the speed of distribution.21 Vaccine logistics can be further simplified by the use of dry vaccine formulations, which, when packaged, can be smaller and lighter than liquid formulations, and might not require a cold chain for storage and distribution.22
Lancet Infect Dis 2010; 10: 699–711 Department of Pharmaceutical Technology and Biopharmacy (J-P Amorij PhD, W L J Hinrichs PhD, Prof H W Frijlink PhD) and Department of Medical Microbiology, Molecular Virology Section, University Medical Centre GrÖningen (Prof J C Wilschut PhD, A Huckriede PhD), University of GrÖningen, GrÖningen, Netherlands; and Research and Development Unit, Netherlands Vaccine Institute, Bilthoven, Netherlands (J-P Amorij) Correspondence to: Dr Anke Huckriede, Department of Medical Microbiology, Molecular Virology Section, PO Box 30001, EB88 9700, RB, GrÖningen, Netherlands
[email protected]
699
Review
Adjuvant system
Vaccine type
Animal model
Immune response detected
Protection (challenge)
Clinically assessed
Joseph47
Ceramide carbamoyl-spermine / cholesterol lipids
Subunit
Mice
HI, IgG(2a), sIgA, interferon γ/ proliferation
Homologous
No
Ko48
α-galactosylceramide
Subunit
Mice
IgG(2a), sIgA, interferon γ/ proliferation
Homologous
No
Youn49
α-galactosylceramide
WIV
Mice
IgG(2a), sIgA, interferon γ, proliferation
Homologous
No
Shim50
Poly[di(sodium carboxylatophenoxy)phosphazene] (PCPP)
WIV
Mice
IgG, sIgA, interferon γ
Homologous
No
Guy51
Cationic cholesterol (DC-Chol)
Split
Mice
HI, IgG(2a), sIgA
··
No
Nishino52
Surfacten (modified bovine pulmonary surfactant)
Split
Minipigs
HI, sIgA, cross-neutralising IgA ··
No
Bracci53,54
Type I interferon (interferon-α)
Subunit
Mice
IgG(2a), sIgA
Homologous
Yes
Arulanandam55
Interleukin 12
Subunit
Mice
IgG(2a), sIgA, interferon γ
Homologous
No
Joseph56
Oligodeoxynucleotides-vaccine encapsulated in large multilamellar liposomes
Subunit
Mice
HI, IgG(2a), sIgA, proliferation
Homologous
NCT00197301
Ichinohe19,57
Poly I:Poly C12U, dsRNA
WIV
Mice
IgG, sIgA, interferon γ, proliferation
Heterologous
No
Song58
Mistletoe lectin C
WIV
Mice
IgG(2a), sIgA, interferon γ, proliferation
Homologous Heterologous (cholera toxin-like)
No
Hasegawa18
Surf clam (microparticles)
Split
Mice
IgG sIgA
Cross-protection against variant No influenza viruses.
Zanvit59
Delipidated bacterium Bacillus firmus
WIV
Mice
IgG sIgA
Homologous and heterologous
No
Plante60
Proteosomes from Neisseria meningitidis
Subunit
Mice
HI, IgG, sIgA, interferon γ
Homologous
No
ISCOMs saponin Quil-A+cholesterol Lovgren,61 Sjolander,62 Coulter,63 Helgeby,64 Hu65
Split
Mice
HI, IgG(2a), sIgA, CTL, interferon γ
Homologous
No
Coulter,63 Scheerlinck66
ISCOMATRIX (empty ISCOM admixed with vaccine)
Split
Mice Sheep
HI, sIgA
··
No
Read,67 Bacon,68 Amidi,69 Illum70
Chitosan or chitosan derivatives
Subunit/ Mice split/WIV
HI, IgG, sIgA, interferon γ
Homologous
Yes
Huang,71 Garmise72
Intranasal powder formulation co-formulated with chitosan
WIV
Rats
HI, IgG(2a), sIgA
··
No
1,2-dipalmitoylphosphatidyl-choline, distearoylphosphatidyl-choline and tyloxapol (spray-dried powder)
WIV or split
Rats
HI, IgG, no sIgA
··
No
mCTA-cholera toxin B
WIV
Mice
IgG(2a), sIgA, interferon γ, CD4 and CD8
Heterologous
No
Nasal
Pulmonary Smith73
Sublingual Song74 Oral McCluskie75
Cpg, non-CpG oligodeoxynucleotides
Subunit
Mice
IgG(2a)/sIgA
··
No
Katz76
LT
Split
Mice
HI, IgG, sIgA, interferon γ, CTL
Homologous
No
Amorij34
LT, gastric vs colonically
Subunit
Mice
HI, IgG(2a), sIgA, interferon γ ·· Intracolonic: Th1-skewing Intragastric: increased T-helper responses; no Th1-skewing
No
Barackman,77 Lu78
LT derivatives
Split
Mice
HI, IgG, sIgA, interferon γ
··
No
Conacher,79 Kunzi80
Deoxcycholate (in bilosomes)
Subunit
Mice/ferret
HI, IgG2a, sIgA, interferon γ
Homologous (ferret)
No
Intradermal Skountzou81
Cholera toxin transcutaneous immunisation
WIV
Mice
IgG, sIgA, interferon γ
Homologous
No
Chen82,83
Cholera toxin B LT-mutants QS-21 adjuvant epidermal powder immunisation
Split
Mice
HI, IgG, sIgA, antibody affinity
Homologous
No
Chen83
QS-21 adjuvant epidermal powder immunisation
Split
Macaques
HI
Homologous Variable virus titres No conclusion
No
WIV=whole inactivated virus. HI=haemagglutination inhibition. LT=heat-labile enterotoxin. CTL=cytotoxic T lymphocytes. ISCOM=immunostimulating complex.
Table 1: Preclinical assessments of adjuvants for needle-free influenza vaccination
700
www.thelancet.com/infection Vol 10 October 2010
Review
Advances in immunological knowledge The identification of pathogen-associated molecular patterns that stimulate and shape the adaptive immune response via interaction with host pattern recognition receptors, has provided a rational basis for the development of new adjuvants.23,24 Moreover, the mechanisms by which traditional adjuvants work are being elucidated.25,26 Understanding of the initiation of adaptive immune responses at mucosal sites and in the skin, antigen processing and presentation, and B-cell activation has also improved.27–29 Topological issues in immunity have received attention and have enabled research into the role of the site of antigen delivery in eliciting an immune response.30–34 These general insights can now be implemented and can guide the development of new vaccines.
Technical progress Progress has been made in the development of nebulisers and inhalers for vaccination via the respiratory tract, and microneedle arrays, patches, and jet injectors for dermal vaccine delivery.35 Dry powder vaccine formulations have high stability and enable new or improved vaccine targeting strategies. The powders are usually produced by freeze-drying in the presence of stabilising sugars.22 Advanced particle engineering techniques allow the generation of powders with predefined physicochemical characteristics (eg, particle size, density, aerodynamic behaviour), which can be tailored to specific methods of vaccination. Moreover, special formulations, such as mucoadhesive compounds and gel-forming powders, can improve the local availability of the delivered vaccine and thereby vaccine efficacy. Formulation of vaccine powder as coated tablets enables delivery of oral vaccination to desired sites in the gastrointestinal tract.36
Limited effectiveness A major drawback of needle-free vaccines is that most have low effectiveness. Dermal vaccination with doses similar to or even lower than those used in parenteral vaccination might be feasible with use of proper delivery methods or suitable adjuvants. However, mucosal vaccination, unless done with a live attenuated vaccine, requires large amounts of antigen and usually several doses to achieve the required haemagglutination inhibition titres in serum. For immunisation programmes against epidemic strains, and especially against pandemic strains, mucosal vaccination with inactivated vaccines is therefore impractical because of financial and logistic reasons.
The absence of approved adjuvants Development of adjuvants for clinical use is a very slow process because of the low predictive value of data from studies in animals. In Europe, the only adjuvants licensed for use in clinical vaccines are aluminium hydroxide, AS04 (aluminium hydroxide with monophosphoryllipid A), the oil-in-water emulsions MF59 and AS03, and www.thelancet.com/infection Vol 10 October 2010
influenza virosomes. In the USA, only aluminium hydroxide and AS04 have obtained approval.37 For intranasal vaccination these adjuvants seem to be ineffective (alum, MF59, influenza virosomes)38–41 or have not been clinically tested (AS03, AS04). An Escherichia coli heat-labile enterotoxin formulated in a nasal influenza vaccine, is one of the few adjuvants that reached the market, but it had to be withdrawn because of risk of Bell’s palsy as an adverse outcome of vaccination.41–44 The clinical assessment of more suitable, detoxified heat-labile enterotoxin variants is underway.45,46 Compared with intranasal vaccination, there has been little assessment of possible adjuvants for transcutaneous vaccination. Many adjuvants for needle-free influenza vaccination are in preclinical assessment (table 1). The process of approving new adjuvants is slow, mainly because of regulatory concerns about safety and universality (ie, whether adjuvant activity and side-effects vary with the type of antigen used and the site of vaccine delivery).84 Consequently, the final combination of adjuvant, antigen, and route of vaccination has to be licensed by a time-consuming, case-by-case process. This cautious approach is warranted because vaccines are given to a large and predominantly healthy population, and toxic adjuvants might put patients at risk.44,85,86 More information about the biological properties of formulations of antigens and adjuvants is urgently needed for successful development and faster licensing of new adjuvants, as are quality control tests, and in-vitro or in-vivo models that predict safety and efficacy.
Restricted predictive value of studies in animals Development of effective and safe needle-free influenza vaccines, whether adjuvanted or not, requires the assessment of vaccine effects and protection in animals. The main species used for the development of influenza vaccines are mice, ferrets, and, to a lesser extent, macaques. Each of these animal species has advantages and disadvantages related to ease of handling, costs, and the availability of reagents for characterisation of their immune response.87,88 However, influenza infection, pathogenicity, and symptomatology in any of these animals only partly overlaps with that in human beings, which limits the predictive value of vaccination results from studies in animals. Nevertheless, results from studies in animals determine whether or not an approach will be tested in clinical trials.88 Ultimately, estimation of vaccine effectiveness in human beings is based on a weight-of-evidence approach87 that takes into account the evidence of protection in one or two animal species and compliance with relevant correlates of protection in the clinical situation.
Poorly defined correlates of protection Haemagglutination inhibition serum titres are currently the only validated correlate of protection for influenza vaccines. On the basis of a titre of 40, which is regarded 701
Review
Seroconversion (HI titre <40 to ≥40 or four-fold rise) Mean fold increase in geometric mean HI titre Seroprotection (HI titre ≥40)
Age 18–60 years old
Age >60 years old
>40%
>30%
>2·5 >70%
>2 >60%
HI=haemagglutination inhibition.
Table 2: European Medicines Agency criteria for influenza vaccines
as protective, the European Medicines Agency formulated criteria that influenza vaccines have to fulfil for registration (table 2).89 Consequently, new influenza vaccines have to induce adequate haemagglutination inhibition serum titres even though other immune responses might be just as or even more relevant for protection from infection. Secretory IgA, rather than nasal or serum IgG or IgM, is a major mediator of protection against nasal challenge with influenza virus.90–92 Accordingly, sufficient protection might be achieved by mucosal IgA induced by oral, pulmonary, or nasal vaccination even in the absence of protective haemagglutination inhibition titres in serum. This issue needs to be urgently addressed in clinical trials. Cellular immune responses can also contribute to protection although they cannot provide sterilising immunity.93 In elderly people, T-cell responses seem to better predict protection from disease than do antibody titres.94 Intranasal vaccination with LAIV and dermal vaccination with WIV induce T cells but are less effective than intramuscularly injected vaccines in eliciting haemagglutination inhibition titres.95,96 Before criteria for these new correlates of protection can be defined, standardised methods have to be developed to reliably sample mucosal tissue and measure T-cell and local antibody responses.97
Mucosal immunisation Because of their large size and immunological competence, mucosal tissues are attractive target sites for vaccination. Moreover, mucosal vaccination can elicit local immune responses, which can protect against infection at the point of virus entry. Because mucosal surfaces are generally exposed to many environmental antigens, tolerance mechanisms prevent an over-reaction of the immune system. For successful vaccination these tolerance mechanisms have to be overcome, which requires the use of strong mucosal adjuvants unless vaccination is done with live attenuated viruses.98
Intranasal antigen delivery So far, vaccination via the intranasal route, which targets the nasal-associated lymphoid tissue, is the only approach for mucosal vaccination that has been successfully used for commercial influenza vaccination. At present, all intranasal influenza vaccines on the market are LAIV vaccines. The attenuated viruses have six RNA segments from a cold-adapted virus strain that encode the viral polymerases, nucleoprotein, the matrix proteins, and 702
non-structural proteins, and two segments from circulating strains that encode haemagglutinin and neuraminidase.99,100 In Russia, LAIVs have been routinely used for many years. In 2003, an LAIV vaccine (FluMist) was approved in the USA for annual influenza vaccination of people aged 2–49 years. The vaccine is given as large droplet aerosol that deposits in the nasopharynx.99,100 By contrast with parenteral vaccines, which induce only systemic immune responses, intranasally delivered LAIV vaccines induce both systemic and broad mucosal immune responses.100 Although inactivated intramuscular vaccines elicit higher haemagglutination inhibition titres than do LAIV vaccines,101 they are both similarly effective in preventing influenza illness from homologous virus infections.90,102 LAIV vaccines can also provide immunity against heterologous virus strains, possibly mediated by mucosal IgA or cytotoxic T lymphocytes.100,103 LAIVs are very effective in priming immune responses in young children and adults.101,104 However, in elderly people, one intranasal LAIV vaccination is not more effective than an intramuscular vaccination.101 The use of inactivated vaccines for intranasal vaccination circumvents any safety concerns associated with LAIVs and has been explored in several clinical studies. Greenbaum and colleagues105 noted that nasal vaccination with 20 μg or 40 μg WIV resulted in a four-fold increase in haemagglutination inhibition titres in 30–40% of the individuals in the study. However, such large increases of haemagglutination inhibition titres were mostly recorded in individuals with low titres before vaccination. In 31–44% of people who received vaccines a local antibody response was recorded.105,106 Samdal and co-workers107 noted protective haemagglutination inhibition titres in 80% of volunteers after four intranasal doses of WIV (total dose 84 μg of haemagglutinin). In elderly people, two nasal vaccinations with WIV resulted in an increase in mucosal antibody titres by up to three times, but only small serum response rates.108 When compared with WIV vaccines, subunit and split influenza vaccines seem to be poorly immunogenic in clinical trials and will require the use of special delivery systems or mucosal adjuvants to be effective.40,109 Accordingly, recent studies have focused on adjuvanted or specially formulated vaccines. The most potent mucosal adjuvants are bacterial enterotoxins such as E coli heat-labile enterotoxin and cholera toxin. Use of these adjuvants in human beings is, however, hampered by their toxic properties. Detoxified variants of cholera toxin and heat-labile enterotoxin retain their adjuvant properties and improve the immunogenicity of influenza vaccine when given intranasally to animals.45,110–112 Stephenson and colleagues46 showed induction of mucosal immune responses but not serum IgG titres in human beings by use of a new nasal influenza vaccine formulation that consisted of a subunit vaccine, a detoxified heat-labile enterotoxin, and a nano-sized biovector (a positively charged polysaccharide core enclosed by a phospholipid-cholesterol double layer). In view of the www.thelancet.com/infection Vol 10 October 2010
Review
Pulmonary antigen delivery Lungs are highly vascularised, have a large absorptive surface area because of the alveoli structure,116,125 and contain bronchoalveolar lymphoid tissue.126 Furthermore, local antigen-presenting cells, such as different types of macrophages and dendritic cells, are ideally located for antigen sampling and subsequent presentation to www.thelancet.com/infection Vol 10 October 2010
A Syringe
Housing
Rupture film
Rupture film
Capsule
B 10 9 8 7 Frequency (%)
adverse outcomes seen with vaccines adjuvanted with the wild-type E coli heat-labile enterotoxin, vaccines containing endotoxin adjuvants might not receive approval easily. Several experimental adjuvants for intranasal influenza vaccination have been assessed in clinical trials. MF59 is inactive as an intranasal adjuvant in human beings, despite its strong adjuvant activity when injected intramuscularly and its mucosal adjuvant function in mice.40,113 Two doses of lipid or polysaccharide carrier molecules with split influenza vaccine stimulated only slight serum antibody responses but substantial increases in nasal antibody responses in healthy adults.114 Two nasal vaccinations that combined chitosan with inactivated influenza vaccine induced sufficient seroprotection and seroconversion rates, but low haemagglutination inhibition titres.67,70 Chitosan-like molecules have mucoadhesive properties, which increase the time that the vaccine antigen remains in the patient’s nose and improves mucosal barrier penetration.115 Johnson and colleagues116 have investigated intranasal vaccination with type 1 interferon in combination with WIV in a clinical trial, but antibody titres in sera and nose swabs were lower than or similar to those achieved by intranasal vaccination with non-adjuvanted WIV. Various other approaches for nasal vaccination against influenza have been assessed in preclinical experiments (table 1). In addition to liquid formulations, powders have also been developed for nasal delivery of influenza vaccines and have shown promising results in animals.71,117–120 Some powder formulations remain in the nasal cavity for longer periods than do liquid formulations, which might translate into higher bioavailability and immune responses.121,122 However, clinical assessment of intranasal influenza vaccine powders has not started. Because there is a non-ciliated area in the anterior part of the nasal cavity and a ciliated area in the posterior parts, the site of vaccine deposition is important when considering mucociliary clearance of vaccine from the nose.31 The site is determined by factors associated with the delivery device and formulation, such as the velocity at which an aerosol is delivered and the size of the droplets or particles in the formulation (figure 1).32 For aerosols or particles larger than 50 μm, intranasal delivery is highly reproducible and independent of the vaccine recipient’s control of breathing, because the site of deposition is governed by inertial impaction (heavy aerosol particles collide with the nasal mucosa rather than follow the streamline direction of the inhaled air).123 A range of devices has been developed for intranasal delivery.71,117,124
6 5 4 3 2 1 0 1
10
100
1000
Particle size (μm)
Figure 1: Intranasal antigen delivery Device for intranasal delivery of powder (A). The vaccine is held in a capsule sealed with rupturable film. Particle size distribution of freeze-dried whole inactivated virus for intranasal vaccination (B). Adapted from reference 71 with permission from Elsevier.
T cells.127–129 Additionally, pulmonary vaccination has the advantage of inducing both systemic and local immunity (IgA and IgG) in the respiratory tract.130 Pulmonary delivery of influenza vaccine suspensions was investigated in the 1960s and 70s in several clinical studies. Kasel and colleagues131–133 assessed aerosol immunisation with a classic inactivated influenza vaccine. Vaccine droplets ranging in size from 1 μm to 100 μm were sprayed into patients’ posterior oropharynx during rapid, deep inhalation and then sprayed into their noses. After two immunisations, substantial levels of cross-reactive mucosal antibodies and satisfactory protection were achieved.131,133–136 In several studies136–138 the protective efficacy of inhaled influenza vaccines was similar to that of an intradermal vaccine. Moreover, even in a season when the vaccine strain did not match with the circulating influenza strain, pulmonary immunisation resulted in a protection rate of 47% (compared with 60% protection against infection with the homologous strain observed in a previous season).137 Despite these successful clinical trials, no pulmonary influenza vaccines are commercially available, possibly because of the absence of effective inhaler systems and insufficient data on the safety of pulmonary antigen delivery. More recently, the pulmonary route for vaccination against influenza has been explored again in various preclinical studies. These studies show that pulmonary immunisation is more effective than intranasal 703
Review
immunisation and can be done with a small amount of antigen in combination with a strong adjuvant, especially when targeted to the deep lungs.30,73,139 Stable dry-powder vaccine formulations for pulmonary delivery of influenza vaccine might have several advantages. Dry-powder inhalation is more reproducible and efficient than is vaccination with liquid aerosols.35 Furthermore, in a dry-powder vaccine, particle properties can be controlled to ensure optimum inhalation and accurate targeting to the desired lung areas. Recent studies in animals, involving, among other vaccines, an inulin-stabilised, freeze-dried influenza subunit vaccine, showed promising results.73,140 No data are available on the efficacy of influenza vaccine powder inhalation in human beings. Inhalers that can guarantee substantial and reproducible deep lung deposition of vaccine will definitely contribute to the development of pulmonary vaccination. Available inhalation devices for liquid vaccines include nebulisers, medical metered-dose inhalers, and soft-mist inhalers. Of these, soft-mist inhalers might be the most suitable because they produce low-velocity aerosols thus preventing the high shear forces characteristic of nebulisers, and do not require the propellants needed for metered-dose inhalers that might damage vaccines.35,141 However, the denaturing effects of aerosol generation on the vaccine compound, in particular the effect of large air-liquid interfaces, should be assessed before clinical trials are done. Dry-powder inhalers combine the advantages of stable vaccine formulations with simple and fast delivery and, potentially, high lung deposition. A prototype dry-powder inhaler is the single-use disposable inhaler Twincer (University of Groningen, Netherlands; figure 2), an inexpensive device with good
dispersion of powder doses and good moisture protection of the powder formulation.142 The main safety concern for pulmonary influenza vaccination is exacerbation of respiratory diseases, such as chronic obstructive pulmonary disease, allergic asthma, or pneumonia.143 Meyer and colleagues144 showed that pulmonary vaccination with pneumococcal polysaccharide is safe in patients with chronic obstructive pulmonary disease and that it induces rapid serum antibody responses. Minne and co-workers145 showed that in asthmatic (ovalbumin-sensitised) mice, pulmonary vaccination with seasonal influenza vaccine does not lead to asthma exacerbation. However, more progress has to be made in the safety assessment of pulmonary vaccination.
Antigen delivery to the oral cavity Delivery to the oral cavity via the sublingual, buccal, or gingival route provides excellent accessibility and avoids the potential degradation of vaccine compounds that occurs with delivery via the gastrointestinal tract.146 The oral cavity contains intraepithelial and submucosal immune cells and lymphoid tissues (palatine, lingual tonsils, and adenoids), which, like the nasopharyngeal tonsils, are part of Waldeyer’s ring.147 However, the immunological processes within the oral cavity and, in particular, the importance of the tonsils in antibody induction are poorly understood.148–150 Only one clinical study has investigated oral cavity immunisation against influenza. In this phase 1–2 trial, healthy adult volunteers were immunised four times at 1-week intervals with WIV (360 μg hemagglutinin in total), by spraying 100 μL vaccine suspension into their mouths.151 No substantial increase in salivary or nasal IgA antibodies was recorded, but serum haemagglutination inhibition antibodies were induced in 56 of 75 volunteers after only two doses. Whether this response was because of antigen deposited in the oral cavity or by inhaled antigen needs to be investigated. Sublingual antigen delivery has received much attention for use in allergen-specific immunotherapy.152 Clinical trials with sublingual influenza vaccines have not been done. In mice, sublingual vaccination with live influenza virus and adjuvanted WIV was shown to be safe and to induce substantial immune responses.74,153
Gastrointestinal antigen delivery
Figure 2: The Twincer (University of GrÖningen, Netherlands), A disposable dry powder inhaler that consists of three specifically formed plastic plates. A blister containing the vaccine powder is located between the plates and can be opened by removal of a pull-off strip.
704
Oral vaccinations are ideal for immunisation programmes because they are easy to give and well accepted by most of the population. Orally delivered vaccines target the gutassociated lymphoid tissue in the gastrointestinal tract, which consists of Peyer’s patches, the appendix, solitary lymphoid nodules, and isolated lymphoid follicles. Few vaccines, all of which are live attenuated vaccines, are given orally and oral vaccination with inactivated vaccines seems to be very difficult because of vaccine instability in the acidic and proteinase-rich environment of the gastrointestinal tract.7 www.thelancet.com/infection Vol 10 October 2010
Review
In a study by Avtushenko and co-workers,154 volunteers who received either one or two oral vaccinations with split influenza vaccine (140 μg haemagglutinin), which was formulated as an emulsion, only sporadically developed serum antibodies. After one vaccination, levels of secretory IgA in nasal secretions and saliva were increased. However, after two vaccinations this response was no longer seen, possibly because of oral tolerance caused by repeated ingestion of substantial doses of antigen. The researchers concluded that the dose of antigen needs to be optimised to ensure a sufficient immune response. Other clinical studies used high doses of antigen to ensure sufficient absorption to M cells.154–156 However, none of these vaccinations resulted in detectable IgG responses, whereas IgA titres in both saliva and nasal lavage fluids were substantially increased. Whether these IgA antibodies alone could provide adequate protection against influenza infection is not known. Current preclinical research on liquid oral influenza vaccination mainly focuses on the use of adjuvants and complex vaccine formulations to achieve satisfactory serum antibody titres. Moldoveanu and colleagues155 vaccinated healthy adults orally with split influenza vaccine air-dried with D-xylose and formulated as enteric-coated gelatin capsules (ten doses, 150 μg haemagglutinin per strain). Serum antibodies of all isotypes were only slightly increased at 21 days after ingestion of the last vaccine dose. However, in saliva and nasal lavage, antigen-specific IgA and IgG responses were detected. Similarly, oral vaccination of volunteers with enteric-coated alum-absorbed WIV (six doses, 24 μg haemagglutinin in total) did not induce serum antibody titres, but resulted in a slow but substantial rise of IgA-specific antibodies in tears, saliva, and nasal secretions.157 Data from studies in animals indicate that the site of vaccine release within the alimentary tract might be of major importance and imply that the colon could be preferable to the upper gastrointestinal tract for vaccination.33,34 However, no clinical studies have been done to investigate this issue. To reliably target specific sites in the gastrointestinal tract, tablets or capsules need to be developed that contain dried vaccine and have special coatings that allow antigen release at the desired location of the gastrointestinal tract.36,158 Despite many years of research into oral influenza vaccination there has not been convincing data. More information is needed about the mechanisms of oral tolerance and how they can be overcome, and also on the immunological properties of different parts of the alimentary tract. With this information, new adjuvants and vaccine formulations with targeted antigen release can be developed that might eventually enable influenza vaccination via the oral route. Alternatively, live vaccine vectors, either bacterial or viral, that encode relevant influenza proteins might contribute to a successful oral vaccination strategy. www.thelancet.com/infection Vol 10 October 2010
Immunisation via the skin The skin is an easily accessible and highly immunocompetent organ and therefore very attractive for immunisations.7 The skin is divided into the stratum corneum, which consists of cornified keratinocytes, the epidermis, which has live keratinocytes and a large number of antigen-presenting Langerhans cells, and the dermis, which also contains antigen-presenting cells and supplies the epidermis through, among other routes, blood vessels and the lymphatics.159,160 The intact stratum corneum is an effective barrier that prevents penetration of foreign molecules larger than 500 Da. However, once the stratum corneum has been penetrated, antigens as large as 1×10⁶ Da are capable of reaching antigen presenting cells in the epidermis and dermis.161 Accordingly, research efforts have focused on the development of devices suitable for delivery of vaccines through the stratum corneum.
