Thesis Was Financially Supported by the Stichting Sarphati and the Medical Researchh Council

UvA-DARE (Digital Academic Repository)

HIV-2 in West Africa. Epidemiological studies

Schim van der Loeff, M.F.

Publication date 2003 Document Version Final published version

Link to publication

Citation for published version (APA): Schim van der Loeff, M. F. (2003). HIV-2 in West Africa. Epidemiological studies.

General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Download date:07 Oct 2021 JÉfoo JïP^h

V V HIV-2lm m EPIDEMIOLOGICAL L

MAARTENN F. SCHIM VAN DER LOEFF HIV-22 IN WEST AFRICA

EPIDEMIOLOGICALL STUDIES Printingg of this thesis was financially supported by the Stichting Sarphati and the Medical Researchh Council.

ISBNN 90-6464-834-4

Printedd by Ponsen & Looi jen - Wageningen Coverr design: Casper Schim van der Loeff - Utrecht

©© 2003 by M. F. Schim van der Loeff, London, UK. All rights reserved. No part of this publicationn may be reproduced or transmitted in any form or by any means, electronic or mechanical,, including photocopy, recording, or any information storage or retrieval system,, without permission from the copyright owner. Copyrights of chapters 2, 3, 5, 6, 88 are transferred to Lippincott Williams & Wilkins. These chapters are reprinted with kindd permission from Lippincott Williams & Wilkins. HIV-22 in West Africa

Epidemiologicall Studies

ACADEMISCHH PROEFSCHRIFT

terr verkrijging van de graad van doctor aann de Universiteit van Amsterdam opp gezag van de Rector Magnificus prof.. mr. P. F. van der Heijden tenn overstaan van een door het college voor promoties ingestelde commissie,, in het openbaar te verdedigen in de Aula der Universiteit

opp woensdag 26 november 2003, te 12:00 uur.

doorr Maarten Francis cus Schim van der Loef f

geborenn te Venray Promotiecommiss sie

Promotores:: Prof. dr. R. A. Coutinho Prof.. dr. P. Aaby

Co-promotor:: Prof. H. C. Whittle

Overigee leden: Prof. dr. F. Miedema Prof.. dr. J. M. A. Lange Prof.. dr. P. A. Kager Prof.. dr. M. W. Borgdorff Prof.. dr. A. D. M. E. Osterhaus Prof.. dr. G.J. Bonsel

Faculteitt der Geneeskunde Voorr vader en moeder

'Enn nu nog iets,' zei Kaatje Kater toen Buitenrust Hettema en Wiegel in Rotterdam de treinn verlaten hadden en ze met zijn drieën in de coupé waren achtergebleven. 'Hoe staat hett met je proefschrift?' Ze keek hem indringend aan. Hett was zo onverwacht dat Maarten niet meteen een antwoord had. Hij schudde langzaamm zijn hoofd, zonder iets te zeggen. 'Hett wordt nu wel echt tijd.1 'Ikk ben daar nog niet aan toe.' 'Enn wanneer kom je er dan aan toe? Want als je te lang wacht, is het te laat.' Beertaa glimlachte heimelijk. 'Ikk ben daar niet bang voor. Ik vind dat je pas een proefschrift moet schrijven als je daarinn gelooft.'

J.J.J.J. Voskuil, Plankton, Het Bureau deel 3

Contents s

Chapterr 1 Introduction 11

Chapterr 2 Towards a better understanding of the epidemiology of HIV-2 19

Chapterr 3 Regional differences in HIV trends in The Gambia: results from 59 sentinell surveillance among pregnant women

Chapterr 4 Incidence of HIV-2 in a rural community in Guinea-Bissau, West 73 Africa a

Chapterr 5 HIV-2 does not protect against HIV-1 infection in a rural 89 communityy in Guinea-Bissau

Chapterr 6 Mortality of HIV-1, HIV-2 and HIV-1 / HIV-2 dually infected 107 patientss in a clinic-based cohort in The Gambia, West Africa

Chapterr 7 Contrasts in plasma viral load, CD4% and survival in a community- 129 basedd cohort of HIV-1 and HIV-2 infected women in The Gambia

Chapterr 8 Survival of HIV-1 and HIV-2 perinatally infected children in The 147 Gambia a

Chapterr 9 Conclusions 161

Listt of Abbreviations 183

Summaryy 185

Samenvattingg 189

Listt of publications 193

CVV 197

Acknowledgementss 199

1 1

Introduction n

HIV-22 IN WEST AFRICA

Inn 1986, three years after the discovery of HIV-1, another retrovirus was identified, and namedd HIV-2 [1,2]. It was isolated from West African patients with AIDS and its discoveryy caused concern that another devastating epidemic was at hand. Various researchh groups in West Africa started to study the epidemiology, risk factors and the naturall history of this infection. The Medical Research Council Laboratories, a research institutee based in The Gambia since 1947, initiated studies on HIV-2 in 1988 and made importantt findings on the descriptive epidemiology [3-7], risk factors for infection [8-10], clinicall course and mortality [11-13], immune response [14-17] and viral load [18]. The underlyingg strategy has been to compare HIV-1 and HIV-2 infections.

Whenn I joined the MRC in 1995,1 was able to build on the work of my predecessors and colleagues,, who had established three cohorts: a hospital-based cohort, a community- basedd cohort of a village with a high prevalence of HIV-2, and a cohort of women recruitedd during pregnancy. This thesis reports on the epidemiological studies carried outt in these three cohorts between 1995 and 2001.

Thee studies were carried out in two countries in West Africa, The Gambia and Guinea- Bissau.. This part of die world was known to early European settiers as the White Man's Grave,, mainly due to malaria and yellow fever. It certainly is a Black Man's Grave as well,, as the high infant and maternal mortality ratios, and the low life expectancies testify (seee Table 1). Poor health and poverty are closely associated; the populations in both countriess are among the poorest in the world. Poverty does not merely mean low incomes,, but also poor education, poor roads, poor communications, poor electricity supply,, and poor water supply. This affects not only the population, but also makes researchh efforts in these countries challenging. In Table 1 some key characteristics of bothh countries are listed, together with those of my native Netherlands, to put these in perspective. .

Thee first study of this thesis was not conducted in any of the existing cohorts, but was a cross-sectionall study (Chapter 3). The Gambia did not have HIV sentinel surveillance untill 1999, so the government and MRC Laboratories joined forces to start this. The

13 3 HIV-22 IN WEST AFRICA

Tabtee 1. Key indicators of The Gambia, Guinea-Bissau, and The Netherlands

Guinea-- The e Thee Gambia Bissau u Netherlands s Areaa (km^) 11,000 0 36,000 0 42,000 0 Populationn (millions) 1.4 4 1.3 3 16.0 0 Populationn density (people per km2) 125 5 36 6 385 5 GNPP per capita (PPP US$) 1,649 9 755 5 25,657 7 Overseass Development Aid (Euros per 26 6 44 4 -203* * inhabitantt p.a.) UNDPP Human Development Index 2002 0.405 5 0.349 9 0.935 5 Rankk on Human Development List 2002 160 0 167 7 8 8 (outt of 173) Illiteracyy (% of people > 15 year) 65% % 63% % 1% % Numberr of universities 1 1 0 0 13 3 Telephoness per 1000 inhabitants 23 3 7 7 607 7 Mainn religions: Muslim m 95% % 38% % 4% % Christian n 4% % 8% % 52% % Animist t — — 54% % -- Infantt mortality ratio (per 1000 live births) 78 8 110 0 4 4 Maternall mortality ratio (per 100,000 live 1100 0 910 910 7 7 births) ) Lifee expectancy in years (women) 49 9 47 7 81 1 Lifee expectancy in years (men) 45 5 44 4 75 5 Vaccinationn coverage** 90% % 6% % 95% % Prevalencee of HIV-1 in general adult 1.0% % 4.4% % 0.2% % population n Prevalencee of HIV-2 in general adult 0.8% % 3.9% % -- population n

** Minus indicates aid is given, not received. ** % of children fully immunised with 6 vaccines beforee age 1 year. GNP = Gross National Product; p.a. = per annum; PPP = Purchasing powerr parity; UNDP = United Nations Development Program. Sources:: Wolters-Noordhoff. De Grote Bosatlas. 52nd Ed. Groningen, Wolters-Noordhoff Atlasprodukties;; 2002. UNAIDS. Report on the Global HIV/AIDS Epidemic 2002. Geneva: UNAIDS;; 2002. Norrgren H, Da Silva ZJ, Biague AJ, Andersson S, Biberfeld G. Trends of HIV-11 and HIV-2 in Guinea-Bissau before and after the civil war 1998-9. XII International Conferencee on AIDS and STDs in Africa. Ouagadougou, December 2001 [Abstract 11DT5- 5].. UNDP. Human Development Report 2002. New York: OUP; 2002.

firstfirst round of the sentinel surveillance among antenatal attendees was conducted in 2000-

11 and the prevalence was compared to the prevalence found during tJhe Gambian

Mother-to-Childd HIV Transmission study of 1993-5 [19].

Thee first cohort reported here is located in Guinea-Bissau. In 1988 Dr Dominique

Ricard,, while working in clinics for commercial sex workers in Ziguinchor (Senegal),

noticedd that a disproportionate number of the women who were HIV-2 infected came

14 4 HIV-22 IN WEST AFRICA

West t Africa a

Thee Gambia

Guinea a

fromm Guinea-Bissau, and that most of those came from one particular rural area in north- westernn Guinea-Bissau. A study was set up in that area (Caio) by him and Dr Andrew Wilkinss to study the epidemiology of HIV-2. A demographic census of the total populationn and a sero-survey among the adults were carried out [20]. This was followed byy a case control study of HIV-2 infected villagers and a number of uninfected controls too determine clinical features associated with HIV-2 infection [21]. Since then a re- surveyy was conducted in 1996-8, and the cases and controls have been re-examined in 19966 [22] and in 2003. The re-survey in 1996-8 made it possible to estimate the incidencee rate of HIV-2 infections and examine risk factors for HIV-2 infection (Chapter 4).. It also enabled an analysis of the intriguing hypothesis that HIV-2 infection could be protectivee against subsequent HIV-1 infection (Chapter 5).

Thee second cohort is a group of patients infected with HIV-1, HIV-2, or both viruses, whoo attend clinics at the MRC facilities in Fajara, The Gambia. The patients were initiallyy tested for a variety of reasons: a sexually transmitted disease, clinical signs of HIV

15 5 HIV-22 IN WEST AFRICA disease,, tuberculosis, blood donation, being a commercial sex worker, or being a partner off an AIDS patient. Consequently this cohort consists of both asymptomatic and very ill patients,, and of recent seroconverters as well as patients infected for many years. As of Januaryy 2003 the database of this cohort contains data on about 2800 adult patients, of whomm 1350 have died. This cohort has formed the basis of many of the immunological andd virological studies done on HIV in The Gambia [23-31]. In this cohort we examined thee mortality rates of patients infected with either or both of the two viruses, stratified by CD44 counts and by clinical stage (Chapter 6).

Thee third cohort is a group of women who were recruited at antenatal clinics in The Gambiaa between 1993 and 1995. A total of 101 were HIV-1 infected, 243 HIV-2 infected,, and 10 HIV-1 and HIV-2 dually infected (HIV-D); 468 were HIV uninfected. Thiss cohort was set up to examine and compare mother-to-child transmission rates of HIV-11 and HIV-2 [19,32]. Since the completion of that study the women have been followedd up again in 2001, to investigate and compare the mortality associated with the twoo virus infections in the women (Chapter 7) and their children (Chapter 8).

Thee chapters describing these studies are preceded by a review of the epidemiology of HIV-22 (Chapter 2), and followed by a final chapter (9) that draws conclusions from the presentedd studies.

References s

1.. Clavel F, Guetard D, Brun-Vezinet F, Chamaret S, Rey MA, Santos-Ferreira MO, Laurentt AG, Dauguet C, Katiama C, Rouzioux C, et al. Isolation of a new human retroviruss from West African patients with AIDS. Science 1986; 233:343-6. 2.. Brun-Vezinet F, Rey MA, Katiama C, Girard PM, Roulot D, Yeni P, Lenoble L, Clavel F,, Alizon M, Gadelle S, et al. Lymphadenopathy-associated virus type 2 in AIDS and AIDS-relatedd complex. Clinical and virological features in four patients. Lancet 1987;1:128-32. . 3.. Wilkins A, Oelman B, Pepin J, Cham K, Corrah T, Hughes A, Manneh K, Mannen K, Njaii R, Rolfe M, et al. Trends in HIV-1 and HIV-2 infection in The Gambia. AIDS 1991;5:1529-30. . 4.. Wilkins HA, Alonso P, Baldeh S, Cham MK, Corrah T, Hughes A, Jaiteh KO, Oelman B,, Pickering H. Knowledge of AIDS, use of condoms and results of counselling

16 6 HIV-22 IN WEST AFRICA

subjectss with asymptomatic H1V2 infection in The Gambia. AIDS Care 1989;1:247- 56. . 5.. Pickering H, Todd J, Dunn D, Pepin J, Wilkins A. Prostitutes and their clients: a Gambiann survey. Soc Sci Med 1992;34:75-88. 6.. Pickering H, Quigley M, Pepin J, Todd J, Wilkins A. The effects of post-test counsellingg on condom use among prostitutes in The Gambia. AIDS 1993;7:271-3. 7.. Pickering H, Quigley M, Hayes RJ, Todd J, Wilkins A. Determinants of condom use in 24,0000 prostitute/client contacts in The Gambia. AIDS 1993;7: 1093-8. 8.. Wilkins A, Hayes R, Alonso P, Baldeh S, Berry N, Cham K, Hughes A, Jaiteh K, Oelmann B, Tedder R, et al. Risk factors for HIV-2 infection in The Gambia. AIDS 1991;5:1127-32. . 9.. Pepin J, Dunn D, Gaye I, Alonso P, Egboga A, Tedder R, Piot P, Berry N, Schellenbergg D, Whittle H, et al. HIV-2 infection among prostitutes working in The Gambia:: association with serological evidence of genital ulcer diseases and with generalizedd lymphadenopathy. AIDS 1991;5:69-75. 10.. Pepin J, Quigley M, Todd J, Gaye I, Janneh M, Van Dyck E, Piot P, Whittle H. Associationn between HIV-2 infection and genital ulcer diseases among male sexually transmittedd disease patients in The Gambia. AIDS 1992;6:489-93. 11.. Pepin J, Morgan G, Dunn D, Gevao S, Mendy M, Gaye I, Scollen N, Tedder R, Whittlee H. HIV-2-induced immunosuppression among asymptomatic West African prostitutes:: evidence that HIV-2 is pathogenic, but less so than HIV-1. AIDS 1991; 5:1165-72. . 12.. Whittle H, Egboga A, Todd J, Corrah T, Wilkins A, Demba E, Morgan G, Rolfe M, Berryy N, Tedder R. Clinical and laboratory predictors of survival in Gambian patients withh symptomatic HIV-1 or HIV-2 infection. AIDS 1992;6:685-9. 13.. Whittle H, Morris J, Todd J, Corrah T, Sabally S, Bangali J, Ngom PT, Rolfe M, Wilkinss A. HIV-2-infected patients survive longer than HIV-1-infected patients. AIDS 1994;8:1617-20. . 14.. Egboga A, Corrah T, Todd J, Wilkins A, Whittle H, Bouchier V, Rolfe M, Seaton A, Morgann G. Immunological findings in African patients with pulmonary tuberculosis and HIV-22 infection. AIDS 1992;6:1045-6. 15.. Whittle H, Egboga A, Todd J, Morgan G, Rolfe M, Sabally S, Wilkins A, Corrah T. Immunologicall responses of Gambians in relation to clinical stage of HIV-2 disease. Clinn Exp Immunol 1993;93:45-50. 16.. Gotch F, McAdam SN, AHsopp CE, Gallimore A, Elvin J, Kieny MP, Hill AV, McMichaell AJ, Whittle HC. Cytotoxic T cells in HIV-2 seropositive Gambians. Identificationn of a virus-specific MHC-restricted peptide epitope. J Immunol 1993; 151:3361-9. . 17.. Ariyoshi K, Cham F, Berry N, Jaffar S, Sabally S, Corrah T, Whittle H. HIV-2-specific cytotoxicc T-lymphocyte activity is inversely related to proviral load. AIDS 1995;9:555- 9. . 18.. Berry N, Ariyoshi K, Jobe O, Ngum PT, Corrah T, Wilkins A, Whittle H, Tedder R. HIVV type 2 proviral load measured by quantitative polymerase chain reaction correlatess with CD4+ lymphopenia in HIV type 2-infected individuals. AIDS Res Hum Retrovirusess 1994;10:1031-7. 19.. O'Donovan D, Ariyoshi K, Milligan P, Ota M, Yamuah L, Sarge-Njie R, Whittle H. Maternall plasma viral RNA levels determine marked differences in mother-to-child transmissionn rates of HIV-1 and HIV-2 in The Gambia. MRC/Gambia Government / Universityy College London Medical School working group on mother-child transmissionn of HIV. AIDS 2000;14:441-8. 20.. Wilkins A, Ricard D, Todd J, Whittle H, Dias F, Paulo Da Silva A. The epidemiology of HIVV infection in a rural area of Guinea-Bissau. AIDS 1993;7:1119-22. 21.. Ricard D, Wilkins A, N'Gum PT, Hayes R, Morgan G, Da Silva AP, Whittle H. The effectss of HIV-2 infection in a rural area of Guinea-Bissau. AIDS 1994;8:977-82.

17 7 HIV-22 IN WEST AFRICA

22.. Ariyoshi K, Schim van der Loeff M, Berry N, Jaffar S, Whittle H. Plasma HIV viral load inn relation to season and to Plasmodium falciparum parasitaemia. AIDS 1999; 13:1145-6. . 23.. Ariyoshi K, Cheingsong-Popov R, Wilkins A, Corrah T, Weber J, Whittle H. HIV-1 subtypee B in west Africa. Lancet 1996;347:328. 24.. Jaffar S, Wilkins A, Ngom PT, Sabally S, Corrah T, Bangali JE, Rolfe M, Whittle HC. Ratee of decline of percentage CD4+ cells is faster in HIV-1 than in HIV-2 infection. J Acquirr Immune Defic Syndr Hum Retrovirol 1997;16:327-32. 25.. Berry N, Ariyoshi K, Jaffar S, Sabally S, Corrah T, Tedder R, Whittle H. Low peripherall blood viral HIV-2 RNA in individuals with high CD4 percentage differentiatess HIV-2 from HIV-1 infection. J Hum Virol 1998;1:457-68. 26.. Ariyoshi K, Cham F, Berry N, Harding E, Sabally S, N'Gom PT, Ishikawa K, Corrah T, Tedderr R, Whittle H. Diagnosis of HIV-1/2 dual infection using dilution analysis of type-specificc antibody. AIDS 1998;12:2504-5. 27.. Bertoletti A, Cham F, McAdam S, Rostron T, Rowland-Jones S, Sabally S, Corrah T, Ariyoshii K, Whittle H. Cytotoxic T cells from human immunodeficiency virus type 2- infectedd patients frequently cross-react with different human immunodeficiency virus typee 1 clades. J Virol 1998;72:2439-48. 28.. Cham F, Heyndrickx L, Janssens W, Van der Auwera G, Vereecken K, De Houwer K, Coppenss S, Whittle H, van der Groen G. Study of HIV type 1 gag/env variability in Thee Gambia, using a multiplex DNA polymerase chain reaction. AIDS Res Hum Retrovirusess 2000;16:1915-9. 29.. Cham F, Heyndrickx L, Janssens W, Vereecken K, De Houwer K, Coppens S, Van derr Auwera G, Whittle H, van der Groen G. Development of a one-tube multiplex reversee transcriptase-polymerase chain reaction assay for the simultaneous amplificationn of HIV type 1 group M gag and env heteroduplex mobility assay fragments.. AIDS Res Hum Retroviruses 2000;16:1503-5. 30.. Ariyoshi K, Jaffar S, Alabi AS, Berry N, Schim van der Loeff M, Sabally S, N'Gom PT, Corrahh T, Tedder R, Whittle H. Plasma RNA viral load predicts the rate of CD4 T cell declinee and death in HIV-2-infected patients in West Africa. AIDS 2000;14:339-44. 31.. Berry N, Ariyoshi K, Balfe P, Tedder R, Whittle H. Sequence specificity of the human immunodeficiencyy virus type 2 (HIV-2) long terminal repeat u3 region in vivo allows subtypingg of the principal HIV-2 viral subtypes A and B. AIDS Res Hum Retroviruses 2001;17:263-7. . 32.. Ota MO, O'Donovan D, Alabi AS, Milligan P, Yamuah LK, N'Gom PT, Harding E, Ariyoshii K, Wilkins A, Whittle HC. Maternal HIV-1 and HIV-2 infection and child survivall in The Gambia. AIDS 2000;14:435-9.

18 8 2 2

Towardss a better understanding of the epidemiologyy of HIV-2

AIDSAIDS 1999; 13 (supplA) :S69-S84.

Maartenn F. Schim van der Loeff ' and Peter Aaby '

(1)(1) Medical Research Council laboratories, Banjul, The Gambia.

(2)(2) Projecto de Saüde de Bandim, Bissau, Guinea-Bissau.

(3)(3) Department of Epidemiology Research, Danish Epidemiology Science Centre, Statens Serum

Institut,Institut, Copenhagen, Denmark.