Delivery of liquid vaccine formulations Traditional needles have been used for intradermal delivery of various vaccines in several clinical trials.21,80,162–172 In general, these trials show that a low-dose intradermal vaccine (two-thirds to one-fifth of the normal dose) is as effective as the conventional full dose intramuscular vaccine in terms of seroconversion, fold increase in geometric mean haemagglutination inhibition titre and seroprotection. The few studies that compare intradermal with intramuscular injection at the same dose do not suggest that intradermal is any more efficient (in terms of dose-sparing) than intramuscular. Irrespective of dose-sparing potential, the technique for intradermal delivery with traditional needles is technically demanding and is therefore not ideal for routine vaccinations.173 Intradermal vaccination can be done with specially designed needles that allow a controlled depth of skin penetration174 and that are suitable for intradermal vaccination with influenza vaccines.175 One such device, the BD Soluvia (BD Medical Pharmaceutical Systems, Franklin Lakes, NJ, USA), proved especially promising in phase 2 and phase 3 clinical trials.165,176,177 In Europe, an intradermal influenza vaccine with this device has been approved by the European Medicines Agency.178 Another approach for intradermal vaccine delivery makes use of arrays of pointed microneedles (25–50 μm) that can penetrate the stratum corneum. These arrays are designed to target Langerhans cells in the epidermis, but do not reach the nerves in the underlying tissue and therefore elicit little sensation and no pain.179,180 Because of their ease of use, microneedle arrays can be given by people with little medical training.8 In a recent clinical trial, individuals who received doses of 3 μg or 6 μg haemagglutinin per influenza strain via a four-needle array, the MicronJet device (NanoPass Technologies Ltd, Rehovot, Israel), developed similar haemagglutinin inhibition titres as did those who received 15 μg by conventional intramuscular injection.181 705
Review
Actuation button Helium microcylinder
Nozzle
patch-assisted immunisation with a standard dose of influenza subunit vaccine induced not only CD4 but also CD8 cellular responses. However, neutralising antibody titres were recorded in only two of seven vaccine recipients.188 Further study is being done in animals to develop more effective immunisation patches.189
Delivery of dry vaccine formulations
Vaccine cassette Silencer
Figure 3: PowderJect ND5.2, a prototype single-use delivery device for epidermal powder immunisation Vaccine particles entrapped between polycarbonate membranes are propelled via the nozzle by release of helium from a microcylinder. Adapted from reference 192 with permission from Elsevier.
A
B
500 μm
100 μm
Figure 4: Microneedle array uniformly coated with a model drug (riboflavin) Imaging by brightfield microscopy shows coating of microneedle shafts of an array of 50 microneedles (A) and a single microneedle (B). Adapted from reference 197 with permission from Springer.
Jet injectors are an alternative option to deliver vaccine to epidermal tissue (and also to deeper subcutaneous or intramuscular tissue). This method is especially suitable for mass vaccination.182,183 Older, multiuse nozzle jet injectors were associated with rare inadvertent transmission of blood-borne infections. However, new generation jet injectors, like Biojector 2000 (Bioject Medical Technologies, Tualatin, OR, USA), Pharmajet (Pharmajet, Golden, CO, USA), and Injex (Equidyne Systems, Fullerton, CA, USA) use disposable cartridges and thus avoid the risk of cross-contamination. Immunisation with these devices has the same efficacy as the traditional intramuscularly injected influenza vaccine.184 Topical immunisation via the skin can also be done by use of patches.11,185,186 The stratum corneum can be penetrated by means of hydration, mechanical disruption, electroporation, or the use of carrier systems; each of these techniques can be used alone or in combination.8 Hydration results in swelling of keratinocytes and pooling of fluid in the intracellular spaces, which allows antigens to pass the stratum corneum more easily.187 In a small phase 1 clinical trial, skin surface stripping before 706
A jet injector for powders, the PowderJect (PowderJect Pharmaceuticals, Oxford, UK; figure 3), accelerates powder particles to such a speed that they can perforate the stratum corneum and reach the epidermis.190,191 The physical properties of the powder, especially the size and density of the particles, are of high importance to ensure delivery to the epidermis. Production of vaccine powders with the right physical properties is challenging.22 In a phase 1 trial, a powdered trivalent subunit influenza vaccine (15 μg per strain) was safe and elicited similar rates of seroconversion (75–100%) and geometric mean titres as standard intramuscular vaccines.192 However, clinical trials on a larger scale are needed to prove the effectiveness and safety of dry vaccine formulations. In animals, epidermal powder immunisations elicit mucosal immune responses;83,193 their potential to do so in human beings should be assessed. Studies in animals show that the inclusion of adjuvants might improve the efficacy of vaccination with PowderJect.82,83 Microprojection arrays have also been developed for delivery of solid antigens, including influenza WIV.194–196 A simple controllable method can be used to coat microneedles197 with influenza vaccines for delivery into the skin (figure 4). Kim and co-workers196 showed that microneedles coated with trehalose-stabilised WIV induced longer immunological memory in mice than did conventional vaccination.196 Also, dissolvable microneedles that consist of a dried hydrogel are a safe and painless alternative to hypodermic needle injection for proteins and vaccines.198
Conclusion Influenza vaccines that can be given via the respiratory or the alimentary tract, or that can be delivered through the skin without use of needles, are alternatives to traditional, intramuscularly injected vaccines. Mucosal vaccine delivery can induce local immune responses at the point of virus entry, but, if not done with live attenuated virus, requires large amounts of antigen and several vaccinations. Immunisation via the skin might have the potential to induce robust systemic immune responses with just one low dose of influenza vaccine. For both mucosal and transcutaneous delivery, the use of dry vaccine powders seems especially promising. So far, the only commercially available needle-free influenza vaccines are intranasal vaccines with live attenuated virus. For other needle-free vaccination strategies, additional clinical research is needed before any decision can be made on their usefulness in human www.thelancet.com/infection Vol 10 October 2010
Review
Search strategy and selection criteria We searched Medline, Current Contents, and Scopus with the search terms “vaccine”, “influenza”, “delivery or administration”, “needle-free”, “non-parenteral”, “nasal”, “oral”, “rectal”, “pulmonary”, “sublingual”, “buccal”, “mucosal”, “dermal”, “intradermal”, “cutaneous”, and “transcutaneous”. From the retrieved papers we selected those describing alternative routes for influenza vaccine delivery. Preference was given to clinical studies. Studies on alternative routes of influenza vaccination based on recombinant proteins, recombinant viruses, or plasmid DNA were excluded.
17
18
19
20
21
22
beings. Re-evaluation of old clinical studies, the use of smart delivery technologies, site-specific vaccine delivery, introduction of new criteria for vaccine immunogenicity, and the design of more effective vaccine formulations together with new, safe adjuvants will further aid the development of needle-free influenza vaccines. References 1 WHO. WHO Media Influenza Factsheet N°211 (2003). http://www. who.int/mediacentre/factsheets/2003/fs211/en/(accessed July 1, 2010). 2 The European agency for the evaluation of medicinal products (EMEA): committee for proprietary medicinal products (CPMP). Guideline on dossier structure and content for pandemic influenza vaccine marketing authorisation application (2004). http://archives. who.int/prioritymeds/report/append/62EMEAguidelines.pdf (accessed July 1, 2010). The European Agency for the Evaluation of Medicinal Products 3 (EMEA). http://www.ema.europa.eu/docs/en_GB/document_ library/Scientific_guideline/2009/09/WC500003869.pdf (accessed July 1, 2010). 4 The European Agency for the Evaluation of Medicinal Products (EMEA). http://www.ema.europa.eu/ema/index.jsp?curl=pages/ special_topics/general/general_content_000267.jsp&murl=menus/ special_topics/special_topics.jsp&mid=WC0b01ac058004b634 (accessed July 1, 2010). 5 Wilschut J, McElhaney JE, Palache AM. Rapid Reference Influenza: 2nd edn. London: Mosby/Elsevier Science, 2006. 6 Kersten G, Hirschberg H. Needle-free vaccine delivery. Expert Opin Drug Deliv 2007; 4: 459–74. 7 Mitragotri S. Immunization without needles. Nat Rev Immunol 2005; 5: 905–16. 8 Giudice EL, Campbell JD. Needle-free vaccine delivery. Adv Drug Deliv Rev 2006; 58: 68–89. 9 Azad N, Rojanasakul Y. Vaccine delivery—current trends and future. Curr Drug Deliv 2006; 3: 137–46. 10 O’Hagan DT, Rappuoli R. Novel approaches to vaccine delivery. Pharm Res 2004; 21: 1519–30. 11 Kersten G, Hirschberg H. Antigen delivery systems. Expert Rev Vaccines 2004; 3: 453–62. 12 Nugent J, Po AL, Scott EM. Design and delivery of non-parenteral vaccines. J Clin Pharm Ther 1998; 23: 257–85. 13 WHO. WHO position paper (macroepidemiology). Wkly Epidemiol Rec 2005; 80: 279–87. 14 Sendi P, Locher R, Bucheli B, Battegay M. Intranasal influenza vaccine in a working population. Clin Infect Dis 2004; 38: 974–80. 15 Liew FY, Russell SM, Appleyard G, Brand CM, Beale J. Crossprotection in mice infected with influenza A virus by the respiratory route is correlated with local IgA antibody rather than serum antibody or cytotoxic T cell reactivity. Eur J Immunol 1984; 14: 350–56. 16 Tumpey TM, Renshaw M, Clements JD, Katz JM. Mucosal delivery of inactivated influenza vaccine induces B-cell-dependent heterosubtypic cross-protection against lethal influenza A H5N1 virus infection. J Virol 2001; 75: 5141–50.
www.thelancet.com/infection Vol 10 October 2010
23
24
25
26
27 28 29 30
31
32 33
34
35
36
37 38
39
Asahi-Ozaki Y, Yoshikawa T, Iwakura Y, et al. Secretory IgA antibodies provide cross-protection against infection with different strains of influenza B virus. J Med Virol 2004; 74: 328–35. Hasegawa H, Ichinohe T, Strong P, et al. Protection against influenza virus infection by intranasal administration of hemagglutinin vaccine with chitin microparticles as an adjuvant. J Med Virol 2005; 75: 130–36. Ichinohe T, Tamura S, Kawaguchi A, et al. Cross-protection against H5N1 influenza virus infection is afforded by intranasal inoculation with seasonal trivalent inactivated influenza vaccine. J Infect Dis 2007; 196: 1313–20. Ekwueme DU, Weniger BG, Chen RT. Model-based estimates of risks of disease transmission and economic costs of seven injection devices in sub-Saharan Africa. Bull World Health Organ 2002; 80: 859–70. Sugimura T, Ito Y, Tananari Y, et al. Improved antibody responses in infants less than 1 year old using intradermal influenza vaccination. Vaccine 2008; 26: 2700–05. Amorij JP, Huckriede A, Wilschut J, Frijlink HW, Hinrichs WL. Development of stable influenza vaccine powder formulations: challenges and possibilities. Pharm Res 2008; 25: 1256–73. van Vliet SJ, den Dunnen J, Gringhuis SI, Geijtenbeek TB, van Kooyk Y. Innate signaling and regulation of dendritic cell immunity. Curr Opin Immunol 2007; 19: 435–40. Geeraedts F, Goutagny N, Hornung V, et al. Superior immunogenicity of inactivated whole virus H5N1 influenza vaccine is primarily controlled by Toll-like receptor signalling. PLoS Pathog 2008; 4: e1000138. Mosca F, Tritto E, Muzzi A, et al. Molecular and cellular signatures of human vaccine adjuvants. Proc Natl Acad Sci USA 2008; 105: 10501–06. Marrack P, McKee AS, Munks MW. Towards an understanding of the adjuvant action of aluminium. Nat Rev Immunol 2009; 9: 287–93. Jensen PE. Recent advances in antigen processing and presentation. Nat Immunol 2007; 8: 1041–48. Harwood NE, Batista FD. New insights into the early molecular events underlying B cell activation. Immunity 2008; 28: 609–19. Neutra MR, Kozlowski PA. Mucosal vaccines: the promise and the challenge. Nat Rev Immunol 2006; 6: 148–58. Minne A, Louahed J, Mehauden S, Baras B, Renauld JC, Vanbever R. The delivery site of a monovalent influenza vaccine within the respiratory tract impacts on the immune response. Immunology 2007; 122: 316–25. Tafaghodi M, Abolghasem Sajadi Tabassi S, Jaafari MR, Zakavi SR, Momen-Nejad M. Evaluation of the clearance characteristics of various microspheres in the human nose by gamma-scintigraphy. Int J Pharm 2004; 280: 125–35. Vidgren MT, Kublik H. Nasal delivery systems and their effect on deposition and absorption. Adv Drug Deliv Rev 1998; 29: 157–77. Meitin CA, Bender BS, Small PA Jr. Enteric immunization of mice against influenza with recombinant vaccinia. Proc Natl Acad Sci USA 1994; 91: 11187–91. Amorij JP, Westra TA, Hinrichs WL, Huckriede A, Frijlink HW. Towards an oral influenza vaccine: Comparison between intragastric and intracolonic delivery of influenza subunit vaccine in a murine model. Vaccine 2007; 26: 67–76. Frijlink HW, De Boer AH. Trends in the technology-driven development of new inhalation devices. Drug Discovery Today: Technologies 2005; 2: 47–57. Schellekens RC, Stellaard F, Mitrovic D, Stuurman FE, Kosterink JG, Frijlink HW. Pulsatile drug delivery to ileo-colonic segments by structured incorporation of disintegrants in pHresponsive polymer coatings. J Control Release 2008; 132: 91–98. Glenn GM, O’Hagan DT. Adjuvants: progress, regress and pandemic preparedness. Expert Rev Vaccines 2007; 6: 651–52. Smith DJ, King WF, Barnes LA, Trantolo D, Wise DL, Taubman MA. Facilitated intranasal induction of mucosal and systemic immunity to mutans streptococcal glucosyltransferase peptide vaccines. Infect Immun 2001; 69: 4767–73. Oien NL, Brideau RJ, Walsh EE, Wathen MW. Induction of local and systemic immunity against human respiratory syncytial virus using a chimeric FG glycoprotein and cholera toxin B subunit. Vaccine 1994; 12: 731–35.
707
Review
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
708
Boyce TG, Hsu HH, Sannella EC, et al. Safety and immunogenicity of adjuvanted and unadjuvanted subunit influenza vaccines administered intranasally to healthy adults. Vaccine 2000; 19: 217–26. Gluck U, Gebbers JO, Gluck R. Phase 1 evaluation of intranasal virosomal influenza vaccine with and without Escherichia coli heat-labile toxin in adult volunteers. J Virol 1999; 73: 7780–86. Glueck R. Pre-clinical and clinical investigation of the safety of a novel adjuvant for intranasal immunization. Vaccine 2001; 20 (suppl 1): S42–44. Huckriede A, Bungener L, Stegmann T, et al. The virosome concept for influenza vaccines. Vaccine 2005; 23 (suppl 1): S26–38. Mutsch M, Zhou W, Rhodes P, et al. Use of the inactivated intranasal influenza vaccine and the risk of Bell’s palsy in Switzerland. N Engl J Med 2004; 350: 896–903. Pine S, Barackman J, Ott G, O’Hagan D. Intranasal immunization with influenza vaccine and a detoxified mutant of heat labile enterotoxin from Escherichia coli (LTK63). J Control Release 2002; 85: 263–70. Stephenson I, Zambon MC, Rudin A, et al. Phase I evaluation of intranasal trivalent inactivated influenza vaccine with nontoxigenic Escherichia coli enterotoxin and novel biovector as mucosal adjuvants, using adult volunteers. J Virol 2006; 80: 4962–70. Joseph A, Itskovitz-Cooper N, Samira S, et al. A new intranasal influenza vaccine based on a novel polycationic lipid—ceramide carbamoyl-spermine (CCS) I: immunogenicity and efficacy studies in mice. Vaccine 2006; 24: 3990–4006. Ko SY, Ko HJ, Chang WS, Park SH, Kweon MN, Kang CY. alpha-Galactosylceramide can act as a nasal vaccine adjuvant inducing protective immune responses against viral infection and tumor. J Immunol 2005; 175: 3309–17. Youn HJ, Ko SY, Lee KA, et al. A single intranasal immunization with inactivated influenza virus and alphagalactosylceramide induces long-term protective immunity without redirecting antigen to the central nervous system. Vaccine 2007; 25: 5189–98. Shim DH, Ko HJ, Volker G, et al. Efficacy of poly[di(sodium carboxylatophenoxy)phosphazene] (PCPP) as mucosal adjuvant to induce protective immunity against respiratory pathogens. Vaccine; 28: 2311–17. Guy B, Pascal N, Francon A, et al. Design, characterization and preclinical efficacy of a cationic lipid adjuvant for influenza split vaccine. Vaccine 2001; 19: 1794–805. Nishino M, Mizuno D, Kimoto T, et al. Influenza vaccine with Surfacten, a modified pulmonary surfactant, induces systemic and mucosal immune responses without side effects in minipigs. Vaccine 2009; 27: 5620–27. Bracci L, Canini I, Puzelli S, et al. Type I IFN is a powerful mucosal adjuvant for a selective intranasal vaccination against influenza virus in mice and affects antigen capture at mucosal level. Vaccine 2005; 23: 2994–3004. Bracci L, Canini I, Venditti M, et al. Type I IFN as a vaccine adjuvant for both systemic and mucosal vaccination against influenza virus. Vaccine 2006; 24 (suppl 2): S2 56–57. Arulanandam BP, O’Toole M, Metzger DW. Intranasal interleukin-12 is a powerful adjuvant for protective mucosal immunity. J Infect Dis 1999; 180: 940–49. Joseph A, Louria-Hayon I, Plis-Finarov A, et al. Liposomal immunostimulatory DNA sequence (ISS-ODN): an efficient parenteral and mucosal adjuvant for influenza and hepatitis B vaccines. Vaccine 2002; 20: 3342–54. Ichinohe T, Kawaguchi A, Tamura S, et al. Intranasal immunization with H5N1 vaccine plus poly I: poly C12U, a toll-like receptor agonist, protects mice against homologous and heterologous virus challenge. Microbes Infect 2007; 9: 1333–40. Song SK, Moldoveanu Z, Nguyen HH, et al. Intranasal immunization with influenza virus and Korean mistletoe lectin C (KML-C) induces heterosubtypic immunity in mice. Vaccine 2007; 25: 6359–66. Zanvit P, Havlickova M, Tacner J, et al. Protective and crossprotective mucosal immunization of mice by influenza virus type A with bacterial adjuvant. Immunol Lett 2008; 115: 144–52.
60
61
62
63 64
65
66
67 68
69
70
71
72
73
74
75
76
77
78
79
80
81
Plante M, Jones T, Allard F, et al. Nasal immunization with subunit proteosome influenza vaccines induces serum HAI, mucosal IgA and protection against influenza challenge. Vaccine 2001; 20: 218–25. Lovgren K, Kaberg H, Morein B. An experimental influenza subunit vaccine (iscom): induction of protective immunity to challenge infection in mice after intranasal or subcutaneous administration. Clin Exp Immunol 1990; 82: 435–39. Sjolander S, Drane D, Davis R, Beezum L, Pearse M, Cox J. Intranasal immunisation with influenza-ISCOM induces strong mucosal as well as systemic antibody and cytotoxic T-lymphocyte responses. Vaccine 2001; 19: 4072–80. Coulter A, Harris R, Davis R, et al. Intranasal vaccination with ISCOMATRIX adjuvanted influenza vaccine. Vaccine 2003; 21: 946–49. Helgeby A, Robson NC, Donachie AM, et al. The combined CTA1-DD/ISCOM adjuvant vector promotes priming of mucosal and systemic immunity to incorporated antigens by specific targeting of B cells. J Immunol 2006; 176: 3697–706. Hu KF, Lovgren-Bengtsson K, Morein B. Immunostimulating complexes (ISCOMs) for nasal vaccination. Adv Drug Deliv Rev 2001; 51: 149–59. Scheerlinck JP, Gekas S, Yen HH, et al. Local immune responses following nasal delivery of an adjuvanted influenza vaccine. Vaccine 2006; 24: 3929–36. Read RC, Naylor SC, Potter CW, et al. Effective nasal influenza vaccine delivery using chitosan. Vaccine 2005; 23: 4367–74. Bacon A, Makin J, Sizer PJ, et al. Carbohydrate biopolymers enhance antibody responses to mucosally delivered vaccine antigens. Infect Immun 2000; 68: 5764–70. Amidi M, Romeijn SG, Verhoef JC, et al. N-Trimethyl chitosan (TMC) nanoparticles loaded with influenza subunit antigen for intranasal vaccination: Biological properties and immunogenicity in a mouse model. Vaccine 2007; 25: 144–153. Illum L, Jabbal-Gill I, Hinchcliffe M, Fisher AN, Davis SS. Chitosan as a novel nasal delivery system for vaccines. Adv Drug Deliv Rev 2001; 51: 81–96. Huang J, Garmise RJ, Crowder TM, et al. A novel dry powder influenza vaccine and intranasal delivery technology: induction of systemic and mucosal immune responses in rats. Vaccine 2004; 23: 794–801. Garmise RJ, Staats HF, Hickey AJ. Novel dry powder preparations of whole inactivated influenza virus for nasal vaccination. AAPS PharmSciTech 2007; 8: e81. Smith DJ, Bot S, Dellamary L, Bot A. Evaluation of novel aerosol formulations designed for mucosal vaccination against influenza virus. Vaccine 2003; 21: 2805–12. Song JH, Nguyen HH, Cuburu N, et al. Sublingual vaccination with influenza virus protects mice against lethal viral infection. Proc Natl Acad Sci USA 2008; 105: 1644–49. McCluskie MJ, Davis HL. Oral, intrarectal and intranasal immunizations using CpG and non-CpG oligodeoxynucleotides as adjuvants. Vaccine 2000; 19: 413–22. Katz JM, Lu X, Young SA, Galphin JC. Adjuvant activity of the heat-labile enterotoxin from enterotoxigenic Escherichia coli for oral administration of inactivated influenza virus vaccine. J Infect Dis 1997; 175: 352–63. Barackman JD, Ott G, Pine S, O’Hagan DT. Oral administration of influenza vaccine in combination with the adjuvants LT-K63 and LT-R72 induces potent immune responses comparable to or stronger than traditional intramuscular immunization. Clin Diagn Lab Immunol 2001; 8: 652–57. Lu X, Clements JD, Katz JM. Mutant Escherichia coli heat-labile enterotoxin [LT(R192G)] enhances protective humoral and cellular immune responses to orally administered inactivated influenza vaccine. Vaccine 2002; 20: 1019–29. Conacher M, Alexander J, Brewer JM. Oral immunisation with peptide and protein antigens by formulation in lipid vesicles incorporating bile salts (bilosomes). Vaccine 2001; 19: 2965–74. Kunzi V, Klap JM, Seiberling MK, et al. Immunogenicity and safety of low dose virosomal adjuvanted influenza vaccine administered intradermally compared to intramuscular full dose administration. Vaccine 2009; 27: 3561–67. Skountzou I, Quan FS, Jacob J, Compans RW, Kang SM. Transcutaneous immunization with inactivated influenza virus induces protective immune responses. Vaccine 2006; 24: 6110–19.
www.thelancet.com/infection Vol 10 October 2010
Review
82
83
84 85
86
87
88
89
90
91
92
93
94
95
96
97
98
99 100
101
102
103
Chen D, Endres RL, Erickson CA, Maa YF, Payne LG. Epidermal powder immunization using non-toxic bacterial enterotoxin adjuvants with influenza vaccine augments protective immunity. Vaccine 2002; 20: 2671–79. Chen D, Endres R, Maa YF, et al. Epidermal powder immunization of mice and monkeys with an influenza vaccine. Vaccine 2003; 21: 2830–36. Sesardic D, Rijpkema S, Patel BP. New adjuvants: EU regulatory developments. Expert Rev Vaccines 2007; 6: 849–61. Kanerva M, Mannonen L, Piiparinen H, Peltomaa M, Vaheri A, Pitkaranta A. Search for Herpesviruses in cerebrospinal fluid of facial palsy patients by PCR. Acta Otolaryngol 2007; 127: 775–79. van Ginkel FW, Jackson RJ, Yoshino N, et al. Enterotoxin-based mucosal adjuvants alter antigen trafficking and induce inflammatory responses in the nasal tract. Infect Immun 2005; 73: 6892–902. van der Laan JW, Herberts C, Lambkin-Williams R, Boyers A, Mann AJ, Oxford J. Animal models in influenza vaccine testing. Expert Rev Vaccines 2008; 7: 783–93. Bodewes R, Rimmelzwaan GF, Osterhaus AD. Animal models for the preclinical evaluation of candidate influenza vaccines. Expert Rev Vaccines 2010; 9: 59–72. The European agency for the evaluation of medicinal products (EMEA) committee for proprietary medicinal products (CPMP). Note for guidance on harmonization of requirements for influenza vaccines (1997). http://www.ema.europa.eu/docs/en_GB/ document_library/Scientific_guideline/2009/09/WC500003945.pdf (accessed July 1, 2010). Clements ML, Betts RF, Tierney EL, Murphy BR. Serum and nasal wash antibodies associated with resistance to experimental challenge with influenza A wild-type virus. J Clin Microbiol 1986; 24: 157–60. Renegar KB, Small PA Jr. Immunoglobulin A mediation of murine nasal anti-influenza virus immunity. J Virol 1991; 65: 2146–48. Mizuno D, Ide-Kurihara M, Ichinomiya T, Kubo I, Kido H. Modified pulmonary surfactant is a potent adjuvant that stimulates the mucosal IgA production in response to the influenza virus antigen. J Immunol 2006; 176: 1122–30. Rimmelzwaan GF, Fouchier RA, Osterhaus AD. Influenza virusspecific cytotoxic T lymphocytes: a correlate of protection and a basis for vaccine development. Curr Opin Biotechnol 2007; 18: 529–36. McElhaney JE, Xie D, Hager WD, et al. T cell responses are better correlates of vaccine protection in the elderly. J Immunol 2006; 176: 6333–39. He XS, Holmes TH, Zhang C, et al. Cellular immune responses in children and adults receiving inactivated or live attenuated influenza vaccines. J Virol 2006; 80: 11756–66. Chen D, Weis KF, Chu Q, et al. Epidermal powder immunization induces both cytotoxic T-lymphocyte and antibody responses to protein antigens of influenza and hepatitis B viruses. J Virol 2001; 75: 11630–40. Treanor J, Wright PF. Immune correlates of protection against influenza in the human challenge model. Dev Biol (Basel) 2003; 115: 97–104. Wilschut J, de Jonge J, Huckriede A, Amorij JP, Hinrichs WL, Frijlink HW. Preservation of influenza virosome structure and function during freeze-drying and storage. J Liposome Res 2007; 17: 173–82. Abramson JS. Intranasal, cold-adapted, live, attenuated influenza vaccine. Pediatr Infect Dis J 1999; 18: 1103–04. Belshe R, Lee MS, Walker RE, Stoddard J, Mendelman PM. Safety, immunogenicity and efficacy of intranasal, live attenuated influenza vaccine. Expert Rev Vaccines 2004; 3: 643–54. Cox RJ, Brokstad KA, Ogra P. Influenza virus: immunity and vaccination strategies. Comparison of the immune response to inactivated and live, attenuated influenza vaccines. Scand J Immunol 2004; 59: 1–15. Beyer WE, Palache AM, de Jong JC, Osterhaus AD. Cold-adapted live influenza vaccine versus inactivated vaccine: systemic vaccine reactions, local and systemic antibody response, and vaccine efficacy. A meta-analysis. Vaccine 2002; 20: 1340–53. Glezen WP. Control of influenza. Tex Heart Inst J 2004; 31: 39–41.
www.thelancet.com/infection Vol 10 October 2010
104 Fleming DM, Crovari P, Wahn U, et al. Comparison of the efficacy and safety of live attenuated cold-adapted influenza vaccine, trivalent, with trivalent inactivated influenza virus vaccine in children and adolescents with asthma. Pediatr Infect Dis J 2006; 25: 860–69. 105 Greenbaum E, Furst A, Kiderman A, et al. Mucosal [SIgA] and serum [IgG] immunologic responses in the community after a single intra-nasal immunization with a new inactivated trivalent influenza vaccine. Vaccine 2002; 20: 1232–39. 106 Greenbaum E, Engelhard D, Levy R, Schlezinger M, Morag A, Zakay-Rones Z. Mucosal (SIgA) and serum (IgG) immunologic responses in young adults following intranasal administration of one or two doses of inactivated, trivalent anti-influenza vaccine. Vaccine 2004; 22: 2566–77. 107 Samdal HH, Bakke H, Oftung F, et al. A non-living nasal influenza vaccine can induce major humoral and cellular immune responses in humans without the need for adjuvants. Hum Vaccin 2005; 1: 85–90. 108 Muszkat M, Greenbaum E, Ben-Yehuda A, et al. Local and systemic immune response in nursing-home elderly following intranasal or intramuscular immunization with inactivated influenza vaccine. Vaccine 2003; 21: 1180–86. 109 Atmar RL, Keitel WA, Cate TR, Munoz FM, Ruben F, Couch RB. A dose-response evaluation of inactivated influenza vaccine given intranasally and intramuscularly to healthy young adults. Vaccine 2007; 25: 5367–73. 110 Haan L, Verweij WR, Holtrop M, et al. Nasal or intramuscular immunization of mice with influenza subunit antigen and the B subunit of Escherichia coli heat-labile toxin induces IgA- or IgG-mediated protective mucosal immunity. Vaccine 2001; 19: 2898–907. 111 McCluskie MJ, Weeratna RD, Clements JD, Davis HL. Mucosal immunization of mice using CpG DNA and/or mutants of the heat-labile enterotoxin of Escherichia coli as adjuvants. Vaccine 2001; 19: 3759–68. 112 Hagiwar Y, Tsuji T, Iwasaki T, et al. Effectiveness and safety of mutant Escherichia coli heat-labile enterotoxin (LT H44A) as an adjuvant for nasal influenza vaccine. Vaccine 2001; 19: 2071–79. 113 Barchfeld GL, Hessler AL, Chen M, Pizza M, Rappuoli R, Van Nest GA. The adjuvants MF59 and LT-K63 enhance the mucosal and systemic immunogenicity of subunit influenza vaccine administered intranasally in mice. Vaccine 1999; 17: 695–704. 114 Halperin SA, Smith B, Clarke K, Treanor J, Mabrouk T, Germain M. Phase I, randomized, controlled trial to study the reactogenicity and immunogenicity of a nasal, inactivated trivalent influenza virus vaccine in healthy adults. Hum Vaccin 2005; 1: 37–42. 115 Ugwoke MI, Agu RU, Verbeke N, Kinget R. Nasal mucoadhesive drug delivery: background, applications, trends and future perspectives. Adv Drug Deliv Rev 2005; 57: 1640–65. 116 Robert Johnson. Interferon as a mucosal adjuvant for influenza vaccine given intranasally (NCT00436046). http://clinicaltrials.gov/ ct2/show/study/NCT00436046?sect=X501 (accessed July 1, 2010). 117 Sullivan VJ, Mikszta JA, Laurent P, Huang J, Ford B. Noninvasive delivery technologies: respiratory delivery of vaccines. Expert Opin Drug Deliv 2006; 3: 87–95. 118 Jiang G, Joshi SB, Peek LJ, et al. Anthrax vaccine powder formulations for nasal mucosal delivery. J Pharm Sci 2006; 95: 80–96. 119 Huo Z, Sinha R, McNeela EA, et al. Induction of protective serum meningococcal bactericidal and diphtheria-neutralizing antibodies and mucosal immunoglobulin A in volunteers by nasal insufflations of the Neisseria meningitidis serogroup C polysaccharide-CRM197 conjugate vaccine mixed with chitosan. Infect Immun 2005; 73: 8256–65. 120 Garmise RJ, Mar K, Crowder TM, et al. Formulation of a dry powder influenza vaccine for nasal delivery. AAPS PharmSciTech 2006; 7: e19. 121 Soane RJ, Frier M, Perkins AC, Jones NS, Davis SS, Illum L. Evaluation of the clearance characteristics of bioadhesive systems in humans. Int J Pharm 1999; 178: 55–65. 122 McInnes FJ, Thapa P, Baillie AJ, et al. In vivo evaluation of nicotine lyophilised nasal insert in sheep. Int J Pharm 2005; 304: 72–82.
709
Review
123 Newman SP, Pitcairn GR, Dalby RN. Drug delivery to the nasal cavity: in vitro and in vivo assessment. Crit Rev Ther Drug Carrier Syst 2004; 21: 21–66. 124 Laube BL. Devices for aerosol delivery to treat sinusitis. J Aerosol Med 2007; 20 (suppl 1): s5–17. 125 Groneberg DA, Paul H, Welte T. Novel strategies of aerosolic pharmacotherapy. Exp Toxicol Pathol 2006; 57 (suppl 2): 49–53. 126 Holmgren J, Czerkinsky C. Mucosal immunity and vaccines. Nat Med 2005; 11 (suppl 4): s45–53. 127 Lambrecht BN, Prins JB, Hoogsteden HC. Lung dendritic cells and host immunity to infection. Eur Respir J 2001; 18: 692–704. 128 von Garnier C, Filgueira L, Wikstrom M, et al. Anatomical location determines the distribution and function of dendritic cells and other APCs in the respiratory tract. J Immunol 2005; 175: 1609–18. 129 Holt PG. Pulmonary dendritic cells in local immunity to inert and pathogenic antigens in the respiratory tract. Proc Am Thorac Soc 2005; 2: 116–20. 130 Holmgren J, Czerkinsky C, Lycke N, Svennerholm AM. Mucosal immunity: implications for vaccine development. Immunobiology 1992; 184: 157–79. 131 Mann JJ, Waldman RH, Togo Y, Heiner GG, Dawkins AT, Kasel JA. Antibody response in respiratory secretions of volunteers given live and dead influenza virus. J Immunol 1968; 100: 726–35. 132 Kasel JA, Fulk RV, Togo Y, et al. Influenza antibody in human respiratory secretions after subcutaneous or respiratory immunization with inactivated virus. Nature 1968; 218: 594–95. 133 Waldman RH, Mann JJ, Kasel JA. Influenza virus neutralizing antibody in human respiratory secretions. J Immunol 1968; 100: 80–85. 134 Waldman RH, Mann JJ, Small PA Jr. Immunization against influenza: prevention of illness in man by aerosolized inactivated vaccine. JAMA 1969; 207: 520–24. 135 Waldman RH, Jurgensen PF, Olsen GN, Ganguly R, Johnson JE 3rd. Immune response of the human respiratory tract—I: Immunoglobulin levels and influenza virus vaccine antibody response. J Immunol 1973; 111: 38–41. 136 Waldman RH, Wigley FMSPA Jr. Specificity of respiratory secretion antibody against influenza virus. J Immunol 1970; 105: 1477–83. 137 Haigh W, Howell RW. The efficacy of the A2/Aichi/68 strain in inhaled influenza immunisation against the A/England/42/72 variant. J Soc Occup Med 1973; 23: 125–7. 138 Haigh W, Howell RW, Meichen FW. A comparative trial of influenza immunization by inhalation and hypojet methods. Practitioner 1973; 211: 365–70. 139 Wee JL, Scheerlinck JP, Snibson KJ, et al. Pulmonary delivery of ISCOMATRIX influenza vaccine induces both systemic and mucosal immunity with antigen dose sparing. Mucosal Immunol 2008; 1: 489–96. 140 Amorij JP, Saluja V, Petersen AH, Hinrichs WL, Huckriede A, Frijlink HW. Pulmonary delivery of an inulin-stabilized influenza subunit vaccine prepared by spray-freeze drying induces systemic, mucosal humoral as well as cell-mediated immune responses in BALB/c mice. Vaccine 2007; 25: 8707–17. 141 Dalby R, Spallek M, Voshaar T. A review of the development of respimat soft mist inhaler. Int J Pharm 2004; 283: 1–9. 142 de Boer AH, Hagedoorn P, Westerman EM, Le Brun PP, Heijerman HG, Frijlink HW. Design and in vitro performance testing of multiple air classifier technology in a new disposable inhaler concept (Twincer) for high powder doses. Eur J Pharm Sci 2006; 28: 171–78. 143 Lu D, Hickey AJ. Pulmonary vaccine delivery. Expert Rev Vaccines 2007; 6: 213–26. 144 Meyer P, Menzel M, Muellinger B, Weber N, Haeussinger K, Ziegler-Heitbrock L. Inhalative vaccination with pneumococcal polysaccharide in patients with chronic obstructive pulmonary disease. Vaccine 2006; 24: 5832–38. 145 Minne A, Huaux F, Jaworska J, Rha RD, Hamelmann E, Vanbever R. Safety evaluation of pulmonary influenza vaccination in healthy and asthmatic mice. Vaccine 2008; 26: 2360–68. 146 Senel S, Kremer M, Nagy K, Squier C. Delivery of bioactive peptides and proteins across oral (buccal) mucosa. Curr Pharm Biotechnol 2001; 2: 175–86.
710
147 Challacombe S, Coogan M, Williams D, Greenspan J. Overview and research agenda arising from the 5th World Workshop on Oral Health and Disease in AIDS. Adv Dent Res 2006; 19: 5–9. 148 Otten K, Wang HC, Wyde PR, Klein JR. Modulation of gamma delta T cells in mouse buccal epithelium following antigen priming. Biochem Biophys Res Commun 2002; 294: 626–29. 149 Gebert A, Pabst R. M cells at locations outside the gut. Semin Immunol 1999; 11: 165–70. 150 Childers NK, Powell WD, Tong G, Kirk K, Wiatrak B, Michalek SM. Human salivary immunoglobulin and antigen-specific antibody activity after tonsillectomy. Oral Microbiol Immunol 2001; 16: 265–69. 151 Bakke H, Samdal HH, Holst J, et al. Oral spray immunization may be an alternative to intranasal vaccine delivery to induce systemic antibodies but not nasal mucosal or cellular immunity. Scand J Immunol 2006; 63: 223–31. 152 Canonica GW, Passalacqua G. Noninjection routes for immunotherapy. J Allergy Clin Immunol 2003; 111: 437–48. 153 Cuburu N, Kweon MN, Song JH, et al. Sublingual immunization induces broad-based systemic and mucosal immune responses in mice. Vaccine 2007; 25: 8598–610. 154 Avtushenko SS, Sorokin EM, Zoschenkova NY, Zacharova NG, Naichin AN. Clinical and immunological characteristics of the emulsion form of inactivated influenza vaccine delivered by oral immunization. J Biotechnol 1996; 44: 21–28. 155 Moldoveanu Z, Clements ML, Prince SJ, Murphy BR, Mestecky J. Human immune responses to influenza virus vaccines administered by systemic or mucosal routes. Vaccine 1995; 13: 1006–12. 156 Lazzell V, Waldman RH, Rose C, Khakoo R, Jacknowitz A, Howard S. Immunization against influenza in humans using an oral enteric-coated killed virus vaccine. J Biol Stand 1984; 12: 315–21. 157 Bergmann KC, Waldman RH, Tischner H, Pohl WD. Antibody in tears, saliva and nasal secretions following oral immunization of humans with inactivated influenza virus vaccine. Int Arch Allergy Appl Immunol 1986; 80: 107–09. 158 Gazzaniga A, Maroni A, Sangalli ME, Zema L. Time-controlled oral delivery systems for colon targeting. Expert Opin Drug Deliv 2006; 3: 583–97. 159 Huang CM. Topical vaccination: the skin as a unique portal to adaptive immune responses. Semin Immunopathol 2007; 29: 71–80. 160 Yu RC, Abrams DC, Alaibac M, Chu AC. Morphological and quantitative analyses of normal epidermal Langerhans cells using confocal scanning laser microscopy. Br J Dermatol 1994; 131: 843–48. 161 Glenn GM, Kenney RT. Mass vaccination: solutions in the skin. Curr Top Microbiol Immunol 2006; 304: 247–68. 162 Chiu SS, Peiris JS, Chan KH, Wong WH, Lau YL. Immunogenicity and safety of intradermal influenza immunization at a reduced dose in healthy children. Pediatrics 2007; 119: 1076–82. 163 Belshe RB, Newman FK, Cannon J, et al. Serum antibody responses after intradermal vaccination against influenza. N Engl J Med 2004; 351: 2286–94. 164 Kenney RT, Frech SA, Muenz LR, Villar CP, Glenn GM. Dose sparing with intradermal injection of influenza vaccine. N Engl J Med 2004; 351: 2295–301. 165 Holland D, Booy R, De Looze F, et al. Intradermal influenza vaccine administered using a new microinjection system produces superior immunogenicity in elderly adults: a randomized controlled trial. J Infect Dis 2008; 198: 650–58. 166 Chi RC, Rock MT, Neuzil KM. Immunogenicity and safety of intradermal influenza vaccination in healthy older adults. Clin Infect Dis 2010; 50: 1331–38. 167 Nicholson KG, Thompson CI, Klap JM, et al. Safety and immunogenicity of whole-virus, alum-adjuvanted whole-virus, virosomal, and whole-virus intradermal influenza A/H9N2 vaccine formulations. Vaccine 2009; 28: 171–78. 168 Chiu SS, Chan KH, Tu W, Lau YL, Peiris JS. Immunogenicity and safety of intradermal versus intramuscular route of influenza immunization in infants less than 6 of age: a randomized controlled trial. Vaccine 2009; 27: 4834–39. 169 Gelinck LB, van den Bemt BJ, Marijt WA, et al. Intradermal influenza vaccination in immunocompromized patients is immunogenic and feasible. Vaccine 2009; 27: 2469–74.
www.thelancet.com/infection Vol 10 October 2010
Review
170 Jo YM, Song JY, Hwang IS, et al. Dose sparing strategy with intradermal influenza vaccination in patients with solid cancer. J Med Virol 2009; 81: 722–27. 171 Belshe RB, Newman FK, Wilkins K, et al. Comparative immunogenicity of trivalent influenza vaccine administered by intradermal or intramuscular route in healthy adults. Vaccine 2007; 25: 6755–63. 172 Auewarakul P, Kositanont U, Sornsathapornkul P, Tothong P, Kanyok R, Thongcharoen P. Antibody responses after dose-sparing intradermal influenza vaccination. Vaccine 2007; 25: 659–63. 173 Hunsaker BD, Perino LJ. Efficacy of intradermal vaccination. Vet Immunol Immunopathol 2001; 79: 1–13. 174 Vandervoort J, Ludwig A. Microneedles for transdermal drug delivery: a minireview. Front Biosci 2008; 13: 1711–15. 175 Alarcon JB, Hartley AW, Harvey NG, Mikszta JA. Preclinical evaluation of microneedle technology for intradermal delivery of influenza vaccines. Clin Vaccine Immunol 2007; 14: 375–81. 176 Arnou R, Eavis P, Pardo JR, Ambrozaitis A, Kazek MP, Weber F. Immunogenicity, large scale safety and lot consistency of an intradermal influenza vaccine in adults aged 18-60 years: Randomized, controlled, phase III trial. Hum Vaccin 2010; 6: 346–54. 177 Arnou R, Icardi G, De Decker M, et al. Intradermal influenza vaccine for older adults: a randomized controlled multicenter phase III study. Vaccine 2009; 27: 7304–12. 178 The European Agency for the Evaluation of Medicinal Products (EMEA). http://www.ema.europa.eu/humandocs/Humans/EPAR/ intanza/intanza.htm, 2010 (accessed July 1, 2010). 179 McAllister DV, Wang PM, Davis SP, et al. Microfabricated needles for transdermal delivery of macromolecules and nanoparticles: fabrication methods and transport studies. Proc Natl Acad Sci USA 2003; 100: 13755–60. 180 Mikszta JA, Alarcon JB, Brittingham JM, Sutter DE, Pettis RJ, Harvey NG. Improved genetic immunization via micromechanical disruption of skin-barrier function and targeted epidermal delivery. Nat Med 2002; 8: 415–19. 181 Van Damme P, Oosterhuis-Kafeja F, Van der Wielen M, Almagor Y, Sharon O, Levin Y. Safety and efficacy of a novel microneedle device for dose sparing intradermal influenza vaccination in healthy adults. Vaccine 2009; 27: 454–59. 182 Baxter J, Mitragotri S. Needle-free liquid jet injections: mechanisms and applications. Expert Rev Med Devices 2006; 3: 565–74. 183 Mitragotri S. Current status and future prospects of needle-free liquid jet injectors. Nat Rev Drug Discov 2006; 5: 543–48. 184 Jackson LA, Austin G, Chen RT, et al. Safety and immunogenicity of varying dosages of trivalent inactivated influenza vaccine administered by needle-free jet injectors. Vaccine 2001; 19: 4703–09.
www.thelancet.com/infection Vol 10 October 2010
185 Glenn GM, Flyer DC, Ellingsworth LR, et al. Transcutaneous immunization with heat-labile enterotoxin: development of a needle-free vaccine patch. Expert Rev Vaccines 2007; 6: 809–19. 186 Mahor S, Gupta PN, Rawat A, Vyas SP. A needle-free approach for topical immunization: antigen delivery via vesicular carrier system(s). Curr Med Chem 2007; 14: 2898–910. 187 Glenn GM, Kenney RT, Hammond SA, Ellingsworth LR. Transcutaneous immunization and immunostimulant strategies. Immunol Allergy Clin North Am 2003; 23: 787–813. 188 Vogt A, Mahe B, Costagliola D, et al. Transcutaneous anti-influenza vaccination promotes both CD4 and CD8 T cell immune responses in humans. J Immunol 2008; 180: 1482–89. 189 Garg S, Hoelscher M, Belser JA, et al. Needle-free skin patch delivery of a pandemic influenza vaccine protects mice from lethal viral challenge. Clin Vaccine Immunol 2007; 14: 926–28 190 Sarphie DF, Johnson B, Cormier M, Burkoth TL, Bellhouse BL. Bioavailabillity following transdermal powdered delivery (TDP) of radiolabeled insulin to hairless guinea pigs. J Control Release 1997; 47: 61–69. 191 Chen D, Endres RL, Erickson CA, et al. Epidermal immunization by a needle-free powder delivery technology: immunogenicity of influenza vaccine and protection in mice. Nat Med 2000; 6: 1187–90. 192 Dean HJ, Chen D. Epidermal powder immunization against influenza. Vaccine 2004; 23: 681–86. 193 Chen D, Periwal SB, Larrivee K, et al. Serum and mucosal immune responses to an inactivated influenza virus vaccine induced by epidermal powder immunization. J Virol 2001; 75: 7956–65. 194 Matriano JA, Cormier M, Johnson J, et al. Macroflux microprojection array patch technology: a new and efficient approach for intracutaneous immunization. Pharm Res 2002; 19: 63–70. 195 Koutsonanos DG, del Pilar Martin M, Zarnitsyn VG, et al. Transdermal influenza immunization with vaccine-coated microneedle arrays. PLoS ONE 2009; 4: e4773. 196 Kim YC, Quan FS, Yoo DG, Compans RW, Kang SM, Prausnitz MR. Enhanced memory responses to seasonal H1N1 influenza vaccination of the skin with the use of vaccine-coated microneedles. J Infect Dis 2010; 201: 190–98. 197 Gill HS, Prausnitz MR. Coating formulations for microneedles. Pharm Res 2007; 24: 1369–80. 198 Lee JW, Park JH, Prausnitz MR. Dissolving microneedles for transdermal drug delivery. Biomaterials 2008; 29: 2113–24.
711
Review
Intrinsic antibody-dependent enhancement of microbial infection in macrophages: disease regulation by immune complexes Scott B Halstead, Suresh Mahalingam, Mary A Marovich, Sukathida Ubol, David M Mosser Lancet Infect Dis 2010; 10: 712–22 Faculty of Applied Science, University of Canberra, Bruce, Canberra, Australia (Prof S Mahalingam PhD); Department of Vaccine Research and Development, Walter Reed Army Institute of Research, Rockville, MD, USA (M A Marovich MD); Microbiology Department, Mahidol University, Bangkok, Thailand (S Ubol PhD); Department of Cell Biology and Molecular Genetics, University of Maryland, College Park, MD, USA (Prof D M Mosser PhD); Supportive Research and Development Program, Pediatric Dengue Vaccine Initiative, International Vaccine Institute, Seoul, Korea (S B Halstead MD) Correspondence to: Dr Scott B Halstead, 5824 Edson Lane, North Bethesda, MD 20852, USA
[email protected]
A wide range of microorganisms can replicate in macrophages, and cell entry of these pathogens via non-neutralising IgG antibody complexes can result in increased intracellular infection through idiosyncratic Fcγ-receptor signalling. The activation of Fcγ receptors usually leads to phagocytosis. Paradoxically, the ligation of monocyte or macrophage Fcγ receptors by IgG immune complexes, rather than aiding host defences, can suppress innate immunity, increase production of interleukin 10, and bias T-helper-1 (Th1) responses to Th2 responses, leading to increased infectious output by infected cells. This intrinsic antibody-dependent enhancement (ADE) of infection modulates the severity of diseases as disparate as dengue haemorrhagic fever and leishmaniasis. Intrinsic ADE is distinct from extrinsic ADE, whereby complexes of infectious agents with non-neutralising antibodies lead to an increased number of infected cells. Intrinsic ADE might be involved in many protozoan, bacterial, and viral infections. We review insights into intracellular mechanisms and implications of enhanced pathogenesis after ligation of macrophage Fcγ receptors by infectious immune complexes.