Thiss work was supported by the Medical Research Council of the United Kingdom, DANIDA, thee Danish Medical Research Council, and die Science & Technology for Development Program off the European Community

HIV-22 IN WEST AFRICA

Introduction n

Threee years after the discovery of HIV in 1983, a related but distinct virus was isolated in Westt African AIDS patients and subsequendy named Human Immunodeficiency Virus typee 2 (HIV-2) [1,2]. This virus has not spread much outside West Africa, its presumed regionn of origin. This review highlights key developments in HIV-2 epidemiology since thee last major review on the topic in this journal in 1996 [3]. Among these are the demonstrationn of limited excess mortality due to HIV-2, the development of a Polymerasee Chain Reaction (PCR) based method to quantify plasma viral load, studies of subtypee distribution, and the possible protective effect of HIV-2 against subsequent HIV-11 infection. We have emphasised comparisons between HIV-1 and HIV-2 (see table 1)--

Virology y

HIV-22 was first isolated from two AIDS patients from West Africa (Guinea-Bissau and Capee Verde) in 1986 [1,2]. The genetic structure of HIV-2 is similar to HIV-1, although HIV-22 has a ipx gene instead of the vpu gene. The nucleotide and amino acid homology betweenn the viruses is about 60% for die more conserved^ and pol genes, but only 30- 40%% for the other viral genes, including env [4]. The structure of the virion of HIV-2 is veryy similar to HIV-1. The differences in the envelope and other proteins induce the host too form antibodies, which can be distinguished from HIV-1 induced antibodies. This formss the basis of the serological tests which differentiate between HIV-1 and HIV-2 infection. .

Sequencingg of pol, env, and gag genes has led to the distinction of 6 different HIV-2 subtypess (A to F)[5,6]. The subtypes have up to 25 % nucleotide diversity in gag [5,6].

HIV-22 viruses are genetically closer to SIVsm than to HIV-1. SIVsm is a retrovirus found inn sooty mangabeys (primates indigenous to West Africa) and not pathogenic to these

21 1 HIV-22 IN WEST AFRICA

animals;; some subtypes of HIV-2 cluster phylogenetically closer to SIV,m than to other HIV-22 strains [7,8].

Tablee 1. Differences and similarities in the epidemiology of HIV-1 and HIV-2

HIV-1 1 HIV-2 2

Mother-to-child d 20-35% % 00 - 4% transmissionn rate

Geographicall spread World-wide Westt Africa; areass with links to Portugal or Westt Africa; India a

Agee peak 200 - 34 years 45-55 5

Excesss mortality 10-fold d 2-fold d

Transmissionn routes Heterosexual; homosexual; Heterosexual;; homosexual; IVDU; IVDU;; BT; mother-to-child; BT;; mother-to-child needlee stick

Rural-urban n Higherr prevalences urban Similarr prevalences urban and differences s rural l

Trend d Prevalencee rising in most Stablee prevalence in most countries s countries s

Notes:: IVDU = intravenous drug user; BT = blood transfusion.

Basedd on the divergence of the HIV-2 subtypes, the very close genetic relationship of

HIV-22 and SIVsm, the geographical overlap between the natural habitat of sooty mangabeyss and the endemic regions of HIV-2, and the common close contacts between humanss and sooty mangabeys (as household pets or as game), it is thought that HIV-2 is aa zoonosis, caused by various independent transmission events from sooty mangabeys to humanss [5-7].

HIV-22 strains, like HIV-1 isolates, can be characterised as slow/low or rapid/high phenotypess [9], or as syncytium-inducing and non-syncytium inducing [10]. The biologicall phenotypes are related to the ability of an HIV-2 isolate to use certain co-

22 2 HIV-22 IN WEST AFRICA receptorss for entry into the cell (e.g. CXCR4 and CCR5); HIV-2 may use a wider range off co-receptors than HIV-1 [10-12]. For a recent review of the immune responses to HIV-22 see Whittle eta/. [13].

Detectionn of HIV-2 infection

Detectionn of HIV-2 infection is based on the demonstration of virus-specific antibodies (anti-HIV-2)) reacting to viral antigens, mostly using enzyme-linked immunosorbent assay (ELISA)) based techniques. In HIV-2 endemic areas die ability to differentiate between singlee infection with either HIV-1 alone, HIV-2 alone, or dual infection with both viruses,, is important. The frequent occurrence of cross-reactions makes virus speciation onn serological grounds problematic, though antibodies to Env, particularly in the regions withinn the transmembrane part of the envelope protein (glycoprotein 41 [gp41] in HIV-1, andd gp36 in HIV-2) elicit the most type-specific response [14-18].

Monospecificc ELISA's are designed to detect only antibodies directed to antigens of one viruss type, but cross-reactivity in these assays is not uncommon [19]. Synthetic peptide- basedd indirect ELISA's, like PEPTILAV 1-2, or INNOLIA HIV Confirmation, both appearr capable of discriminating between HIV-1 and HIV-2 in the majority of individuals,, and are used in many West African research settings in lieu of Western Blot [19,20].. The Western Blot (WB) itself, which was at one time regarded as the gold standardd for HIV diagnosis, has been shown to have severe limitations due to the high degreee of cross-reactivity between HIV-1 and HIV-2 [15,16,21,22]. Adoption of more recentt recommendations for interpretation of HIV-2 WB may lead to less false-positive HIV-22 WB results in HIV-1 positive samples [16,22].

Evenn by the more specific line immuno-assays, by a combination of ELISA's [19,22], or byy dilutional analysis [23], dual sero-reactiviry iss not proof of dual infection. Only the isolation of both viruses from the same individual,, or demonstration of proviral HIV-1 and HIV-2 sequences in peripheral blood

23 3 HIV-22 IN WEST AFRICA mononuclearr cells (PBMC's) by PCR constitutes definitive evidence of dual infection [24].. True dual infection is not uncommon in several West African countries [24-29]. However,, it is difficult to isolate HIV-2, especially from asymptomatic individuals [5,30- 33].. The proviral load in asymptomatic subjects with single HIV-2 infection can be very loww [31,34,35], and nested PCR can be repeatedly negative in some subjects [33]. Consequently,, a failure to isolate virus or to demonstrate the presence of HIV-2 DNA in duallyy sero-reactive patients does not necessarily exclude infection with HIV-2 and hence indicatee HIV-1 single infection. Recent studies showed that a failure to detect proviral DNAA was associated w?ith low CD4+ counts [23,27]. More sensitive PCR techniques are noww able to detect viral sequences of HIV-2 in 95 to 98.4% of HIV-2 singly infected subjectss [33-35]. In two recent studies employing new PCR methods DNA of both HIV- 11 and HIV-2 was detected in 72% [29] and 81 - 90% [36] of dually sero-reactive subjects.

Inn summary, in the majority of individuals HIV-2 single infection can be diagnosed confidentlyy with serological assays in an appropriate algorithm. Dual sero-reactivity is mostt suggestive of dual infection though subsequent confirmation with sensitive genome detectionn techniques is required. In some cases confirmation cannot be secured.

Epidemiology y

Geography Geography Thee first cases of HIV-2 infection were found in people from West-Africa [1,2,37], and largerr studies confirmed that West Africa is the epicentre of HIV-2 [38-45]. There is prooff that HIV-2 existed in West Africa already in the mid 1960s [46-48]. The highest HIV-22 prevalences have been found in Guinea-Bissau, both urban (8.9% among adult residentss in the capital Bissau)[40], and rural (7,9%) in a community-based survey in north-westt Guinea-Bissau [49|). 'Origin of Guinea-Bissau' was a significant risk factor for HIV-22 infection in female CSW's in Ziguinchor, Senegal [44], Many places outside West Africaa with HIV-2 cases have historical links with Portugal (Brazil, Angola, Mozambique, India).. The recently observed decline in prevalence in Bissau suggests that previous levels

24 4 HIV-22 IN WEST AFRICA mayy have been boosted by the Portuguese colonial war in Guinea-Bissau (1963-1974) [50].. A geographical analysis of HIV-2 cases detected in Europe, showed that a massive majorityy of patients were West Africans, or Europeans with West African sexual partners [51]. .

Guinea-Bissauu continues to report the highest prevalences in pregnant women (4.3% in thee capital in 1995 [52]; 9.2% in mothers attending modier-and-child heakh clinics in Bafataa province [53,54]). In other West African countries prevalences among pregnant womenn range from 0.5% to 1.6% in recent studies [55-58]. In female commercial sex workerss (CSW's) in West Africa recent prevalences of HIV-2 and dual seropositivity rangee from a low 8.1% in Mali [59] to a high 27.5% in The Gambia [60-63]. It is striking thatt Sierra Leone, the country widi the highest diversity of HIV-2 subtypes [5,6], has the lowestt seroprevalence of the sub-region: 0.02% in a survey in 1993 among 9309 rural subjectss [6]; among 359 adult TB patients in 1994, only one case of HIV-2 was detected [64].. Recent large surveys from Cameroon [65], Equatorial Guinea [66], Gabon [67,68] , andd the Democratic Republic of the Congo [69], found no HIV-2 and dual infections, or sporadicc cases only; this suggests that HIV-2 has not spread substantially to other regions off Africa.

Inn North America and Europe HIV-2 has not got a strong foothold: in die United States 677 cases have been identified up to 1996 [70] quoted by [71]). In Europe most countries hadd a cumulative count of less than 100 cases by the mid 90's (England and Wales 33 casess [72] quoted by [73], France 98 [74], Spain 72 [75]). In Portugal HIV-2 prevalence is considerablyy higher: 13% of all HIV positive STD patients and 29% of all HIV positive TBB patients are HIV-2 infected [76]. In South America HIV-2 remains rare [77-79]. A surveyy among 3525 prisoners in Pakistan in 1994 failed to detect any HIV-2 infections [80]. .

Theree are two possible exceptions to this observation: Zimbabwe and India. A study amongg pregnant women in Harare, Zimbabwe, found a prevalence of 22.8% for HIV-1, andd 7.6 % for dual infection, but not one single HIV-2 infection was detected [81].

25 5 ra ra SS o v) (DD > > ra -Il l CDD c >« «« E ©© C O)) o STJU U cc to g ££ , SS S tótó (B "O rara O t»» c -ZZ ® © O'S S COO ra * * > > ~~ 3 e "OO .&> m (DD E ©CDD a ww .2 -9ïuu|| . SS 3 » o. o. fJJ « *i es gg « S >.. i © aarara <^o *sr o o o^^ a x:: p ra ro 5 cc o (—— ifc O. © CO

,—.. co r-- coo ^j-" „ coo 2L OO Oï O) CDD m T- T- en LO

££ h- co CNN £!

oo o o OOOO i-OOO oco

oo o o oo o o oo o eg co

coo m *

o o © © 'S 'S ra ra > > 3 3 © © 3 3 (0 0 «0 0 10 0 CL L o o T3 3 3 3 O O MII co © © J22 o OO co rara ~ co > ra ra Q.T3 3 ©© 9--Sc C/ï ï CL L o>> E © CN N ra ra ©© E cIoos o s >. O O rara >. F ~ © © »» =J — == « >,co > > 1> > rara co 99 ra < < coo < << « X >> < w a I I T3 3Li -- ü-- < << > _>O.» O.» n n 3 3 22 >>>^ (/5) -—. > (Tl > > O)) O) «« ^cn cn X X

ra ra © © SS m ra ra ©"" o 5 © © §Ü.. « '5 58 8 CO O ®® c £ -1 -1 _jj t ra > > O O mmrara © ^ CD o o ë ë 2. 2. © ©< < c c ja a ^^ -e = to oS ïï co o o (O O « -J = -^ O 3 « -cc o o o© © © © l-- 33 co £DÜÜ OQÜZÜ CD ©© O CO OO O O O Ï2.S S CD D COO Z S 5 5 > ra < < T3 3 T7 7 CC 3 o o > F >> QJ .> > o o a>> ra O O o.. co Iel lÜ Ü CLL C c c .C C oo J> 5 rara £ 03 3 CUU § 5 5 22 * F F OO Q r r o o rara "ra <2 E E 00 0 2 2 ra ra cc => E Da" " OOee tet- o o F FO i i LÜ>CQ Q—** —££ Ll_ c c I"-- >. . c c CU UO l l cuu ra cc = ww cu CO O 03 3 EE 3 c c (1) ) ra ra (O O 1 1 D D 81 Em-^ ^ o o O) ),r az> z > 777 P c c S =»-"? ? «tt 0) 33 C T T O O 22 13-5 c c ra ra a> > c c ^^ fc ra cc .c > cc ra 'ra a ^» » O.. CO £ O O "CC P -- c c ra ra 0) ) T>3 3 oo> >fc JS 33 « ra rara ra c c > > 055 ® "raa = c c CL L n n CU U -- J= o x: : o o ra ra Q.. C 0) c cra ra CO O .55 cö c Q.. O O "ra a 33 o g> O O == 3 o "-- ra O O coo m 5 mm ücoo cu <

to to coo ^- 000 co -- co

^^ ^

C C CM M CO O coo o C C o o > > Ol l CD D JU U ra ra X X Q. . ra ra E E CO .e e o o rara -ocQ oo.. -B O O E E t—t— n r— oo c