Introduction Over the past four decades, information from different lines of scientific inquiry has improved our understanding of antibody-mediated mechanisms that modulate severity of infections by diverse microorganisms. Independent studies of pathogenesis of cellular and host responses to acute and chronic human and animal infectious diseases have generated evidence that cross-linking of immune complexes with Fcγ receptors increases cellular infection, contributing to disease severity by a mechanism we call intrinsic antibody-dependent enhancement (ADE) of infection. Intrinsic ADE is distinct from extrinsic ADE, whereby complexes of infectious agents with nonneutralising antibodies lead to an increased number of infected cells. Hawkes1 made early observations in studies of the neutralisation of Murray Valley encephalitis virus by use of the serum-dilution, virus-constant method. Chick embryo fibroblast monolayers exposed to virus mixed with high dilutions of chicken antisera had more plaques than did those exposed to virus alone.1 In follow-up studies,2 this finding seemed to result from antibody stabilisation of infectivity by the Murray Valley encephalitis virus. A different explanation emerged when sequential infections with dengue viruses resulted in dengue haemorrhagic fever.3 When monocytes and macrophages were identified as the main hosts of dengue infection, ADE was implicated.4,5 Infection with the Murray Valley encephalitis virus was increased by ADE in the 2% of functional chicken macrophages identified in chick embryo fibroblasts. Because of the conformational requirement that Fc receptors and Fcγ termini must be of the same phylogenetic class, ADE in chick embryo fibroblasts was reported only when Murray Valley encephalitis virus antibodies were raised in chickens, not in mammals.6,7
Macrophage biology Monocytes are produced in the bone marrow by haemopoietic stem-cell precursors and then circulate in 712
the bloodstream for about 1–3 days; about half of these cells are stored in the spleen. Monocytes mature to replenish resident macrophages and dendritic cells. Macrophages are scavengers that remove worn-out cells and other debris. They also present antigens that initiate the immune response. Macrophages have receptors for lymphokines that enhance their function. Two signals are needed to produce macrophages activated by the classic pathway: interferon γ and tumour necrosis factor (TNF) or interferon γ and exogenous toll receptor ligands such as lipopolysaccharide, resulting in macrophages that are able to kill intracellular organisms.8 Macrophages activated by the classic pathway are essential components of the host defence system, but their activation must be tightly controlled because the cytokines and mediators that they produce can lead to host tissue damage and immunopathological disorders such as rheumatoid arthritis and inflammatory bowel disease. Macrophages have great plasticity and can respond to various environmental cues by adopting many different states of activation.8 T-helper-2 (Th2) immune responses, commonly elicited by disturbances at mucosal surfaces, result in the production of interleukin 4 and interleukin 13, which stimulate macrophages to secrete components of the extracellular matrix, making the cells effective in wound healing.8 These activated or wound-healing macrophages produce minimum amounts of proinflammatory cytokines and are less efficient at killing intracellular microbes than are macrophages activated by the classic pathway. Regulatory macrophages typically arise during late stages of adaptive immune responses with a primary role of reducing immune responses and restricting inflammation. These cells can occur in response to tolllike receptor stimulation in the presence of glucocorticoids, prostaglandin E2, interleukin 10, or immune complexes. Regulatory macrophages are www.thelancet.com/infection Vol 10 October 2010
Review
programmed to produce high concentrations of interleukin 10 and to suppress production of interleukin 12.9 Many other unique populations of activated macrophages with distinct phenotypes are likely to arise in response to specific diseases.8
Protozoan parasitism of macrophages Leishmaniasis is caused by protozoan parasites of the genus Leishmania that are transmitted by the bite of sandflies. About 21 of 30 Leishmania species that infect mammals can cause human infection. The disease exists as two major forms: cutaneous and visceral leishmaniasis. Cutaneous leishmaniasis is endemic to many parts of the world, and is closely linked to specific geographical regions—with villages as little as 15 miles apart having different infection rates. Visceral leishmaniasis, also known as kala-azar, is the most severe form of leishmaniasis and, after malaria, is the second-leading cause of death from parasitic infection, with an estimated 500 000 cases each year worldwide.10 Some Leishmania species are closely linked to human beings and are commonly reported in cities (eg, L chagasi, L tropica), whereas others are more traditionally associated with animal species and are classified as zoonoses (eg, L major). Leishmania species are transmitted as promastigotes, motile forms that infect macrophages, and spread within hosts as amastigotes, the obligate parasites of macrophages. In human hosts, the responses to infection vary with Leishmania species and the patients’ immune reactions. Patients whose lymphocytes produce high amounts of interferon γ from Th1-type T cells usually recover from cutaneous infections without treatment; after recovery, these patients are immune to reinfection.11 Patients infected with visceral forms of the parasite produce high titres of antibodies, which do not contribute to host defence. Without treatment, these patients are not likely to recover from visceral leishmaniasis.11 Mouse models exist for the study of cutaneous and visceral leishmaniasis. For example, BALB/c mice are susceptible to this disease and develop progressive nonhealing lesions with many intracellular parasites within macrophages.12 Despite the presence of high concentrations of antiparasite antibodies, these parasites usually disseminate to the liver, spleen, and bone marrow, thus mimicking human visceral leishmaniasis. In-vitro models are also available as promastigote or amastigote infections in cultures of bone marrow or peritoneal macrophages from mice or in human differentiated peripheral blood mononuclear cells (PBMCs).13 During the 1980s and 1990s, several mammalian cytokines were discovered. Among these cytokines was mouse cytokine synthesis inhibitory factor produced by Th2 cells, later renamed interleukin 10.14,15 Interleukin 10 is a type 2 cytokine and the first identified member of a family of cytokines that include interleukins 19, 20, 22, 24, 26, 28, www.thelancet.com/infection Vol 10 October 2010
and 29.16 All these cytokines have similar genomic organisation, bind to receptors with similar structures and, in some cases, shared components, and all activate Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signalling pathways. Interleukin 10 is capable of inhibiting synthesis of proinflammatory cytokines such as interferon γ, interleukins 2, 3, and 12, TNF, and granulocyte macrophage colony-stimulating factor. Interleukin 10 also has a potent ability to suppress the antigen-presentation capacity of antigen-presenting cells. However, interleukin 10 is not always immunosuppressive. The cytokine can promote survival of natural killer cells and B cells and production of B-cell antibodies.17 Interleukin 10 is made by macrophages and various T cells, including Th1, Th2, Th17, and T-regulatory cells.18 Interleukin 10 has important roles in the regulation of immune responses and can also increase host susceptibility to intracellular infections. In a mouse model of cutaneous leishmaniasis, susceptibility to L major is associated with Th2 responses. CD4 T cells from susceptible BALB/c mice produce interleukins 4 and 10 when infected with L major, whereas CD4 T cells from resistant C57BL/6 mice express interferon γ and interleukin 2.19 The transient depletion of CD4 cells or the in-vivo neutralisation of interleukin 4 in BALB/c mice promotes death of intracellular parasites and recovery from leishmanial lesions. Interleukin 4 is not the only determinant of susceptibility, however, because susceptible BALB/c mice that do not express interleukin 10 are fully resistant to infection.20 Furthermore, in people in whom the Th1 to Th2 dichotomy is not as pronounced, interleukin 10 seems to be a major inducer of susceptibility. In people with visceral leishmaniasis, interleukin 10 titres in plasma directly correlate with disease severity.21 Interleukin 10 is found in lymph nodes taken from patients with visceral leishmaniasis.11 Furthermore, PBMCs from patients with acute visceral leishmaniasis respond to stimulation with leishmania lysate by producing interleukin 10 mRNA. When added to PBMCs, interleukin 10 suppresses production of interferon γ and interleukin 2, whereas, after treatment with anti-interleukin 10, PBMCs from patients with acute visceral leishmaniasis have a substantial increase in proliferative response to leishmania lysate. Results of invitro studies of peritoneal and bone marrow macrophages infected with Leishmania species show that the intracellular killing of this organism by macrophages activated by the classic pathway can be inhibited by treatment with exogenous interleukin 1022,23 or by endogenous macrophage production of interleukin 10.22 Interleukin 12 has a complementary role to interleukin 10, acting directly on CD4 T cells to enhance priming for interferon-γ production and to reverse interleukin-4 priming.24 The most potent cytokine for the induction of leishmanicidal activity in macrophages is interferon γ.12 The sustained production of interferon γ in response to infection is commonly associated with the development of specific Th1 responses. In mice infected with L major, 713
Review
treatment with interleukin 12 increases interferon-γ production, reduces the severity of disease, potentiates vaccine-derived immunity, and suppresses interleukin-10 production.25 However, exposure of macrophages from susceptible mice to opsonised leishmania promastigotes suppresses the expression of interleukin 12.26 In bone-marrow-derived macrophages from BALB/c mice, the production of interleukin 12 in response to lipopolysaccharide is suppressed after ligation of Fc, complement, or scavenger receptors.27 Both mRNA synthesis and protein secretion are diminished to almost undetectable levels after receptor ligation. TNF production is not inhibited, so suppression is specific to interleukin 12. Additionally, the ligation of mouse Fcγ receptor with immune complexes increases the production of interleukin 10.28 Stimulation of mouse bone-marrow macrophages by lipopolysaccharide results in some interleukin-10 production, but the addition of red blood cells opsonised with IgG antibodies substantially increases interleukin-10 production. Immune complexes not only induce activated macrophages to produce interleukin 10, but they also induce both macrophages and dendritic cells to stop production of interleukin 12.27,29 Interleukins 10 and 12 in macrophages are modulated by different mechanisms.30 The abrogation of interleukin-12 biosynthesis is a property shared by ligation of several macrophage receptors, the induction of interleukin 10 is specific to Fcγ receptors. Additional evidence for the role of interleukin 10 in promoting chronic infection is provided by the observation that normal BALB/c mice develop progressive non-healing lesions with many L major parasites, whereas BALB/c mice that do not express interleukin 10 control disease progression and have small lesions with 1000-times fewer parasites at the 5th week of infection.20 Furthermore, in established L donovani visceral infection in wild-type mice, treatment with monoclonal antibodies for either interleukin 10 or interleukin-10 receptor successfully induces intracellular parasite killing within liver macrophages.31 These findings and those from other studies32,33 indicate that amastigotes of leishmania exploit an unusual and unexpected virulence factor: host IgG. When the surface of Leishmania amastigotes are coated with IgG, the resultant immune complexes allow them to ligate Fcγ receptors on inflammatory macrophages, preferentially inducing the production of high amounts of interleukin 10.16 This induction of interleukin 10 by the IgG-amastigotes complex did not occur in macrophages derived from mice that did not have the common γ chain that signals through Fcγ receptors 1, 3, and 4, indicating that one or all of these three receptors were involved.30 Results from subsequent studies that used defined immune complexes indicated that all three of the Fcγ receptors that signal through the gamma chain were capable of signalling for interleukin-10 production in macrophages.34 Therefore, in some settings, IgG itself seems to bias the immune response towards a Th2-type response. For some species of Leishmania, the 714
persistence of infection depends on whether amastigotes are coated with IgG.23,35–37 Production of interleukin 10 by ligation of Fcγ receptors is a generic process that does not require a non-microbial antigen.38 When exposed to ovalbumin alone, lipopolysaccharide-treated BALB/c mouse macrophages develop Th1-biased T-cell responses, characterised by the production of interferon γ.38 When the same antigen is complexed with IgG anti-ovalbumin, Th2 responses predominate with production of interleukin 4. This Th2 phenotype is stable and retained when T cells are subsequently stimulated again under non-biasing conditions. Mice vaccinated with IgG-opsonised ovalbumin produce high titres of IgG antibody of the IgG1 isotype.38 The T-cell biasing and its reversal via Fcγ receptor ligation is also seen in vivo.38 In macrophages from knockout mice, the production of interferon γ by T cells is controlled by the macrophage cytokine interleukin 12, and the production of interleukin 4 by interleukin 10. These findings show that the ligation of Fc receptors on activated macrophages reverses Th1 biasing that accompanies innate immune responses to microbial products. In patients with visceral leishmaniasis, high titres of leishmanial antibodies correlate with peak parasitaemia and with negative delayed-type hypersensitivity responses.23 Successful treatment of leishmaniasis with amphotericin B results in decreased antibody titres and restoration of delayed-type hypersensitivity responses. Earlier observations identified polyclonal B-cell activation and high concentrations of immune complexes, as well as rheumatoid factor in patients with visceral leishmaniasis.39 In a small series, individuals infected with L donovani were more likely to develop rheumatoid arthritis than were noninfected controls.40 In experimental models of visceral leishmaniasis, infected hamsters develop immune complex glomerulonephritis.41 In addition to L donovoni and L major, humoral immune responses against L mexicana were not effective at killing organisms hiding in parasitophorous vacuoles because host IgG-coated amastigotes generated immunosuppressive interleukin-10 responses by infected macrophages.32,33 Although perhaps caused by a different mechanism, a similar process might be involved in JhD mice infected with L amazonensis: in one study,42 infections were minimised in the absence of B cells or antibodies; when these immune elements were restored, lesions progressed by a process thought to involve CD4 T cells. How do antibody-coated amastigotes result in the production of interleukin 10 by macrophages? Ligation of macrophage Fcγ receptors produces a rapid and enhanced activation of two mitogen-activated protein kinases (MAPKs): extracellular signal-regulated kinase (ERK) and p38.43 The activation of ERK leads to the phosphorylation of serine 10 on histone H3 on the gene encoding interleukin 10, increasing the accessibility of the promoter to transcription factors generated in response to p38 activation (figure 1). Activation of both MAPKs is needed for interleukin-10 synthesis. In addition to ERK activation, www.thelancet.com/infection Vol 10 October 2010
Review
A
B
TLR
Fcγ receptor
MyD88
γ chain
IRAK
Syk
TRAF6
PKC?
IκK complex NEMO
MEK1/2
NFκB
ERK
p38
Activate transcription factors
Phosphorylate the chromatin on gene that encodes interleukin 10
Figure 1: Pathways by which toll-like-receptor signals and ligation of Fcγ receptor lead to the production of interleukin 10 (A) Toll-like receptors (TLRs) signal through the adaptor molecule myeloid differentiation primary response gene (MyD88) to activate a kinase cascade, resulting in the activation of nuclear factor κB (NFκB) and p38. This pathway leads to the activation of transcription factors that have the potential to bind to the interleukin 10 promoter. In resting cells, this promoter is inaccessible to these transcription factors because it is tightly packed in chromatin. (B) The ligation of Fcγ receptors signals through a different protein kinase, Syk, to activate extracellular signal-regulated kinase (ERK). This pathway leads to the phosphorylation of chromatin, making the promoter accessible to transcription factors. IRAK= interleukin-1 receptor associated kinase. TRAF6=TNF receptorassociated factor 6. IκK=IκB kinase. NEMO=NFκB essential modulator. PKC=protein kinase C.
an inflammatory stimulus, such as low-molecular-weight hyaluronic acid from the extracellular matrix, is needed. The combination of these two signals results in the superinduction of interleukin 10.44 Macrophages that do not have Fcγ receptors, or macrophages treated with an inhibitor of spleen tyrosine kinase (which is activated after Fcγ-receptor ligation) do not activate ERK and fail to produce interleukin 10 after infection with Leishmania species amastigotes.44
The role of Fc receptors or immune complexes has not been studied exhaustively for most of these organisms. Although studies on resistance to tuberculosis have focused on T-cell immunity, C57BL/6 mice deficient in inhibitory Fcγ receptor IIB have improved bacterial control and diminished pathological changes at 30 days but not 20 days after aerosol challenge with M tuberculosis.46 Mice deficient in Fcγ receptor IIB have increased production of the p40 subunit of interleukin 12. Interleukins 12 and 23 have the p40 subunit in common and both promote polarisation of naive CD4 T cells into Th1 effectors.47 Treatment of human macrophages with exogenous interleukin 12 combined with blockade of interleukin 27 reduced the burden of mycobacterial infection.48 These infections were characterised by large protective interferon-γ responses that coincided with increased activation of macrophages. In human beings, high concentrations of interleukin 10 are related to the suppression of host defence mechanisms and exacerbation of infection.49 In a model of reactivation tuberculosis, the presence of macrophage-derived interleukin 10 in the lungs of infected transgenic mice enables Th1 cells to efficiently express effector functions and secrete sufficient interferon γ to induce activation of macrophages by the classic pathway, characterised by the expression of inducible nitric oxide synthase (NOS2) and interferon-γ inducible protein member 1 (IRGM1; formerly LRG-47).50 However, mycobacteria survived and successfully proliferated within mouse macrophages with interleukin-10 production under control of a human CD68 promoter. Macrophage-derived interleukin 10 seems to override interferon-γ-dependent activation of macrophages by the classic pathway and other effector mechanisms against M tuberculosis by inducing an alternatively activated phenotype. So far, the specific contribution of M tuberculosis– IgG antibody complexes towards modulation of infections in model systems is unknown.
Viruses Other intracellular parasites and bacteria Many bacteria replicate partly or solely in human macrophages. Many organisms that do infect macrophages produce chronic infections. One of the criteria of successful parasitism by microorganisms that produce systemic infections might be the ability to evade microbicidal mechanisms of macrophages. These defensive mechanisms have been much studied.45 Chronic infections themselves or successive infection by microorganisms with similar antigenic structures might contribute to the formation of pathogenic IgG immune complexes that control microbial survival. Infections that might be analogous to the leishmania model include: Mycobacterium tuberculosis, Mycobacterium leprae, Legionella pneumophila, Listeria monocytogenes, Brucella spp, Salmonella spp, Shigella spp, Coxiella burnetii, Anaplasma phagocytophilum, Ehrlichia chaffeensis, another protozoan, Toxoplasma gondii, and fungi (eg, Histoplasma capsulatum). www.thelancet.com/infection Vol 10 October 2010
The ADE mechanism has attracted wide interest in virology because many viruses replicate in macrophages in vivo and cause enhanced infections or disease.51 Not all severe viral infections have been irrefutably linked to antibodies. Enhanced infections need an initial immunological event, termed sensitisation, and viral infections can be categorised into two groups. Some viruses cause immune responses that sensitise hosts as a result of sequential infection by more than one antigenic type or as a result of the rapid evolution of antigenic diversity during the course of a chronic infection (table 1). Hosts can be sensitised by other viruses that are given as killed vaccines (table 2). Respiratory syncytial and measles viruses infect respiratory epithelial cells in vivo. and the enhanced infections observed in individuals given killed viral vaccines are not caused by ADE but have been attributed to poor antibody quality resulting from aberrant antigen presentation.71 715
Review
During initial ADE research, increased yields of virus were assumed to result from higher numbers of cells infected in the presence of antibodies than in the absence of antibodies.51 This increased infection was shown to be mediated by Fcγ receptors.5,72 For example, more West Nile virus particles were shown to be attached to the surface of mouse macrophages as immune complexes than as naked virus.73,74 In feline infectious peritonitis virus, more peritoneal macrophages are infected in vitro in the presence of antibody than in the absence of antibody.75 Alternatively, immune complexes might be internalised more rapidly than virus alone. With HIV-1, viral replication is initiated and progeny viruses released sooner from cells pretreated with HIV-specific antibodies than in cells exposed to untreated virus.76 Protein and RNA synthesis are increased in cells infected with HIV-1 immune complexes than with virus alone.
Ross River virus Acute infections with Ross River virus (RRV) often result in post-infection arthritis and arthralgia for many months. Synovial cells typically stain for RRV antigens, whereas synovial fluid often contains interferon γ and TNF. Attempts have been made to model this finding by establishing chronic RRV infections in mouse macrophage lines.77 Infection in these cells might be a model of how In-vitro ADE
In-vivo enhanced disease
RRV persists in synovial tissues and produces arthritis and arthralgia.78 The incubation of RRV with diluted RRV antiserum results in increased infection in these cells.79 This finding has been confirmed in a human monocytic cell line, Mono mac 6, and in primary human monocytes and macrophages.79 RRV-infected RAW 264.7, continuous mouse macrophage cells, produced antiviral TNF when treated with lipopolysaccharide but, when these cells are infected via the ADE pathway, TNF production is abrogated and RRV infection greatly enhanced.79 Increased ADE production of virus was correlated with a downregulation of nitric oxide production and an inhibition of the transcription of genes for TNF, NOS2, interferon regulatory factor 1 (IRF-1), and nuclear factor κB (NFκB). ADE-mediated infections inhibit lipopolysaccharideinduced reactive nitrogen intermediates and TNF gene transcription and protein synthesis, but do not decrease general protein synthesis or increase transcription of non-antiviral control genes. In one study, ADE-mediated infections in lipopolysaccharide-treated RAW 264.7 cells resulted in a substantial increase in RRV-infected cells compared with virus-only controls at and after 12 h.80 Although 14·2% of ADE-infected cells stained for RRV antigen by immunofluorescence microscopy, a greater percentage of cells were positive by PCR and electron microscopy. RRV ADE infection suppresses signals for interferon-inducible protein 10 (IP-10), NOS2, IRF-1, TNF, and interferon β, but not for Sp1 (a human genetic In-vitro ADE In-vivo enhanced disease
RNA virus group RNA virus group
Picornaviridae +
Coxsackie B52
+ (animal model)
+
+ (mouse model)
Respiratory syncytial virus65
+
+
Measles66
+
+
+
+ (accelerated disease onset)
Influenza A64
Flaviviridae Dengue4
+
+
Lactate dehydrogenase53
+
–
Coronaviridae
Paramyxoviridae
Feline infectious peritonitis
+
+
Rhabdoviridae
PRRSV56
+
+
Rabies67,68
+
+
+
?
+
+
+
+
54,55
Simian haemorrhagic fever
51
Retroviridae HIV57 Caprine arthritis
58
Equine infectious anaemia
59
DNA virus group Parvoviridae Aleutian disease of mink60–62
+
+
Asfarviridae African swine fever
716
Orthomyxoviridae
63
··
··
Coronaviridae Feline infectious peritonitis54 +
+
+
+
Simian haemorrhagic fever69 +
+
PRRSV56 Retroviridae HIV70
+
?
Equine infectious anaemia59
+
+
Caprine arthritis58
+
+
+
+
DNA virus group Parvoviridae
Sensitisation results from sequential or chronic infection with several naturally occurring antigenic types. Evidence of in-vivo enhancement of dengue and HIV effects is from human beings; the other evidence is from domestic or wild animals. ADE=antibody-dependent enhancement of infection. PRRSV=porcine reproductive and respiratory syndrome virus. +=positive effect. –=no effect. ?=unknown effect. ··=no data.
Evidence of in-vivo enhancement of respiratory syncytial virus, measles, rabies effects is from human beings; the other evidence is from domestic or wild animals. PRRSV=porcine reproductive and respiratory syndrome. +=positive effect. ?=unknown effect.
Table 1: Viruses that cause in-vitro antibody-dependent enhancement of infection or antibody-enhanced disease
Table 2: Viral diseases that are enhanced after sensitisation of hosts by vaccines
Aleutian disease of mink60
www.thelancet.com/infection Vol 10 October 2010
Review
transcription factor that regulates immune responses). A substantial increase in interleukin-10 gene transcription and protein production in these cells was reported. Crucially, the inhibition of antiviral factor transcription requires infectious virus in immune IgG complexes since the ligation of Fcγ receptors by zymosan–antibody complexes does not ablate antiviral transcription.81
Dengue virus Dengue viruses are a group of four closely related members of the Flavivirus genus that share 60–70% genetic homology and are inoculated by the bite of infected Aedes aegypti. Initial infections raise crossreactive non-neutralising antibodies that can enhance an infection with a different virus type.82 Data from in-vitro studies in Fc receptor-bearing cells indicate that any specificity of monoclonal dengue antibodies might form infectious immune complexes, the major requirement being antibody concentration below that needed for neutralisation.83,84 In practice, antibodies directed at surface epitopes not involved in virus entry efficiently produce ADE.83,85 The fact that infants might acquire severe dengue disease during a first dengue infection when placentally acquired maternal polyclonal dengue antibodies have diminished to below protective concentrations is a unique example of ADE.86 Results from studies of human tissues have established that monocytes, macrophages, and immature and mature dendritic cells are major targets for dengue virus infection.87,88 In human beings, secondary dengue infections follow a stereotypical course with severe outcomes, such as shock or gastrointestinal haemorrhage, accompanying vascular collapse that results from capillary permeability around the time of defervescence.89 Indirect evidence suggests that cytokines mediate vascular permeability in patients with dengue infection. Much work has focused on the measurement of blood cytokine concentrations in patients late in the acute phase, just before onset of shock.90 Elevated viraemia and high concentrations of proinflammatory and immunomodulatory cytokines, including interleukin 10, are associated with severe disease.91 During in-vitro ADE infection of THP-1 cells (a human monocytic Fcγ receptor-bearing continuous cell line), intracellular dengue virus production is increased as a result of idiosyncratic Fcγ-receptor signalling.92 Infections with polyclonal antibodies plus dengue virus 2 result in a suppression of innate responses. After ligation of Fcγ receptor I and Fcγ receptor IIA, entry of infectious immune complexes activates expression of the negative regulators dihydroxyacetone kinase and the autophagy proteins Atg5–Atg12 of retinoic acid-inducible protein I/melanoma differentiation-associated gene 5 (RIG-I/MDA5), resulting in downregulation of the RIG-I/MDA5 signalling pathway and decreased production of type I interferon and interferon-activated antiviral molecules.93 After ligation of Fc receptors, www.thelancet.com/infection Vol 10 October 2010
dengue virus infection activates interleukin-10 production at an early phase of infection.92 The suppressive role of interleukin 10 during ADE infection has been confirmed and not only induces Th2 biasing but also inhibits a potent suppressor of the JAK–STAT signalling pathway via the suppressor of cytokine signalling (SOCS) system.93 As a result of these two suppression pathways, ADEinfected THP-1 cells secrete low concentrations of type I interferon, thus suppressing transcription and translation of interleukin 12, interferon γ, and TNF and facilitating expression and synthesis of anti-inflammatory cytokines. This pathway of infection also suppressed an innate dengue virus mediator, nitric oxide radicals, by disrupting the transcription of the NOS2 gene transcription factor, IRF-1. Thus, ADE infection not only facilitates virus entry (extrinsic ADE), but also modifies innate and adaptive intracellular antiviral mechanisms and enhances replication (intrinsic ADE; figure 2). Human in-vivo correlations can be made with these invitro responses. During the acute severe illness stage of secondary infections, increased production of interleukin 10 is accompanied by downregulation of several interferon regulatory genes.94–96 Genome-wide transcriptomes from PBMCs collected during the acute phase of mild dengue fever and dengue haemorrhagic fever in children were compared by use of microarray analysis.97 Compared with patients with milder illness, those with dengue haemorrhagic fever during secondary infections had low concentrations of nitric oxide, low interferon transcript in PBMCs, and high interleukin-10 blood concentrations. A cDNA array indicated that 47% of immune response genes strongly upregulated in PBMCs from children with mild illness were interferon-inducible and interferon-induced genes. The robust upregulation of type I interferon genes in PBMCs from children with mild illnesses was accompanied by increased plasma interferon-α concentrations. Interferon gene upregulation and production were substantially increased in patients with mild dengue illness compared with patients with severe dengue illness. The notable abundance of interferon and interferoninducible factors in patients with mild disease is consistent with a protective role of interferon in the control of dengue infection, as indicated in a mouse model and suggested for human beings with dengue fever.98–100 Patients who survive dengue haemorrhagic fever have higher concentrations of circulating interferon β than do those who die.94,95 Patients with dengue fever have lower peak viraemia titres than do those with dengue haemorrhagic fever.101,102 A role for another interferon-related mediator that reduces viral load in dengue fever has been suggested. Production of human IP-10 was upregulated in the PBMCs of patients with dengue fever.93 In a mouse model, IP-10 protected against dengue infection by serving as a chemoattractant for natural killer and T cells and by blocking interactions between dengue and its putative receptor, heparan sulphate.103,104 717
Review
Wound-healing macrophages readily support infection with all four dengue viruses in vitro.105 Human monocytes, activated macrophages, and mature dendritic cells express ADE, whereas immature dendritic cells do not (Marovich MA, unpublished). In macrophages, dengue virus 2 infection produced high concentrations of type I interferons (α/β) but these were downregulated under ADE conditions and replaced with the secretion of proinflammatory cytokines (interleukin 6 and TNFα) at peak enhancement titres. Mature dendritic cells secreted interferon β coincident with peak ADE. Immature dendritic cells are readily infected by dengue virus alone, a process efficiently mediated by the attachment molecule dendritic cell-specific intercellular adhesion molecule-3grabbing nonintegrin (DC-SIGN).106 Monocytes secreted interleukin 10 during ADE infection at the same serum dilution that produced peak virus (Marovich MA, unpublished). Production of interleukin-10 protein was controlled by interleukin-10 promoter polymorphisms. Three single nucleotide polymorphisms at the 5' flanking region of interleukin 10 (position –1082 A/G, –819 C/T, and –592 C/A) were studied. Dengue infection rates at enhancement titres were remarkably consistent in primary monocytes (10–15%) from 20 different donors; homozygous GCC haplotype was associated with high interleukin-10 secretion, whereas donors with the ACC haplotype produced an intermediate concentration of interleukin 10, and the ATA haplotype secreted the lowest interleukin-10 concentrations. The precise role of interleukin-10 production induced from immune complexes on the clinical events of dengue infections is not known. The reported shift from Th1 response to Th2 response in severe secondary dengue DENV infection
ADE infection Activation
DENV receptor
Suppression
Fc receptor
Block
PRR signalling
Interleukin 10
Th2 bias Type I interferon proinflammatory cytokine
SOCS
Figure 2: Intrinsic ADE of dengue virus infection in THP-1 human monocytic cells Attachment and entry of infectious virus–antibody complexes into Fc receptor-bearing cells results in increased production of virus by inhibition of type I interferon and production of proinflammatory cytokines. Additionally, interleukin-10 biosynthesis is activated, stimulating members of the suppressor of cytokine signalling (SOCS) family, which results in suppression of the Janus kinase–signal transducer and activator of transcription (JAK–STAT) signalling pathway and bias towards a Th2 response. By contrast, infection by naked virus via virus-specific receptor signals through pattern recognition receptor (PRRs) produces intracellular antiviral molecules. ADE=antibody-dependent enhancement. Th2=T helper 2. DENV=dengue virus.
718
illnesses could be attributed to interleukin 10.107 Use of a mouse model has indicated an ADE-related early vascular permeability that is mediated by TNF.108,109 The contribution of Th1 or Th2 responses to vascular permeability in these mice is unknown. In Asia and the Americas, preceding infection or immunisation with a non-dengue flavivirus—Japanese encephalitis in Asia and yellow fever or yellow fever vaccine in the Americas—does not predispose individuals to dengue haemorrhagic fever with subsequent dengue virus infection, although non-dengue flavivirus antibodies form infectious immune complexes with dengue viruses.110 In an in-vitro study with U937 cells, late convalescent sera from individuals who had had infections with Japanese encephalitis virus did not lead to ADE infections with dengue virus 2.111 Non-neutralising heterotypical flaviviral antibodies are generally non-neutralising and, therefore, attach to parts of the virion that are not crucial to neutralisation. Might the site of attachment of antibodies to virions control the outcome of intrinsic ADE?