"ra a 3 3 COO c EE .2 cc "i i E E CN N .® o o rara -o«2 . . rara =e t t « « °88 W "533 T3 J^"c> c OO O CC CU -2> 33 o co ZZ CD E< < SX^X X -- 3 C << s Q.. < CO
55 3 22 5 *2 33 3 "raa ra rara ra rara ra "raa "ra -oo -o -o oo -o Oï ï r-- co o O) ) oo o o oo o o> > en n ccc c d d
OO C "O s an >> c- ra b e
.fcc '5 <2 o << O o mm Pari th
cc -O f ,, d C C Bom
o** ra ra ra ra Bost
0) ) CU U o
Q.. O) O) O) ) Netherlan ,, ££ §>=> o o ,, ww <ü (0 oo => = c c -o o tt 0)) c « e e ee
cc -c -c ^ ^ co o co o co o dia otterda es SA nglan h ^^ CD .se 33 O O o o CL L CL L rance cu u rwco o LUU CL O. CL L to o CO O LL L 1-- CL LL LU CO O KD£ £ CO O >i i ü ü CO O o o CD D >, >, CO O c c D Do o CM Mo o > > X X
CMM CD i- cons s inn en oqooojl^ n n OOT-- öö ö o CMM T^ kri
UJ J Üz Üz Ui i _l l Ifii-BS S oHs^coomO'-sn n ^^ N-u ;;[ƒ) O) CM CMM T- T- T~ < < Ti-icnVo o -- CO ^ > > od^rîôdcbiricb^rcci'* *C MM CM i- CM irii --^CftCD^-0OCM M __ 00 UJ J co o 6^^ ddöd^rdfjuiai
SOJOKOSCO^CMCOTtmOOmO55 coi, o t(DO>MnSOOOrNm(DS a3oommoooOffiCBC»01o00105<'i01010iO>SSS0;,000)<'3050)t'5050)0)fJ>05 5 (JT-T-r-T-^^gjCSÖlr^T-T-T-T-r-T-T-T-T-T--
CMM CO 0>CO N-g^ g^g&lfi'*sr>-K)co r»» m CO 'T w n- UJ u>loôp'-co'j^sptolfi'-SSS ,,-* T-- 00 Tf i S S ,11 CO CM CMT-T-TTT 2 ^^ «o S E ló
o o 2 2 ra ra o. .
Cl) ) E E co o 5 5 o o c c co o -s s 0 0 o o co o o o co o Q. . co o c c Q Q E E u u 0 0 c\-\-o o LL L LL L
n n o o < < JÉ É _ > _ co o n n J3 J3 co o >-- CD D E E .o^ o ^ ct ctm m F F E E co o ra a co o co o CD D T3 3 Z Z fii i _> > CS S O O ._r r 3 3 c c a> > a> > a> > Ü Ü3 3 CO O o o ü üO O O O 2» » ü ü o o 2" "
o o coCT>5a>coa>a>r'^cDr*-(o£i^L£2 2 cd d 000 o co TJ- CM co COO CO CM 1- T- T- SII i LU U O O
00 0 ^•COlO'^-'^-COCOCMCMCNICMh-lO^-Tl-- Tff CO CM CM CM i- CM en n 5 5 co o UI I fl fl 0£ £ o o O. O. o o I I o o O O c c Q£ £ (ONfflNioiniotoscocosonmco o O!! h- CO h- C7J 00 _g g COO > "tff rf T* CO CO Tf n n X X T-T-CMCOcOcOCOCOCOCOCOCMCO-fl-rJ-ir) ) > > 0 0
NOOC50T-(N(*)^miDSN(00 0'-CMlO(Nrtînï)S cooooocAO>cncncbo)c7)o>cococj)CJ)a?a)cnc7)cncnc})cn n cncnCTO)0>cno)
XT T ÖÖ§Tro)coco<<<< ^ w V UM'J tu 's, M, M, . - [J (\l CM iri f« zo> zzzz ss<<<<<< < (0 0 O O sss!;fcE3 ^^^^^cM M O. . T—— co CO CO CO ' CO CM ^ 0> >
m m
3 3 -2 2 Oa. O. TJ J 0. . < <
c c Q> > ca a 3 3 o o <** * -Q Q C C < < oO Oo a> > K>-- K ~o o~ 1-- Z Z x> > 3 3 CD D O O O O ü ü O O HIV-22 IN WEST AFRICA
Althoughh the authors mention that Western Blot and recombinant ELISA were used for confirmation,, the precise criteria for diagnosing dual infection are unclear. The startling findingg of 89 dual and no HIV-2 infections, combined with the absence of any other recentt reports from Zimbabwe on HIV-2 cases (e.g. on the 12th World AIDS
Conferencee in Geneva), suggest that cross-reacting HIV-1 antibodies may have been responsiblee for the high number of dually reactive samples, and that HIV-2 is very uncommonn in Zimbabwe.
Studiess from various states and cities of India have documented HIV-2 infections, mostlyy in high risk groups like STD patients and professional blood donors [82-85].
Somee authors claim that HIV-2 is spreading rapidly in India [86-88]. However, there are noo longitudinal studies demonstrating a rising prevalence or documenting HIV-2 incidencee [89-92]. A phylogenetic study suggesting a 'recent entry' of HIV-2 into India wass based on only five samples [86]. Therefore a conclusion on rapid spread of HIV-2 in
Indiaa is unjustified: the prevalence is low, and the incidence is unknown.
HI\HI\ '-2 Subtypes
Dataa on the prevalence of HIV-2 subtypes are scarce. There are only three epidemiologicall studies, one from Mali [59], and two from Guinea-Bissau [93-95]. These studiess only found subtype A (see table 2). Apart from the original studies describing subtypess C, D, E, and F for the first time [5,6], only one other case of D was found [96].
Fromm table 2 can be seen that subtype B has been found only in the eastern part of West
Africa,, or in subjects with links to that area.
Alll currently identified HIV-1 subtypes are pathogenic. In HIV-2 subtypes A and B are pathogenic;; for the other subtypes this is not known. Subtype A is compatible with fast progressionn [32], as well as with long-term non-progression [94]. Therefore factors other thann subtype must be important in determining the course of the infection. A study amongg 131 HIV-2 infected people in Guinea-Bissau found a correlation between multiplee mutations in the viral genome and mortality [95]. Prospective studies are needed whichh describe the relative distribution of the subtypes, and which compare the natural
30 0 HIV-22 IN WEST AFRICA
history,, as well as the virological and immunological course of the infections with differentt subtypes. Ideally these studies should be done in settings where more than one subtypee occurs.
StableStable prevalence and incidence HIV-22 has been endemic in West Africa for an unknown period. In countries where repeatedd cross-sectional data are available, HIV-2 prevalence has been stable or declining; inn the same populations HIV-1 prevalence has increased (see table 3). These trends have beenn observed in low risk groups like pregnant women in Guinea-Bissau [52] and Cote d'lvoiree [97,98], and in high risk groups like CSW's in The Gambia [43,60,99,100]. Projectt RETRO-CI, in Abidjan, Cote d'lvoire has collected an impressive set of longitudinall data since 1987 from 10 risk groups [98]. The HIV-2 prevalence has been remarkablyy stable in most. At the same time, HIV-1 has risen and more recendy plateaued,, for example among TB patients and pregnant women. The decline of dual sero-reactivityy among female CSWs over the years is striking (from 38% in 1992 to 13% inn 1997).
Incidencee data are available from a cohort of female CSWs in Dakar, Senegal, over the periodd 1985-1994. This demonstrates a stable HIV-2 incidence and a clearly rising HIV-1 incidencee [101,102]. Two low risk cohorts in Guinea-Bissau experienced a decline in HIV-22 incidence and a rising [103] respectively stable [52,104] HIV-1 incidence (see table 4). .
Thesee data show that HIV-2 is an endemic infection in West Africa, which may be slowly declining,, and that HIV-1, previously unknown in many parts of West Africa, is increasingg and overtaking HIV-2 as the dominant immunodeficiency virus. This suggests aa lower transmissibility of HIV-2 by the heterosexual route [105,106], and agrees with the outcomess of a mathematical model, which suggested that in a population where both virusess are being transmitted sexually, HIV-1 will competitively displace HIV-2 in the longerr term [107].
31 1 HIV-22 IN WEST AFRICA
Tablee 4. Incidence rates of HIV-1 and HIV-2 in three cohorts
INCIDENCE E expressed d per r10 00 pyo
COUNTRY Y POPULATION N SIZE* * PERIOD D HIV-1 1 HIV-2 2 REFS S Guinea-Bissau u 1000 houses in 330 0 1987-89 9 NA A 1.0 0 [137] ] semi-urban n areas s 209 9 1989-92 2 NA A 0.3 3 20500 pyo 1987-91 1 0.04 4 NA A [103] ] 20899 pyo 1991-6 6 0.64 4 NA A
Guinea-Bissau u Policee officers 819 9 1990-2 2 0.7 7 1.6 6 [104] ] (Bissau) ) NA A 1990-5 5 0.8 8 0.8 8 [52] ] Senegal l Femalee CSWs 1877 pyo 1985 5 0 0 1.1 1 [101] ]
(Dakar) ) 4488 pyo 1986 6 0 0 1.1 1 5522 pyo 1987 7 0.7 7 0.9 9 6099 pyo 1988 8 0.7 7 0.7 7 7044 pyo 1989 9 0.7 7 1.1 1 7322 pyo 1990 0 1.2 2 1.6 6 6400 pyo 1991 1 2.8 8 1.3 3 2744 pyo 1992 2 2.2 2 0.7 7 NA A 1993 3 1.3 3 0.3 3 [102] ] NA A 1994 4 2.1 1 0.9 9 NA A 1996 6 3.9 9 0 0 [187] ]
Notes:: pyo = person-years of observation; NA = not available; CSW = commercial sex worker; REFSS = references. * in numbers of subjects, unless indicated otherwise.
A.geA.ge distribution Thee age peak of HIV-2 prevalence is distinctly higher than that of HIV-1. The highest prevalencee in women is usually seen in age groups 35 — 45 [42,49] or 50 - 59 years [40], andd in men around the age of 50 [40,42,49]. A recent study among residents older than 500 years in a semi-urban area of Bissau, GB, found the peak ages to be as high as 60 to 699 years in men (13.1% positive), and 50 to 59 years in women (18.0% positive) (Larsen, manuscriptt in preparation). This is in contrast to the peak age for HIV-1 in sub-Saharan Africa,, which is 20 to 34 years for both men and women [108-110]. The higher peak age off HIV-2 reflects several aspects of HIV-2: lower heterosexual transmissibility, lower mortalityy rate, and earlier introduction in the populations studied. It is possible that a cohortt effect is responsible for the gradual decline of HIV-2 among men in Bissau [50]
32 2 HIV-22 IN WEST AFRICA
Tablee 5. Strength of the evidence for HIV-2 transmission routes
Modee of Qualityy of Efficiency y Transmission n supporting g compared d References s evidence e too HIV-1 Heterosexual l ++ + 1/55 to 1/9 [188] [62,101,111,117,175] ] Homosexual l + + ? ? [74,75,77,189-192] ] Mother-to-child d ++ + 1/7 7 [119,121,122] ] IVDU U ? ? [75,76] ] Bloodd Transfusions + + ? ? [40,134,136,157] ] Needlee stick Injury -- ? ? [193] ]
Notes:: IVDU = intravenous drug use; - - = no evidence; = circumstantial; + = some evidence;; ++ = convincing evidence; ? = unknown
Ann analysts of risk factors for concordance of wives to their husband's HIV-2 seropositivityy in a rural area in Guinea-Bissau, concluded that age > 45 years of the wife wass the only significant predictor [111]. This finding, plus the observed higher incidence amongg women > 45 years in an urban setting in Guinea-Bissau, raises the question whetherr older women are more susceptible to retroviruses in general.
ModesModes of transmission <& risk factors Evidencee for the classical transmission routes is strong for vertical and heterosexual transmission,, good for blood transfusion, circumstantial for intravenous drug use (IVDU)) and the homosexual route, and lacking for needle stick injuries (see table 5). It is plausiblee that also the latter three modes of transmission are operating, but formal studies demonstratingg this are lacking.
Thee efficiency of heterosexual transmission is theoretically dependent on the amount of viruss in cervico-vaginal secretions or semen, and the local specific and non-specific immunityy in the female and male reproductive tracts [112], In Cote d'lvoire investigators amplifiedd viral RNA in cervico-vaginal secretions (CVS) of female CSWs [62]. In 24% of HIV-11 infected women (96 / 404) virus was found in CVS, but only in 5% of HIV-2 infectedd women (1 / 21). In dually infected women 26% (53 / 205) shed HIV-1, and onlyy 3% (6 / 205) HIV-2. In HIV-1 women the shedding of virus was significandy more frequentt in those with low CD4+% and higher plasma viral load. Due to the low
33 3 HIV-22 ÏN WEST AFRICA numberss this could not be analysed for HIV-2. Two smaller studies done in female CSWss in Dakar, Senegal, gave similar results [63,113]. These findings may explain the differentiall transmissibility between HIV-1 and HIV-2. It is likely that the lower shedding inn HIV-2 is correlated to the lower plasma viral load of most HIV-2 infected women. Thee serological concordance of female spouses of a group of 48 hospitalised HIV-2 infectedd men having advanced disease, was 44% [114]. Comparative studies on viral sheddingg in semen are needed.
Amongg the established risk factors for HIV- 2 infection are older age [49,115](Larsen, manuscriptt in preparation), history of sexually transmitted diseases (STD's) [42,45,116,117],, serological proof of genital ulcer diseases [99,116], having had sex with a femalee CSW [117], having been or being a female CSW, lack of circumcision in males (Larsen,, manuscript in preparation) [117], and various proxies for increased number of lifetimee sexual partners [44,115].
Mother-to-cbildMother-to-cbild transmission AA serological diagnosis of HIV-2 infection in a child cannot be made before the age of 155 months due to the presence of maternal antibodies; with PCR techniques or virus isolationn a diagnosis can be made before that age, though the sensitivity of these methods inn infants is not known. Mother-to-child transmission of HIV-2 had been found in severall epidemiological studies [118,119], and has recendy been convincingly demonstratedd in a molecular study [120,121], Most comparative prospective studies indicatee transmission is a much rarer event than in HIV-1: less than 5% versus 20 - 25% (seee table 6).
Thee largest community-based study, which was conducted in The Gambia, found a mother-to-childd transmission of 4% (8 infants of 201 HIV-2 infected women; 95% CI: 1.77 - 7.7) [122]. Two children who were HIV-negative by PCR at 9 months of age, becamee HIV-2 positive by PCR at 18 months of age; late postnatal transmission thus mayy have accounted for 2 / 8 of transmission events (H. Whittle, personal communication).. No cases of late (> age 6 months) postnatal transmission of HIV-2
34 4 HIV-22 IN WEST AFRICA weree detected among 122 children of HIV-2 infected, and 49 children of dually infected womenn in a study from Cote d'lvoire [123]. These late post-natal transmission rates of 1.0%% and 0% compare to 3% for HIV-1, calculated in a pooled analysis of three African studiess [124].
Thee Gambian study also showed that the HIV-2 plasma viral load in mothers who infectedd their children was significandy higher than in those who did not [122].
Naturall history & prognosis
NaturalNatural history in adults HIV-22 infection can lead to CD4+ decline [125,126], opportunistic infections [127-130], HIV-associatedd malignancies [129,131,132], and early death.
Thee median time from infection to AIDS is longer than in HIV-1: in a cohort of women inn Senegal with known date of seroconversion (n=131) the incidence rate of AIDS was 0.955 (95% CI 0.2 - 3.8) per 100 person-years of observation (pyo) among HIV-2 infected women,, and 5.6 (3.3 - 9.8) in HIV-1 infected women [125]. Cases with an incubation periodd of 14 years, and even 27 years, have been reported [133-136]. It is thought that onlyy a small proportion of HIV-2 infected subjects will develop immunodeficiency and AIDSS [137]; a large proportion of HIV-2 subjects may be categorised as long-term non- progressorss [136,138,139].
Inn a community-based study from Guinea-Bissau with 9 years of follow-up, the mortality rate,, adjusted for age, among HIV-2 infected subjects was twice as high as among HIV- negativee subjects (mortality rate ratio 2.3, 95% CI 1.2 - 4.6) [137]. This compares to a mortalityy rate among HIV-1 infected subjects of 9.7 times higher than in HIV-negative people,, in a study in a rural area of Uganda [140].
35 5 s s CNN C x: : o o o> > cb b XX co .c c EE 2 «I I oo o "c c coo c 99 03 ^ COO X o o »» Q. C 2 CDD (0 COO CD O O Q^^ ca coo Q- 22 & gg 3 (0 D) == Ë rara Ög1S EE 1 « —— 5 CLT- OO ? 0)) * coo c o££ CC oo"«$( o 55 CL-gEE © £=§§ o> DD 51 .oo o S3 3 CNN ifc! raX Jll £ ?
,,w o ^^ ir» *77»» '5CA5 CDD «O £có ó (NN ^.'
óó 2 2 ss vpp CN 0 I O*"" T— 11 2 = CNN CO oo *- o OO i "** , *** £ TJ '55 "** r-- >> o==.!2 II Eo E
ZZ o X o n o
~~ $ © a
ïï CB E E o cc S 5
00 0 CUU lO 00 0 O O ir»» > i- CN N oo c T~~ .« a> == 2 >> O "CD D 'üi'üi CO "" o P «« o ^3 3 coo ^J 9>> CD &^© © »» o» (BB 5)5 A) £ 3 .22 § ^^ .2 22 ra >> Q i- .> «« 2 o»» 2 oo 2 OO 01 cc S o> OO 0> OO 03 Q.. CO isilsig gfc:fc: CL co OO .E o
T3 3 1(^0- 10 1 ^j j 0) ) (O O OJ J CDD r CP Pro ro 922 ra 55 Q_ c c c c il l > >c c ,o> o> (DD ^ o oa > > i5 5o o0 0 ass ra O O CLL C n nE E0 0 CU U n nE EC D DCL' -? ? E E (/)) «J O O o oo o coo d co oo o ">> 9- OO £i o oc c1 >) )> oo o"ö >j j>o oo oc c7> 5 5>22 ^ CN N Q.33 2< CL L ü ü -C CC L L0 0 T3 3 > > o o u u o.^: :
co o o oC N N r^-- CD Drj -- 01 1 9 9 O) ) CT) CT) CT) 00 0 00 0 CD D O) ) CD Da > > co o a> > ,cn n o> > cI—n n— S S '" " '" "'~ ~ ^ ^ ^ ^~ ~ ' V) )
c c CD D "CD D il l ra ra 2 2 D D CD D 03 3 ra raCD D CD D c c .o o 03 3 Si Si'ö ö co o c cC O O 3 3 E E co o CO O ra ra CD D "o o '3 3 1r-- rO O << OT >3 3> inn O m Li. . COL OL CDD ni c c -00 co o .o o488 c OO . befo CD D oo ® ma E E CO O EE CO CO *;; o o c jZ —
exclu COO CD
yy CO O E« « yy == "O'c o o t > > i
d CO O 1§Ü> > r «« CO C Ü CJ)) -o —3 3 — o oC U U E E CO O CD D CD DC O O5»' cc 'to 55 IS°° c « w to a>> a>c c stu o o o o n nc c —— .. CD » 52
co o ff CO O CQ Q c c en n too CD O oo Q. CO o .2 mothe mortal
o .Soi2f f §1 1 CD D 3 3 CL L .fc cDT 33 a- QQ CO fc .E CC 3 CD Q.JD
>> m I .2 ïï 5 E « Ê^ ^
*** " <ïï i ^ » ,2 e mm i ^^ £ -C (O tf) W EE e ö ö
'' c 3> o -o OO S O O O
3 3 vv 8 J
CC — "Ö C. ïï"Q . a3 OO X CU © E T
33 > 3
00 0 oo c *** CD CD D CDD CD A A CO O 0>> TO CDD S; »» en C>O. O. f-»».. O>») .11 1 OO A ^^ .£> .>> _o a> > SS o > >o o .>> o o 'èfll 2 CD aa 2 "'SS 2 0 cö ö? ? ipt t ill * ODD) ) OO CD 99 <» c oo =: O OC D D 'SS 2 o CLL CO CO Q.. CO onCLL COs Ts- CLL CO C CL Lco o ee <» c "Si i Q.. (0 t "o , oo .E E E CO O CD D o o -- CD it ."Ü CDD -a CD E E I? ? o o SS ra 2? ? OO > CD CDD V COO ï- c£ £11 % E >** CD a>> co ii CD too = CDD CL oo « s 5 fS-S S CÖÖ CL~ ~ 22 ^. 22 O O £ ÜÜ C ££ OO"" -O ^ ** t- .fa n öö ~ o 2 § 33 Cl 22 g uu 1:
l-v-- co coo o O)) CD
CD D foo co 33 10 3 c CD D 3"" W 3 = In n CDD nn g co oo . „ O O ^JDD O coo .E co ^^ Ü o coo 92 co ] coo 3 co CC i_ CO CO O COO .E CO ai c .cc co 55 O > COO 3 CO «« 3 CO ^ CO I-O O inn O S COO 00 LL LL L COO CD CO I- CD HIVV 2 IN WEST AFRICA
Thee clinical manifestations of symptomatic HIV-2 disease are broadly similar to those of HIV-11 [141](reviewed recently by Grant et al. [129]). The only marked difference may be thee lower frequency of Kaposi's sarcoma (KS) among HIV-2 patients [142].
Hospital-basedd studies in The Gambia and Guinea-Bissau have demonstrated a high mortalityy among symptomatic HIV-2 patients [143-145]; however, in a Gambian study thee survival for HIV-2 patients with AIDS was 3 times longer than for HIV-1 AIDS patientss [143]. A case-control study among African patients living in Paris, demonstrated alsoo significantly longer survival for HIV-2 patients after AIDS diagnosis: 4.1 years versuss 3.1 years for HIV-1 (p=0.036) [131].
Mostt studies demonstrated a decreased CD4+% or CD4+ count in HIV-2 subjects, also inn clinically asymptomatic people [139,146,147]. Lisse et al. found no decline over a 4 year periodd among 20 HIV-2 infected subjects [139]. Jaffar et al. compared the rate of CD4+ declinee between HIV-1 and HIV-2 patients in The Gambia; the proportion showing no declinee in CD4+% was significandy higher in the HIV-2 group (47% versus 32%); and thee rate of decline was slower in HIV-2 patients (-1.08 vs. -2.1 CD4+% per annum) [126]. .
Deletionss in the nej gene render HIV-1 non-virulent, and lead to long term non- progression.. A recent study of HIV-2 samples from Cote d'lvoire, Spain, and Portugal, foundd iief deletions in 14% of asymptomatic, but only in 4% of AIDS cases [148]. Comparativee studies of HIV-1 and HIV-2 are needed to determine whether tiej deletions mayy be partly responsible for the lower virulence of HIV-2.
Thee range of HIV-2 proviral load is similar to that in HIV-1 infection [35], and the provirall load is inversely correlated with CD4+% [31,34,35] [93] [149]. More advanced clinicall stage is also correlated with higher proviral load [34,93]. High proviral load and loww CD4+% are strongly associated with mortality [35].
38 8 HIV-22 IN WEST AFRICA
Inn 1993 Simon et ai. demonstrated a correlation between plasma viral load and CD4 count,, using culture methods [31]. A novel RT PCR assay, specific and sensitive for the quantificationn of HIV-2 plasma viral load, was recently employed in a study of Gambian patients.. Both HIV-1 and HIV-2 plasma and cellular viral load were inversely related to CD4++ count. The proviral load of the two infections did not differ in any stage of infection.. Although the plasma viral load of both infection was similar in the lower CD4+%% categories (<14% and 14-28%), it was undetectable in three quarters of HIV-2 patientss with a normal CD4+%; in contrast, HIV-1 plasma viral load could be detected inn all HIV-1 patients with normal CD4+% [149]. The threshold of detection was 500 copies/ml.. No study has formally examined the prognostic value of plasma viral load for survivall in HIV-2 infection.
Thesee findings are compatible with the thesis that a minority of subjects have progressive HIV-22 disease. They present to hospital and are characterised by detectable provirus, CD4++ decline, detectable plasma viral load, clinical symptoms, and early death. The majorityy are long-term non-progressors, who live in the community unnoticed by health services,, and who have detectable provirus, stable CD4+%, no detectable plasma viral load,, are asymptomatic, and have no increased mortality risk. What determines the peacefull co-existence between HIV-2 and the human host has yet to be elucidated, but couldd be related to genetic host factors (HLA type, co-receptor polymorphisms), or to mutationss [95] or truncations [148] of the virus genome.
Occasionall HIV-2 patients in Europe, are treated with HAART [32,150]. However, no reportss of clinical trials with anti-retroviral drugs in HIV-2 infected patients have been published. .
PaediatricPaediatric HÏV-2 infection HIV-22 infection is rare in children; in 12 references we found data on only 63 children [118-120,137,151-158].. The two main modes of transmission are from mother to child andd through unscreened blood transfusions [118,137,157]. The course of HIV-2 infectionn in children seems to be slower than that of HIV-1 infection, with a lower
39 9 HIV-22 IN WEST AFRICA mortality.. Among eight perinatally infected children in The Gambia, Ota observed no deathss until the end of follow up, at 18 months of age. This compared to the death of 6 /
177 HIV-1 infected children [158]. Several children have been described who were still alivee 16 years after infection [156,157].
HIV-22 and HIV-1 interactions
Couldd HIV-2, seen as an 'attenuated' HIV-1, provide protection against subsequent infectionn by HIV-1 [159]? This challenging question has led to considerable research efforts,, both in the lab and in the field.
InIn vitro studies
Severall laboratory7 studies have found indications that HIV-2 infection may lead to immunee or other responses against HIV-1. Neutralising antibodies from HIV-2 infected subjectss were shown to cross-neutralise HIV-1 strains in early studies [160,161],
Cross-reactivee Cytotoxic T-lymphocyte (CTL) activity between HIV-1 and HIV-2 epitopess has been demonstrated [162,163]. Bertoletti et at. found that 9 out of 11 HIV-2 infectedd patients had cross-reactive CTL, which was able to recognise Gag proteins from variouss HIV-1 clades. Four out of five patients with broad cross-reactivity were HLA
B58011 positive, which is a very common allele in The Gambia and Senegal [164].
InIn vitro experiments showed that HIV-2 was able to inhibit HIV-1 expression in a dose- dependentt manner; this was not due to an immune response but to down-regulation at cellularr level. This finding points more towards a mitigating effect after established infectionn of HIV-1, than towards an effect of preventing infection [165].
Schwartzz et al. [166] found that 11 of 18 PBMC cultures of 9 HIV-2 infected subjects weree protected against subsequent infection with HIV-1 BM_, though only 3 out of 18 were resistantt against a challenge with HIV-1 xtN.
40 0 HIV-22 IN WEST AFRICA
Tablee 7. Impact of HIV-2 infection on subsequent HIV-1 infection in three West African cohortt studies
Studyy site (reference)
Dakarr [170] Thee Gambia [173] Bissauu [103]*
Population n Femalee CSWs Femalee CSWs Loww risk M & F Cohortt Size HIV-22 positive 199 9 53 3 }total l HIVV seronegative 398 8 189 9 }729 9 Incidentt HIV-1 cases perr pyo HIV-22 positive 7/780 0 8/92 2 77 / 388.4 HIVV negative 444 /2020 19/369 9 77 / 3751 Powerr calculations Ratee Ratio limits detectablee at 80% power r Riskk decrease <0.27 7 <0.01 1 <0.01 1 Riskk increase >2.0 0 >2.9 9 >7.5 5 HIV-11 incidence rate perr 100 pyo (95% CI) HIV-22 positive 0.9(0.4-1.9) ) 8.7(4.3-17.4) ) 1.89(0.45-3.33) ) HIVV seronegative 2.2(1.6-2.9) ) 5.11 (3.3-8.1) 0.19(0.05-0.33) ) Incidencee Rate Ratio 0.26(0.09-0.7)** * 1.7(0.7-3.9) ) 10.11 (3.5-28.9) Weaknesses s Biass in selection of Noo adjustment for Noo proxies of HIV-negative e confounding;; large exposure e CSWss [172] lossess to follow-up
** This paper is reporting on the same study as Aaby ef al. [198], but with a longer follow-up; *** adjusted for age, nationality, years of registered prostitution, gonorrhoea infection status, andd calendar year, pyo = person-years of observation; 95% CI = 95% confidence interval; CSWW = commercial sex worker; M & F = males and females. This table is an adapted and updatedd version of a table from Travers et al. [199].
Kankii et al. [167] found that PBMC's of HIV-2 infected patients demonstrated >90% inhibitionn of one particular HIV-1 strain (JRCSF); none of the 13 seronegative controls showedd a similar inhibition. However all HIV-2 positive and HIV-negative patients were equallyy susceptible to HIV-1 strain Illb. The protection was mediated by beta- chemokines. .
Belecc and colleagues studied the ability of antibodies of the IgG and IgA type in cervico- vaginall secretions (CVS) and serum of HIV infected women to act against envelope proteinss of HIV. They found diat IgG of HIV-2 patients were significandy more often
41 1 HIV-22 IN WEST AFRICA reactingg to the HIV-1 antigen gpl60 than antibodies of HIV-1 patients to the HIV-2 antigenn gpl05 both in serum, and in CVS. IgA of HIV-2 was also reacting more often to thee HIV-1 antigen than the reverse, in both serum and CVS [168].
CohortCohort studies Itt should be noted that the correlates of protective immunity against HIV in humans are nott elucidated, and the relevance of the above findings to real life situations is therefore unclear.. Epidemiological studies done so far have yielded contrasting results. Three such studiess have been done, and all employed PCR methods to diagnose dual infection; data fromm a fourth study can be analysed as well for a protective effect, but this study did not employy PCR.
Traverss et al. studied a cohort of 756 female CSW's in Senegal, from 1985 till 1994 [169]. Thee HIV-1 incidence rate among the initially seronegative women was 2.53 per 100 personn years of observation (pyo), and among the HIV-2 positive women 1.06 per 100 pyo.. The crude incidence rate ratio (IRR) was 0.43 and after adjusting for calendar year, gonorrhoeaa infection status, age, nationality, and years of registered prostitution, the IRR wass 0.32 (95% CI: 0.20 - 0.59). This suggests a protective effect of 68% by preceding HIV-22 infection. Later updates on this study with longer follow-up time confirmed the findings;; one suggested an even stronger protective effect (74%) by HIV-2 [170], and anotherr a less strong but still significant effect (58%) [171].
Thiss study was criticised because of a probable selection bias, which could have caused, partiallyy or completely, the results. From a large group of initially seronegative CSW's thatt the investigators were studying [101], only about half were selected for this analysis. Thiss subgroup contained all the women who later seroconverted to HIV-1 positivity, whichh seems a highly unlikely outcome of a truly random sampling [172]. The findings of thee Senegalese study could also be due to different HIV-1 prevalences among the clients off HIV-2 infected and uninfected women.
42 2 HIV-22 IN WEST AFRICA
AA smaller study among 242 CSWs in The Gambia (1988 - 1995) found a non-significant, increasedincreased incidence of HIV-1 among HIV-2 infected women: the crude incidence rate ratio waswas 1.7 (0.7 - 3.9) [173]. A striking difference between these two studies in neighbouring countriess was the much higher HIV-1 incidence rate in The Gambia (see table 7). This contrastt may reflect the different segments of the population of CSWs included in the twoo studies (low risk in Senegal, high risk in The Gambia).
AA study from 1987 - 1996 among 729 people in three semi-urban areas of Bissau, the capitall of neighbouring Guinea-Bissau, also found an increased incidence of HIV-1 infectionn in those already infected with HIV-2: IRR = 10.10 (95% CI: 3.5 - 28.9) [103]. In aa multivariate analysis HIV-2 seropositivity was the single most important risk factor for HIV-11 seroconversion (OR = 7.5; 95% CI: 1.95 - 27.8; p=0.005).
AA study of HIV vertical transmission among women in Cote d'lvoire found coincidentallyy that 7/132 (5.3%) HIV-2 positive women and 5 / 274 (1.8%) HIV- negativee women became infected with HIV-1 in the observation period [123]. This translatess into a risk ratio of 2.91 (95% CI 0.87 - 9.7), indicating a higher risk for those alreadyy infected with HIV-2. Although this is a crude ratio, the dual status was confirmed byy Western Blot instead of PCR, and no data on exposure or other risk factors were available,, this again points away from a protective effect.
Althoughh conceptually appealing, and supporting evidence from several in vitro studies is suggestive,, the protective effect of HIV-2 in vivo remains a hypothesis. If the protective effectt occurs at all, it is by no means absolute. What would determine its efficacy in any individual?? Possible factors are HLA type, length of preceding HIV-2 infection, subtype off the 'immunising' HIV-2, stage of immunodeficiency, co-infection with STD's, and dosee and subtype of infecting virus. The finding of an increased risk of HIV-1 infection inn the three smaller studies could be due to residual confounding. In some women HIV-2 inducedd damaged to the immune system could have facilitated HIV-1 infection. More longitudinall studies in populations with a high HIV-2 baseline prevalence are needed to clarifyy this issue.
43 3 HIV-22 IN WEST AFRICA
Iff HIV-2 is unable to prevent a subsequent HIV-1 infection, it is conceivable that it mightt modify the course of disease. Studies done so far suggest that dual infection followss a course similar to HIV-1 (174]; however no proper longitudinal comparative studiess have been reported.
Conclusions s
HIV-22 is less pathogenic, less transmissible, and less wide-spread than HIV-1. The prognosiss for an asymptomatic HIV-2 infected individual is fundamentally different from thatt for an asymptomatic HIV-1 patient. Counselling of patients should take account of thesee differences.
Thee development of HIV-2 specific Reverse Transcriptase PCR techniques has opened neww opportunities for the study of the natural history of HIV-2 infection and tlie responsee to antiviral drugs.
Inn the next few years clinical and epidemiological studies should provide more informationn on the question why HIV-2 is a commensal in some, and a virulent pathogenn in others. Elucidating this question will enhance the understanding of protectivee immunity in HIV in general. Because HIV-2 is a slow pathogen, cohort studies mustt count in decades, not in years.
Acknowledgements s
Thee authors would like to thank Koya Ariyoshi, Neil Berry, Roel Coutinho, Olav Larsen, Janee Rowley, and Hilton Whittie, for their helpful comments on earlier drafts of this review,, and Neil Berry and Olav Larsen for sharing unpublished data.
44 4 HIV-22 IN WEST AFRICA
References s
1.. Clavel F, Guetard D, Brun-Vezinet F et al.: Isolation of a new human retrovirus from Westt African patients with AIDS. Science 1986, 233:343-346.
2.. Brun-Vezinet F, Rey MA, Katlama C et al.: Lymphadenopathy-associated virus type 2 inn AIDS and AIDS-related complex. Clinical and virological features in four patients. Lanceff 1987, 1:128-132.
3.. Marlink R: Lessons from the second AIDS virus, HIV-2 [editorial]. [Review] [217 refs]. AIDSAIDS 1996, 10:689-699.
4.. Guyader M, Emerman M, Sonigo P, Clavel F, Montagnier L, Alizon M: Genome organizationn and transactivation of the human immunodeficiency virus type 2. Nature 1987,, 326:662-669.
5.. Gao F, Yue L, Robertson DL et al.: Genetic diversity of human immunodeficiency viruss type 2: evidence for distinct sequence subtypes with differences in virus biology. JournalJournal of Virology 1994, 68:7433-7447.
6.. Chen Z, Luckay A, Sodora DL et al. : Human immunodeficiency virus type 2 (HIV-2) seroprevalencee and characterization of a distinct HIV-2 genetic subtype from the naturall range of simian immunodeficiency virus-infected sooty mangabeys. Journal of VirologyVirology 1997, 71:3953-3960.
7.. Chen Z, Telfier P, Gettie A et al. : Genetic characterization of new West African simiann immunodeficiency virus SIVsm: geographic clustering of household- derived SIVV strains with human immunodeficiency virus type 2 subtypes and genetically diversee viruses from a single feral sooty mangabey troop. Journal of Virology 1996, 70:3617-3627. .
8.. Rey-Cuille M-A, Berthier J-L, Bomsel-Demontoy M-C et al.: Simian Immunodeficiency Viruss replicates to high levels in Sooty Mangabeys without inducing disease. Journal ofof Virology 1998, 72:3872-3886.
9.. Albert J, Stalhandske P, Marqutna S et al.: Biological phenotype of HIV type 2 isolatess correlates with V3 genotype. AIDS Research & Human Retroviruses 1996, 12:821-828. .
10.. Guillon C, van der Ende ME, Boers PH, Gruters RA, Schutten M, Osterhaus AD: Coreceptorr usage of human immunodeficiency virus type 2 primary isolates and biologicall clones is broad and does not correlate with their syncytium-inducing capacities.. Journal of Virology 1998, 72:6260-6263.
11.. Owen SM, Ellenberger D, Rayfield M et al.: Genetically divergent strains of human immunodeficiencyy virus type 2 use multiple coreceptors for viral entry. Journal of VirologyVirology WW, 72:5425-5432.
12.. McKnight A, Dittmar MT, Moniz-Periera J et al.: A broad range of chemokine receptorss are used by primary isolates of human immunodeficiency virus type 2 as coreceptorss with CD4. Journal of Virology 1998, 72:4065-4071.
45 5 HIV-22 IN WEST AFRICA
13.. Whittle HC, Ariyoshi K, Rowland-Jones S: HIV-2 and T cell recognition. Current OpinionOpinion in Immunology 1998,10:382-387.
14.. Norrby E, Biberfeld G, Chiodi F et a/.: Discrimination between antibodies to HIV and too related retroviruses using site-directed serology. Nature 1987, 329:248-250.
15.. Tedder RS, O'Connor T, Hughes A, N'jie H, Corrah T, Whittle H: Envelope cross- reactivityy in Western blot for HIV-1 and HIV-2 may not indicate dual infection. Lancet 1988,, ii:927-930.
16.. De Cock KM, Porter A, Kouadio J er al.: Cross-reactivity on western blots in HIV-1 andd HIV-2 infections [published erratum appears in AIDS 1991 Sep;5(9):following 1155].. AIDS 1991, 5:859-863.
17.. M'Boup S, Kanki PJ, Mhalu FS: Laboratory diagnosis of HIV-1 and HIV-2 in Africa. [Review]] [30 refs]. AIDS 1991, 5 Suppl 1:S93-101.
18.. Kanki PJ, Peeters M, Gueye-Ndiaye A: Virology of HIV-1 and HIV-2: implications for Africa.. [Review] [138 refs]. AIDS 1997, 11 Suppl B:S33^2.
19.. Nkengasong JN, Maurice C, Koblavi S et al.: Field evaluation of a combination of monospecificc enzyme-linked immunosorbent assays for type-specific diagnosis of humann immunodeficiency virus type 1 (HIV-1) and HIV-2 infections in HIV- seropositivee persons in Abidjan, Ivory Coast. Journal of Clinical Microbiology 1998, 36:123-127. .
20.. Ariyoshi K, Cham F, Berry N et al.: al.: HIV-2-specific cytotoxic T-lymphocyte activity is inverselyy related to proviral load. AIDS 1995, 9:555-559.
21.. De Cock KM, Porter A, Kouadio J er a/.: Rapid and specific diagnosis of HIV-1 and HIV-22 infections: an evaluation of testing strategies. AIDS 1990, 4:875-878.
22.. Andersson S, da Silva Z, Norrgren H, Dias F, Biberfeld G: Field evaluation of alternativee testing strategies for diagnosis and differentiation of HIV-1 and HIV-2 infectionss in an HIV-1 and HIV-2-prevalent area. AIDS 1997, 11:1815-1822.
23.. Ariyoshi K, Cham F, Harding E et al.: Diagnosis of HIV-1/2 dual infection using dilutionall analysis of type specific antibody. AIDS 1998, In press:
24.. Peeters M, Gershy-Damet GM, Fransen K er al:. Virological and polymerase chain reactionn studies of H1V-1/HIV-2 dual infection in Cote d'lvoire. Lancet 1992, 340:339- 340. .
25.. George JR, Ou CY, Parekh B et al. : Prevalence of HIV-1 and HIV-2 mixed infections inn Cote d'lvoire. Lancet 1992, 340:337-339.
26.. Leonard G, Chaput A, Courgnaud V, Sangare A, Denis F, Brechot C: Characterizationn of dual HIV-1 and HIV-2 serological profiles by polymerase chain reaction.. AIDS 1993, 7:1185-1189.
27.. Sarr AD, Hamel DJ, Thior I et al. : HIV-1 and HIV-2 dual infection: lack of HIV-2 proviruss correlates with low CD4+ lymphocyte counts. AIDS 1998, 12:131-137.
46 6 HIV-22 IN WEST AFRICA
28.. Ariyoshi, K. The role of viral load in the pathogenesis of HIV-2 infection in West Africa.. 1998. MRC Laboratories, Fajara, The Gambia. PhD Thesis.
29.. Ishikawa K, Fransen K, Ariyoshi K et al.: Improved detection of HIV-2 proviral DNA in duallyy seroreactive individuals by PCR. AIDS 1998,12:1419-1425.
30.. Albert J, Naucler A, Bottiger B et al.: Replicative capacity of HIV-2, like HIV-1, correlatess with severity of immunodeficiency. AIDS 1990, 4:291-295.
31.. Simon F, Matheron S, Tamalet C ef a/.: Cellular and plasma viral load in patients infectedd with HIV-2. AIDS 1993, 7:1411-1417.
32.. van der Ende ME, Schutten M, Ly TD, Gruters RA, Osterhaus AD: HIV-2 infection in 122 European residents: virus characteristics and disease progression. AIDS 1996, 10:1649-1655. .
33.. Damond F, Loussert-Ajaka I, Apetrei C er al.: Highly sensitive method for amplificationn of human immunodeficiency virus type 2 DNA. Journal of Clinical MicrobiologyMicrobiology 1998, 36:809-811.
34.. Berry N, Ariyoshi K, Jobe O et al.: HIV type 2 proviral load measured by quantitative polymerasee chain reaction correlates with CD4+ lymphopenia in HIV type 2- infected individuals.. AIDS Research & Human Retroviruses 1994, 10:1031-1037.
35.. Ariyoshi K, Berry N, Wilkins A et al.: A community-based study of human immunodeficiencyy virus type 2 provirus load in rural village in West Africa. Journal of InfectiousInfectious Diseases 1996,173:245-248.
36.. Walther L, Andersson S, Dias F, da Silva Z, Biberfeld G: Discrimination between HIV- 11 and HIV-2 by PCR of serologically double reactive samples. 12th World AIDS Conference.Conference. Geneva, 1998 [Abstract 41160]. .
37.. Clavel F, Mansinho K, Chamaret S ef a/.: Human immunodeficiency virus type 2 infectionn associated with AIDS in West Africa. New England Journal of Medicine 1987,316:1180-1185. .
38.. Fultz PN, Switzer WM, Schable C, Desrosiers RC, Silva DP, McCormick JB: Seroprevalencee of HIV-1 and HIV-2 in Guinea Bissau in 1980. AIDS 1988, 2:129-132.
39.. Odehouri K, De Cock KM, Krebs JW et al.: HIV-1 and HIV-2 infection associated with AIDSS in Abidjan, Cote d'lvoire. AIDS 1989, 3:509-512.
40.. Poulsen AG, Kvinesdal B, Aaby P et al.: Prevalence of and mortality from human immunodeficiencyy virus type 2 in Bissau, West Africa [see comments]. Lancet 1989, 1:827-831. .
41.. Romieu I, Marlink R, Kanki P, M'Boup S, Essex M: HIV-2 link to AIDS in West Africa. [Review].. Journal of Acquired Immune Deficiency Syndromes 1990, 3:220-230.
42.. Wilkins A, Hayes R, Alonso P et al.: Risk factors for HIV-2 infection in The Gambia. AIDSAIDS 1991, 5:1127-1132.