Other viruses ADE of virus infection might be a disease-enhancing factor for several human and animal viral diseases. Attributes of these diseases have been summarised in tables 1 and 2. In this section, we briefly review three of these diseases in particular: feline infectious peritonitis, Aleutian disease of mink, and porcine reproductive and respiratory syndrome (PRRS). Feline infectious peritonitis is caused by a coronavirus that replicates in macrophages and leads to peritonitis and occasionally a fatal pyogranulomatous disease in kittens and cats with ADE incriminated as a disease-enhancing factor.54,112 Cats with active disease or who have acquired passive maternal antibodies develop a more rapid and fulminant disease after challenge with this virus than do seronegative cats. Disease enhancement has been reported in kittens that had vaccine-derived humoral immunity directed against the spike protein of feline infectious peritonitis virus.113 Similarly, kittens immunised with a recombinant vaccinia virus expressing the spike protein of this virus die earlier than do control animals. Feline infectious peritonitis virus derives from chronic infections with feline coronaviruses, during the course of which the virus mutates to present new antigens to the host.114 These kittens die earlier than do control animals. Aleutian disease of mink is caused by a parvovirus that is pathogenic for mink that replicates in macrophages and circulates in the blood mainly as fully infectious immune complexes, both in vivo and in vitro.115 Deposition of these soluble immune complexes on renal glomerular membranes or walls of capillary blood vessels leads to fatal glomerulonephritis. ADE might be an important contributing factor in the pathogenesis of this disease. Aleutian disease virus replicates in macrophages, and infected mink produce large amounts of non-neutralising antibodies.116 These antibodies subsequently produce www.thelancet.com/infection Vol 10 October 2010
Review
infectious immune complexes, which lead to increased infection of macrophages. Passive transfer of virus-specific antibody at the peak of viral replication results in foci of necrosis around virus-infected cells, a reaction thought to be due to complement-mediated cytolysis.60 Infectious immune complexes result in increased infection of mink peritoneal macrophages.61 In one study, inoculation of an experimental killed Aleutian disease virus vaccine did not produce detectable neutralising antibodies, although eight of ten of vaccinated mink, but no control animals, developed the disease on challenge with live virus.60 PRRS is caused by a member of the Arterivirus genus of the coronavirus family that was first isolated in the Netherlands and classified in 1991. The syndrome was recognised in the USA in the mid-1980s and called mystery swine disease and has also been called blue ear disease. There are two prototype strains of PRRS virus (European and North American) that cause similar clinical symptoms but are distinct viral genotypes whose genomes diverge by about 40%.56,117,118 The genetic variation among the viruses varies geographically, increasing the difficulty of vaccine development.119 For reasons not understood, disease presentation varies between one herd and another. For every three herds exposed to PRRS, for the first exposure, one herd will show no recognisable disease, the second might show mild disease, and the third will show moderate to severe disease: the better the health status of the herd, the less severe the disease.117 Evidence exists that PRRS virus mutates as it multiplies, leading to some strains that are highly virulent and some that are not.117,118 PRRS virus has a particular affinity for macrophages, particularly those in the lungs.56,120,121 The virus multiplies in macrophages and kills these cells. PRRS virus infections are often chronic and, once a herd is infected, the virus tends to remain present and active indefinitely.56
Conclusions Most of the macrophage-tropic organisms discussed in this Review have evolved various offensive mechanisms. In some cases, these mechanisms are sufficient to reduce host defence mechanisms but, in other cases, Fcγ receptor-mediated intrinsic ADE might also result in either severe or sustained infections. In cases when host T-cell responses quickly predominate, effective control of infection with robust resistance to reinfection is achieved. Why there are these differential outcomes is not clear. Complexity can characterise interactions of infectious pathogens with antibodies. For example, when Mahalingam and Lidbury80 added an irrelevant non-infectious immune complex to mouse macrophage cell line and then infected these cells with RRV, ADE did not occur—despite reports that Th2 responses after ligation of Fcγ receptors by IgG immune complexes were a generic process.38 ADE can lead to a Th2-biased proliferation of lymphocytes. The specific outcome of ligation of each human Fcγ receptor by immune complexes needs careful study. Many years ago, an antibody-like molecule was recognised www.thelancet.com/infection Vol 10 October 2010
to attach to primary monocytes obtained from dengueimmune donors.5,51,72 From our present understanding of human Fcγ receptors, could cytophilic antibodies attach to high-affinity Fcγ receptor I?122 An interesting question is whether cytophilic or non-cytophilic immune complexes, or both, mediate enhanced and severe dengue disease. Furthermore, does intrinsic ADE result from interactions of immune complexes with Fcγ receptor IIb, the inhibitory Fc receptor? Several lines of evidence suggest that monocyte/ macrophage interactions with infectious virus or infectious virus–antibody complexes differ at an early stage of infection. The mutated form of the Fc portion of the polyvalent dengue virus monoclonal antibody E60-hIgG1-N297Q, which abolished Fc-receptor binding, had neutralising activity instead of enhancing activity.109 Moreover, not only the functional Fc portion is needed but the complete cytoplasmic tail of Fc receptor is essential for dengue virus–antibody complexes to promote enhancing activity, as indicated by the fact that disruption of the immune tyrosine activating motif (ITAM) or removal of the sequence between the two ITAM regions abolishes ADE activity.123 However, the innate immune responses cannot be suppressed by simply cross-linking Fc receptors with anti-CD32 or anti-CD64 antibodies (Ubol S, unpublished). Hence, the enhancing effect of virus–antibody complexes might start once the Fc portion is engaged with the Fc receptor, which in turn switches on the negative regulatory innate immune response. These events do not occur during the entry of naked virus. Intrinsic ADE is linked to the immunobiology of interleukin 10, a complex type II cytokine that has a key role in many infectious processes.16 Results from in-vitro experiments suggest that early production of interleukin 10 in dengue infections might have an important role in promoting ADE, whereas interleukin 10 produced in the latter half of secondary dengue infections might be immunoregulatory. Careful studies that follow up individuals through the course of primary mild and secondary severe dengue infections are needed. Events that simply accompany secondary immune responses should be distinguished from those responses that might contribute to immunopathological changes.124 Several questions concern the essential and nonessential components of intrinsic ADE. Where, exactly, does interleukin-10 production fit into this process? Does interleukin-10 production lead to increased infection in the initial infected cell, or does interleukin 10 mainly function on bystander cells? Are there functional differences between interleukin 10 of T-cell and macrophage origin? Another profound effect of interleukin 10 is to inhibit the production of proinflammatory cytokines and mediators from macrophages and dendritic cells. The major inflammatory cytokines, interleukins 1, 6, and 12, and TNF, are all substantially repressed after exposure to interleukin 10. 719
Review
Search strategy and selection criteria References were identified from searches of PubMed from January, 1960, to August, 2010, for articles on in-vitro studies of all macrophage-tropic microorganisms. Searches included crossreferences to “macrophages”, “macrophage biology”, “immune complexes”, “IgG gamma receptors”, “IL-10”, “interferon”, and “toll-like receptors”. Additional references from the authors’ own files were also identified. Authors of several papers that were difficult to access were directly contacted.
14
15
16 17
18 19
Many, perhaps most, of the fundamental characteristics of intrinsic ADE are unknown. Studies into intrinsic ADE should expand our understanding of the mechanisms of pathogenesis for a broad range of infectious diseases and open many approaches for improvements to the treatment and prevention of such diseases. Contributors SBH was involved in the literature review, data interpretation, and writing and editing of the paper. SM was involved in writing the paper and providing intellectual input. MAM was involved with data analysis, writing the paper, and preparing the figures. SU was involved in writing the paper and preparing the figures. DMM was involved in data collection and interpretation and development of the figures. Conflicts of interest All authors declare that they have no conflicts of interest. References 1 Hawkes RA. Enhancement of the infectivity of arboviruses by specific antisera produced in domestic fowls. Aust J Exp Biol Med Sci 1964; 42: 465–82. 2 Hawkes RA, Lafferty KJ. The enhancement of virus infectivity by antibody. Virology 1967; 33: 250–61. 3 Halstead SB, Nimmannitya S, Yamarat C, Russell PK. Hemorrhagic fever in Thailand; recent knowledge regarding etiology. Jpn J Med Sci Biol 1967; 20: 96–103. 4 Halstead SB, Chow J, Marchette NJ. Immunologic enhancement of dengue virus replication. Nat New Biol 1973; 243: 24–26. 5 Halstead SB, O’Rourke EJ. Dengue viruses and mononuclear phagocytes. I. Infection enhancement by non-neutralizing antibody. J Exp Med 1977; 146: 201–17. 6 Kliks S, Halstead SB. An explanation for enhanced virus plaque formation in chick embryo cells. Nature 1980; 285: 504–05. 7 Kliks S, Halstead SB. Role of antibodies and host cells in plaque enhancement of Murray Valley encephalitis virus. J Virol 1983; 46: 394–404. 8 Mosser DM, Edwards JP. Exploring the full spectrum of macrophage activation. Nat Rev Immunol 2008; 8: 958–69. 9 Gerber JS, Mosser DM. Reversing lipopolysaccharide toxicity by ligating the macrophage Fc gamma receptors. J Immunol 2001; 166: 6861–68. 10 Centers for Disease Control and Prevention, Division of Parasitic Diseases. Factsheet: Leishmania infection. http://www.cdc.gov/ ncidod/dpd/parasites/leishmania/factsht_leishmania.htm (accessed Aug 31, 2010). 11 Ghalib HW, Piuvezam MR, Skeiky YA, et al. Interleukin 10 production correlates with pathology in human Leishmania donovani infections. J Clin Invest 1993; 92: 324–29. 12 Murray HW, Masur H, Keithly JS. Cell-mediated immune response in experimental visceral leishmaniasis. I. Correlation between resistance to Leishmania donovani and lymphokine-generating capacity. J Immunol 1982; 129: 344–50. 13 Pearson RD, Romito R, Symes PH, Harcus JL. Interaction of Leishmania donovani promastigotes with human monocyte-derived macrophages: parasite entry, intracellular survival, and multiplication. Infect Immun 1981; 32: 1249–53.
720
20 21
22
23
24
25
26
27
28
29 30
31
32 33
34
35
36
37
Fiorentino DF, Bond MW, Mosmann TR. Two types of mouse T helper cell. IV. Th2 clones secrete a factor that inhibits cytokine production by Th1 clones. J Exp Med 1989; 170: 2081–95. Moore KW, Vieira P, Fiorentino DF, Trounstine ML, Khan TA, Mosmann TR. Homology of cytokine synthesis inhibitory factor (IL-10) to the Epstein-Barr virus gene BCRFI. Science 1990; 248: 1230–34. Mosser DM, Zhang X. Interleukin-10: new perspectives on an old cytokine. Immunol Rev 2008; 226: 205–18. Rousset F, Garcia E, Defrance T, et al. Interleukin 10 is a potent growth and differentiation factor for activated human B lymphocytes. Proc Natl Acad Sci USA 1992; 89: 1890–93. O’Garra A, Vieira P. T(H)1 cells control themselves by producing interleukin-10. Nat Rev Immunol 2007; 7: 425–28. Heinzel FP, Sadick MD, Mutha SS, Locksley RM. Production of interferon gamma, interleukin 2, interleukin 4, and interleukin 10 by CD4+ lymphocytes in vivo during healing and progressive murine leishmaniasis. Proc Natl Acad Sci USA 1991; 88: 7011–15. Kane MM, Mosser DM. The role of IL-10 in promoting disease progression in leishmaniasis. J Immunol 2001; 166: 1141–47. Karp CL, el-Safi SH, Wynn TA, et al. In vivo cytokine profiles in patients with kala-azar. Marked elevation of both interleukin-10 and interferon-gamma. J Clin Invest 1993; 91: 1644–48. Vieth M, Will A, Schroppel K, Rollinghoff M, Gessner A. Interleukin-10 inhibits antimicrobial activity against Leishmania major in murine macrophages. Scand J Immunol 1994; 40: 403–09. Miles SA, Conrad SM, Alves RG, Jeronimo SM, Mosser DM. A role for IgG immune complexes during infection with the intracellular pathogen Leishmania. J Exp Med 2005; 201: 747–54. Seder RA, Gazzinelli R, Sher A, Paul WE. Interleukin 12 acts directly on CD4+ T cells to enhance priming for interferon gamma production and diminishes interleukin 4 inhibition of such priming. Proc Natl Acad Sci USA 1993; 90: 10188–92. Heinzel FP, Schoenhaut DS, Rerko RM, Rosser LE, Gately MK. Recombinant interleukin 12 cures mice infected with Leishmania major. J Exp Med 1993; 177: 1505–09. Carrera L, Gazzinelli RT, Badolato R, et al. Leishmania promastigotes selectively inhibit interleukin 12 induction in bone marrow-derived macrophages from susceptible and resistant mice. J Exp Med 1996; 183: 515–26. Sutterwala FS, Noel GJ, Clynes R, Mosser DM. Selective suppression of interleukin-12 induction after macrophage receptor ligation. J Exp Med 1997; 185: 1977–85. Sutterwala FS, Noel GJ, Salgame P, Mosser DM. Reversal of proinflammatory responses by ligating the macrophage Fcgamma receptor type I. J Exp Med 1998; 188: 217–22. Anderson CF, Lucas M, Gutierrez-Kobeh L, Field AE, Mosser DM. T cell biasing by activated dendritic cells. J Immunol 2004; 173: 955–61. Gerber JS, Mosser DM. Stimulatory and inhibitory signals originating from the macrophage Fcgamma receptors. Microbes Infect 2001; 3: 131–39. Murray HW, Moreira AL, Lu CM, et al. Determinants of response to interleukin-10 receptor blockade immunotherapy in experimental visceral leishmaniasis. J Infect Dis 2003; 188: 458–64. Buxbaum LU. A detrimental role for IgG and FcgammaR in Leishmania mexicana infection. Immunol Res 2008; 42: 197–209. Buxbaum LU, Scott P. Interleukin 10- and Fcgamma receptor-deficient mice resolve Leishmania mexicana lesions. Infect Immun 2005; 73: 2101–08. Ioan-Facsinay A, de Kimpe SJ, Hellwig SM, et al. FcgammaRI (CD64) contributes substantially to severity of arthritis, hypersensitivity responses, and protection from bacterial infection. Immunity 2002; 16: 391–402. Kima PE, Constant SL, Hannum L, et al. Internalization of Leishmania mexicana complex amastigotes via the Fc receptor is required to sustain infection in murine cutaneous leishmaniasis. J Exp Med 2000; 191: 1063–68. Padigel UM, Farrell JP. Control of infection with Leishmania major in susceptible BALB/c mice lacking the common gamma-chain for FcR is associated with reduced production of IL-10 and TGF-beta by parasitized cells. J Immunol 2005; 174: 6340–45. Mosser DM. The many faces of macrophage activation. J Leukoc Biol 2003; 73: 209–12.
www.thelancet.com/infection Vol 10 October 2010
Review
38
39
40 41
42
43
44
45
46
47
48
49
50
51 52
53
54
55
56
57 58
59
60
Anderson CF, Gerber JS, Mosser DM. Modulating macrophage function with IgG immune complexes. J Endotoxin Res 2002; 8: 477–81. Pearson RD, Naidu TG, Young AC, de Alencar JE, Romito R, Davis JS. Circulating immune complexes and rheumatoid factors in visceral leishmaniasis. J Infect Dis 1983; 147: 1102. Salama MM. Rheumatoid factor among several cases of visceral leishmaniasis. J Egypt Soc Parasitol 1990; 20: 837–39. Oliveira AV, Rossi MA, Rogue-Barreira MC, et al. The potential role of Leishmania antigens and immunoglobulins in the pathogenesis of glomerular lesions of hamsters infected with Leishmania donovani. Ann Trop Med Parasitol 1985; 79: 539–43. Wanasen N, Xin L, Soong L. Pathogenic role of B cells and antibodies in murine Leishmania amazonensis infection. Int J Parasitol 2008; 38: 417–29. Lucas M, Zhang X, Prasanna V, Mosser DM. ERK activation following macrophage FcgammaR ligation leads to chromatin modifications at the IL-10 locus. J Immunol 2005; 175: 469–77. Yang Z, Mosser DM, Zhang X. Activation of the MAPK, ERK, following Leishmania amazonensis infection of macrophages. J Immunol 2007; 178: 1077–85. Flannagan RS, Cosio G, Grinstein S. Antimicrobial mechanisms of phagocytes and bacterial evasion strategies. Nat Rev Microbiol 2009; 7: 355–66. Maglione PJ, Xu J, Casadevall A, Chan J. Fc gamma receptors regulate immune activation and susceptibility during Mycobacterium tuberculosis infection. J Immunol 2008; 180: 3329–38. Watford WT, Moriguchi M, Morinobu A, O’Shea JJ. The biology of IL-12: coordinating innate and adaptive immune responses. Cytokine Growth Factor Rev 2003; 14: 361–68. Robinson CM, Nau GJ. Interleukin-12 and interleukin-27 regulate macrophage control of Mycobacterium tuberculosis. J Infect Dis 2008; 198: 359–66. Almeida AS, Lago PM, Boechat N, et al. Tuberculosis is associated with a down-modulatory lung immune response that impairs Th1-type immunity. J Immunol 2009; 183: 718–31. Schreiber T, Ehlers S, Heitmann L, et al. Autocrine IL-10 induces hallmarks of alternative activation in macrophages and suppresses antituberculosis effector mechanisms without compromising T cell immunity. J Immunol 2009; 183: 1301–12. Halstead SB. Immune enhancement of viral infection. Prog Allergy 1982; 31: 301–64. Sauter P, Hober D. Mechanisms and results of the antibody dependent enhancement of viral infections and the role in the pathogenesis of coxsackie virus B-induced diseases. Microbes Infect 2009; 11: 443–51. Stueckemann JA, Holth M, Swart WJ, et al. Replication of lactate dehydrogenase-elevating virus in macrophages. 2. Mechanism of persistent infection in mice and cell culture. J Gen Virol 1982; 59: 263–72. Weiss RC, Scott FW. Antibody-mediated enhancement of disease in feline infectious peritonitis: comparisons with dengue hemorrhagic fever. Comp Immune Microbiol Infect Dis 1981; 4: 175–88. Olsen CW. A review of feline infectious peritonitis virus: molecular biology, immunopathogenesis, clinical aspects, and vaccination. Vet Microbiol 1993; 36: 1–37. Yoon KJ, Wu LL, Zimmerman JJ, Hill HT, Platt KB. Antibodydependent enhancement (ADE) of porcine reproductive and respiratory syndrome virus (PRRSV) infection in pigs. Viral Immunol 1996; 9: 51–63. Burke DS. Human HIV vaccine trials: does antibody-dependent enhancement pose a genuine risk? Perspect Biol Med 1992; 35: 511–30. Mdurvwa EG, Ogunbiyi PO, Gakou HS, Reddy PG. Pathogenic mechanisms of caprine arthritis-encephalitis virus. Vet Res Commun 1994; 18: 483–90. Mealey RH, Leib SR, Littke MH, Wagner B, Horohov DW, McGuire TC. Viral load and clinical disease enhancement associated with a lentivirus cytotoxic T lymphocyte vaccine regimen. Vaccine 2009; 27: 2453–68. Porter AD, Larsen AE, Porter HG. The pathogenesis of Aleutian disease of mink. II Enhancement of tissue lesions following the administration of a killed virus vaccine or passive antibody. J Immunol 1972; 109: 1–7.
www.thelancet.com/infection Vol 10 October 2010
61
62
63
64
65
66
67
68 69 70
71
72
73
74
75
76
77
78 79
80
81
82 83 84
Kanno H, Wohlinbarger JB, Bloom ME. Aleutian mink disease parvovirus infection of mink macrophages and human macrophage cell line *937: demonstration of antibody-dependent enhancement of infection. J Virol 1993; 67: 7017–24. Best SM, Bloom ME. Aleutian mink disease parvovirus. In: Kerr JR, Cotmore SF, Bloom ME, Linden RM, Parrish CR, eds. The parvoviruses. London: Hodder Arnold, 2006: 457–71. Gomez-Villamandos JC, Carrasco L, Bautista MJ, et al. African swine fever and classical swine fever: a review of the pathogenesis. Dtsch Tierarztl Wochenschr 2003; 110: 165–69. Webster RG, Askonas BA. Cross-protection and cross reactive cytotoxic T cells induced by influenza virus vaccines in mice. Eur J Immunol 1980; 10: 396–401. Chanock RM, Parrott RH, Kapakian AZ. Possible role of immunological factors in pathogenesis of RS virus in lower respiratory tract disease. Perspec Virol 1968; 6: 125–35. Fulginiti FA, Eller JJ, Downie AW, Kempe CH. Altered reactivity to measles virus. Atypical measles in children previously immunized with inactivated measles virus vaccines. JAMA 1967; 202: 1075–80. King AA, Sands JJ, Porterfield JS. Antibody-mediated enhancement of rabies virus infection in a mouse macrophage cell line (P388D1). J Gen Virol 1984; 65: 1091–93. Prabhakar BS, Nathanson N. Acute rabies deaths mediated by antibody. Nature 1981; 290: 590–5991. Tirado SM, Yoon KJ. Antibody-dependent enhancement of virus infection and disease. Viral Immunol 2003; 16: 69–86. Burke DS. Human HIV vaccine trials: does antibody-dependent enhancement pose a genuine risk? Perspect Biol Med 1992; 35: 511–30. Delgado MF, Coviello S, Monsalvo AC, et al. Lack of antibody affinity maturation due to poor Toll-like receptor stimulation leads to enhanced respiratory syncytial virus disease. Nat Med 2009; 15: 34–41. Halstead SB, O’Rourke EJ, Allison AC. Dengue viruses and mononuclear phagocytes: II. Identity of blood and tissue leukocytes supporting in vitro infection. J Exp Med 1977; 146: 218–28. Gollins SW, Porterfield JS. Flavivirus infection enhancement in macrophages: radioactive and biological studies on the effect of antibody on virus fate. J Gen Virol 1984; 65: 1261–72. Gollins SW, Porterfield JS. Flavivirus infection enhancement in macrophages: an electron microscopic study of viral cellular entry. J Gen Virol 1985; 66: 1969–82. Olsen C, Scott F. Evaluation of antibody-dependent enhancement of feline infectious peritonitis virus infectivity using in situ hybridization. Microb Pathog 1993; 14: 275–85. Robinson WE, Montefiori DC, Gillespie DH, Mitchell WM. Complement-mediated, antibody-dependent enhancement of HIV-1 infection in vitro is characterized by increased protein and RNA syntheses and infectious virus release. J Acquir Immune Defic Syndr 1989; 2: 33–42. Linn ML, Aaskov JG, Suhrbier A. Antibody-dependent enhancement and persistence in macrophages of an arbovirus associated with arthritis. J Gen Virol 1996; 77: 407–11. Suhrbier A, Mahalingam S. The immunobiology of viral arthritides. Pharmacol Ther 2009; 124: 301–08. Lidbury BA, Mahalingam S. Specific ablation of antiviral gene expression in macrophages by antibody-dependent enhancement of Ross River virus infection. J Virol 2000; 74: 8376–81. Mahalingam S, Lidbury BA. Suppression of lipopolysaccharideinduced antiviral transcription factor (STAT-1 and NF-kB) complexes by antibody-dependent enhancement of macrophage infection by Ross River virus. Proc Natl Acad Sci USA 2002; 99: 13819–24. Suhrbier A, La Linn M. Suppression of antiviral responses by antibody-dependent enhancement of macrophage infection. Trends Immunol 2003; 24: 165–68. Halstead SB, O’Rourke EJ. Antibody-enhanced dengue virus infection in primate leukocytes. Nature 1977; 265: 739–41. Halstead SB. Neutralization and antibody dependent enhancement of dengue viruses. Adv Virus Research 2003; 60: 421–67. Halstead SB, Venkateshan CN, Gentry MK, Larsen LK. Heterogeneity of infection enhancement of dengue 2 strains by monoclonal antibodies. J Immunol 1984; 132: 1529–32.
721
Review
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100 101
102
103
104
722
Dejnirattisai W, Jumnainsong A, Onsirisakul N, et al. Cross-reacting antibodies enhance dengue virus infection in humans. Science 2010; 328: 745–48. Kliks SC, Nimmannitya S, Nisalak A, Burke DS. Evidence that maternal dengue antibodies are important in the development of dengue hemorrhagic fever in infants. Am J Trop Med Hyg 1988; 38: 411–19. Jessie K, Fong MY, Devi S, Lam SK, Wong KT. Localization of dengue virus in naturally infected human tissues, by immunohistochemistry and in site hybridization. J Infect Dis 2004; 189: 1411–18. Balsitis SJ, Coloma J, Castro G, et al. Tropism of dengue virus in mice and humans defined by viral nonstructural protein 3-specific immunostaining. Am J Trop Med Hyg 2009; 80: 416–24. Cohen SN, Halstead SB. Shock associated with dengue infection. I. Clinical and physiologic manifestations of dengue hemorrhagic fever in Thailand, 1964. J Pediatrics 1966; 68: 448–56. Bethell DB, Flobbe K, Cao XT, et al. Pathophysiologic and prognostic role of cytokines in dengue hemorrhagic fever. J Infect Dis 1998; 177: 778–82. Green S, Vaughn DW, Kalayanarooj S, et al. Elevated plasma interleukin-10 levels in acute dengue correlate with disease severity. J Med Virol 1999; 59: 329–34. Chareonsirisuthigul T, Kalayanarooj S, Ubol S. Dengue virus (DENV) antibody-dependent enhancement of infection upregulates the production of anti-inflammatory cytokines, but suppresses antiDENV free radical and pro-inflammatory cytokine production, in THP-1 cells. J Gen Virol 2007; 88: 365–75. Ubol S, Phuklia W, Kalayanarooj S, Modhiran N. Mechanisms of immune evasion induced by a complex of dengue virus and preexisting enhancing antibodies. J Infect Dis 2010; 201: 923–35. Chen LC, Lei HY, Liu CC, et al. Correlation of serum levels of macrophage migration inhibitory factor with disease severity and clinical outcome in dengue patients. Am J Trop Med Hyg 2006; 74: 142–47. Simmons CP, Popper S, Dolocek C, et al. Patterns of host genome-wide gene transcript abundance in the peripheral blood of patients with acute dengue hemorrhagic fever. J Infect Dis 2007; 195: 1097–107. Nguyen TH, Lei HY, Nguyen TL, et al. Dengue hemorrhagic fever in infants: a study of clinical and cytokine profiles. J Infect Dis 2004; 189: 221–32. Ubol S, Masrinoul P, Chaijaruwanich J, Kalayanarooj S, Charoensirisuthikul T, Kasisith J. Differences in global gene expression in peripheral blood mononuclear cells indicate a significant role of the innate responses in progression of dengue fever but not dengue hemorrhagic fever. J Infect Dis 2008; 197: 1459–67. Shresta S, Kyle JL, Snider HM, Basavapatna M, Beatty PR, Harris E. Interferon-dependent immunity is essential for resistance to primary dengue virus infection in mice, whereas T- and B-celldependent immunity are less critical. J Virol 2004; 78: 2701–10. Shresta S, Sharar KL, Prigozhin DM, Snider HM, Beatty PR, Harris E. Critical roles for both STAT1-dependent and STAT1independent pathways in the control of primary dengue virus infection in mice. J Immunol 2005; 175: 3946–54. Navarro-Sanchez E, Despres P, Cedillo-Barron L. Innate immune responses to dengue virus. Arch Med Res 2005; 36: 425–35. Vaughn DW, Green S, Kalayanarooj S, et al. Dengue viremia titer, antibody response pattern, and virus serotype correlate with disease severity. J Infect Dis 2000; 181: 2–9. Libraty DH, Endy TP, Houng HS, et al. Differing influences of virus burden and immune activation on disease severity in secondary dengue-3 virus infections. J Infect Dis 2002; 185: 1213–21. Hsieh MF, Lai SL, Chen JP, et al. Both CXCR3 and CXCL10/IFNinducible protein 10 are required for resistance to primary infection by dengue virus. J Immunol 2006; 177: 1855–63. Chen JP, Lu HL, Lai SL, et al. Dengue virus induces expression of CXC chemokine ligand 10/IFN-gamma-inducible protein 10, which competitively inhibits viral binding to cell surface heparan sulfate. J Immunol 2006; 177: 3185–92.