43.. Wilkins A, Oelman B, Pepin J et al.: Trends in HIV-1 and HIV-2 infection in The Gambiaa [letter]. AIDS 1991, 5:1529-1530.
47 7 HIV-22 IN WEST AFRICA
44.. Kanki P, M'Boup S, Marlink R et at.; Prevalence and risk determinants of human immunodeficiencyy virus type 2 (HIV-2) and human immunodeficiency virus type 1 (HIV-1)) in west African female prostitutes. American Journal of Epidemiology 1992, 136:895-907. .
45.. Poulsen AG, Aaby P, Gottschau A et a/.: HIV-2 infection in Bissau, West Africa, 1987- 1989:: incidence, prevalences, and routes of transmission. Journal of Acquired ImmuneImmune Deficiency Syndromes 1993, 6:941-948.
46.. Bryceson A, Tomkins A, Ridley D et a/.: HIV-2-associated AIDS in the 1970s [letter]. LanceMM 988, 2:221-221.
47.. Kawamura M, Yamazaki S, Ishikawa K, Kwofie TB, Tsujimoto H, Hayami M: HIV-2 in westt Africa in 1966 [letter]. Lancet 1989, 1:385-385.
48.. Le Guenno B: HIV1 and HIV2: two ancient viruses for a new disease? Transactions ofof the Royal Society of Tropical Medicine & Hygiene 1989, 83:847-847.
49.. Wilkins A, Ricard D, Todd J, Whittle H, Dias F, Paulo Da Silva A: The epidemiology of HIVV infection in a rural area of Guinea- Bissau. AIDS 1993, 7:1119-1122.
50.. Larsen O, da Silva Z, Sandstrom A et a/.: Declining HIV-2 prevalence and incidence amongg men in a community study from Guinea-Bissau. AIDS 1998, 12:1707-1714.
51.. Smallman-Raynor M, Cliff A: The spread of human immunodeficiency virus type 2 into Europe:: a geographical analysis. International Journal of Epidemiology 1991, 20:480- 489. .
52.. Norrgren H, Andersson S, Dias F, Naucler A, Biberfeld G: Trends of incidence and prevalencee of HIV-1 and HIV-2 in Guinea-Bissau, West Africa. Xth International ConferenceConference on AIDS and STD in Africa. Abidjan, 1997 [Abstract B.045].
53.. Cassama L, Alvarenga I, Drouin M, Laroque I, Milord F, Frost EH: Detection of antibodiess to HIV-1 and HIV-2 dried blood spots using liquid and solid phase enzyme immunoassays.. Xth International Conference on AIDS and STD in Africa. Abidjan, 19977 [Abstract A.632].
54.. Cassama L, Alvarenga I, Drouin M ef a/.: Prevalence of HIV-1, HIV-2, Neisseria gonorrhoeae,, Chlamydia trachomatis, and geniatl warts in women living in Bafata province,, Guinea-Bissau. Xth International Conference on AIDS and STD in Africa. Abidjan,, 1997 [Abstract B.492].
55.. Whittle H, Yamuah LK, O'Donovan D, Milligan P, Harding E, Wilkins HA: HIV-1 and HIV-22 infection in pregnant women in The Gambia: risk factors. 12th World AIDS Conference.Conference. Geneva, 1998 [Abstract 23164].
56.. Kassrm S, Wiktor SZ, Maurice C, Lawson A, Coulibaly IM: Prevalence of HIV and syphiliss infections among pregnant women attending urban antenatal clinics in Cote d'lvoire,, 1997. 12th World AIDS Conference. Geneva, 1998 [Abstract 13111],
57.. Obisesan KA, Olaleye OD, Adeyemo AA: The increasing prevalence of HIV-1 and HIV-22 infections in a low risk antenatal population in south west Nigeria. International JournalJournal of Gynaecology & Obstetrics 1997, 56:271-273.
48 8 HIV-22 IN WEST AFRICA
58.. UNAIDS and WHO. Report on the global HIV/AIDS epidemic. 1998. Geneva, UNAIDSS & WHO.
59.. Peeters M, Koumare B, Mulanga C et ai: Genetic subtypes of HIV type 1 and HIV typee 2 strains in commercial sex workers from Bamako, Mali. AIDS Research & HumanHuman Retroviruses 1998, 14:51-58.
60.. Hawkes S, West B, Wilson S, Whittle H, Mabey D: Asymptomatic carriage of Haemophiluss ducreyi confirmed by the polymerase chain reaction. Genitourinary MedicineMedicine 1995, 71:224-227.
61.. Langley CL, Benga-De E, Critchlow CW et ai: HIV-1, HIV-2, human papillomavirus infectionn and cervical neoplasia in high-risk African women. AIDS 1996, 10:413-417.
62.. Ghys PD, Fransen K, Diallo MO et at.: The associations between cervicovaginal HIV shedding,, sexually transmitted diseases and immunosuppression in female sex workerss in Abidjan, Cote d'lvoire. AIDS 1997,11 :F85-F93.
63.. Samb ND, Seek K, Samb B et ai: Cervico-vaginal shedding of HIV and STD among prostitutess and AIDS patients in Senegal. Xth International Conference on AIDS and STDSTD in Africa. Abidjan, 1997 [Abstract B.178].
64.. Boillot F, Peeters M, Kosia A, Delaporte E: Prevalence of the human immunodeficiencyy virus among patients with tuberculosis in Sierre Leone, established fromm dried blood spots on filter paper. International Journal of Tuberculosis & Lung DiseaseDisease 1997, 1:493-497.
65.. Kouinche Matagne A, Tapko JBT, Kamdem TK, Zokong BZ, Tookam MT, Kaptue LNK:: Evolution of HIV seroprevalence in the General Hospital Yaounde. 12th World AIDSAIDS Conference. Geneva, 1998 [Abstract 13126].
66.. Zekeng L, Obiang Sima J, Hampl H et a/.: Update on HIV-1 group O infection in Equatoriall Guinea, Central Africa. AIDS 1997, 11:1410-1412.
67.. Tevi-Benissan C, Okome M, Makuwa M et ai: HIV-2 infection and HIV-1/HIV-2 dual reactivityy in patients with and without AIDS-related symptoms in Gabon [letter]. EmergingEmerging Infectious Diseases 1998, 4:130-131.
68.. Delaporte E, Janssens W, Peeters M et ai.: Epidemiological and molecular characteristicss of HIV infection in Gabon, 1986-1994. AIDS 1996,10:903-910.
69.. Mulanga-Kabeya C, Nzilambi N, Edidi B et ai: Evidence of stable HIV seroprevalencess in selected populations in the Democratic Republic of the Congo. AIDSAIDS 1998, 12:905-910.
70.. Anonymous.: HIV-1 and HIV-2 infections in England and Wales. Communicable DiseaseDisease Report 1993, CDR Weekly. 3:181-181.
71.. Sullivan MT, Guido EA, Metier RP et ai: Identification and characterization of an HIV- 22 antibody- positive blood donorr in the United States. Transfusion 1998, 38:189-193.
72.. Anonymous. Facts about human immunodeficiency virus type 2. 1997. Atlanta, Ga, Centerss for Disease Control and Prevention.
49 9 HIV-22 IN WEST AFRICA
73.. Nandwani R, Shanson D, Coleman JC: Missed HIV-2 infection in a bisexual man [letter].. International Journal of STD & AIDS 1998, 5:155-Apr.
74.. Matheron S, Puyeo S, Lepretre A et a/.: HIV-2 infection in adult patients: the French nationall cohort. XI International Conference on AIDS. Vancouver, 1996 [Abstract Mo.C.1665]. .
75.. Soriano V, Machuca A, Gutierrez M, Aguilera A, Caballero E: HIV-2 in Spain: Data fromm the National Registry up to December 1997. 12th World AIDS Conference. Geneva,, 1998 [Abstract 11207].
76.. Cazein F, Hamers FF, Alix J, Brunet JB: HIV-2 seroprevalence in Europe: a review of existingg data. XI International Conference on AIDS. Vancouver, 1996 [Abstract Mo.C.1664]. .
77.. Pieniazek D, Peralta JM, Ferreira JA et a/.: Identification of mixed HIV-1/HIV-2 infectionss in Brazil by polymerase chain reaction. AIDS 1991, 5:1293-1299.
78.. Carvalho MB, Hamerschlak N, Vaz RS, Ferreira OC, Jr.: Risk factor analysis and serologicall diagnosis of HIV-1/HIV-2 infection in a Brazilian blood donor population: validationn of the World Health Organization strategy for HIV testing. AIDS 1996, 10:1135-1140. .
79.. Libonatti O, Avila M, Zlatkes R et at: First report of HIV-2 infection in Argentina [letter].. Journal of Acquired Immune Deficiency Syndromes & Human Retrovirology 1998,, 18:190-191.
80.. Baqi S, Nabi N, Hasan SN et a/.: HIV antibody seroprevalence and associated risk factorss in sex workers, drug users, and prisoners in Sindh, Pakistan. Journal of AcquiredAcquired Immune Deficiency Syndromes & Human Retrovirology 1998,18:73-79.
81.. Mbizvo MT, Mashu A, Chipato T, Makura E, Bopoto R, Fottrell PF: Trends in HIV-1 andd HIV-2 prevalence and risk factors in pregnant women in Harare, Zimbabwe. CentralCentral African Journal of Medicine 1996, 42:14-21.
82.. Rubsamen-Waigmann H, Briesen HV, Maniar JK, Rao PK, Scholz C, Pfutzner A: Spreadd of HIV-2 in India [letter]. Lancer 1991, 337:550-551.
83.. Pfutzner A, Dietrich U, von Eichel U et a/.: HIV-1 and HIV-2 infections in a high-risk populationn in Bombay, India: evidence for the spread of HIV-2 and presence of a divergentt HIV-1 subtype. Journal of Acquired Immune Deficiency Syndromes 1992, 5:972-977. .
84.. Rodrigues JJ, Mehendale SM, Shepherd ME er a/.: Risk factors for HIV infection in peoplee attending clinics for sexually transmitted diseases in India. BMJ 1995, 311:283-286. .
85.. Neogi D, Chakrabarti T, Bhattacharya N: Prevalence of HIV-2 infection in Calcutta (India)) with a special reference to blood donors. 12th World AIDS Conference. Geneva,, 1998 [Abstract 42130].
86.. Grez M, Dietrich U, Balfe P et a/.: Genetic analysis of human immunodeficiency virus typee 1 and 2 (HIV-1 and HIV-2) mixed infections in India reveals a recent spread of
50 0 HIV-22 IN WEST AFRICA
HIV-11 and HIV-2 from a single ancestor for each of these viruses. Journal of Virology 1994,68:2161-2168. .
87.. Rubsamen-Waigmann H, Maniar J, Gerte S ef ai: High proportion of HIV-2 and HIV- 1/22 double-reactive sera in two Indian states, Maharashtra and Goa: first appearance off an HIV-2 epidemic along with an HIV-1 epidemic outside of Africa. Zentralblatt fur BakteriologieBakteriologie 1994, 280:398-402.
88.. Dietrich U, Maniar JK, Rubsamen-Waigmann H: The epidemiology of HIV in India. [Review].. Trends in Microbiology 1995, 3:17-21.
89.. Joshi SH, Patil RS, Chipkar SS, Apte SV: Seroprevalence of HIV-1 and HIV-2 infectionn in Western India. VIII International Conference on AIDS. Amsterdam, 1992 [Abstract[Abstract PuC 8110],
90.. Mehendale SM, Rodrigues JJ, Brookmeyer RS et a/.: Incidence and predictors of humann immunodeficiency virus type 1 seroconversion in patients attending sexually transmittedd disease clinics in India. Journal of Infectious Diseases 1995, 172:1486- 1491. .
91.. Pais P: HIV and India: looking into the abyss. [Review] [54 refs]. Tropical Medicine & InternationalInternational Health 1996, 1:295-304.
92.. De Cock, KM, Brun-Vezinet, F: HIV-2 infection: current knowledge and uncertainties. 1996,, 2nd: 171-176.
93.. Norrgren H, Marquina S, Leitner T et ai: HIV-2 genetic variation and DNA load in asymptomaticc carriers and AIDS cases in Guinea-Bissau. Journal of Acquired ImmuneImmune Deficiency Syndromes & Human Retrovirology 1997, 16:31-38.
94.. Xiang Z, Ariyoshi K, Wilkins A, Dias F, Whittle H, Breuer J: HIV type 2 pathogenicity is nott related to subtype in rural Guinea Bissau. AIDS Research & Human Retroviruses 1997,, 13:501-505.
95.. Grassly NC, Xiang Z, Ariyoshi K et ai: Mortality amongst Human Immunodeficiency Viruss Type 2-positive villagers in rural Guinea-Bissau is correlated with viral genotype.. Journal of Virology 1998, 72:7895-7899.
96.. Hunt JC, Brennan CA, Golden AM et ai: Molecular analyses of HIV-1 group O and HIV-22 variants from Africa. Leukemia 1997,11 Suppl 3:138-141.
97.. Greenberg A, Coulibaly IM, Kadio A et ai: Trends in the HIV-1 and HIV-2 epidemics inn Abidjan, Cote d'lvoire: 11 years of HIV serosurveillance at Project Retro-CI. Xth InternationalInternational Conference on AIDS and STD in Africa. Abidjan, 1997 [Abstract B.041].
98.. Djomand G, Greenberg AE, Sassan-Morokro M et a/.: The epidemic of HIV/AIDS in Abidjan,, Cote d'lvoire: a review of data collected by Projet RETRO-CI from 1987 to 1993.. Journal of Acquired Immune Deficiency Syndromes & Human Retrovirology 1995,, 10:358-365.
99.. Pepin J, Dunn D, Gaye I et ai: HIV-2 infection among prostitutes working in The Gambia:: association with serological evidence of genital ulcer diseases and with generalizedd lymphadenopathy. AIDS 1991, 5:69-75.
51 1 HIVV 2 IN WEST AFRICA
100.. Pepin J, Morgan G, Dunn D et al. : HIV-2-induced immunosuppression among asymptomaticc West African prostitutes: evidence that HIV-2 is pathogenic, but less so thann HIV-1. AIDS 1991, 5:1165-1172.
101.. Kanki PJ, Travers KU, Mboup S et a/.: Slower heterosexual spread of HIV-2 than HIV- 1.. Lancet 1994, 343:943-946.
102.. Dia MC, M'Boup S, Traore I et a/.: 10-year incidence trends of HIV-1 and HIV-2 in Senegal.. XI International Conference on AIDS. Vancouver, 1996 [Abstract We.C.223]. .
103.. Larsen O, da Silva Z, Sandstrom A et a/.: Rapid spread of HIV-1 infection among adultss in Guinea-Bissau: does previous HIV-2 infection offer protection? Submitted 1998, ,
104.. Norrgren H, Andersson S, Naucler A, Dias F, Johansson I, Biberfeld G: HIV-1, HIV-2, HTLV-I/III and Treponema pallidum infections: incidence, prevalence, and HIV-2- associatedd mortality in an occupational cohort in Guinea-Bissau. Journal of Acquired ImmuneImmune Deficiency Syndromes & Human Retrovirology 1995, 9:422-428.
105.. De Cock KM, Adjorlolo G, Ekpini E er al.: Epidemiology and transmission of HIV-2. Whyy there is no HIV-2 pandemic [published erratum appears in JAMA 1994 Jan 19;271{3):: 196] [see comments]. JAMA 1993, 270:2083-2086.
106.. Kanki PJ, De Cock KM: Epidemiology and natural history of HIV-2. AIDS 1994, 8:S85-S93. .
107.. Anderson RM, May RM: The population biology of the interaction between HIV-1 and HIV-- 2: coexistence or competitive exclusion? AIDS 1996, 10:1663-1673.
108.. Killewo J, Nyamuryekunge K, Sandstrom A et a/.: Prevalence of HIV-1 infection in the Kageraa region of Tanzania: a population-based study. AIDS 1990, 4:1081-1085.
109.. Wawer MJ, Serwadda D, Musgrave SD, Konde-Lule JK, Musagara M, Sewankambo NK:: Dynamics of spread of HIV-I infection in a rural district of Uganda. BMJ 1991, 303:1303-1306. .
110.. Wagner HU, Kamali A, Nunn AJ, Kengeya-Kayondo JF, Mulder DW: General and HIV-1-associatedd morbidity in a rural Ugandan community. AIDS 1993, 7:1461-1467.
111.. Aaby P, Ariyoshi K, Buckner M ef a/.: Age of wife as a major determinant of male-to- femalee transmission of HIV-2 infection: a community study from rural West Africa. /A/DS1996,, 10:1585-1590.
112.. Critchlow CW, Kiviat NB: Detection of human immunodeficiency virus type 1 and type 22 in the female genital tract: implications for the understanding of virus transmission. [Review]] [73 refs]. Obstetrical & Gynecological Survey 1997, 52 :315-324.
113.. Sankale JL, Mboup S, Essex ME, Kanki PJ: Genetic characterization of viral quasispeciess in blood and cervical secretions of HIV-1 and HIV-2 infected women. 12th12th World AIDS Conference. Geneva, 1998 [Abstract 60652].
52 2 HIV-22 IN WEST AFRICA
114.. N'Gbichi JM, De Cock KM, Batter V et al.: HIV status of female sex partners of men reactivee to HIV-1, HIV- 2 or both viruses in Abidjan, Cote d'lvoire. AIDS 1995, 9:951- 954. .
115.. Abbott RC, NDour-Sarr A, Diouf A et al.: Risk factors for HIV-1 and HIV-2 infection in pregnantt women in Dakar, Senegal. Journal of Acquired Immune Deficiency SyndromesSyndromes 1994, 7:711-717.
116.. Pepin J, Quigley M, Todd J et a/.: Association between HIV-2 infection and genital ulcerr diseases among male sexually transmitted disease patients in The Gambia. AIDSAIDS 1992, 6:489-493.
117.. Sassan-Morokro M, Greenberg AE, Coulibaly IM et a/.: High rates of sexual contact withh female sex workers, sexually transmitted diseases, and condom neglect among HIV-infectedd and uninfected men with tuberculosis in Abidjan, Cote d'lvoire. Journal ofof Acquired Immune Deficiency Syndromes & Human Retrovirology 1996, 11:183- 187. .
118.. Morgan G, Wilkins HA, Pepin J, Jobe O, Brewster D, Whittle H: AIDS following mother-to-childd transmission of HIV-2. AIDS 1990, 4:879-882.
119.. Adjorlolo-Johnson G, De Cock KM, Ekpini E et a/.: Prospective comparison of mother- to-childd transmission of HIV-1 and HIV-2 in Abidjan, Ivory Coast [see comments] [publishedd erratum appears in JAMA 1994 Nov 16;272(19):1482]. JAMA 1994, 272:462-466. .
120.. Cavaco-Silva P, Taveira NC, Lourenco MH, Santos Ferreira MO, Daniels, RS.: Verticall transmission of HIV-2 [letter]. Lancet 1997, 349:177-178.
121.. Cavaco-Silva P, Taveira NC, Rosado L et a/.: Virological and molecular demonstrationn of human immunodeficiency virus type 2 vertical transmission. Journal ofof Virology 1998, 72:3418-3422.
122.122. Whittle H, Ariyoshi KA, Ngom PTN et a/.: A community based study of perinatal transmissionn of HIV-1 and HIV-2 in The Gambia. 12th World AIDS Conference. Geneva,, 1998 [Abstract 23284].
123.. Ekpini ER, Wiktor SZ, Satten GA et al:. Late postnatal mother-to-child transmission of HIV-11 in Abidjan, Cote d'lvoire [see comments]. Lancet 1997, 349:1054-1059.
124.. Leroy V, Newell ML, Dabis F er al.: International multicentre pooled analysis of late postnatall mother-to-child transmission of HIV-1 infection. Lancet 1998, 352:597-600.
125.. Marlink R, Kanki P, Thior I er al.: Reduced rate of disease development after HIV-2 infectionn as compared to HIV-1. Science 1994, 265:1587-1590.
126.. Jaffar S, Wilkins A, Ngom PT er al.: Rate of decline of percentage CD4+ cells is faster inn HIV-1 than in HIV-2 infection. Journal of Acquired Immune Deficiency Syndromes & HumanHuman Retrovirology 1997, 16:327-332.
127.. De Cock KM, Gnaore E, Adjorlolo G et al.: Risk of tuberculosis in patients with HIV-I andd HIV-II infections in Abidjan, Ivory Coast [see comments]. BMJ 1991, 302:496- 499. .
53 3 HIV-22 IN WEST AFRICA
128.. Norrgren H, Cardoso AN, da Sifva ZJ et a/.: Increased prevalence of HIV-2 infection inn hospitalized patients with severe bacterial diseases in Guinea-Bissau. ScandinavianScandinavian Journal of Infectious Diseases 1997, 29:453-459.
129.. Grant AD, Djomand G, De Cock KM: Natural history and spectrum of disease in adultss with HIV/AIDS in Africa. [Review] [146 refs]. AIDS 1997, 11 Suppl B:S43-54.
130.. Grant AD, Djomand G, Smets P er a/.: Profound immunosuppression across the spectrumm of opportunistic disease among hospitalized HIV-infected adults in Abidjan, Cotee d'lvoire. AIDS 1997, 11:1357-1364.
131.. Matheron S, Mendoza-Sassi G, Simon F, Olivares R, Coulaud JP, Brun-Vezinet F: HIV-11 and HIV-2 AIDS in African patients living in Paris [letter]. AIDS 1997, 11:934- 936. .
132.132. Colebunders R, De Vuyst H, Verstraeten T, Schroyens W, Van Marck E: A non- Hodgkin'ss lymphoma in a patient with HIV-2 infection [letter]. Genitourinary Medicine 1995,, 71:129-129.
133.. Ancelle R, Bletry O, Baglin AC, Brun-Vezinet F, Rey MA, Godeau P: Long incubation periodd for HIV-2 infection [letter]. Lancet 1987, 1:688-689.
134.. Dufoort G, Courouce AM, Ancelle-Park R, Bletry O: No clinical signs 14 years after HIV-22 transmission via blood transfusion [letter]. Lancet 1988, 2:510-
135.. Mota-Miranda A, Gomes H, Marques R et a/.: HIV-2 infection with a long asymptomaticc period. Journal of Infection 1995, 31:163-164.
136.. Mota-Miranda A, Gomes MH, Serrao MR et a/.: Long term nonprogressive HIV-2 infection.. XI International Conference on AIDS. Vancouver, 1996 [Abstract We.C.3469]. .
137.. Poulsen AG, Aaby P, Larsen O ef a/.: 9-year HIV-2-associated mortality in an urban communityy in Bissau, west Africa. Lancet 1997, 349:911-914.
138.. Marlink R, Traore I, Thior I et at:. HIV-2 as a model for long term non-progression. XI InternationalInternational Conference on AIDS. Vancouver, 1996 [Abstract Tu.B.313].
139.. Lisse IM, Poulsen AG, Aaby P, Knudsen K, Dias F: Serial CD4 and CD8 T- lymphocytee counts and associated mortality in an HIV-2-infected population in Guinea-Bissau.. Journal of Acquired Immune Deficiency Syndromes & Human RetrovirologyRetrovirology 1996, 13:355-362.
140.. Nunn AJ, Mulder DW, Kamali A, Ruberantwari A, Kengeya-Kayondo JF, Whitworth J: Mortalityy associated with HIV-1 infection over five years in a rural Ugandan population:: cohort study. BMJ 1997, 315:767-771.
141.. Greenberg AE, Kadio A, Grant AD ef a/.: Clinical manifestations of advanced HIV diseasee using hospital surveillance, clinical and autopsy data in Abidjan, Cote d'lvoire.. 12th World AIDS Conference. Geneva, 1998 [Abstract 12146].
142.. Ariyoshi K, Schim van der Loeff M, Cook P et a/.: Kaposi's sarcoma in The Gambia, Westt Africa is less frequent in Human Immunodeficiency Virus Type 2 than in Human
54 4 HIV-22 IN WEST AFRICA
Immunodeficiencyy Virus Type 1 infection despite a high prevalence of Human Herpesviruss 8. Journal of Human Virology 1998,1:193-199.
143.. Whittle H, Morris J, Todd J et at.: HIV-2-infected patients survive longer than HIV-1- infectedd patients. AIDS 1994, 8:1617-1620.
144.. Whittle H, Egboga A, Todd J et a/.: Clinical and laboratory predictors of survival in Gambiann patients with symptomatic HIV-1 or HIV-2 infection. AIDS 1992, 6:685-689.
145.. Grankvist O, Bredberg-Raden U, Gustafsson A ef a/.: Improved detection of HIV-2 DNAA in clinical samples using a nested primer-based polymerase chain reaction. JournalJournal of Acquired Immune Deficiency Syndromes 1992, 5:286-293.
146.. Whittle H, Egboga A, Todd J et a/.: Immunological responses of Gambians in relation too clinical stage of HIV-2 disease. Clinical & Experimental Immunology 1993, 93:45- 50. .
147.. Ricard D, Wilkins A, N'Gum PT ef a/.: The effects of HIV-2 infection in a rural area of Guinea-Bissau.. AIDS 1994, 8:977-982.
148.. Switzer WM, Wiktor S, Soriano V et a/.: Evidence of Nef truncation in human immunodeficiencyy virus type 2 infection. Journal of Infectious Diseases 1998, 177:65- 71. .
149.. Berry N, Ariyoshi K, Jaffar S ef a/.: Low peripheral blood viral HIV-2 RNA in individualss with high CD4% differentiates HIV-2 from HIV-1 infections. Journal of HumanHuman Virology 1998, In press:
150.. Soriano V, Gomes P, Holguin A ef a/.: HIV-2 in Portugal: circulating subtypes, virus isolation,, plasma viral load, and clinical spectrum (the HIV-2 Iberian project). 12th WorldWorld AIDS Conference. Geneva, 1998 [Abstract 21154].
151.. Matheron S, Courpotin C, Simon F et ai: Vertical transmission of HIV-2 [letter]. LancetLancet 1990, 335:1103-1104.
152.. De Cock KM, Zadi F, Adjorlolo G et a/.: Retrospective study of maternal HIV-1 and HIV-22 infections and child survival in Abidjan, Cote d'lvoire [see comments]. BMJ 1994,308:441-443. .
153.. Prazuck T, Yameogo JM, Heylinck B et ai: Mother-to-child transmission of human immunodeficiencyy virus type 1 and type 2 and dual infection: a cohort study in Banfora,, Burkina Faso. Pediatric Infectious Disease Journal 1995,14:940-947.
154.. Lucas SB, Peacock CS, Hounnou A et ai.: Disease in children infected with HIV in Abidjan,, Cote d'lvoire [see comments]. BMJ 1996, 312:335-338.
155.. Vetter KM, Djomand G, Zadi F et ai: Clinical spectrum of human immunodeficiency viruss disease in children in a west African city. Project RETRO-CI. Pediatric Infectious DiseaseDisease Journal 1996, 15:438-442.
156.. Rosado L, Neves C, Bessa A et ai: Survival among HIV-2 perinatally infected infants. XIXI International Conference on AIDS. Vancouver, 1996 [Abstract We.C.3436],
55 5 HIVV 2 IN WEST AFRICA
157.. Faye A, Burgard M, Crosnier H, Retbi JM, Blanche S: Human immunodeficiency virus typee 2 infection in children. Journal of Pediatrics 1997, 130:994-997.
158.. Ota M, O'Donovan D, Yamuah LK et a/.: Influence of maternal HIV-1 and HIV-2 on childd survival in Gambia. 12th World AIDS Conference. Geneva, 1998 [Abstract 12153]. .
159.. Cohen J: Can one type of HIV protect against another type? [news; comment]. ScienceScience 1995, 268:1566-1566.
160.. Weiss RA, Clapham PR, Weber JN et at.: HIV-2 antisera cross-neutralize HIV-1. AIDSAIDS 1988, 2:95-100.
161.. Bottiger B, Karlsson A, Andreasson PA et a/.: Envelope cross-reactivity between humann immunodeficiency virus types 1 and 2 detected by different serological methods:: correlation between cross-neutralization and reactivity against the main neutralizingg site. Journal of Virology 1990, 64:3492-3499.
162.162. Nixon DF, Huet S, Rothbard J ef a/.: An HIV-1 and HIV-2 cross-reactive cytotoxic T- celll epitope. AIDS 1990, 4:841-845.
163.. Rowland-Jones S, Sutton J, Ariyoshi K et a/.: HIV-specific cytotoxic T-cells in HIV- exposedd but uninfected Gambian women [published erratum appears in Nat Med 19955 Jun;1(6): 598]. Nature Medicine 1995, 1:59-64.
164.. Bertoletti A, Cham F, McAdam S et a/.: Cytotoxic T cells from human immunodeficiencyy virus type 2- infected patients frequently cross-react with different humann immunodeficiency virus type 1 clades. Journal of Virology 1998, 72:2439- 2448. .
165.. Arya SK, Gallo RC: Human immunodeficiency virus (HIV) type 2-mediated inhibition off HIV type 1: a new approach to gene therapy of HIV-infection. Proceedings of the NationalNational Academy of Sciences of the United States of America 1996, 93:4486-4491.
166.. Schwartz DH, Castillo R, Arango-Jaramillo S, Kanki P, M'Boup S: Peripheral blood mononuclearr cells (PBMCs) from HIV-2+ individuals resist in vitro infection with macrophagee tropic HIV-W. 12th World AIDS Conference. Geneva, 1998 [Abstract 31154]. .
167.. Kanki P, Kokkotov E, Gueye-Noiaye A, Schwartz D, Mboup S, Essex ME: In vitro evidencee of HIV-2 protection from subsequent HIV-1 infection. 12th World AIDS Conference.Conference. Geneva, 1998 [Abstract 21136].
168.. Belec L, Tevi-Benissan C, Dupre T ef at.: Comparison of cervicovaginal humoral immunityy in clinically asymptomatic (CDC A1 and A2 category) patients with HIV-1 andd HIV-2 infection. Journal of Clinical Immunology 1996, 16:12-20.
169.. Travers K, Mboup S, Marlink R ef a/.: Natural protection against HIV-1 infection providedd by HIV-2 [see comments] [published erratum appears in Science 1995 Jun 30;; 268(5219):1833]. Science 1995, 268:1612-1615.
170.. Kanki PJ, Eisen G, Travers KU ef a/.: HIV-2 and natural protection against HIV-1 infection.. Science 1996, 272:1959-1960.
56 6 HIV-22 IN WEST AFRICA
171.. Travers K, Eisen G, Hsieh C et ai: HIV-2 provides natural protection against HIV-1 infection.. Xth International Conference on AIDS and STD in Africa. Abidjan, 1997 [Abstractt A.072].
172.172. Greenberg AE, Wiktor SZ, DeCock KM, Smith P, Jaffe HW, Dondero TJ, Jr.: HIV-2 andd natural protection against HIV-1 infection [letter; comment]. Science 1996, 272:1959-1960. .
173.. Ariyoshi K, Schim van der Loeff M, Sabally S, Cham F, Corrah T, Whittle H: Does HIV-22 infection provide cross-protection against HIV-1 infection? [letter]. AIDS 1997, 11:1053-1054. .
174.. Wiktor S, Sassan-Morokko M, Abouya L era/.: HIV-1 plasma viral load in tuberculosis patientss with single HIV-1 and dual HIV-1/2 infections, Abidjan, Cote d'lvoire. 12th WorldWorld AIDS Conference. Geneva, 1998 [Abstract 13328].
175.. Gao F, Yue L, Sharp PM, Hahn BH: Genetic typing of HIV-2 from a Senegalese/Germann heterosexual transmission. AIDS Research & Human RetrovirusesRetroviruses 1993, 9:703-704.
176.. Takehisa J, Osei-Kwasi M, Ayisi NK et ai. Phylogenetic analysis of HIV type 2 in Ghanaa and intrasubtype recombination in HIV type 2. AIDS Research & Human RetrovirusesRetroviruses 1997, 13:621-623.
177.. Brennan CA, Yamaguchi J, Vallari AS, Hickman RK, Devare SG: Genetic variation in humann immunodeficiency virus type 2: identification of a unique variant from human plasma.. AIDS Research & Human Retroviruses 1997, 13:401-404.
178.. Heredia A, Vallejo A, Soriano V et ai: Genetic analysis of human immunodeficiency viruss type 2 strains from Spain. AIDS Research & Human Retroviruses 1998, 14:91- 94. .
179.. Molguin A, Rodes B, Soriano V, Gonzalez-Lahoz J: Circulating HIV-1 and HIV-2 geneticc subtypes in Spain. 12th World AIDS Conference. Geneva, 1998 [Abstract 60681]. .
180.. Simon F, Damond F, Loussert-Ajaka I et ai: Variability of HIV-2 from patients living in France.. 12th World AIDS Conference. Geneva, 1998 [Abstract 11205].
181.. Smith NA, Kennedy J, Bewley S et ai: HIV-2 in pregnancy: to treat or not to treat? IntemationalJournalIntemationalJournal of STD & AIDS 1998, 9:246-246.
182.. Christiansen CB, Albert J, Machuca R, Nielsen C: Differential transmission of HIV-1 andd HIV-2 [letter]. Lancet *\997, 350:562-562.
183.. Sankale JL, De La Tour RS, Marlink RG et ai: Distinct quasi-species in the blood and thee brain of an HIV-2- infected individual. Virology 1996, 226:418-423.
184.. Mabey DC, Tedder RS, Hughes AS et ai: Human retroviral infections in The Gambia: prevalencee and clinical features. British Medical Journal 1988, 296:83-86.
185.. Pichard E, Guindo A, Grosette G, et al.: [.'infection par le virus de rimmunodeficience humainee (VIH) au Mali. Medecine Tropicale 1988, 48:345-349.
57 7 HIV-22 IN WEST AFRICA
186.. Djomand G, Greenberg AE, Sassan-Morokro M et a/.: The epidemic of HIV/AIDS in Abidjan,, Cote d'lvoire: a review of data collected by Projet RETRO-CI from 1987 to 1993.. Journal of Acquired Immune Deficiency Syndromes & Human Retrovirology 1995,, 10:358-365.
187.. Marlink R, Mboup S, Thior I et a/.: Incidence of disease after HIV-2 infection as comparedd to HIV-1 infection. 12th World AIDS Conference. Geneva, 1998 [Abstract 13319]. .
188.. Donnelly C, Leisenring W, Kanki P, Awerbuch T, Sandberg S: Comparison of transmissionn rates of HIV-1 and HIV-2 in a cohort of prostitutes in Senegal. Bulletin of MathematicalMathematical Biology 1993, 55:731-743.
189.. Brucker G, Brun-Vezinet F, Rosenheim M, Rey MA, Katlama C, Gentilini: HIV-2 infectionn in two homosexual men in France [letter]. Lancet 1987, 1:223-
190.. Veronesi R, Mazza CC, Santos Ferreira MO, Lourenco MH: HIV-2 in Brazil (letter). LancetLancet 1987, ii:402-402.
191.. Soriano V, Gutierrez M, Caballero E et al.: Epidemiology of HIV-2 infection in Spain. Thee HIV-2 Spanish Study Group. European Journal of Clinical Microbiology & InfectiousInfectious Diseases 1996, 15:383-388.
192.. Marquart K-H, Muller HAG, Brede HD: HIV-2 in West Germany. AIDS 1988, 2:141- 141. .
193.. Heredia A, Vallejo A, Soriano V et at.: Genetic analysis of an HIV type 2 subtype B viruss from a Spanish individual with AIDS. AIDS Research & Human Retroviruses 1997,, 13:899-900.
194.. Andreasson PA, Dias F, Naucler A, Andersson S, Biberfeld G: A prospective study of verticall transmission of HIV-2 in Bissau, Guinea-Bissau. AIDS 1993, 7:989-993.
195.. Anonymous.: Comparison of vertical human immunodeficiency virus type 2 and humann immunodeficiency virus type 1 transmission in the French prospective cohort. Thee HIV Infection in Newborns French Collaborative Study Group. Pediatric Infectious DiseaseDisease Journal 1994,13:502-506.
196.. Poulsen AG, Kvinesdal BB, Aaby P et a/.: Lack of evidence of vertical transmission of humann immunodeficiency virus type 2 in a sample of the general population in Bissau.. Journal of Acquired Immune Deficiency Syndromes 1992, 5:25-30.
197.. Del Mistro A, Chotard J, Hall AJ, Whittle H, De Rossi A, Chieco-Bianchi L: HIV-1 and HIV-22 seroprevalence rates in mother-child pairs living in The Gambia (west Africa). JournalJournal of Acquired Acquired Immune Deficiency Syndromes 1992, 5:19-24.
198.. Aaby P, Poulsen AG, Larsen O ef a/.: Does HIV-2 protect against HIV-1 infection? [letter].. AIDS 1997, 11:939-940.
199.. Travers KU, Eisen GE, Marlink RG et al.\ Protection from HIV-1 infection by HIV-2 [letter].. AIDS 1998, 12:224-225.
58 8 3 3
Regionall differences in HIV trends inn The Gambia: results from sentinel surveillancee among pregnant women
AIDSAIDS 2003; 17 (12): 1841-6.
Maartenn F Schim van der Loeff '*2, Ramu Sarge-Njie ', Saihou Ceesay 3, Akum Aveika
Awasanaa l, Pa Jaye 3, Omar Sam 3, Kebba O Jaiteh \ David Cubitt 4, Paul Milligan ',
Hütonn C Whitde '
(1)(1) the Medical Research Council laboratories, The Gambia, (2) the London School of Hygiene <&
TropicalTropical Medicine, UK, (3) the Department of State for Health, The Gambia, and (4) the Institute of
ChildChild Health, London, UK.
Thiss study was funded by the MRC (UK), the Gambia Government, UNDP, and WHO. Murex Diagnosticss kindly provided the ELISA kits for the study.
Regionall differences in HIV trends inn The Gambia: results from sentinel surveillancee among pregnant women
ABSTRACT T
Objective:: To monitor HIV-1 and HIV-2 trends in The Gambia, West Africa Methods:: In 1993-5 a nation-wide survey among 29,670 pregnant women attendingg eight antenatal clinics estimated the seroprevalence of HIV-1 at 0.6%,, and of HIV-2 at 1.1%. Five years later sentinel surveillance in pregnant womenn was established, using unlinked anonymous testing in four clinics. A driedd blood spot on filter paper was obtained and tested for HIV antibodies. Results:: Between May 2000 and August 2001, 8054 analysable samples weree collected at four sites. HIV-1 prevalence rose sharply in one rural area fromm 0.6% to 3.0% (p < 0.0001), but the rise was small and non-significant in twoo other rural sites and in the urban site. The HIV-2 prevalence did not changee significantly in any of the sites. The overall prevalence of HIV-1 was 1.0%% (95% Confidence Interval (CI) 0.8 - 1.3%), and of HIV-2 0.8% (CI 0.6 - 1.0%).. Site, nationality and higher age were significantly associated with HIV- 11 infection, and higher parity and site with HIV-2 infection. Conclusions:: Fifteen years after the first case of HIV-1 was described in The Gambia,, the epidemic is still at a low level. There is heterogeneity within the country,, with one rural area experiencing a 5-fold rise in 6 years. HIV-2 prevalencee in The Gambia is stable. HIV-22 IN WEST AFRICA
Introduction n
Inn most West African countries the prevalence of HIV-1 is rising, and that of HIV-2 is fallingg or stable [1-6]. In Senegal HIV-1 prevalence has been remarkably stable, a fact thatt has been ascribed to that country's effective AIDS control programme [7,8].
Becausee the epidemiology and clinical impact of the two HIV infections is so different [9,10,11],, it is important to distinguish between them in surveillance: a decline in HIV-2 withh a concurrent rise in HIV-1 could be interpreted erroneously as a stable HIV prevalencee [12].
AA sentinel surveillance system has never been in place in The Gambia. The most recent prevalencee data are from a nation-wide study among pregnant women attending antenatal clinicss in 1993-5 [11]. HIV-1 prevalence was 0.6% (95% Confidence interval (CI) 0.5- 0.7%.)) and HIV-2 prevalence 1.1% (CI 1.0-1.3%).
AA national HIV sentinel surveillance among pregnant women was started in 2000. In orderr to evaluate whether a significant rise of HIV-1 had occurred since the 1993-5 survey,, surveillance was done in four of the antenatal clinics used in the previous survey.
Methods s
SurveySurvey in 1993-5 Inn a study of mother-to-child transmission of HIV in 1993-5 pregnant women were recruitedd at eight health centres throughout The Gambia. HIV diagnosis was based on a combinationn of ELISA's using serum obtained from venous blood samples [11].
Subjects Subjects Forr the present study, four of the original health facilities were selected as recruitment sites.. Serekunda health centre is located in the largest urban area of the country, and is
62 2 HIV-22 IN WEST AFRICA
thee busiest clinic. Sibanor health centre is in a rural area on the south bank of the river Gambia,, 88 km from the capital. The majority of the people in the area belong to the Jolaa ethnic group. Farafenni hospital is on the north bank, 194 km inland, where over 50%% of people are from the Wollof ethnic group. Basse is a small town in the eastern tip off the country (371 km inland), and there are three major ethnic groups: Fula, Serahuli andd Mandinka. These four centres provide antenatal care to over 20% of all pregnant womenn in The Gambia. Staff from these facilities conduct mobile clinics in villages beyondd walking distance of the health facility.
Thee health worker who completed the routine antenatal card, also filled a study data form.. Names or other unique personal identifiers were not recorded.
LaboratoryLaboratory methods AA circle of 2 cm diameter on filter paper (Whatman grade BFC 180, Whatman Internationall Ltd, Maidstone, UK) was filled with a few drops of blood. The filter paperss with the blood spots were allowed to dry separately overnight in a compartmentalisedd wooden rack, and collected once every 7 — 14 days.
Onn reception in the laboratory, a 5.5 mm diameter blood spot was punched out. The spotss were eluted in an elution buffer overnight. Eluates were combined in pools of five andd screened using ICEHIV 1.0.2 (Murex Diagnostics Ltd, Dartford, UK). If a pool wass positive, each individual sample was tested with ICEHIV I.O.2. Positive samples weree then tested for HIV-1 antibodies by Wellcozyme HIV-recombinant-1 (Murex) and forr HIV-2 antibodies by ICEHIV-2 (Murex). If the sample was reactive in only one of thee ELISA's, it was assigned the corresponding HIV diagnosis. If it was strongly reactive inn both it was considered a dual infection with both HIV-1 and HIV-2. These methods weree piloted first in the MRC clinic. In the pilot study Wellcozyme-1 (but not ICEHIV-2 orr ICEHIVl.O.2) was shown to be slightly less sensitive when testing the eluate than the serumm (R Sarge-Njie, manuscript in preparation), so false-negative results for HIV-1 were possible.. Therefore, in cases where ICEHIVl.O.2 was positive and both monospecific ELISA'ss were negative, an HIV-1 particle agglutination test was done (Serodia HIV-1,
63 3 HII V-2 IN WEST AFRICA
Fujirebioo Inc, Tokyo, Japan). The results of the Serodia HIV-1 were regarded as final, irrespectivee of the titre. Most dual serological patterns are due to true dual infections.
[13]] Therefore women whose samples were dually reactive, were regarded as being infectedd with both HIV-1 and HIV-2, and analysed both in the HIV-1 group and in the
HIV-22 group.
Threee types of controls were used: negative, low and high positive DBS filter paper discs
(suppliedd by CDC, Atlanta, US); known HIV-1 and HIV-2 positive sera diluted in freshly drawnn HIV negative blood spotted on filter paper and dried; serum controls (provided byy test manufacturers).
StatisticalStatistical methods
Inn the analysis each health facility was given a weight proportional to the population size off the region that the health facility was regarded to represent. Using the weighted prevalences,, risk ratios (RR) comparing the 2000-1 prevalence with that in 1993-5 were calculated. .
Htbics Htbics
Thee study employed Unlinked Anonymous Testing [14,15,16]. The women in the study providedd a blood sample for other purposes (haemoglobin or syphilis serology) and a few drops,, considered to be left-overs, were used for the study. Women were not explicitly toldd that part of their blood was to be used for HIV surveillance. A parallel system for womenn to be tested for HIV if they wished so was in place. The study was approved by thee Gambia Government / MRC Laboratories Ethics Committee.
Results s
Samplee were collected between May 2000 and August 2001. In total 8054 samples were availablee for analysis. HIV infection was found in 170: 94 HIV-1, 61 HIV-2, and 15
HIV-11 and HIV-2 dually positive.
64 4 ü ü .. . OJ
<*}} co oo ra.. «> OO — coo *- CD D o o rara —'; CMM csi 1 i i .. .. ra kss *** oo ^- öö CM "" £ " .122 o> o o ei i ci i coo o OO o cP P ££ O 22 >> rara -Q u-- -o Ö Ö .. - CO oo öö Ï5 CO O h-- CM M en n (O O CoM> M> 00 0 CM M Ol l CM M CM M 3 3 «>> » öö nj- COO Ü) öö E ^^ £ (IJJ _>. o o O«*>) ) > coo ra O) ) > o CDD ~ rara cu IS -Soo . > cc .+; roro co .. o ojj ro CAA CL .033 g Ö3> > =:: -> .T22^ cCoü .. n- 1 1 JDD O .. - JS f ££ ro11 = ra ra "55 £ in n CUU QJ có ó 00 0 N-- f? ? en n CO O CM M — * «« 0) CO O (O O att m oo Ü Oï ï CM M oo o £ CDD C O O Oi i T-- 00 22 CO O t t oo s cc c —— o 9 9 ô o ro ro vPP 00 roro N?OO -o "OO = > > OO Ui Q. . coo p o o «I I CM M |I Iroro o > > 1 in n - -- © rara co X X t t T33 O CC C COO II 5 5 (00 J» CDD _ i-- ra go o >£ rara£ to XX CL>> co n c c <*>> ^