105 Miller JL, deWet BJ, Martinez-Pomares L, et al. The mannose receptor mediates dengue virus infection of macrophages. PLoS Pathog 2008; 4: e17. 106 Boonnak K, Slike BM, Burgess TH, et al. Role of dendritic cells in antibody-dependent enhancement of dengue virus infection. J Virol 2008; 82: 3939–51. 107 Chaturvedi UC, Raghupathy R, Pasca AS, et al. Shift from a Th1-type response to Th1-type in dengue haemorrhagic fever. Curr Sci 1999; 76: 63–69. 108 Zellweger RM, Prestwood TR, Shresta S. Enhanced infection of liver sinusoidal endothelial cells in a mouse model of antibodyinduced severe dengue disease. Cell Host Microbe 2010; 7: 128–39. 109 Balsitis SJ, Williams KL, Lachica R, et al. Lethal antibody enhancement of dengue disease in mice is prevented by Fc modification. PLoS Pathog 2010; 6: e1000790. 110 Halstead SB, Porterfield JS, O’Rourke EJ. Enhancement of dengue virus infection in monocytes by flavivirus antisera. Am J Trop Med Hyg 1980; 29: 638–42. 111 Putvatana R, Yoksan S, Chayayodhin T, Bhamarapravati N, Halstead SB. Absence of dengue 2 infection enhancement in human sera containing Japanese encephalitis antibodies. Am J Trop Med Hyg 1984; 33: 288–94. 112 Petersen NC, Boyle JF. Immunologic phenomena in the effusive form of feline infectious peritonitis. Am J Vet Res 1980; 41: 868–76. 113 Vennema H, DeGroot RJ, Harbour DA, et al. Early death after feline infectious peritonitis challenge due to recombinant vaccinia virus immunization. J Virol 1990; 64: 1407–09. 114 Vennema H, Poland A, Foley J, Pedersen NC. Feline infectious peritonitis viruses arise by mutation from endemic feline enteric coronaviruses. Virology 1998; 243: 150–57. 115 Porter DDL, Larsen AE, Porter HG. Aleutian disease of mink. Adv Immunol 1980; 29: 261–86. 116 Porter DD, Larsen AE, Porter HG. The pathogenesis of Aleutian disease of mink. I. In vivo viral replication and the host antibody response to viral antigen. J Exp Med 1969; 130: 575–93. 117 Cancel-Tirado SM, Evans RB, Yoon KJ. Monoclonal antibody analysis of porcine reproductive and respiratory syndrome virus epitopes associated with antibody-dependent enhancement and neutralization of virus infection. Vet Immunol Immunopathol 2004; 102: 249–62. 118 Vanhee M, Costers S, Van Breedam W, Geldhof MF, Van Doorsselaere J, Nauwynck HJ. A variable region in GP4 of European-type porcine reproductive and respiratory syndrome virus induces neutralizing antibodies against homologous but not heterologous virus strains. Viral Immunol 2010; 23: 403–13. 119 Kimman TG, Cornelissen LA, Moormann RJ, Rebel JM, Stockhofe-Zurwieden N. Challenges for porcine reproductive and respiratory syndrome virus (PRRSV) vaccinology. Vaccine 2009; 27: 3704–18. 120 Van Breedam W, Van Gorp H, Zhang JQ, Crocker PR, Delputte PL, Nauwynck HJ. The M/GP(5) glycoprotein complex of porcine reproductive and respiratory syndrome virus binds the sialoadhesin receptor in a sialic acid-dependent manner. PLoS Pathog 2010; 6: e1000730. 121 Jung K, Renukaradhya GJ, Alekseev KP, Fang Y, Tang Y, Saif LJ. Porcine reproductive and respiratory syndrome virus modifies innate immunity and alters disease outcome in pigs subsequently infected with porcine respiratory coronavirus: implications for respiratory viral co-infections. J Gen Virol 2009; 90: 2713–23. 122 Nimmerjahn F, Ravetch JV. Fcgamma receptors as regulators of immune responses. Nat Rev Immunol 2008; 8: 34–47. 123 Moi ML, Lim CK, Takasaki T, Kurane I. Involvement of the Fc{gamma} receptor IIA cytoplasmic domain in antibodydependent enhancement of dengue virus infection. J Gen Virol 2010; 91: 103–11. 124 Fink J, Gu F, Vasudevan SG. Role of T cells, cytokines and antibody in dengue fever and dengue haemorrhagic fever. Rev Med Virol 2006; 16: 263–75.
www.thelancet.com/infection Vol 10 October 2010
Review
Tuberculosis case-contact research in endemic tropical settings: design, conduct, and relevance to other infectious diseases Philip C Hill, Martin O C Ota
The study of the contacts of patients with tuberculosis has a long history. Where tuberculosis is endemic, regular recruitment of tuberculosis cases and their household contacts can be done for research and strategic intervention. This recruitment provides a platform whereby host, pathogen, and environmental factors related to tuberculosis can be investigated and new interventions can be assessed. We describe the types of study possible within a tuberculosis casecontact study platform and its essential components, including recruitment and follow-up of the patients with tuberculosis, their household contacts and community controls, assessments and sampling, and data management and processing. Sample handling and storage, local engagement, ethical challenges, and the strengths and weaknesses of study design are all important issues in case-contact research. A case-contact study platform is a powerful research tool to answer fundamental questions in tuberculosis and has relevance to the study of other major infectious diseases.
Introduction The study of the close contacts of individuals with an infectious disease is a classic public health field-study paradigm, most commonly used in the investigation of outbreaks. The study of household contacts of tuberculosis cases has a long history, with reports as early as the 1930s.1,2 Where tuberculosis is endemic, the paradigm can be expanded to regular recruitment of patients with tuberculosis and their household contacts for purposes of research and strategic intervention. A tuberculosis casecontact study platform can be used to explore a large number of research questions. Indeed, a Bill & Melinda Gates Foundation Grand Challenge tuberculosis casecontact project to identify biomarkers of tuberculosis pathogenesis is now established across sub-Saharan Africa.3 In this Review, we describe the types of study that can be done and how to establish a case-contact research platform. In addition, we discuss how specific considerations and challenges can be approached.
Possible specific study types The basic operational concept of a tuberculosis casecontact study is the recruitment of patients with sputumsmear-positive tuberculosis and their household contacts. Specific substudies benefit from the recruitment of community controls. Those recruited are then followed over time—in particular, household contacts are followed for the development of tuberculosis disease. Such secondary tuberculosis cases can then be compared with non-progressors, according to a wide range of factors from basic public health to high-tech molecular issues. Many different types of interconnected studies are possible, including those focused on tuberculosis cases or their contacts and comparisons between them. Both cross-sectional and longitudinal assessments can be done and randomised controlled trial, cohort, case-control, and qualitative methods can be used (panel 1). Whereas assessment of new tools, such as diagnostic tests, is very clearly appropriate in a case-contact study, www.thelancet.com/infection Vol 10 October 2010
assessment of postexposure vaccines is also possible. One possible approach for the assessment of preexposure vaccines is as follows: after vaccination of study participants, adult tuberculosis cases could be prospectively identified at clinics, and study participants who happen to be their household contacts could be subject to intense active investigation and follow-up. Other participants might have less active follow-up. In this way, a case-contact study may be nested within a larger vaccine trial.
Recruitment and selection Panel 2 shows the key definitions of participants in a case-contact study. Contacts of sputum-smear-positive tuberculosis cases are more likely to have Mycobacterium tuberculosis infection than those exposed to a smearnegative case.2,27,28 Typical tuberculosis radiography is advised because sputum-smear positivity may be due to the presence of non-tuberculous mycobacteria. More than 95% of sputum smear-positive tuberculosis cases can be expected to have radiographic changes typical of tuberculosis.8 Because of the time to obtain a culture result, use of culture positivity as a recruitment criterion is not possible. Culture-negative, smear-positive, and chest-radiograph-positive individuals are extremely rare and require repeat culture if sputum samples are stored. Ideally, consecutively diagnosed adult patients with tuberculosis should be recruited. If too many cases are available for the capacity of the researchers, a purposeful selection process is advised (ie, every second case diagnosed). Restriction of recruitment (ie, to a specific geographical area) might also be needed for logistical reasons. Where there is reliable census information and the source population is well described, population-level epidemiological studies may be possible. This is further reason to select tuberculosis cases that are representative of all sputum-smear-positive cases in the population. Some households may be extraordinarily large (eg, >20). In such situations, it may be necessary to recruit
Lancet Infect Dis 2010; 10: 723–32 Centre for International Health, Department of Preventive & Social Medicine, University of Otago School of Medicine, Dunedin, New Zealand (Prof P C Hill MD); and Bacterial Diseases Programme, MRC Laboratories, Banjul, The Gambia (M O C Ota MD) Correspondence to: Prof Philip C Hill, Centre for International Health, University of Otago School of Medicine, Dunedin, New Zealand
[email protected]
For more on Bill & Melinda Gates Foundation Grand Challenge http://www.grandchallenges.org/
723
Review
Panel 1: Research study designs possible within a tuberculosis case-contact study platform
Panel 2: Key definitions for participants in a tuberculosis case-contact study6,16,26
Cross-sectional • Study of aerosols from tuberculosis cases in relation to their case contacts4 • Assessment of a new test for M tuberculosis infection against a gradient of exposure to a tuberculosis case5 • Prevalence of tuberculosis in tuberculosis case contacts6 • Prevalence of drug resistance in tuberculosis cases7 • Clinical and radiological description of tuberculosis cases8 • Population structure of M tuberculosis isolates9 • Immunological and nutritional comparisons of tuberculosis cases and infected contacts10
Index tuberculosis case • Aged ≥15 years • Sputum smear positive • Typical tuberculosis chest radiograph • First tuberculosis case identified in a household • Has at least one household contact
Case control and nested case control • Risk factors for tuberculosis disease11,12 • Risk factors for M tuberculosis infection13 • Assessment of markers for progression to disease comparing progressors to randomly selected non-progressors, using stored samples Cohort • Risk factors for defaulting from tuberculosis treatment and for treatment failure14 • Duration of infectivity of tuberculosis cases15 • Longitudinal assessment of a test for M tuberculosis infection16–18 • Incidence of disease in tuberculosis case contacts on follow-up16,19 • Assessment of a marker of treatment response in tuberculosis cases20 • Assessment of immune correlates of new infection21 • Assessment of likelihood of case-contact progression to disease according to index case factors, including infecting M tuberculosis strain22 Randomised controlled trial • Comparison of prophylactic treatments against M tuberculosis infection23 • Assessment of pre-exposure and post-exposure vaccines Qualitative • Identification of barriers to accessing tuberculosis treatment24,25 • Assessment of informed consent procedures
only a proportion of the contacts. A simple random selection process is the most robust method. However, there may be a large excess of individuals in a particular age, sex, or exposure category. In such a situation, sampling should be done more purposefully while maintaining a policy of random selection within strata. Some substudies might require all available contacts to be sampled in at least some households. Laboratory tests on fresh samples may be run in batches of varying sizes. The numbers of individuals for such tests may therefore need to be limited and a random selection process used at the time of sampling. This process can become complicated if more than one household is sampled at a time, and a certain amount of wasted space in batching of samples may be necessary. Some tests on fresh samples will have a maximum time window from sampling to analysis (table). Time of collection, receipt of specimen, and processing need to be documented. Those eligible but not participating at any stage of the study should be recorded as such. 724
Tuberculosis case contact • Aged ≥6 months • Living most of the time at the same property as the respective index case, sharing meals, and identifying a common household head Community control • Matched by age and sex to a tuberculosis case contact • No symptoms of tuberculosis • Living most of the time at a particular property, sharing meals, and identifying a common household head with other members • No history of tuberculosis in his/her household Co-prevalent tuberculosis case • Diagnosis of tuberculosis in which symptoms commenced <2 months after the respective index case was diagnosed Secondary tuberculosis case • Diagnosis of tuberculosis in which symptoms commenced ≥2 months after the respective index case was diagnosed
Selection of community controls for either index cases or contacts may be appropriate.26 The selection of a control is difficult if there is limited or no census information. A technique to deal with this problem was first used in tuberculosis studies by Crampin and colleagues:39 outside the case house, the field researcher spins a pen on the ground and proceeds in the direction in which it points for a specified number of households. A suitable control is then selected from within that household. The procedure is repeated in the event of a refusal or if a suitable control is not found. If there is more than one suitable control, then a random selection process is initiated; the researcher should not select, for example, an older child who might be easier to sample. Certain settings suit other sampling processes. For example, in Indonesia, people tend to live in an administrative unit called a rukun tetangga, where all the residents are identified and known. Therefore, an external control can be selected from individuals outside a case household, but inside the next level of the community. The control may be matched to prospectively recruited household contacts or by frequency matching to avoid conditional analyses. The latter can be achieved if the frequencies of specific age-groups and sex in tuberculosis case contacts from previous work are known. Controls should be recruited over the same timeframe as the study contacts. www.thelancet.com/infection Vol 10 October 2010
Review
Examples of research
Sampling volume (mL) Tube type
Processing issues
Storage issues
Sputum9,29
Pathogen genetics and fingerprinting with subsequent comparisons
≥2
Sterile container
A quality specimen from trained field staff
−70°C freezer (sputum and M tuberculosis)
Red blood cells30
Haematological reference values
≥1
EDTA
Within 6 h
..
Peripheral blood mononuclear cells10,31,32
Cytokine profiles; short and long incubation T-cell-based assays for infection and disease
≥3
Nunc
Time to cell separation ideally <2 h
−70°C freezer and liquid nitrogen
Serum and plasma33
Association of soluble markers with tuberculosis disease
≥1
Sardstedt
Within 4 h
−20°C freezer
Urine34
Assessment of urine-based diagnostic tests, profiling of chemotherapeutic agents
≥5
Bijou
Within 4 h
−70°C freezer
Saliva
Tuberculosis-specific antibodies
≥2
Sardstedt
Within 4 h
−20°C freezer
Breath condensate
Identification of diagnostic or prognostic factors
..
Bijou
Within 2 h
−70°C freezer
Stool
Worms and immune response to M tuberculosis
≥5
Bijou
Within 6 h
−20°C freezer
RNA35,36
Gene expression profiles of tuberculosis cases and contacts
≥2
Eppendorff
Reverse transcription necessary
−70°C freezer
DNA37,38
Genetic susceptibility to tuberculosis disease
≥1 or buccal swab
Eppendorff
DNA extraction required
−20°C freezer
Table: Samples that can be obtained in a tuberculosis case-contact study and their volume requirements, processing, and storage issues
Assessment and sampling The various assessments of tuberculosis cases include clinical and radiologically assessed disease severity.8,40 Tuberculosis contacts require examination for co-prevalent disease as defined in panel 2,16 or alternatively defined as the presence of active tuberculosis disease at the baseline household investigation.19 Field researchers should be able to detect reliably a BCG scar, which less commonly forms if given before 1 month of age. Although BCG scar size may vary according to certain factors (ie, sex),41 it is not an indicator of protective immunity.42,43 If the presence of a BCG scar is uncertain, it should be documented as such. The nutritional status of tuberculosis cases and contacts can be assessed using anthropometry and micronutrient status.40 Vitamin A and D statuses of individuals have been related to the pathogenesis of tuberculosis.44,45 A hair sample, processed appropriately, may be used for the assessment of micronutrients, offering retrospective insights.46,47 We and others have also measured bio-impedance. The table shows which samples may be obtained, how quickly they need to be delivered to the laboratory, and process and storage issues. Ideally, study participants should be brought close to the clinic and laboratory for initial assessment and sampling. The application of appropriate genetic, cellular, metabolomic, proteomic, transcriptomic, and other assays to the vast numbers of samples collected should allow identification of relevant biosignatures in the pathogenesis of tuberculosis.3 Proper sample processing, storage, maintenance, and documentation through specific standard operating procedures are crucial because stored samples provide an opportunity to apply new technologies over time and are required for nested case-control studies. A chest radiograph should be treated as a sample and stored after initial reporting. A digital photograph of each radiograph, obtained with a standardised method, can be useful and facilitates external consultation.48 An HIV test should be requested from all study participants after specific counselling and consenting www.thelancet.com/infection Vol 10 October 2010
procedures. However, in certain situations where HIV prevalence is very low and HIV testing is a sensitive issue with a high refusal rate, a test should not be requested.
Follow-up Case finding Continuing assessment of tuberculosis contacts for the development of tuberculosis disease may involve the following: repeated tests for M tuberculosis infection, followed by a clinical and radiological assessment in those who have test conversion; active regular follow-up (usually every 6 months) for symptoms with clinical assessment; and passive case finding from self-referral to clinics. Some contacts, when unwell, may attend their nearest open health facility, so regular surveillance at health facilities is necessary. This can be achieved by placing staff in clinics or by intermittently matching name and age-group in clinic records against the study database and visiting those who are a possible match.16 Tuberculin skin test (TST) conversion may be defined as a change from positive to negative with at least a 6 mm increase in induration.49 Enzyme-linked immunospot (ELISPOT) conversion may be defined as a change from positive to negative, plus an increase of at least six spots per well.17,20 Others have calculated that a tolerance of three spots either way in a T-spot test explains 95% of short-term variation.50 Similarly, most within-subject variability was identified within 15% of an interferon-γ ELISA-derived value.51 An ELISPOT test could be used to expose TST conversion due to boosting.52 More than one definition of a case of secondary tuberculosis disease may be appropriate because different studies will require more or less certainty of diagnosis. The most stringent criterion is the presence of culture-positive tuberculosis with fingerprint confirmation of a match between the index tuberculosis case and secondary case isolates. Spoligotyping is the most user-friendly technique, but is not as specific as IS6110 restriction fragment length polymorphism typing. The latter is appropriate in population-based studies in 725
Review
Symptoms and/or signs of tuberculosis
Positive/converted Mantoux or IGRA test
Clinical assessment
Definitely not tuberculosis
Routine follow-up
Possible tuberculosis
Probable tuberculosis
Active follow-up
Definite tuberculosis
Tuberculosis treatment
Figure 1: Flow diagram for the diagnostic examination of tuberculosis case contacts who are suspected of having tuberculosis disease in a case-contact study IGRA=interferon-γ release assay.
For more on the Stop TB Partnership see http://www.stoptb.org/
726
which pretest probability is relatively low.53 The diagnosis of tuberculosis can be difficult. In general, it is important to be able to label each tuberculosis case contact at each follow-up visit as either having definite tuberculosis, probable tuberculosis, possible tuberculosis, or definitely not having tuberculosis. This is best achieved by a formal process of consensus decision making among clinical investigators and independent review. If one excludes probable or possible tuberculosis cases in certain comparative analyses, the chances of contamination of one study group are reduced, with little chance of selection bias, because the numbers excluded are relatively small. Figure 1 summarises the process for identifying tuberculosis disease in tuberculosis case contacts. Symptoms, signs, a positive test, or test conversion for M tuberculosis infection trigger a clinical assessment. The symptoms that are appropriate for further investigation should be clearly defined for field researchers. The key symptom is cough of at least 2 weeks’ duration with or without the presence of chest pain, fever, haemoptysis, night sweats, or fatigue. In the absence of cough, weight loss (preferably confirmed by use of a growth chart in children) is important.54 Clinical assessment is limited by locally available tools. In general, one would hope to have sputum smear and culture facilities, chest radiograph, urine microscopy and culture, plus gastric lavage. At first assessment, a large proportion of suspected cases will clearly not have tuberculosis and can revert to routine follow-up. A few contacts will be diagnosed with definite tuberculosis at their first clinical assessment. Some will have probable tuberculosis, but will be unwell enough to be prescribed tuberculosis treatment. Others with probable or possible tuberculosis should have active follow-up. The results of all investigations should be clearly documented. This will enable a final decision at the end of follow-up as to whether each contact definitely did not have, or possibly, probably, or definitely had tuberculosis. The diagnosis of tuberculosis can be especially challenging in young children and HIV-infected individuals. Several approaches to the diagnosis of
tuberculosis in children are available. These include a guidance document from the Stop TB Partnership55 and scoring systems, although these lack uniform symptom definitions and validation.56 One recently developed approach is a symptom-based method,54,57 which starts with referral of children with a cough of at least 2 weeks’ duration. Those with resolving symptoms by the time of assessment are simply followed for symptom resolution with no further assessment. Others without resolution are investigated for tuberculosis disease. Those not treated for tuberculosis are followed according to symptom resolution or persistence. This approach is particularly appropriate for children aged 2–10 years. Younger children tend to progress to disease more rapidly without early symptoms. A key issue in asymptomatic healthy tuberculosis case contacts is whether children who have a positive test for M tuberculosis infection and hilar adenopathy on chest radiograph should automatically be treated for tuberculosis disease. Transient hilar adenopathy is common after primary infection and only a small proportion of such cases are likely to develop progressive disease.58 Furthermore, preventive chemotherapy may be adequate for successful treatment of such children.59
Duration of follow-up The highest number of secondary tuberculosis cases in a case-contact study, even after accounting for co-prevalent cases, occur in the first year of follow-up.16,19 However, secondary cases will continue to accumulate with time. Therefore, the duration of follow-up should be determined according to the number of case-contact households that can be recruited in a particular location and whether there are time limitations on the study. In addition, household members will be increasingly lost to the study with time, through migration in particular. The degree to which this occurs should be taken into account for each study population.
Longitudinal sampling A major barrier to the development of effective new tuberculosis vaccines is the lack of understanding of what constitutes protective immunity during natural infection with M tuberculosis. The case-contact model provides a unique opportunity to obtain a better understanding of these mechanisms through longitudinal collection and storage of samples that can be retrospectively analysed after classification of the participants’ phenotype (progressor or non-progressor) at the end of the study. Longitudinal sampling should be undertaken to answer specific research questions (table). Identified factors will facilitate the development of better diagnostic, treatment, and protective measures against tuberculosis. The circumstances that lead to development of disease may be present only transiently before and quite close to the time that disease develops. Therefore, multiple longitudinal sampling is advised. www.thelancet.com/infection Vol 10 October 2010
Review
Matching field and laboratory capacity Field and laboratory research teams need to understand each others’ capacity to handle samples during the followup and recruitment phases. Importantly, the period when recruitment of some households and follow-up of others occur at the same time may not be ideal for taking samples for a particular substudy that does not require all study participants. These samples could be strategically collected close to the end of the recruitment phase, according to the size of the substudy. Specific substudies should be powered appropriately and be no larger than necessary.
Data management and statistical issues All field data should be collected with standardised forms that are pretested on tuberculosis cases and contacts. Ideally, study participants should each have only one computer-generated identification number given to them at recruitment. Furthermore, detachable labels should be available for the various tubes with collected specimens. The number should be clearly linked to laboratory processes, noting that separate laboratory numbers may be required if specimens are processed in batches. All forms and other data that require manual entry should be entered into a database using double data entry and validation. The sheer diversity of field and laboratory data means that this information needs to be managed with an appropriate system. One approach is to build relational databases that can extract data from various electronic sources.60 Combination of electronic transmission with supplementary manual data entry might be necessary, but the core data should remain unaffected by risk of data-entry error. Paper files should be stored securely, not just for data verification, but as a requirement for particular studies, such as vaccine trials done according to good clinical practice regulations.61 The database should be backed up with copies that are stored electronically in different physical locations. Data quality can be monitored and improved through regular training of staff. Substudies should have a clear hypothesis and analysis plan, with a sample size that is appropriate for the particular question being addressed. Study design and statistical analyses need to take into account the effect of household clustering and, for specific studies, multiple testing. A clustered sample has less variation than an unclustered sample, and thus reduces the effective sample size; the most common approach for analysing clustered data is the inclusion of a random effect for the clustering level in a regression model.62 A case-contact study includes large numbers of unrelated individuals suitable for genetic studies. Furthermore, several analytical methods and software packages exist that account and correct for more or less cryptic relatedness, inbreeding, and population structure.63,64
Local acceptance and engagement and ethics Researchers running case-contact studies need to engage with national tuberculosis control programmes and seek www.thelancet.com/infection Vol 10 October 2010
to do interlocking studies of relevance to them. This may include studies of drug resistance,7 health-seeking behaviour,24,65 social and environmental factors amenable to relevant practical interventions,66 and studies of defaulting and treatment failure that may lead to changes in service provision.14 All case-contact studies have ethical considerations. These include informed consent, procedures for collection and storage of samples, and processes required for further analysis of stored specimens onsite or through external collaboration. All newly diagnosed tuberculosis cases should be referred to the national tuberculosis control programme for a full course of treatment. One particular issue is whether all contacts with evidence of latent tuberculosis infection should be given prophylactic antitubercular treatment. A 6–12-month course of isoniazid offers 60–90% protection against progression to disease,67 including in HIV-positive individuals.68–70 The case for treatment of latent infection is increasing in the light of a global target for tuberculosis elimination,71 and current WHO guidelines advise that children aged less than 5 years who are tuberculosis contacts should be given preventive therapy if possible.55 Pressure on national tuberculosis programmes, local policy, inability to do active case finding, difficulties of excluding active tuberculosis, risk of reinfection, and low adherence are often given as reasons why the policy is not implemented.72,73 Therefore, those planning a case-contact study in any setting should take careful note of this changing situation when considering their obligations in relation to preventive therapy. The obligations of researchers to HIV-positive individuals identified in projects in low-income countries have been the subject of complex debate.74 However, most low-income countries now have established antiretroviral treatment programmes, and study investigators are obliged to refer all HIV-positive study participants for formal assessment, follow-up, and consideration for treatment.