> .22 > COO JC t/J J CD D O O oo o l l == a> .. -C rara 3 rara I- HIV-22 IN WEST AFRICA
HIV-1 HIV-1 Inn all four sites the prevalence of HIV-1 rose over the six years, but this rise was small andd not significant in three of the sites (Table 1). The only significant and substantial rise occurredd in a rural site (Sibanor), from 0.6 to 3.0% (p<0.0001). The risk ratio in Sibanor wass 4.9 (CI 2.9-8.2); the risk ratio in each of the other sites was around 1.4. The overall prevalencee of HIV-1 rose from 0.7% (CI: 0.6-0.8) in 1993-5 to 1.0 % (0.8-1.3%) in 2000- 11 (p=0.004). There was no significant difference in the mean age of participating women betweenn the two studies (p-0.6), but the mean age of HIV-1 infected women was 2.5 yearss higher in 2000-1 compared to 1993-5 (p=0.0001).
Tablee 2 shows that the HIV-1 prevalence was lower among pregnant teenagers compared too older women; this phenomenon was restricted to Sibanor (data not shown). HIV-1 infectedd women were on average 1.4 years older than seronegative women (25.8 vs. 24.4 years;; p=0.01). HIV-1 infection was not associated with marital status or parity. The prevalencee was higher among Senegalese women compared to other non-Gambians (p=0.016).. Significant differences were observed between ethnic groups, the prevalence beingg highest among die Jola (2.7%), and lowest among the Wollof (0.6%). Among womenn recruited at Sibanor die HIV-1 prevalence was significandy higher among Senegalesee women compared to Gambian women (5.6% vs. 2.5%; p=0.006). This differencee was not observed at the other sites.
Inn a multivariate analysis only site and nationality were significant risk factors. Compared too Basse, the OR for women recruited at Serekunda was 0.7 (CI: 0.4-1.3), at Sibanor 2.2 (CII 1.3-3.9), and at Farafenni 0.3 (CI: 0.1-0.7). The OR for Senegalese nationality was 1.7 (CII 1.0-1.7; p=0.03) and for other non-Gambian nationality (mainly Guinea-Conakry) 0.5 (CI:: 0.2-1.7; p=0.3). There was significant interaction (p=0.01) between ethnic group andd site: among women recruited in Sibanor, Jola's had a 3.4-fold (CI: 1.0-11) higher oddss of having HIV-1 than Mandinka's (baseline group), but in the other sites there was noo significant association between Jola ethnicity and HIV-1 infection.
HIV-2 HIV-2 Thee prevalence of HIV-2 rose in two and decreased in the other two sites (Table 1); all of thesee changes were small and only one decline was significant (Farafenni; p=0.05). The
66 6 .22 tü £ roro « -o E?? ü
COO CD N- cFCMM ^ ir>> oo coo oo m co m .EE « =5- irii r^ oo VV ^ V COO CM Tf CO CM OHH « t^-- có oó lOO o» i-; cöö -4 cöö cö o in có coo s- 2,2.2 2 2.2. . 2,2.112,2, , 33 U) o>> io co CT>CT> co lO I»-- T- CDD O) *- CO Oï _>» » 1 =5*811 1 T^^ CM CM COO CM lO öö T- CC C ©© o < < Is s c^£ £ CMM "O os o s >,-- CD oo o ö)™-. . c c oo .ü 000 CD -3- 0>> CO »fr IT)) O) O CM O) t/>> c COO CM tfi >«« 3 Q. . "oo ** CMM CO -^ ^-- Ö CM ^ Ó O O O) ) caa co Ü Ü c c .c c -99 co CD D o o T3 3 COO QJ- c c r^.. i- 2* i^- 5^ S^ c c 0 0 oo o CO O CMM sS ^^ O r- ^ ,-;; " oo co ö ö ^vSvoo o 18 8 i~ruii en h- ^TT CO _ CM —"— £>> S CO * f-- 00 ooino o T-- J CM oo co !T cocos il l CMM CO J- 2:: co in co —— r\J O5 OO CD S ifi O CM OO %1 CM CM T^ l£) ^^ £M T- T- CO OC MM LO £v|| T- J- j- CD h;- OO ü :?:: CM 5> ^ Ol *-- CM 555 o 55 55 -- -0 0.. 0. 55 ^^ 55 rs 55 «5 co *"*" CM CM T- 100 CO T- CM CO O T-- T- m CM
0)) «o II I O O CDD CD CDD Tt * s I -- oo Tff T- O 00 00 ^^ o i—— 1— CM CMM TJ CMM CM CO CM ^ o o 44 CO 4 ÓÓ CM CDD co oo m co tS55 5 T". .2H , , ooo o 33 tf> ILL 2. 2,2,2.2,2 2 000 CO 000 f- o s O O r>-- CD CMM OO CD - I - CM M*-- CM T^T^ Ó T^ T^ Ó < < iss % CDD t- CD caa o » Ell ei T»» .2 O)) (-- CM O)) m coo m T-- - CM S2 2 OS S JOO -g coo o o o gg f c c 000 lO CO O) ©^ ö^CO^- CMCO^-O (O O ï OO N ?„ S &- è- 1 ^jj m * ,_: E-I I r-r- n ÖÖ; T- ^ T-^^T-^CM^ ^ T-- CM m m OO => — —^ o, o, ^—-o ,o > > ££ E CMM m en s mm ^r —' 0OO IZTcM -' — coo in co TJ- oioomoo o T-- *- CM S il lc oo io m CM coo T- COO CO O) -* i-- CM 05 cnn IZ co CD h- CDD T- LO oo co S m o CM ZZ ai OO ^ CM CM t;; O £ÜÜ ^1 T- ?Z SS t: II I CMM r*- oo r-- oo o s? CM f^ at CMM Z CMM CM ^t 1-- CO CM cc > EE c E E a> CC> CC "^^ co > > r r TwJ Jw o« « C" "T J J I I CDD «t O O OO II "'SS oc TT Ó 55 75 >> S a> > rara o —— _ (11 ^ O)) JÜ ,_ _ oo .E rara c CN|, coo ai o o 99 = ©TIIII i + CCraj-raj- o) »- O o'E~ = = îr> 2?? 2 ZZ o TOOTOO momom si i « "5 ' 'il UiUi
2000-11 (p=0.09yl. The average age of HIV-2 infected women was identical in the two studiess (26.5 years).
Thee prevalence of HIV-2 rose with age (test for trend: p-0.002; Table 2). HIV-2 infectedd women were on average 2.1 years older than seronegative women (26.5 vs. 24.4 years;; p=0.001). There were no significant differences in prevalence by nationality or maritall status. Prevalence of HIV-2 increased with parity (test for trend: p=0.001). Significantt differences were observed between ethnic groups, the prevalence being highestt among the Jola (1.7%), and lowest among the Wollof (0.4%).
Inn a multivariate analysis age and ethnic group were no longer significant, but higher parityy was a significant risk factor, as was site (Table 2). Subjects recruited at Farafenni weree less likely, and subjects recruited at Sibanor were more likely to have HIV-2.
Discussion n
Comparedd to 1993-5 the HIV-1 prevalence among pregnant women has significandy risenn in one rural area, while the overall prevalence has remained low at 1%. The HIV-2 prevalencee showed a non-significant decrease from 1.0% to 0,8%. In a period of 6 years HIV-22 has been overtaken by HIV-1 as the predominant virus.
Theree were some differences in study design between the 1993-5 and the current study thatt could have lead to bias. The refusal rate in the first study was high (10%) in Serrekundaa but low at the other sites. Thus selection bias may have influenced the risk ratioo in Serrekunda. The current study used dried blood spots as opposed to serum and considerablee delays occurred before they were eluted and tested, so decay of antibodies couldd possibly have occurred. We analysed the possible effect that delays before testing mightt have had on the results (data not shown), and concluded that this was unlikely to havee lead to an appreciable underestimate [17].
68 8 HIV-22 IN WEST AFRICA
Inn 1993-5, HIV-2 prevalence was higher in Sibanor than elsewhere. Although HIV-2 did nott rise further, HIV-1 increased 5-fold and was significantly higher among the Jola's comparedd to other ethnic groups. Prevalence among Senegalese women attending the Sibanorr clinic (less than 10 km from the Senegalese border), most of whom were Jola andd living in the Cassamance, the southern province of Senegal, was more than twice as highh as that in Gambian women. In this area, where there is an armed conflict, there has beenn a significant rise in HIV prevalence among female sexworkers [7)12]. The Jola ethnicc group in the area around Sibanor celebrates initiation ceremonies, during which marriagee bonds are considered temporarily invalid [18]. These ceremonies, which are uniquee to this area, are popular and attended by Jola's from far afield; [19] the promiscuityy during these ceremonies may be an explanation for the rising HIV-1 prevalencee in this area.
Fifteenn years after the first case of HIV-1, the prevalence is still low in The Gambia, and thee same is true in Senegal. Male circumcision, which was shown to be a protective factorr in a recent meta-analysis [20], is nearly universal in The Gambia. The prevalence off herpes simplex virus type 2, an important co-factor for HIV infection [21], is lower in Thee Gambia [19] than in East Africa [22]. HIV-1 prevalences have been lowest in Muslimm countries of North Africa and the Middle East [23], and this is likely to be linked too sexual behaviour. As over 95% of the Gambian population is Muslim [24], the same reasonss could underlie the low prevalence in Senegal and The Gambia. Studies on sexual behaviourr in The Gambia are needed to further interpret the low prevalence.
Acknowledgements s
Wee thank Sophie Njai, Dr Jamie Erskine, Sisawo Konteh, Ivan Cooker, and their teams inn the health facilities for collecting data and samples. We thank all lab staff and Ishbell Millerr for laboratory work; Malang Fofana, Ba Fodeh Jawara, Abi Kahn, and the Divisionall Health Teams for support in planning; Kalifa Manneh, Nellie Lloyd-Evans,
69 9 HIV-22 IN WEST AFRICA
Sanaa Sambou, Alhagie Kolley, Momodou Lamin Cham, Clement Mboto, and Pa Ousman
Bahh for their help in planning and supervision; Shabbar Jaffar and Jim Todd for statistical advicee and suggestions; Richard Hayes and Roel Coutinho for providing comments on ann earlier draft; and Joanne Mei (CDC), Stefano Lazarri and Bernhard Schwartlaender
(WHO),, Catherine Peckham (Institute of Child Health, London) for suggestions. This studyy builds on earlier work by Dr Dairmuid O'Donovan and his team.
Thiss study was funded by the MRC (UK), the Gambia Government, UNDP, and WHO.
Murexx Diagnostics kindly provided the ELISA kits for the study.
Partt of this study was presented at the Xllth International Conference on AIDS and
STD'ss in Africa (ICASA), Ouagadougou, 9-13 December 2001, abstracts number 12BT5-
33 and 12PT5-424.
References s
1.. Djomand G, Greenberg AE, Sassan-Morokro M, Tossou O, Diallo MO, Ekpini E, et al. Thee epidemic of HIV/AIDS in Abidjan, Cöte d'lvoire: a review of data collected by Projett RETRO-CI from 1987 to 1993. J Acquir Immune Defic Syndr 1995, 10: 358- 365. . 2.. Larsen O, da Silva Z, Sandstrom A, Andersen PK, Andersson S, Poulsen AG, et al. Decliningg HIV-2 prevalence and incidence among men in a community study from Guinea-Bissau.. AIDS 1998, 12:1707-1714. 3.. Norrgren H, Andersson S, Biague AJ, da Silva ZJ, Dias F, Naucler A, Biberfeld G. Trendss and interaction of HIV-1 and HIV-2 in Guinea-Bissau, west Africa: no protectionn of HIV-2 against HIV-1 infection. AIDS 1999, 13: 701-708. 4.. Bouckenooghe A, Shandera W. HIV trends in African Blood Donors. Journal of InfectionInfection 1999, 39: 122-128. 5.. UNAIDS/WHO. Epidemiological Fact Sheets by Country. On: http:// www.unaids.org / hivaidsinfoo / statistics / fact_sheets / index_en.htm. Accessed 11 April 2002. 6.. Schim van der Loeff MF, Aaby P. Towards a better understanding of the epidemiologyy of HIV-2. AIDS 1999, 13 (suppl A): S69-S84. 7.. Meda N, Ndoye I, M'Boup S, Wade A, Ndiaye S, Niang C, et al. Low and stable HIV infectionn rates in Senegal: natural course of the epidemic or evidence for success of prevention?? AIDS 1999, 13: 1397-1405. 8.8. UNAIDS. Acting early to prevent AIDS: The case of Senegal. UNAIDS Best Practice Collection.. Geneva: UNAIDS; 1999. 9.. De Cock KM, Brun-Vézinet F, Soro B. HIV-1 and HIV-2 infections and AIDS in West Africa.. AIDS 1991, 5 (suppl 1): S21-S28. 10.. Marlink R, Mboup S, Thior I, Traore I, Essex M, Ndoye I, Kanki P. Incidence of diseasee after HIV-2 infection as compared to HIV-1 infection. 12th World AIDS Conference.Conference. Geneva, July 1998 [Abstract 13319].
70 0 HIV-22 IN WEST AFRICA
11.. O'Donovan D, Ariyoshi K, Milligan P, Ota M, Yamuah L, Sarge-Njie R, Whittle H for thee MRC/Gambia Government/University College London Medical School working groupp on mother-child transmission of HIV. Maternal plasma viral RNA levels determinee marked differences in mother-to-child transmission rates of HIV-1 and HIV- 22 in The Gambia. AIDS 2000, 14: 441^448. 12.. Schim van der Loeff MF, Corrah T, Whittle HC. Low and stable HIV infection rates in Senegal?? [Letter] AIDS 2000, 14: 1276-1277. 13.. Ishikawa K, Fransen K, Ariyoshi K, Ngenkasong JN, Janssens W, Hendrickx L, era/. Improvedd detection of HIV-2 proviral DNA in dually seroreactive individuals by PCR. AIDSAIDS 1998; 12: 1419-1425. 14.. Chin J. Public health surveillance of AIDS and HIV infections. Bull. WHO 1990, 68: 529-536. . 15.. WHO. Unlinked Anonymous Screening for the Public Health Surveillance of HIV infections.infections. Proposed International Guidelines. Geneva: WHO; 1989 (document GPA/SFI/89.3). . 16.. UNAIDS / WHO. Second Generation Surveillance for HIV. CD ROM. Geneva: UNAIDS/WHO;; 2001. 17.. Behets F, Kashamuka M, Pappaioanou M, Green TA, Ryder RW, Batter V, ef a/. Stabilityy of Human Immunodeficiency Virus Type 1 Antibodies in Whole Blood Dried onn Filter Paper and Stored under Various Tropical Conditions in Kinshasa, Zaire. J ClinClin Microbiology 1992, 30: 1179-1182. 18.. Enel C. Social and behavioural risk factors for STD and HIV transmission in the SibanorareaSibanorarea of The Gambia. Fajara: Medical Research Council, 1995. 19.. Shaw M, van der Sande M, West B, Paine K, Ceesay S, Bailey R, ef a/. Prevalence off herpes simplex type 2 and syphilis serology among young adults in a rural Gambiann community. Sex Transm Infect 2001, 77:358-365. 20.. Weiss HA, Quigley MA, Hayes RJ. Male circumcision and risk of HIV infection in sub- Saharann Africa: a systematic review and meta-analysis. AIDS 2000, 14: 2361-2370. 21.. Weiss H, Buve A, Robinson NJ, van Dyck E, Kahindo M, Anagonou S, et al. The epidemiologyy of HSV-2 infection and its association with HIV infection in four urban Africann populations. AIDS 2001, 15 (suppl 4):S97-S108. 22.. Obasi A, Mosha F, Quigley M, Sekirassa 2, Gibbs T, Munguti K, et al. Antibody to herpess simplex virus type 2 as a marker of sexual risk behavior in rural Tanzania. J InfectInfect Dis 1999, 179:16-24. 23.. Walker N, Garcia-Calleja JM, Heaton L, Asamoah-Odei E, Poumerol G, Lazzari S, et al.al. Epidemiological analysis of the quality of HIV serosurveillance in the world: how welll do we track the epidemic? AIDS 2001,15:1545-1554. 24.. Central Statistics Department. Department of State for Finance and Economic Affairs PopulationPopulation and Housing Census 1993. Banjul: Republic of The Gambia; 1996.
71 1 HIV-22 IN WEST AFRICA
72 2 4 4
Incidencee of HIV-2 in a rural community inn Guinea-Bissau, West Africa
Submitted Submitted
Maartenn F Schim van der Loeff u, Peter Aaby M, Jim Todd 2, Akum Aveika Awasana ',
Carloss da Costa \ Koya Ariyoshi l-5, Tim Vincent3, Elizabeth Harding ',
Franciscoo Dias 6, Hilton Whittle '
(1)(1) the Medical Research Council'Laboratories; Fajara, The Gambia; (2) London School of Hygiene
<&<& Tropical Medicine, London, UK; (3) Projecto de Saiide de Bandim, Guinea-Bissau;
(4)(4) Statens Serum Institut, Copenhagen, Denmark; (5) National Institute of Infectious Diseases, Tokyo,
Japan;Japan; (6) Laboratório Nacional de Saüde Püblica, Bissau, Guinea-Bissau
Thiss study was funded by the Medical Research Council (UK).
Incidencee of HIV-2 in a rural community inn Guinea-Bissau, West Africa
ABSTRACT T
Background:: HIV-2 is thought to be transmitted by the same routes as HIV-1, butt there are few studies examining the incidence of HIV-2 in the community. Methods:: Sero-surveys of HIV infection were conducted in 1989-92 and in 1996-88 in the same rural population of Guinea-Bissau, West Africa. We estimatedd the incidence rate of HIV-2 infection and examined risk factors for incidentt infection. Results:: The prevalence among 3109 adults in 1996-8 was 2.7% for HIV-1 andd 7.9% for HIV-2. Of 2257 adult subjects who were not infected with HIV-2 inn the first survey and still resident in the area during the second survey, 1522 (67%)) provided a second blood sample. Fifty-one incident HIV-2 infections weree identified, 13 in men and 38 in women. The overall incidence rate (IR) of HIV-22 was 4.8 per 1000 person-years (95% confidence interval (CI) 3.7-6.4). Thee IR among men was 3.8 (95%CI 2.2-6.5) and among women 5.3 (95%CI 3.9-7.3).. The IR was significantly higher in men younger than 30 years (P=0.02);; in women there was no variation by age. Conclusions:: There is continuing heterosexual transmission of HIV-2 in this rurall area of West Africa, where new infections occur in women of all ages andd in young men. HIV-22 IN WEST AFRICA
Introduction n
HIV-22 can be transmitted by the same routes as HIV-1, but transmission rates are lower [1,2].. This has been attributed to the generally lower plasma viral load (PVL) [3-5]. Cross-sectionall studies have shown that risk factors for HIV-2 infection are similar to riskk factors for HIV-1 infection, among them history of sexually transmitted diseases (STDs),, serological evidence of genital ulcer disease, having had sex with a commercial sexx worker (CSW), high number of lifetime sexual partners, blood transfusion, and lack off circumcision in men [6-16]. The only consistent difference is that HIV-2 is more commonn in older adults [7,17]. Studies of seroprevalent cases may be biased, as participantss who die more rapidly from the infection will be under-represented.
Incidencee rates of HIV-1 have been estimated in several prospective community-based orr occupational cohort studies in sub-Saharan Africa, and range between 5 and 15 per 10000 person-years of observation (pyo) [18-23]. Six cohort studies have reported incidencee rates of HIV-2, all from West Africa: two studies among CSWs [1,24], one occupationall cohort [21,25,26], one cohort of women recruited at an antenatal clinic (ANC)) [27], and two community-based studies [14,28-31]. The largest study had 46 sero incidentt cases [1]; the smallest zero [24,27]. The incidence rate (IR) was 0 and 1.1 per 10000 pyo in the CSW cohorts, and varied between 0 and 8.3 per 1000 pyo in the other cohorts. .
Threee of the studies with sero-incident data have reported on risk factors. In a community-basedd study from Bissau, Guinea-Bissau, a history of a sexually transmitted diseasee (STD) and higher age were risk factors among women [14,28,29]. In a cohort of policee officers from Guinea-Bissau, a positive Treponema pallidum hemagglutination assay (TPHA)) and history of genital ulcer disease (GUD) were associated with HIV-2 incidence,, though not-significandy so (P=0.06 and P=0.08 respectively) [25]. In a cohort off CSWs in Dakar, Senegal, sex workers from other countries tlian Senegal and Ghana weree more likely to acquire HIV-2 than Senegalese women (P < 0.05) [1].
76 6 HIV-22 IN WEST AFRICA
Thee objectives of this study are to estimate the incidence rate of HIV-2, and to examine riskk factors for incident HIV-2 infection in a high-prevalence community in a rural area inn West Africa.
Methods s
Subjects Subjects Thee study was conducted in a rural area in North-western Guinea-Bissau. The methods havee been described in detail elsewhere [32]. In brief, in 1989-1992 a first survey among thee adult population was conducted, consisting of a short questionnaire and HIV serologyy [17,33]. During die second survey in 1996-1998 [32], a questionnaire was administeredd to all participants and a blood sample obtained. Questions were asked regardingg a range of possible risk factors for STDs, including HIV. The blood sample waswas tested for syphilis and HIV serology, haemoglobin, and malaria parasites. Samples withh dual sero-reactivity (positive for both HIV-1 and HIV-2) were subjected to further testingg by a line-immuno assay and polymerase chain reaction (PCR). Diagnosis of dual infectionn with both HIV-1 and HIV-2 (HIV-D) was based on PCR in 16 of the 17 cases; onee sample was insufficient and the diagnosis was based on a convincing serological pattern.. Active syphilis was defined as positive reactions of any titre by both TPHA (Microsyphh TP, Porton Cambridge, Newmarket, UK) and the Rapid Plasma Reagin test (Bectonn Dickinson, Cockeysville, Maryland, USA).
TreatmentTreatment <&'feedback Subjectss with healdi problems were referred by the field worker to the project physician forr examination and treatment the same day. Field workers re-visited all homes to providee the results of malaria, haemoglobin, and syphilis tests, and provide treatment if indicated.. Anti-retroviral therapy was not available. The study was approved by the Gambiaa Government / MRC Joint Ethics Committee, and by the Science Committee of thee Ministry of Health in Guinea-Bissau. All subjects gave verbal informed consent for
77 7 HIV-22 IN WEST AFRICA theirr participation. Participants could obtain their HIV test result from the study counsellor,, in accordance with Guinea-Bissau HIV policy.