Potential weaknesses of case-contact studies in tuberculosis-endemic settings In Africa, the time to diagnosis of tuberculosis disease is commonly at least 2 months, which is related to delays in health seeking and the inadequacies of current diagnostic methods.75,76 In Gambian case-contact studies, more than half of the secondary tuberculosis cases diagnosed in the first 2 years already had disease at the first meeting (figure 2).6,16 Therefore, the ability to identify factors early in infection that predict progression to disease in contacts is compromised by late presentation of index cases. By contrast with controlled animal models, a human case-contact model in a tuberculosis-endemic setting is subject to contamination from outside the household. In The Gambia, a match was found between the isolates of the index cases and their respective contacts who became secondary cases two thirds of the time (figure 3).16 In a study of children in South Africa, only two of six secondary cases had an isolate that matched their respective index 727
Review
18 tuberculosis case pairs
2381 tuberculosis case contacts Secondary cases Screening for tuberculosis at recruitment
33 tuberculosis cases
9 co-prevalent
9 secondary Fingerprinting
2 years of follow-up
26 tuberculosis cases 6 identical
Figure 2: Problem of late presentation of tuberculosis index cases The yield from initial screening of tuberculosis cases in household tuberculosis contacts in The Gambia.6,16
case on fingerprinting.77 In a comparison of the molecular fingerprints of 765 tuberculosis cases according to household, the proportion of transmission in the community that took place inside households in Cape Town was only 19%.78 Therefore, the ideal setting for a case-contact study is not necessarily one where the incidence of tuberculosis is highest, because the balance of community versus household spread is more likely to be in favour of the community. For nested case-control studies in particular, a casecontact study platform depends on properly maintained and stored laboratory samples. Sample collections represent a huge investment of resources, and are relied on heavily as a crucial resource for understanding tuberculosis pathogenesis. The state of samples is important in interpreting the procedures done on them, including whether they are analysed when fresh or after a period of freezing.79 Constancy of supply of electricity or liquid nitrogen, reliability of freezers, and a robust samplemonitoring system are crucial. The possibility of setting up a sample repository at another facility should be considered as a safeguard. Thorough documentation is important, including an appropriate electronic record system for all samples (as described above). Although BCG has little or no effect on the interpretation of the TST in Africa in particular,5,80,81 and when given in infancy in general,82 major uncertainties remain about the M tuberculosis infection phenotype. Latent M tuberculosis infection is multifaceted and could easily be classified further into at least four distinct categories,83 including the possibility of early clearance by the innate immune system and transient infection of longer duration.84 Furthermore, these categories are indistinguishable from each other by any of the currently available tests.85 In a gene-expression study in The Gambia that used both TST and ELISPOT to define the M tuberculosis infection phenotype, nearly a third of tuberculosis case contacts were excluded because of discordant test results.35 Interestingly, those with discordant results had the same probability of progression to disease as those with concordant positive results.16 Those who convert from a negative test for M tuberculosis infection to a positive one on follow-up are most likely to have acquired new infection from their known exposure, with relatively more certainty about the timeline. However, boosting of the immune response to a previous exposure 728
3 different
6 identical
3 different
Figure 3: Problem of an open versus a closed model Results from molecular fingerprinting, by spoligotype, of 18 tuberculosis cases and their respective secondary cases among household contacts in The Gambia.16
is estimated to account for about a third of TST conversions.52 More precision around the M tuberculosis phenotype is crucial to link the molecular, genetic, and immunological laboratory tools now available. There are many unknowns in relation to timelines in the relation between the immune system and M tuberculosis. The interval between initial exposure and TST conversion is a maximum of 6 weeks after BCG vaccination and 3–7 weeks after M tuberculosis exposure.86 However, this interval has not been clearly identified for T-cell-based tests. The dynamic nature of the relation between the immune system and M tuberculosis has implications for longitudinal assessments. Timelines for shifting from one M tuberculosis infection category to another are likely to vary, involving host and organism factors.22 Furthermore, a specific immune signal for risk of progression to disease may not necessarily occur in the initial stages of new infection.87 Therefore, repeated longitudinal sampling in tuberculosis case contacts may be necessary to understand protective immunity and susceptibility. HIV co-infection has a profound effect on a case-contact study. The risk of tuberculosis in TST-negative HIV-infected individuals in Spain was 2·4 cases per 100 person-years and 16·2 per 100 person-years in TST-positive individuals who did not take preventive treatment or in whom it was discontinued early.88 In South African gold miners, the risk of tuberculosis was 2·9 cases per 100 person-years in HIVpositive individuals and 0·8 cases per 100 person-years in HIV-negative individuals.89 In Harare, Zimbabwe, the risk of tuberculosis was 0·13 cases per 100 person-years in HIV-negative business people and 2·53 per 100 person-years in those who were HIV positive.90 The risk increases with falling CD4-cell count, but also remains raised with antiretroviral treatment for those with CD4-cell counts that remain above 500×10⁶ cells/L.91 With known tuberculosis case contacts, the risk of tuberculosis in HIV-positive patients is particularly high, although precise incidence estimates are not available. 14 (13%) of 109 HIV-positive tuberculosis contacts in the USA were identified as having active tuberculosis at an initial screen, compared with 120 (2%) of 6116 HIV-negative contacts or contacts with unknown HIV status.92 HIV-positive tuberculosis case contacts should always be considered separately in analyses, at least in the first instance. Specific studies are www.thelancet.com/infection Vol 10 October 2010
Review
useful in HIV-positive case-contact study participants to provide further clues as to mechanisms of progression from infection to disease.93 Finally, HIV-positive tuberculosis case contacts are an especially high-risk group for rapid assessment of interventions such as prophylaxis and post-exposure vaccines.
Potential for a case-contact research model in other infectious diseases Pneumococcal disease Although there are many published studies of pneumococcal spread in families and households, these are largely focused on pneumococcal carriage.94,95 Pneumococci from diseased cases spread within households and cause disease in family members,96 and a case may infect a substantial proportion of the household. In a Gambian study, pneumococci of the same serotype that caused invasive disease in a child were obtained from 8·5% of family members, most commonly from young siblings of the index patient.97 No studies have documented the proportion of these new carriers that progress to disease. A small study found that invasive disease in young children occurs most commonly in the first month after new acquisition of pneumococcus.98 In a study of 43 134 relatives of 10 597 cases of invasive pneumococcal disease from a total population of over 1·7 million Danish people, Hjuler and colleagues99 found that household relatives were more than seven times more likely to develop invasive disease than the general population in the first year after exposure to their respective case, with the highest incidence of 52 cases per 100 000 per year in those aged less than 2 years. A much higher incidence would be expected in households in sub-Saharan Africa, where the incidence of invasive pneumococcal disease in young children in the community is commonly over 200 per 100 000 per year.100 There are good reasons why a pneumococcal casecontact study may offer significant new insights. The short time to culture positivity for pneumococcus and the ease of access of nasopharyngeal carriage offer significant advantages over M tuberculosis in relation to precise phenotyping. From the epidemiological studies that are available, a focus on household contacts who have acquired the same serotype as identified in the case seems sensible, although closer matching of strains is possible through molecular typing.101 Such subtyping may be necessary because of the high prevalence of nasopharyngeal carriage in many low-income countries, with the likelihood that pneumococcus of the same serotype is identified but from a previous acquisition. Early progression to disease might facilitate the identification of immunological and other markers of susceptibility and protection through comparison with non-progressors, similar to such studies in carriers versus non-carriers.102
Meningococcal disease Meningococcal epidemics in sub-Saharan Africa may reach an incidence of 1000 per 100 000 per year.103 www.thelancet.com/infection Vol 10 October 2010
Epidemics vary according to type—for example, type B meningococcal epidemics have a timespan of a decade or more, and the incidence is relatively low (<50 cases per 100 000 per year), at least in high-income country settings.104,105 The attack frequency of meningococcal disease in household contacts of a case can be over a thousand times higher than in the general community, in the region of four per 1000 contacts,106 and about half of the secondary cases occur within 2 months of diagnosis of the index case.107,108 Antibiotic prophylaxis is nearly 90% effective at preventing secondary disease.109 In non-endemic, non-epidemic settings, meningococcal carriage can be expected in a third of household contacts of children with meningococcal disease over 2 months of surveillance without chemoprophylaxis.110 The prevalence of carriage in household contacts of meningococcal disease cases seems to vary by meningococcal type.111 The ease of culture and access to the carrier state through pharyngeal swabs make meningococcus very suitable for case-contact study. In particular, the study of the innate and acquired immune response in contacts of cases of meningococcal disease has great potential.112 As with the pneumococcus, a focus on those who carry the same strain as their respective index case would be advisable and feasible in light of the short time to culture positivity. Because most secondary cases occur in the first 2 months after exposure, longitudinal studies in meningococcal case contacts also may not require long periods of follow-up.
Conclusions A case-contact study platform is a powerful research tool to answer fundamental questions related to environment, pathogen, and host factors in tuberculosis. Routine recruitment of cases and their household contacts can be a focal point for the merging of multiple scientific disciplines to address key hypotheses. Several epidemiological study designs can be applied to the study platform with the aim of linking field and laboratory most effectively. The case-contact model is not one that should be restricted to tuberculosis research, but clearly has potential to answer key questions in other major infectious diseases that affect human beings.
Search strategy and selection criteria We searched PubMed without date restrictions up to April, 2010, by use of the terms “tuberculosis”, “case”, and “contact”, alone and in combination for articles in English. We placed special focus on articles based on work in tuberculosisendemic settings. Additional references were identified from citations in retrieved articles. Articles were selected for their relevance to the design, conduct, and analysis of a casecontact study. Preference was given, but not restricted to, studies published after 1975.
729
Review
730
Conflicts of interest We declare that we have no conflicts of interest.
19
Contributors PH searched the published work, collated the data, and wrote the paper. MO contributed to writing the paper and refined it for submission.
20
Acknowledgments We would like to thank Richard Adegbola, Chris Whalen, Giorgio Sirugo, Reinout van Crevel, David Jeffries, Bouke de Jong, Roger Brookes, Rod Ellis-Pegler, Martin Pollock, and Merrin Rutherford for critically reading the manuscript. We thank the staff of the Tuberculosis Research Group at the MRC Laboratories in The Gambia for their work on the Gambian tuberculosis case-contact studies. We also thank all tuberculosis patients, their contacts, and community controls who participated in our studies.
21
References 1 Frost WH. Risk of persons in familial contact with pulmonary tuberculosis. Am J Public Health Nations Health 1933; 23: 426–32. 2 McPhedran FM, Opie EL. The spread of tuberculosis in families. Am J Hyg 1935; 22: 565–643. 3 Kaufmann SH, Parida SK. Tuberculosis in Africa: learning from pathogenesis for biomarker identification. Cell Host Microbe 2008; 4: 219–28. 4 Fennelly KP, Martyny JW, Fulton KE, Orme IM, Cave DM, Heifets LB. Cough-generated aerosols of Mycobacterium tuberculosis: a new method to study infectiousness. Am J Respir Crit Care Med 2004; 169: 604–09. 5 Hill PC, Brookes RH, Fox A, et al. Large-scale evaluation of enzyme-linked immunospot assay and skin test for diagnosis of Mycobacterium tuberculosis infection against a gradient of exposure in The Gambia. Clin Infect Dis 2004; 38: 966–73. 6 Jackson-Sillah D, Hill PC, Fox A, et al. Screening for tuberculosis among 2381 household contacts of sputum-smear-positive cases in The Gambia. Trans R Soc Trop Med Hyg 2007; 101: 594–601. 7 Adegbola RA, Hill P, Baldeh I, et al. Surveillance of drug-resistant Mycobacterium tuberculosis in The Gambia. Int J Tuberc Lung Dis 2003; 7: 390–93. 8 Rathman G, Sillah J, Hill PC, et al. Clinical and radiological presentation of 340 adults with smear-positive tuberculosis in The Gambia. Int J Tuberc Lung Dis 2003; 7: 942–47. 9 de Jong BC, Antonio M, Awine T, et al. Use of spoligotyping and large sequence polymorphisms to study the population structure of the Mycobacterium tuberculosis complex in a cohort study of consecutive smear-positive tuberculosis cases in The Gambia. J Clin Microbiol 2009; 47: 994–1001. 10 Sutherland JS, Adetifa IM, Hill PC, Adegbola RA, Ota MO. Pattern and diversity of cytokine production differentiates between Mycobacterium tuberculosis infection and disease. Eur J Immunol 2009; 39: 723–29. 11 Lienhardt C, Fielding K, Sillah JS, et al. Investigation of the risk factors for tuberculosis: a case-control study in three countries in West Africa. Int J Epidemiol 2005; 34: 914–23. 12 Hill PC, Jackson-Sillah D, Donkor SA, Otu J, Adegbola RA, Lienhardt C. Risk factors for pulmonary tuberculosis: a clinicbased case control study in The Gambia. BMC Public Health 2006; 6: 156. 13 Lienhardt C, Sillah J, Fielding K, et al. Risk factors for tuberculosis infection in children in contact with infectious tuberculosis cases in the Gambia, West Africa. Pediatrics 2003; 111: e608–14. 14 Hill PC, Stevens W, Hill S, et al. Risk factors for defaulting from tuberculosis treatment: a prospective cohort study of 301 cases in the Gambia. Int J Tuberc Lung Dis 2005; 9: 1349–54. 15 Rouillon A, Perdrizet S, Parrot R. Transmission of tubercle bacilli: the effects of chemotherapy. Tubercle 1976; 57: 275–99. 16 Hill PC, Jackson-Sillah DJ, Fox A, et al. Incidence of tuberculosis and the predictive value of ELISPOT and Mantoux tests in Gambian case contacts. PLoS One 2008; 3: e1379. 17 Hill PC, Brookes RH, Fox A, et al. Longitudinal assessment of an ELISPOT test for Mycobacterium tuberculosis infection. PLoS Med 2007; 4: e192. 18 Pai M, Joshi R, Dogra S, et al. T-cell assay conversions and reversions among household contacts of tuberculosis patients in rural India. Int J Tuberc Lung Dis 2009; 13: 84–92.
24
22
23
25
26
27 28 29
30
31
32
33
34
35 36
37
38
39
40
41
Guwatudde D, Nakakeeto M, Jones-Lopez EC, et al. Tuberculosis in household contacts of infectious cases in Kampala, Uganda. Am J Epidemiol 2003; 158: 887–98. Aiken AM, Hill PC, Fox A, et al. Reversion of the ELISPOT test after treatment in Gambian tuberculosis cases. BMC Infect Dis 2006; 6: 66. Whalen CC, Chiunda A, Zalwango S, et al. Immune correlates of acute Mycobacterium tuberculosis infection in household contacts in Kampala, Uganda. Am J Trop Med Hyg 2006; 75: 55–61. de Jong BC, Hill PC, Aiken A, et al. Progression to active tuberculosis, but not transmission, varies by Mycobacterium tuberculosis lineage in The Gambia. J Infect Dis 2008; 198: 1037–43. Jasmer RM, Saukkonen JJ, Blumberg HM, et al. Short-course rifampin and pyrazinamide compared with isoniazid for latent tuberculosis infection: a multicenter clinical trial. Ann Intern Med 2002; 137: 640–47. Eastwood SV, Hill PC. A gender-focused qualitative study of barriers to accessing tuberculosis treatment in The Gambia, West Africa. Int J Tuberc Lung Dis 2004; 8: 70–75. Harper M, Ahmadu FA, Ogden JA, McAdam KP, Lienhardt C. Identifying the determinants of tuberculosis control in resourcepoor countries: insights from a qualitative study in The Gambia. Trans R Soc Trop Med Hyg 2003; 97: 506–10. Hill PC, Brookes RH, Fox A, et al. Surprisingly high specificity of the PPD skin test for M. tuberculosis infection from recent exposure in The Gambia. PLoS One 2006; 1: e68. Brailey M. Mortality in the children of tuberculosis households. Am J Public Nations Health 1940; 30: 816–23. Shaw JB, Wynn-Williams N. Infectivity of pulmonary tuberculosis in relation to sputum status. Am Rev Tuberc 1954; 69: 724–32. de Jong BC, Hill PC, Brookes RH, et al. Mycobacterium africanum elicits an attenuated T cell response to early secreted antigenic target, 6 kDa, in patients with tuberculosis and their household contacts. J Infect Dis 2006; 193: 1279–86. Adetifa IM, Hill PC, Jeffries DJ, et al. Haematological values from a Gambian cohort—possible reference range for a West African population. Int J Lab Hematol 2008; 31: 615–22. Hill PC, Jackson-Sillah D, Fox A, et al. ESAT-6/CFP-10 fusion protein and peptides for optimal diagnosis of Mycobacterium tuberculosis infection by ex vivo enzyme-linked immunospot assay in the Gambia. J Clin Microbiol 2005; 43: 2070–74. Cehovin A, Cliff JM, Hill PC, Brookes RH, Dockrell HM. Extended culture enhances sensitivity of a gamma interferon assay for latent Mycobacterium tuberculosis infection. Clin Vaccine Immunol 2007; 14: 796–98. Lienhardt C, Azzurri A, Amedei A, et al. Active tuberculosis in Africa is associated with reduced Th1 and increased Th2 activity in vivo. Eur J Immunol 2002; 32: 1605–13. Gopinath K, Singh S. Urine as an adjunct specimen for the diagnosis of active pulmonary tuberculosis. Int J Infect Dis 2009; 13: 374–79. Burl S, Hill PC, Jeffries DJ, et al. FOXP3 gene expression in a tuberculosis case contact study. Clin Exp Immunol 2007; 149: 117–22. Wassie L, Demissie A, Aseffa A, et al. Ex vivo cytokine mRNA levels correlate with changing clinical status of ethiopian TB patients and their contacts over time. PLoS One 2008; 3: e1522. Bellamy R, Beyers N, McAdam KP, et al. Genetic susceptibility to tuberculosis in Africans: a genome-wide scan. Proc Natl Acad Sci USA 2000; 97: 8005–09. Olesen R, Wejse C, Velez DR, et al. DC-SIGN (CD209), pentraxin 3 and vitamin D receptor gene variants associate with pulmonary tuberculosis risk in West Africans. Genes Immunol 2007; 8: 456–67. Crampin AC, Mwinuka V, Malema SS, Glynn JR, Fine PE. Field-based random sampling without a sampling frame: control selection for a case-control study in rural Africa. Trans R Soc Trop Med Hyg 2001; 95: 481–83. de Jong BC, Hill PC, Aiken A, et al. Clinical presentation and outcome of tuberculosis patients infected by M. africanum versus M. tuberculosis. Int J Tuberc Lung Dis 2007; 11: 450–56. Floyd S, Ponnighaus JM, Bliss L, et al. BCG scars in northern Malawi: sensitivity and repeatability of scar reading, and factors affecting scar size. Int J Tuberc Lung Dis 2000; 4: 1133–42.
www.thelancet.com/infection Vol 10 October 2010
Review
42
43
44
45 46
47 48
49 50
51
52
53
54
55
56
57
58
59
60
61
62 63
64
Sterne JA, Fine PE, Ponnighaus JM, Sibanda F, Munthali M, Glynn JR. Does bacille Calmette-Guerin scar size have implications for protection against tuberculosis or leprosy? Tuber Lung Dis 1996; 77: 117–23. Zodpey SP, Shrikhande SN, Kulkarni SW, Maldhure BR. Scar size and effectiveness of Bacillus Calmette Guerin (BCG) vaccination in the prevention of tuberculosis and leprosy: a case-control study. Indian J Public Health 2007; 51: 184–89. Pakasi TA, Karyadi E, Wibowo Y, et al. Vitamin A deficiency and other factors associated with severe tuberculosis in Timor and Rote Islands, East Nusa Tenggara Province, Indonesia. Eur J Clin Nutr 2009; 63: 1130–35. Chocano-Bedoya P, Ronnenberg AG. Vitamin D and tuberculosis. Nutr Rev 2009; 67: 289–93. Zhang DR. Determination of zinc, copper, iron and zinc/copper ratio in the hair of active pulmonary tuberculosis patients [in Chinese]. Zhonghua Jie He He Hu Xi Za Zhi 1991; 14: 170–72, 192. Gibson RS, Gibson IL. The interpretation of human hair trace element concentrations. Sci Total Environ 1984; 39: 93–101. Szot A, Jacobson FL, Munn S, et al. Diagnostic accuracy of chest X-rays acquired using a digital camera for low-cost teleradiology. Int J Med Inform 2004; 73: 65–73. Menzies R. Interpretation of repeated tuberculin tests. Am J Respir Crit Care Med 1999; 159: 15–21. van Zyl-Smit RN, Pai M, Peprah K, et al. Within-subject variability and boosting of T-cell interferon-gamma responses after tuberculin skin testing. Am J Respir Crit Care Med 2009; 180: 49–58. Veerapathran A, Dakshinamoorthy G, Gnanasekar M, Reddy MV, Kalyanasundaram R. Evaluation of Wuchereria bancrofti GST as a vaccine candidate for lymphatic filariasis. PLoS Negl Trop Dis 2009; 3: e457. Hill PC, Jeffries DJ, Brookes RH, et al. Using ELISPOT to expose false positive skin test conversion in tuberculosis contacts. PLoS One 2007; 2: e183. Diaz R, Kremer K, de Haas PE, et al. Molecular epidemiology of tuberculosis in Cuba outside of Havana, July 1994–June 1995: utility of spoligotyping versus IS6110 restriction fragment length polymorphism. Int J Tuberc Lung Dis 1998; 2: 743–50. Marais BJ, Gie RP, Hesseling AC, et al. A refined symptom-based approach to diagnose pulmonary tuberculosis in children. Pediatrics 2006; 118: e1350–59. Stop TB Partnership Childhood TB Subgroup. Guidance for National Tuberculosis Programmes on the management of tuberculosis in children. Chapter 1: introduction and diagnosis of tuberculosis in children. Int J Tuberc Lung Dis 2006; 10: 1091–97. Hesseling AC, Schaaf HS, Gie RP, Starke JR, Beyers N. A critical review of diagnostic approaches used in the diagnosis of childhood tuberculosis. Int J Tuberc Lung Dis 2002; 6: 1038–45. Kruk A, Gie RP, Schaaf HS, Marais BJ. Symptom-based screening of child tuberculosis contacts: improved feasibility in resourcelimited settings. Pediatrics 2008; 121: e1646–52. Marais BJ, Gie RP, Schaaf HS, et al. The natural history of childhood intra-thoracic tuberculosis: a critical review of literature from the pre-chemotherapy era. Int J Tuberc Lung Dis 2004; 8: 392–402. Marais BJ, Gie RP, Schaaf HS, Beyers N, Donald PR, Starke JR. Childhood pulmonary tuberculosis: old wisdom and new challenges. Am J Respir Crit Care Med 2006; 173: 1078–90. Jeffries DJ, Donkor S, Brookes RH, Fox A, Hill PC. Design and implementation of relational databases relevant to the diverse needs of a tuberculosis case contact study in the Gambia. Int J Tuberc Lung Dis 2004; 8: 1095–99. Acosta CJ, Galindo CM, Ochiai RL, et al. Implementation of good clinical practice guidelines in vaccine trials in developing countries. Vaccine 2007; 25: 2852–57. Graubard BI, Korn EL. Regression analysis with clustered data. Stat Med 1994; 13: 509–22. Aulchenko YS, Ripke S, Isaacs A, van Duijn CM. GenABEL: an R library for genome-wide association analysis. Bioinformatics 2007; 23: 1294–96. Choi Y, Wijsman EM, Weir BS. Case-control association testing in the presence of unknown relationships. Genet Epidemiol 2009; 33: 668–78.
www.thelancet.com/infection Vol 10 October 2010
65
66 67
68
69
70
71 72
73
74 75
76
77
78
79
80
81 82
83 84
85
86 87
88
Godfrey-Faussett P, Kaunda H, Kamanga J, et al. Why do patients with a cough delay seeking care at Lusaka urban health centres? A health systems research approach. Int J Tuberc Lung Dis 2002; 6: 796–805. Chapman JS, Dyerly MD. Social and other factors in intrafamilial transmission of tuberculosis. Am Rev Respir Dis 1964; 90: 48–60. Comstock GW. How much isoniazid is needed for prevention of tuberculosis among immunocompetent adults? Int J Tuberc Lung Dis 1999; 3: 847–50. Whalen CC, Johnson JL, Okwera A, et al. A trial of three regimens to prevent tuberculosis in Ugandan adults infected with the human immunodeficiency virus. N Engl J Med 1997; 337: 801–08. Lim HJ, Okwera A, Mayanja-Kizza H, Ellner JJ, Mugerwa RD, Whalen CC. Effect of tuberculosis preventive therapy on HIV disease progression and survival in HIV-infected adults. HIV Clin Trials 2006; 7: 172–83. Johnson JL, Okwera A, Hom DL, et al. Duration of efficacy of treatment of latent tuberculosis infection in HIV-infected adults. AIDS 2001; 15: 2137–47. Dye C, Williams BG. Eliminating human tuberculosis in the twenty-first century. J R Soc Interface 2008; 5: 653–62. van Zyl S, Marais BJ, Hesseling AC, Gie RP, Beyers N, Schaaf HS. Adherence to anti-tuberculosis chemoprophylaxis and treatment in children. Int J Tuberc Lung Dis 2006; 10: 13–18. Marais BJ, van Zyl S, Schaaf HS, van Aardt M, Gie RP, Beyers N. Adherence to isoniazid preventive chemotherapy: a prospective community based study. Arch Dis Child 2006; 91: 762–65. Angell M. Investigators’ responsibilities for human subjects in developing countries. N Engl J Med 2000; 342: 967–69. Kasse Y, Jasseh M, Corrah T, et al. Health seeking behaviour, health system experience and tuberculosis case finding in Gambians with cough. BMC Public Health 2006; 6: 143. Keeler E, Perkins MD, Small P, et al. Reducing the global burden of tuberculosis: the contribution of improved diagnostics. Nature 2006; 444 (suppl 1): 49–57. Marais BJ, Hesseling AC, Schaaf HS, Gie RP, van Helden PD, Warren RM. Mycobacterium tuberculosis transmission is not related to household genotype in a highly endemic setting. J Clin Microbiol 2009; 47: 1338–43. Verver S, Warren RM, Munch Z, et al. Proportion of tuberculosis transmission that takes place in households in a high-incidence area. Lancet 2004; 363: 212–14. Doherty TM, Demissie A, Menzies D, Andersen P, Rook G, Zumla A. Effect of sample handling on analysis of cytokine responses to Mycobacterium tuberculosis in clinical samples using ELISA, ELISPOT and quantitative PCR. J Immunol Methods 2005; 298: 129–41. Mudido PM, Guwatudde D, Nakakeeto MK, et al. The effect of bacille Calmette-Guerin vaccination at birth on tuberculin skin test reactivity in Ugandan children. Int J Tuberc Lung Dis 1999; 3: 891–95. Gustafson P, Lisse I, Gomes V, et al. Risk factors for positive tuberculin skin test in Guinea-Bissau. Epidemiology 2007; 18: 340–47. Farhat M, Greenaway C, Pai M, Menzies D. False-positive tuberculin skin tests: what is the absolute effect of BCG and non-tuberculous mycobacteria? Int J Tuberc Lung Dis 2006; 10: 1192–204. Young DB, Gideon HP, Wilkinson RJ. Eliminating latent tuberculosis. Trends Microbiol 2009; 17: 183–88. Martineau AR, Newton SM, Wilkinson KA, et al. Neutrophilmediated innate immune resistance to mycobacteria. J Clin Invest 2007; 117: 1988–94. Pai M. Spectrum of latent tuberculosis — existing tests cannot resolve the underlying phenotypes [letter]. Nat Rev Microbiol 2010; 8: 242. Palmer CE, Edwards LB. Tuberculin test in retrospect and prospect. Arch Environ Health 1967; 15: 792–808. Adetifa IM, Brookes R, Lugos MD, et al. Rising ELISPOT count prior to the onset of symptoms of full-blown tuberculosis disease. Int J Tuberc Lung Dis 2007; 11: 350–52. Guelar A, Gatell JM, Verdejo J, et al. A prospective study of the risk of tuberculosis among HIV-infected patients. AIDS 1993; 7: 1345–49.
731
Review
89
Sonnenberg P, Glynn JR, Fielding K, Murray J, Godfrey-Faussett P, Shearer S. How soon after infection with HIV does the risk of tuberculosis start to increase? A retrospective cohort study in South African gold miners. J Infect Dis 2005; 191: 150–58. 90 Corbett EL, Bandason T, Cheung YB, et al. Epidemiology of tuberculosis in a high HIV prevalence population provided with enhanced diagnosis of symptomatic disease. PLoS Med 2007; 4: e22. 91 Lawn SD, Myer L, Edwards D, Bekker LG, Wood R. Short-term and long-term risk of tuberculosis associated with CD4 cell recovery during antiretroviral therapy in South Africa. AIDS 2009; 23: 1717–25. 92 Centers for Disease Control and Prevention. Missed opportunities for prevention of tuberculosis among persons with HIV infection—selected locations, United States, 1996–1997. MMWR Morb Mortal Wkly Rep 2000; 49: 685–87. 93 Hammond AS, McConkey SJ, Hill PC, et al. Mycobacterial T cell responses in HIV-infected patients with advanced immunosuppression. J Infect Dis 2008; 197: 295–99. 94 Sleeman KL, Daniels L, Gardiner M, et al. Acquisition of Streptococcus pneumoniae and nonspecific morbidity in infants and their families: a cohort study. Pediatr Infect Dis J 2005; 24: 121–27. 95 Hendley JO, Sande MA, Stewart PM, Gwaltney JM Jr. Spread of Streptococcus pneumoniae in families. I. Carriage rates and distribution of types. J Infect Dis 1975; 132: 55–61. 96 Shimada J, Yamanaka N, Hotomi M, et al. Household transmission of Streptococcus pneumoniae among siblings with acute otitis media. J Clin Microbiol 2002; 40: 1851–53. 97 Lloyd-Evans N, O’Dempsey TJ, Baldeh I, et al. Nasopharyngeal carriage of pneumococci in Gambian children and in their families. Pediatr Infect Dis J 1996; 15: 866–71. 98 Gray BM, Converse GM 3rd, Dillon HC Jr. Epidemiologic studies of Streptococcus pneumoniae in infants: acquisition, carriage, and infection during the first 24 months of life. J Infect Dis 1980; 142: 923–33. 99 Hjuler T, Poulsen G, Wohlfahrt J, Kaltoft M, Biggar RJ, Melbye M. Genetic susceptibility to severe infection in families with invasive pneumococcal disease. Am J Epidemiol 2008; 167: 814–19. 100 Usen S, Adegbola R, Mulholland K, et al. Epidemiology of invasive pneumococcal disease in the Western Region, The Gambia. Pediatr Infect Dis J 1998; 17: 23–28.
732
101 Antonio M, Hakeem I, Awine T, et al. Seasonality and outbreak of a predominant Streptococcus pneumoniae serotype 1 clone from The Gambia: expansion of ST217 hypervirulent clonal complex in West Africa. BMC Microbiol 2008; 8: 198. 102 Zhang Q, Bagrade L, Bernatoniene J, et al. Low CD4 T cell immunity to pneumolysin is associated with nasopharyngeal carriage of pneumococci in children. J Infect Dis 2007; 195: 1194–202. 103 Riedo FX, Plikaytis BD, Broome CV. Epidemiology and prevention of meningococcal disease. Pediatr Infect Dis J 1995; 14: 643–57. 104 Baker MG, Martin DR, Kieft CE, Lennon D. A 10-year serogroup B meningococcal disease epidemic in New Zealand: descriptive epidemiology, 1991–2000. J Paediatr Child Health 2001; 37: S13–19. 105 Dyet KH, Martin DR. Clonal analysis of the serogroup B meningococci causing New Zealand’s epidemic. Epidemiol Infect 2006; 134: 377–83. 106 The meningococcal disease surveillance group. Analysis of endemic meningococcal disease by serogroup and evaluation of chemoprophylaxis. J Infect Dis 1976; 134: 201–04. 107 Cooke RP, Riordan T, Jones DM, Painter MJ. Secondary cases of meningococcal infection among close family and household contacts in England and Wales, 1984–7. BMJ 1989; 298: 555–58. 108 Olcen P, Kjellander J, Danielsson D, Lindquist BL. Epidemiology of Neisseria meningitidis; prevalence and symptoms from the upper respiratory tract in family members to patients with meningococcal disease. Scand J Infect Dis 1981; 13: 105–09. 109 Purcell B, Samuelsson S, Hahne SJ, et al. Effectiveness of antibiotics in preventing meningococcal disease after a case: systematic review. BMJ 2004; 328: 1339. 110 Marks MI, Frasch CE, Shapera RM. Meningococcal colonization and infection in children and their household contacts. Am J Epidemiol 1979; 109: 563–71. 111 Simmons G, Martin D, Stewart J, Jones N, Calder L, Bremner D. Carriage of Neisseria meningitidis among household contacts of patients with meningococcal disease in New Zealand. Eur J Clin Microbiol Infect Dis 2001; 20: 237–42. 112 van Deuren M, Brandtzaeg P, van der Meer JW. Update on meningococcal disease with emphasis on pathogenesis and clinical management. Clin Microbiol Rev 2000; 13: 144–66.
www.thelancet.com/infection Vol 10 October 2010
Personal View
Schistosomiasis elimination: lessons from the past guide the future Darren J Gray, Donald P McManus, Yuesheng Li, Gail M Williams, Robert Bergquist, Allen G Ross
Schistosomiasis is a major neglected tropical disease, with more than 200 million people infected and close to 800 million at risk. The disease burden is estimated to exceed 70 million disability-adjusted life-years. The anthelmintic drug praziquantel is highly effective in killing adult schistosome worms, but it is unable to kill developing schistosomes and so does not prevent reinfection. As a result, current praziquantel-based control programmes in Asia and subSaharan Africa are not effective or sustainable in the long term. The control of neglected tropical diseases, including schistosomiasis, is a funding priority for several donor agencies, with over US$350 million committed until 2013. Here we put forward an argument that donor funds would be more effectively spent on the development of a multifaceted, integrated control programme, which would have a greater and longer lasting effect on disease transmission than the current chemotherapy-based programmes. The development of a transmission-blocking vaccine is also of great importance. A multi-faceted integrated control programme that incorporates a vaccine, even if only partly effective, has the potential to eliminate schistosomiasis. This integrated-approach model has the potential to improve the health of a billion of the world’s poorest people and its effect cannot be underestimated.
Introduction Schistosomiasis is a major neglected tropical disease. Over 200 million people are infected and close to 800 million are at risk.1 This disease of poverty has proved difficult to control for centuries. Consequently, the current schistosomiasis disease burden remains high and could exceed 70 million disability-adjusted life-years.2,3 The anthelmintic drug praziquantel, discovered in the mid-1970s, has been the cornerstone of schistosomiasis control.4 In the 1980s, a schistosomiasis control programme in Mali funded by the German Agency for Technical Cooperation succeeded in reducing disease prevalence with mass praziquantel treatment, but disease prevalence returned to baseline levels after treatment cycles were stopped.5–7 Similarly, during the 1990s, programmes funded by the World Bank in China and the Philippines successfully reduced the prevalence of Schistosoma japonicum, but there are signs that disease prevalence is increasing now that the programmes have finished.8–10 85% of all schistosomiasis cases occur in sub-Saharan Africa (figure 1), but until 2002, and the introduction of the Schistosomiasis Control Initiative, only sporadic control activities had taken place there.11–13 The high cost of praziquantel and the priority given by funding bodies to other diseases, such as malaria, tuberculosis, and HIV/AIDS, were among the main reasons for this neglect. However, in the past 5–10 years, because of a renewed interest in neglected tropical diseases14–16 and a reduction in the cost of praziquantel, there has been increased funding available for schistosomiasis control. Here we discuss past and present challenges, and propose a strategy that could potentially lead to worldwide elimination of schistosomiasis.
Drug-based control: past and present challenges Praziquantel is highly effective in killing adult worms of all schistosome species that infect human beings4 and, when given to people with schistosomiasis, controls www.thelancet.com/infection Vol 10 October 2010
morbidity well. However, because praziquantel does not kill immature schistosomes and cannot prevent reinfection, praziquantel-based control programmes have only a temporary effect on transmission and are limited in their potential to interrupt disease transmission in the long term.17 In endemic areas, once mass treatment with praziquantel is stopped, disease prevalence can return to baseline levels within 18–24 months.4,7,18 Less well known, but pertinent for control programme design, is the evidence for severe rebound morbidity when chemotherapy campaigns are interrupted.18–22 When mass chemotherapy is interrupted in areas of high transmission, 80% of children have recurrent aggressive inflammation and require frequent praziquantel treatment, otherwise their morbidity would be more severe than before the mass chemotherapy campaign began.19–21 Evidence from Burkina Faso and Niger, where control programmes have focused primarily on mass praziquantel treatment, points to unacceptably high prevalences in high-transmission areas despite extensive donor support.23,24 These outcomes recall the failed attempt to control schistosomiasis in Mali despite almost a decade of mass chemotherapy.5–7 Clearly, the approach based on mass praziquantel treatment needs to be carefully scrutinised in terms of effectiveness and long-term sustainability. The Schistosomiasis Control Initiative has been very successful in establishing national schistosomiasis control programmes and intends to reduce prevalence and morbidity in six countries in sub-Saharan Africa (Burkina Faso, Mali, Niger, Uganda, Tanzania, and Zambia). The initiative, however, primarily uses vertical mass drug distribution,13,24 which has proved unsuccessful in the past.5–7 Furthermore, coverage has been somewhat ineffective. Infants and preschool children are not treated and less than 50% of the high-risk population receive praziquantel, but, paradoxically, many uninfected
Lancet Infect Dis 2010; 10: 733–36 Published Online August 11, 2010 DOI:10.1016/S14733099(10)70099-2 School of Public Health, Griffith University, Meadowbrook, Queensland, Australia (D J Gray PhD, Prof A G Ross MD); School of Population Health, University of Queensland, Brisbane, Australia (D J Gray, Prof G M Williams PhD); Molecular Parasitology Laboratory, Infectious Diseases Division, Queensland Institute of Medical Research, Queensland, Australia (D J Gray, Prof D P McManus DSc, Prof Y Li MD); Hunan Institute of Parasitic Diseases, WHO Collaborating Centre for Research and Control on Schistosomiasis in Lake Region, Yueyang, China (Y Li); and Ingerod, Brastad, Sweden (Prof R Bergquist MD) Correspondence to: Dr Darren J Gray, School of Public Health, Griffith University, Meadowbrook, QLD 4131, Australia d.gray@griffith.edu.au
733
Personal View
Countries or areas at high risk Countries or areas at low risk
Figure 1: Countries or areas at risk for schistosomiasis
individuals are unnecessarily given the drug.13,25,26 This is because of an approach that sets out to treat everyone rather than focusing on case finding and targeted chemotherapy. As a result, the overall reduction in morbidity has been small and there has been limited interruption of disease transmission. Poor compliance is also an issue; in China and the Philippines, compliance declines substantially over prolonged chemotherapy campaigns.27,28 In light of these shortcomings, mass chemotherapy control programmes seem to be not cost effective and, more importantly, only sustainable with continued donor support.12,29 Thus far, roughly US$150 million has been spent on drug-based morbidity control for schistosomiasis and other neglected tropical diseases in sub-Saharan Africa.19,30, This total will rise to $350 million by 2013 after the USA announced an increase in funding for a new global initiative to combat neglected tropical diseases and an increase in the number of targeted countries.19 Roughly 1·2 billion praziquantel tablets will be needed to treat 400 million people per year in sub-Saharan Africa for at least 5 years, at an annual cost of US$100 million.26,31 Countries in sub-Saharan Africa are among the poorest in the world, with roughly 73% of the population living on less than US$2 per day.32 Without continuing financial support, these countries will be unable to finance and sustain mass treatment programmes against schistosomiasis or any of the other neglected tropical diseases.23 The crucial question, especially in times of 734
global economic uncertainty, is what will happen to the current neglected tropical disease control programmes if donor funding stops?
Future integrated, sustainable control Important and urgent decisions must be made on the future approach needed for effective, sustainable control and elimination of schistosomiasis. The strategy used over the past 25 years in sub-Saharan Africa focuses on short-term morbidity control.5–7,13 Because transmission reduction is an important step towards elimination, a control programme that controls morbidity and reduces transmission would be more cost effective because it would not incur the cost of continually treating new infections. Such an approach is being used effectively in China.33–37 Control and elimination of neglected tropical diseases do not involve treatment of the disease alone: many people with neglected tropical diseases are at the mercy of inadequate health and social systems that need to be improved if any long-term benefit is to be achieved. There are two major transmission pathways in the schistosome life cycle that can be targeted by control programmes: the parasites’ path from human beings (the definitive hosts) to snails (the intermediate hosts), and their path from snails to human beings (figure 2).34 Mass praziquantel treatment acts only on the transmission pathway from human being to snail and only for as long as treatment is given. By combining chemotherapy with known and effective control measures, such as the use of www.thelancet.com/infection Vol 10 October 2010
Personal View
Chemotherapy, sanitation, environmental modification, anti-fecundity vaccination
Definitive mammalian host
Intermediate snail host
Health education, use of molluscicides, environmental modification, anti-infection vaccination
Figure 2: Interventions that can target transmission pathways in the schistosome life cycle
molluscicides, environmental modification, improved sanitation, and health education, multifaceted integrated control programmes can target both transmission pathways. Such effective use of donor money could result in the reduction and even elimination of schistosomiasis from endemic areas. Because of the focal nature of schistosome transmission, the multifaceted integrated control programme we are advocating would build on historical and current data, and use case finding, geographic information systems, and remote sensing to develop predictive maps that would allow endemic areas and re-emerging pockets of transmission to be targeted.38,39 Our approach would be tailored to specific endemic settings and be incorporated into national and local health services. This integrated approach is being used with success in China, where the elimination of schistosomiasis is regarded as one of the top public-health priorities along with HIV/AIDS, tuberculosis, and hepatitis B.40 The challenge in subSaharan Africa and other parts of Asia is to develop a sustainable, multiple component schistosomiasis control strategy that is cost effective, and more importantly, not dependent on endless donor funding. Systemic integrated approaches to control communicable diseases, such as malaria, tuberculosis, and HIV/AIDS, are proving successful in sub-Saharan Africa,41 and although an integrated strategy might initially be more expensive than a chemotherapy-based approach, a greater and longer-lasting effect on transmission makes it more cost effective and sustainable in the long term. Vaccines are thought of as silver bullets for public health interventions in the control of infectious diseases; their crucial role in the eradication of smallpox in 1978 substantiates this notion.42 The recent publication of the Schistosoma mansoni43 and S japonicum44 genomes takes us a step closer to the identification of key protective molecules and the development and implementation of effective antischistosome vaccines, although many research questions need to be addressed if we are to successfully achieve this goal.45 Schistosomiasis japonica, caused by S japonicum, is a zoonosis, and in countries, such as China and the Philippines, where animals are major reservoirs for this disease, a transmission-blocking vaccine for livestock would be beneficial and would probably take less time to develop than a human vaccine.34 www.thelancet.com/infection Vol 10 October 2010
In China, randomised, double-blind trials of two DNA vaccines for water buffaloes, encoding well researched S japonicum antigens, reduced worm burden, fecal egg counts, and miracidal hatching by roughly 50%.46 Mathematical modelling predicts that these vaccines, in combination with other control options, will substantially reduce transmission, meaning that a transmissionblocking bovine vaccine could well be available within the next few years.46–48 Vaccine development should be a priority along with other avenues of schistosomiasis research, including the search for alternative drugs to praziquantel, because resistance against the drug in the future cannot be ruled out. We believe that multiple component, integrated control programmes incorporating praziquantel treatment with transmission reduction through other control measures, such as use of molluscicides, environmental modification, health education and promotion, and improved sanitation, is the best option for schistosomiasis control. Even if only partly effective, anti-schistosome vaccines, incorporated as part of an integrated control strategy, will be needed to accelerate efforts to eliminate a disease that has existed for at least two millennia.19 This integratedapproach model has the potential to improve the health of a billion of the world’s poorest people and its effect cannot and should not be underestimated. Contributors DJG, AGR, and DPM had the idea for and prepared the paper. YL, GMW, and RB assisted with the preparation of the paper. Conflicts of interest We declare that we have no conflicts of interest. Acknowledgments The authors’ studies on schistosomiasis have received financial support from various sources including: the UNICEF/UNDP/World Bank/WHO Special Program for Research and Training in Tropical Diseases, the National Health and Medical Research Council of Australia, the Wellcome Trust (UK), the Sandler Foundation (USA); the Dana Foundation (USA); and the National Institute of Allergy and Infectious Diseases. DJG is a Griffith University Postdoctoral Fellow. We would like to thank all our colleagues in Australia and our collaborators in China who have provided us with much invaluable support during the course of our research. Thanks also go to Hawys McManus for comments on the paper. References 1 Steinmann P, Keiser J, Bos R, Tanner M, Utzinger J. Schistosomiasis and water resources development: systematic review, meta-analysis, and estimates of people at risk. Lancet Infect Dis 2006; 6: 411–25. 2 King CH, Dickman K, Tisch DJ. Reassessment of the cost of chronic helminthic infection: a meta-analysis of disability-related outcomes in endemic schistosomiasis. Lancet 2005; 365: 561–69. 3 King CH, Dangerfield-Cha M. The unacknowledged impact of chronic schistosomiasis. Chronic Illn 2008; 4: 65–79. 4 Doenhoff MJ, Cioli D, Utzinger J. Praziquantel: mechanisms of action, resistance and new derivatives for schistosomiasis. Curr Opin Infect Dis 2009; 21: 659–67. 5 Korte R, Schmidt-Ehry B, Kielmann AA, Brinkmann UK. Cost and effectiveness of different approaches to schistosomiasis control in Africa. Trop Med Parasitol 1986; 37: 149–52. 6 Brinkmann UK, Werler C, Traoré M, Korte R. The costs of schistosomiasis control in Sahelian County. Trop Med Parasitol 1988; 39: 175–81. 7 Clements AC, Bosqué-Oliva E, Sacko M, et al. A comparative study of the spatial distribution of schistosomiasis in Mali in 1984–1989 and 2004–2006. PLoS Negl Trop Dis 2009; 3: e431.
735
Personal View
8
9
10
11
12
13
14 15
16
17
18 19
20
21
22
23
24
25
26
27
736
Zhang W, Wong CM. Evaluation of the 1992–1999 World Bank Schistosomiasis Control Project in China. Acta Trop 2003; 85: 303–13. Utzinger J, Zhou XN, Chen MG, Bergquist R. Conquering schistosomiasis in China: the long march. Acta Trop 2005; 96: 69–96. WHO. Prevention and control of schistosomiasis and soil-transmitted helminthiasis: report of a WHO Expert Committee. WHO Tech Rep Ser number 912. Geneva: World Health Organization, 2002. van der Werf MJ, de Vlas SJ, Brooker S, et al. Quantification of clinical morbidity associated with schistosome infection in subSaharan Africa. Acta Trop 2003; 86: 125–39. Utzinger J, Bergquist R, Shu-Hua X, Singer BH, Tanner M. Sustainable schistosomiasis control—the way forward. Lancet 2003; 362: 1932–34. Fenwick A, Webster JP, Bosque-Oliva E, et al. The schistosomiasis control initiative (SCI) rationale, development and implementation from 2002–2008. Parasitology 2009; 136: 1719–30 Hotez PJ, Molyneux DH, Fenwick A, et al. Control of neglected tropical diseases. N Eng J Med 2007; 357: 1018–27. Hotez P. Mass drug administration and integrated control for the world’s high-prevalence neglected tropical diseases. Clin Pharmacol Ther 2009; 85: 659–64. Hotez PJ, Fenwick A, Savioli L, Molyneux DH. Rescuing the bottom billion through control of neglected tropical diseases. Lancet 2009; 373: 1570–75. Doenhoff MJ, Hagan P, Cioli D, et al. Praziquantel: its use in control of schistosomiasis in sub-Saharan Africa and current research needs. Parasitology 2009; 136: 1825–35. Ross AGP, Bartley PB, Sleigh AC, et al. Schistosomiasis. N Eng J Med 2002; 346: 1212–19. Bergquist R, Utzinger J, McManus DP. Trick or treat: the role of vaccines in integrated schistosomiasis control. PLoS Negl Trop Dis 2008; 2: e244. Finkelstein JL, Schleinitz MD, Carabin H, McGarvey ST. Decisionmodel estimation of the age-specific disability weight for schistosomiasis japonica: a systematic review of the literature. PLoS Negl Trop Dis 2008; 2: e158. Olveda RM, Daniel BL, Ramirez BD, et al. Schistosomiasis japonica in the Philippines: the long-term impact of population-based chemotherapy on infection, transmission, and morbidity. J Infect Dis 1996; 174: 163–72. Reimert CM, Tukahebwa EM, Kabatereine NB, Dunne DW, Vennervald BJ. Assessment of Schistosoma mansoni induced intestinal inflammation by means of eosinophil cationic protein, eosinophil protein X and myeloperoxidase before and after treatment with praziquantel. Acta Trop 2008; 105: 253–59. Garba A, Touré S, Dembelé R, Bosque-Oliva E, Fenwick A. Implementation of national schistosomiasis control programmes in West Africa. Trends Parasitol 2006; 22: 322–26. Garba A, Touré S, Dembelé R, et al. Present and future schistosomiasis control activities with support from the schistosomiasis control initiative in west Africa. Parasitology 2009; 136: 1731–37. Stothard JR, Gabrielli AF. Schistosomiasis in African infants and preschool children: to treat or not to treat? Trends Parasitol 2007; 23: 83–86. Utzinger J, Raso G, Brooker S, et al. Schistosomiasis and neglected tropical diseases: towards integrated and sustainable control and a word of caution. Parasitology 2009; 136: 1859–74. Guo JG, Cao CL, Hu GH, et al. The role of passive chemotherapy plus health education for schistosomiasis control in China during maintenance and consolidation phase. Acta Trop 2005; 96: 177–83.
28
29
30
31
32
33 34
35
36
37
38
39
40 41
42 43 44
45 46
47
48
Tallo VL, Carabin H, Alday PP, et al. Is mass treatment the appropriate schistosomiasis elimination strategy? Bull World Health Organ 2008; 86: 765–71. Guyatt H. The cost of delivering and sustaining a control programme for schistosomiasis and soil-transmitted helminthiasis. Acta Trop 2003; 86: 267–74. Stothard JR, Chitsulo L, Kristensen TK, Utzinger J. Control of schistosomiasis in sub-Saharan Africa: progress made, new opportunities and remaining challenges. Parasitology 2009; 136: 1665–75. Hotez PJ, Fenwick A. Schistosomiasis in Africa: an emerging tragedy in our new global health decade. PLoS Negl Trop Dis 2009; 3: e485. Hotez PJ, Kamath A. Neglected tropical diseases in sub-Saharan Africa: review of their prevalence, distribution, and disease burden. PLoS Negl Trop Dis 2009; 3: e412. King CH. Toward the elimination of schistosomiasis. N Engl J Med 2009; 360: 106–09. Gray DJ, Williams GM, Li Y, McManus DP. Transmission dynamics of Schistosoma japonicum in the lakes and marshlands of China. PLoS One 2008; 3: e4058. Wang LD, Chen HG, Guo JG, et al. A strategy to control transmission of Schistosoma japonicum in China. N Engl J Med 2009; 360: 121–28. Zhou XN, Wang LY, Chen MG, et al. The public health significance and control of schistosomiasis in China: then and now. Acta Trop 2005; 96: 97–105. Wang LD, Guo JG, Wu XH, et al. China’s new strategy to block Schistosoma japonicum transmission: experiences and impact beyond schistosomiasis. Trop Med Int Health 2009; 14: 1475–83. Clements ACA, Firth S, Dembelé R, et al. Use of Bayesian geostatistical prediction to estimate local variations in Schistosoma haematobium infection in west Africa. Bull World Health Organ 2009; 87: 921–29. Yang K, Zhou XN, Wu XH, et al. Landscape pattern analysis and Bayesian modelling for predicting Oncomelania hupensis distribution in Eryuan county, People’s Republic of China. Am J Trop Med Hyg 2009; 81: 416–23. Wang LD, Utzinger J, Zhou XN. Schistosomiasis control: experiences and lessons from China. Lancet 2008; 372: 1793–95. de Savigny D, Kasale H, Mbuya, C, Reid G. Fixing Health Systems. Ottawa, Canada: International Development Research Centre, 2004: 108. Geddes AA. The history of smallpox. Clin Dermatol 2006; 24: 152–57. Berriman M, Haas BJ, LoVerde PT, et al. The genome of the blood fluke Schistosoma mansoni. Nature 2009; 460: 352–58. Schistosoma japonicum genome sequencing and functional analysis consortium, Liu F, Zhou Y, et al. The Schistosoma japonicum genome reveals unique features of host-parasite interplay. Nature 2009; 460: 345–51. McManus DP, Loukas A. Current status of vaccines for schistosomiasis. Clin Microbiol Rev 2008; 21: 225–42. Da’Dara AA, Li YS, Xiong T, et al. DNA-based vaccine protects against zoonotic schistosomiasis in water buffalo. Vaccine 2008; 26: 3617–25. Williams GM, Sleigh AC, Li Y, et al. Mathematical modelling of schistosomiasis japonica: comparison of control strategies in the Peoples’ Republic of China. Acta Trop 2002; 82: 253–62. McManus DP, Gray GJ, Li YS, et al. Schistosomiasis in the Peoples’ Republic of China: the era of Three Gorges Dam. Clin Microbiol Rev 2010; 23: 442–66.
www.thelancet.com/infection Vol 10 October 2010