StatisticalStatistical methods Ann analysis of risk factors for incident HIV-1 infection has been published separately [32].. The observation time per subject was calculated as the time between the first and thee last blood sample. HIV infections were assumed to have occurred midway between thee dates of the last seronegative and the first seropositive sample. Incidence rates were calculatedd as the number of HIV-2 infections per 1000 pyo. Poisson regression was used too assess independent significant risk factors for HIV-2 incidence. All analyses were donee separately for men and women. Age at entry was used throughout the analysis, exceptt in the estimation of the age of infection. Age group was kept a priori in the models. .
Results s
Prevalence Prevalence Inn 1996-8, 3109 subjects provided a blood sample: 49 (1.6%) were singly HIV-1 positive, 2111 (6.8%) singly HIV-2 positive, and 36 (1.2%) HIV-D. Considering HIV-D as both HIV-11 and HIV-2 infected, the overall HIV-1 prevalence was 2.7%, and the overall HIV- 22 prevalence 7.9%. The overall prevalence of HIV-2 had not changed significantly comparedd to 1989-91 (from 8.3% to 7.9%; P = 0.62, y_2 test), nor had the gender-specific prevalencess changed significantly (from 9.3% to 9.0% for women [P — 0.81], and from 6.6%% to 6.3%) for men [P = 0.77], test). The prevalence of HIV-1 had increased significantlyy since the first survey, from 0.5% to 2.7% (P < 10"7).
SubjectsSubjects with blood samples in two surveys Inn the first survey, conducted between 1989 and 1992, 3064 subjects provided a blood sample,, but 36 of whom no further details were available and 83 who were younger than 155 years were excluded from the analysis. Of the 2945 adult subjects, 230 had HIV-2
78 8 HIV-22 IN WEST AFRICA
andd 10 HIV-D, and were no longer at risk of HIV-2 infection. Of die remaining 2705 at riskk of HIV-2 infection, 284 died and 164 moved permanendy from the village. Of the 22577 still resident in the village 1522 (67%) provided a valid blood sample in 1996-8. Thee reasons a second blood sample was not provided were: travelled but expected back (nn = 325), not at home when visited (n - 276), refusal (n = 131), and insufficient sample (nn = 3). Those with a second sample were more often female than those with only a baselinee sample (P < 0.0005, x2 test), but there were no significant differences in median age,, in proportion CSWs among women, or in proportion living in the central area (data nott shown).
IncidenceIncidence rate of HIV-2 Fifty-onee incident HIV-2 cases were identified: 34 HIV-2 single and 17 HIV-D infections.. The overall incidence rate was 4.8 per 1000 pyo (95% CI 3.7-6.4), the IR amongg men being 3.8 (95% CI 2.2-6.5) and among women 5.3 (95% CI 3.9-7.3; %2 test> d.f.. = 1, P = 0.28). The median age at injection was 31 years among men, and 34 years amongg women (P = 0.31; ranksum test). All subjects with incident HIV-2 infections had hadd sexual intercourse. Only one incident infection occurred among the 21 subjects who hadd had a blood transfusion since 1989.
RiskRisk factors in women Tablee 1 shows die incidence rates stratified by risk factors for women. The incidence ratee did not vary by age. Various risk factors reflecting mobility, multiple sex partners, andd (history of) STDs were significantly associated with HIV-2 incidence. In a multivariablee model for incident HIV-2 infection in women the following remained independentt significant risk factors after adjusting for age group (see Table 1): having everr been divorced, having had injections in the last 12 months, having ever had foul- smellingg vaginal discharge, having drunk alcohol on day of interview, having lived in Bafataa or Gabu (two towns up-country) for longer than one month, and having active syphilis.. Having 3 or more living children was protective. History of sex work, blood transfusionss or genital ulcers were not significant in the multivariate model.
79 9 cuu co II «
coo 0
c c coo -«- 0 0 coo co 0 0 *** C *J w 22 8 »» «> n cóó có CDD -C > 33 -O >- 00 0 f** Ti- Q Q ** o 00 J2£ £ 00 R OO T3 (00 ® CC CO coo co (UU -O ^^ CU 5* * ww co o o Ü Ü sico oi c c cFF ™ cö1™ OO > 0)) o — CMM Y ?? £ ^ ™ fro-fro- uSS cö .. - CU oo !" T-- lO ?? ,- có "^ t^- PP P —— -°° OO CU OO CO ^^ co ^ en i-:: ö "" TJ ^ CM
(00 CU "OO o CMM E
a a .00 0 h-- ** to K«« « O) O) 0 0 CO O SI I ** o~ co o "> 0? ? ebb m in n a> > COO ^ ~ ^ T3.E E uSS ó o> >0 0 ^CM^ ^ 22 £ö o o >> 4 ? 99 d> ,__ o £2-coo ^ coo 06, o. .jo _ _ £2.2.CMM £2, :: o m co"r r £2.4 4 o>> co i-- 0 0 COO CD JJ^CM << 'S co o ^^ en ^ ">-« f-- co^co o ** 00 CM M 44 OS ,j CO —— >2 "o o XX ^ c c 'ii £ aa o 22 c >=i i (00 o «« *: 100 co h- NN CM CC CU ?*-?*- > coo co -^f OO ^t O)) CM COO ^ 5 ^ == oC 0)0f f coo m COO C-- ïïïï in J£ m CC 3 OO 0) CMM CM CM Ifii T- coo ^ CMM rr V)) CA JQ Q O O
CM Maa o) >> TJ w ™™ co ïgg 8
100 CM M- coo co CMM mroo o If)) CM coo « £ ™ Ttt CM CO NN CM coo co ££ TJ- £CM
OO CM «« <" ^^ o> 0 0 C33 O! r?:i i ï ï CUU JL ^^ 0 X X coo 00 0 0 *-- o> (0 0 (ft t ©© 1- J2288== -"0= 00 « "5 5 3 3 oo ^^ Q.. £.' £ w —— w _ ?-- Cö u u 33 (0 (0 Jü CC CO c 0 0 CUU — — XJ CU cï> a>> ioo O II OO 0 C OO O CU kk 0) 4) Bt0C.y> << T- CO A -
co o^L ^L c^ o^ o CMM Tj- T_ _o o CD D
O) O) O O o o
a> > o o c c coo r>- o o T-- co o O O «>> O s* coo Y co o 22 7 7 'S O cóó CM 1-- *"? o o SLS, , S2& & T-- m r>-- 1-- o o O O T-- o CM M co o CM M CM M i3_ 3_ O O
] 00 0 9 6 *99 IO (N Ncs i i 5)) ^ cd dC M M COO £
12. tDD CM 23. -6.2 inn ó «? CMM CO CT>Tl CT>" " 000 T- r^ ^ CM'.CÓ. . CM'' m' £L coo to i_il_Eüi iLL * " COO f- CM Mo o v T-- CO (Ct-® ® 11 't o o tsJoóó co Tj-- CM * * CO O
V)V) - t-- c »» (0 o>> m 88 5" £:: CM ££ co CJ>> m oo a> Is s "" CM * rnn C»
CMM 5 * coo £ w
CMM h- coo * 100 COO CO T-- CM co tD 000 *- r--- CM
> > I I O O O) ) 88 1 u u O) ) fa a ,C C c c o o «E E > > EE o o o o o > > »» co SI I 315 5 ©a,, o (A A SS S f-cc ©CM E == = C fc ï; o>> o> „ « CC « UU O CD OO CD >> "° w ZZ >- 33 £ O O O >> es o cu ZZ O Z f co LUU > Z >- zz >- fljflj C O -- XX oz >- 0 0 _3 3 CM M CO O O O > > O O Q. . 00 0 O) ) _ _ co o O O 0d0> 0 > OO ?s\~ O) ) +*+* c & \J l»J J ^1 1 «« o> re ^0 33 "O «- SS 11 1 11 1 1 11 1 có ó T—— CM «co o« <<* * co o 1^-- m m ÏJ|8 8 co o co' ' eb b cv v 0 0 in n ,c c _2 2 1^. . 0 0 n n O O 0 0 O O O O ö ö 0 0 0 0 Ö Ö c c > > ö ö 0 0 V V CD D O. . (0 0 0) ) 5 5 CJ J c c 5T T 57 7 cö" " O O Ö Ö 0 0 c\i i co o 4)) O — ^__ r^ ^<9.n n T-- CM T-- CO O O c c CC co o co o CD D «SeS e TT co m m CO O T E E 3 3 ~ ~ *~~ iri ^ ^ co o CO O 0 0 Ü Ü
C^M~ M ~ 0 0 1_ _ T— — 0 0 01 1 a a „„(O" " r—G? r—G? 00 * 0 0 coP P ™in n »» N.- 22 o~ 000 co r*- inn T|- COO tt e e 22 &ö 11 1 1 11 m OJ J 000 Tt 0 cq-A A óó d> tnn có v> v> ^^ 4 CM',, tri co. co'' cd 0 0 SS 0 to £L4 4 ^5 5 » » ^^ CO r--- c\i o> CMM m OJJ co a a 1/11 ^^ co ^^ iri 't ^ co'' ö ^^ r^ £ £ '0 0 CM M 'S S _c c
0 0 £ £ 'm 'm ww .- 3 3 L.. C aa O Cö ö >> ra mm in ^^ * 0 000 h>- 000 r- T3 3 000 00 mm CM 2 && CM C C 000 0 CDD 0 ir OO O CD D coo co CMM "fr ^ COO ^ mm CM CDD ^ soo to CD D T3 3 3 3 11 0 O O TJ J C C — — ca a r-»» ,_ CDD CM co o >> "o 3> t-t- CN ?s« « co o *-- CM S" " ro ro i|| 5 > o> > c c c c 00 0 X X to o en n (0 0 > CC CD QQ c II I u u 22 w « 'SS .2 'SS AA JIJ g.too £ "OO k. ^^ CD 0 OO E ïï O 0 0 rara >. >. == 3 OO c CD O 00 «0 £1 £1 tvv o> 00 00 «E co ra ra o>->-- ^- «raa 0 S >> ra 0 CD 1-- << Al VI ZZ E C O £ ItZ> > 11 « z >- Uii ffi z >- HIV-22 IN WEST AFRICA
RiskRisk factors in men Theree were 13 incident HIV-2 infections among 501 men. The median age at enrolment of thee 13 male seroconverters was 27 years (IQR 21-30). Age less than 30 years (p = 0.02), serologicall evidence of active syphilis (p = 0.01), and history of injections outside clinics (p < 0.0005)) were significandy associated witli incident HIV-2 infection; in a Poisson multivariable modell these three factors remained independendy associated. Sex with a CSW, circumcision, andd history of STD's were not associated with HIV-2 infection.
Discussion n
Thee incidence rate of HIV-2 in this rural West African area was 4.8 per 1000 pyo. This is similarr to die rates found in previous cohort studies in Guinea-Bissau [21,29]. In the same periodd the HIV-1 incidence was 4.4 per 1000 pyo in the study area [32], although the prevalencee of HIV-1 is much lower (2.7%, versus 7.9% for HIV-2). Significant risk factors forr HIV-2 infection were travel, history of vaginal discharge, co-infection with syphilis, historyy of divorce, history of recent injections, drinking alcohol on the day of the interview in women;; and young age, history of injections outside clinics, and syphilis in men.
AA second blood sample was not available from 33% of the subjects enrolled in the first survey.. This was mainly due to the mobility of the population. More men were lost than womenn (P < 0.0005) As incidence was associated with travel (in women at least), the estimatedd incidence rates may be underestimates.
Thee median period between the first sample and the second was very long (7.2 years [interquartilee range (IQR) 6.7 - 7.7]), and therefore the seroconversion dates could not be estimatedd with precision. For the calculation of HIV-2 incidence we assumed a mid-point seroconversionn date. Although this is imprecise on an individual level, this will have had little effectt on the estimation of the incidence rate.
83 3 HIVV 2 IN WEST AFRICA
Becausee a history of infections in the last 12 months could be a consequence of HIV-2 infectionn (seeking health cate for HJV-2 induced illness) rather tlian a cause of then infection, wee repeated the multi-variable analysis leaving out injections in the last 12 months from the model.. The same variables as mentioned above remained in the final model, and the size of thee rate ratios was not materially different (data not shown). As it is unknown which came first,, HIV-2 infection or injections, no conclusions about causality can be drawn.
Bloodd transfusions after 1989 were not a risk factor for HIV-2 in this population. In view of thee significant risk factors, many directly or indirectly related to sexual activity, it appears that transmissionn of HIV-2 in this population is predominantly sexual. This is in agreement with earlierr studies on HIV-2 and similar to what is known about HIV-1 in sub-Saharan Africa. Surprisingly,, commercial sex work was not a risk factor for HIV-2. The majority of known CSWss in the area were already infected with HIV-2 in 1989-92, and were no longer at risk for incidentt HIV-2. This study included only 29 CSWs, and the power to confirm sex work as a riskk factor was therefore small.
Inn this same population significant risk factors for incident HIV-1 infection were history of commerciall sex work, preceding HIV-2 infection, and age 30 years or higher in men [32]. As thee epidemics of HIV-2 (established and stable) and HIV-1 (emergent) are in a different phasee in this area, it is hard to draw firm conclusions about these different patterns of risk factors. .
Amongg men active syphilis was a strong independent risk factor. Ten out of 13 incident infectionss occurred in men who were younger than 30 years at enrolment, and young age was aa strong independent risk factor. It may be the case that young men (often unmarried) more frequentlyy have sexual contacts with CSWs, putting them at higher risk for HIV-2. Sex with sexsex workers was not an independent risk factor, but that may have been under-reported. The powerr to detect significant risk factors in men was limited due to small numbers.
Inn most societies incidence of sexually transmitted infections (including HIV-1) is highest amongg young adults (15-30 years of age). This is also the case for HIV-2 in men in this rural
84 4 HIV-22 IN WEST AFRICA population,, and this observation supports the hypothesis that the high prevalence of HIV-2 foundfound in older men in the area [17] may be due to a cohort effect. Among women no effect off age on incidence was found, and die high prevalence among older women may be due to neww infections, or to a cohort effect or both. It is striking that the incidence of HIV-2 did nott decline with age in women. Assuming a decreasing sexual activity with age, this lends supportt to the hypothesis of an increased susceptibility to retroviral infections in older womenn [33, 34].
Thee adult prevalence of HIV-2 in the study area was stable at around 8.0% during a 7-year period.. Most other studies in West Africa have found a stable prevalence [35,36], although twoo Guinea-Bissauan studies have found a declining prevalence [21,29]. In the same period thee HIV-1 prevalence increased sharply from 0.5% to 2.7%. HIV-1 may soon overtake HIV- 22 as the predominant retrovirus in Guinea-Bissau, as has happened in The Gambia [36].
Inn conclusion, HIV-2 transmission is ongoing in rural Guinea-Bissau at an incidence rate of aroundd 5 per 1000 pyo. Risk factors indicate that transmission is now mosdy sexual.
Acknowledgements s
Wee are grateful to the population of Caio for participating in the studies over the years. We thankk Dr Andrew Wilkins, Dr Dominic Rkard, and Margaret Buckner for the initial work on whichh this study is built; Dr Abraham Alabi, Alhajie Bayang, Dr Tom Blanchard, Dr Assan Jaye,, Pa Tamba N'gom, Mamady Njie, Bakary Sanneh and Ramu Sarge-Njie, for the lab work;; the team of Caio field assistants for the field work; Chad Anderson, Margaret Buckner, Stephaniee Constantine, Bryan Savage, Glyn Taylor and Lucy Pembrey, for supervising the fieldd work; Dr. Mamae Barbosa (Ministério de Saüde Püblica, Bissau) for her support; Dr Robinn Bailey, Prof. Roel Coutinho, Prof. Richard Hayes and Dr Emmanuel Lagarde for advice;; and Peter Langfield and M'bemba Ceesay for transport.
85 5 HIV-22 IN WEST AFRICA
References s
1.. Kanki PJ, Travers KU, Mboup S, Hsieh CC, Marlink Rg, Gueye-NDiaye A, et al. Slower heterosexuall spread of HIV-2 than HIV-1. Lancer 1994, 343:943-946. 2.. Adjorlolo-Johnson G, De Cock KM, Ekpini E, Vetter KM, Sibailly T, Brattegaard K, et al.: Prospectivee comparison of mother-to-child transmission of HIV-1 and HIV-2 in Abidjan, Ivoryy Coast. JAMA 1994, 272:462-466. 3.. De Cock KM, Adjorlolo G, Ekpini E, Sibailly T, Kouadio J, Maran M, et al. Epidemiology andd transmission of HIV-2. Why there is no HIV-2 pandemic. JAMA 1993, 270:2083-2086. 4.. Andersson S, Norrgren H, da Silva Z, Biague A, Bamba S, Kwok S, et al. Plasma viral loadd in HIV-1 and HIV-2 singly and dually infected individuals in Guinea-Bissau, West Africa.. Arch Intern Med 2000,160:3286-3293. 5.. O'Donovan D, Ariyoshi K, Milligan P, Ota M, Yamuah L, Sarge-Njie R, et al. Maternal plasmaa viral RNA levels determine marked differences in mother-to-child transmission ratess of HIV-1 and HIV-2 in The Gambia. AIDS 2000, 14:441-448. 6.. Dufoort G, Courouce AM, Ancelle-Park R, Bletry O: No clinical signs 14 years after HIV-2 transmissionn via blood transfusion [letter]. Lancet 1988, 2:510-511. 7.. Poulsen AG, Kvinesdal B, Aaby P, Molbak K, Frederiksen K, Dias F, Lauritzen E. Prevalencee of and mortality from human immunodeficiency virus type 2 in Bissau, West Africa.. Lancet-\ 989; 1:827-31 8.. Wilkins A, Hayes R, Alonso P, Baldeh S, Berry N, Cham K, et al. Risk factors for HIV-2 infectionn in The Gambia. AIDS 1991;5:1127-32. 9.. Pepin J, Dunn D, Gaye I, Alonso P, Egboga A, Tedder R, et al. HIV-2 infection among prostitutess working in The Gambia: association with serological evidence of genital ulcer diseasess and with generalized lymphadenopathy. AIDS 1991;5:69-75. 10.. Naucler A, Albino P, Da Silva AP, Andreassin PA, Andersson S, Biberfeld G. HIV-2 infectionn in hospitalised patients in Bissua, Guiena-Bissau. AIDS 1991; 5: 301-4. 11.. Naucler A, Albino P, Da Silva AP, Biberfeld G. Sexually transmitted disease and sexual beahviourr as risk factors for HIV-2 infection in Bissau, Guinea-Bissau. Int J STD AIDS 1993;4:217-21. . 12.. Kanki P, M'Boup S, Marlink R et al.: Prevalence and risk determinants of human immunodeficiencyy virus type 2 (HIV-2) and human immunodeficiency virus type 1 (HIV-1) inn west African female prostitutes. Am J Epidemiol 1992, 136:895-907. 13.. Pepin J, Quigley M, Todd J, Gaye I, Janneh M, Van Dyck E, et al. Association between HIV-22 infection and genital ulcer diseases among male sexually transmitted disease patientss in The Gambia. AIDS 1992;6:489-93. 14.. Poulsen AG, Aaby P, Gottschau A, Kvinesdal BB, Dias F, Molbak K, etai: HIV-2 infection inn Bissau, West Africa, 1987-1989: incidence, prevalences, and routes of transmission. J AcquirAcquir Immune Defic Syndr 1993, 6:941-948. 15.. Abbott RC, NDour-Sarr A, Diouf A, NDoye T, Ba D, Tall N, ef al.: Risk factors for HIV-1 andd HIV-2 infection in pregnant women in Dakar, Senegal. J Acquir Immune Defic Syndr 1994,7:711-717. . 16.. Sassan-Morokro M, Greenberg AE, Coulibaly IM et al.: High rates of sexual contact with femalee sex workers, sexually transmitted diseases, and condom neglect among HIV- infectedd and uninfected men with tuberculosis in Abidjan, Cote d'lvoire. J Acquir Immune DeficDefic Syndr Hum Retrovirology 1996, 11:183-187. 17.. Wilkins A, A, Ricard D, Todd J, Whittle H, Dias F, Paulo Da Silva A. The epidemiology of HIVV infection in a rural area of Guinea-Bissau. AIDS 1993; 7: 1119-1122. 18.. Borgdorff MW, Barongo LR, Klokke AH, Newell JN, Senkoro KP, Velema JP, et al. HIV-1 incidencee and HIV-1 associated mortality in a cohort of urban factory workers in Tanzania. GenitourinGenitourin Med 1995; 71: 212-5.
86 6 HIV-22 IN WEST AFRICA
19.. Grosskurth H, Mosha F, Todd J, Mwijarubi E, Klokke A, Senkoro K, et al. Impact of improvedd treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomisedd controlled trial. Lancet 1995; 346:530-6. 20.. Wawer MJ, Sewankambo NK, Serwadda D, Quinn TC, Paxton LA, Kiwanuka N, et al. Controll of sexually transmitted diseases for AIDS prevention in Uganda: a randomised communityy trial. Rakai Project Study Group. Lancet 1999; 353:525-35. 21.. Norrgren H, Andersson S, Biague AJ, Da Silva ZJ, Dias F, Naucler A, et al. Trends and interactionn of HIV-1 and HIV-2 in Guinea-Bissau, west Africa: no protection of HIV-2 againstt HIV-1 infection. AIDS 1999; 13: 701-707. 22.. Senkoro KP, Boerma JT, Klokke AH, Ng'weshemi JZL, Muro AS, Gabone R, et al. HIV incidencee and HIV-associated mortality in a cohort of factory workers and their spouses in Tanzania,, 1991 through 1996. JAcquirImmune Defic Syndr2000, 23: 194-202. 23.. Mbulaiteye SM, Mahe C, Whitworth JAG, Ruberantwari A, Nakiyingi JS, Ojwiya A, et al. Decliningg HIV-1 incidence and associated prevalence over 10 years in a rural population inn south-west Uganda: a cohort study. Lancet 2002; 360: 41-6. 24.. Ghys PD, Diallo MO, Ettiegne-Traore V, Satten GA, Anoma CK, Maurice C, et al. Effect of interventionss to control sexually transmitted disease on the incidence of HIV infection in femalee sex workers. AIDS 2001; 15:1421-31. 25.. Norrgren H, Andersson S, Naucler A, Dias F, Johansson I, Biberfeld G: HIV-1, HIV-2, HTLV-I/III and Treponema pallidum infections: incidence, prevalence, and HIV-2- associatedd mortality in an occupational cohort in Guinea-Bissau. J Acquir Immune Defic SyndrHumSyndrHum Retrovirology 1995, 9:422-428. 26.. Norrgren H, Da Silva ZJ, Biague AJ, Andersson S, Biberfeld G. Trends of HIV-1 and HIV- 22 in Guinea-Bissau before and after the civil war 1998-1999. Xllth International ConferenceConference on AIDS and STDs in Africa. Ouagadougou, December 2001 [abstract 11DT5-5]. . 27.. Wiktor SZ, Nkengasong JN, Ekpini ER, Adjorlolo-Johnson GT, Ghys PD, Brattegaard K, etet al. Lack of protection against HIV-1 infection among women with HIV-2 infection. AIDS 1999;; 13:695-699. 28.. Poulsen AG, Aaby P, Larsen O, Jensen H, Naucler A, Lisse IM, et al. 9-year HIV-2- associatedd mortality in an urban community in Bissau, west Africa. Lancet 1997; 349: 911-914. . 29.. Larsen O, da Silva Z, Sandstrom A, Andersen PK, Andersson S, Poulsen AG, et al. Decliningg HIV-2 prevalence and incidence among men in a community study from Guinea-Bissau.. AIDS 1998, 12:1707-1714. 30.. Diop OM, Pison G, Diouf I, Enel C, Lagarde E. Incidence of HIV-1 and HIV-2 infections in aa rural community in southern Senegal. AIDS 2000; 14:1671-1672. 31.. Holmgren B, da Silva Z, Larsen O, Vastrup P, Andersson S, Aaby P. Dual infections with HIV-1,, HIV-2 and HTLV-I are more common in older women than in men in Guinea- Bissau.. AIDS 2003; 17: 241-253. 32.32. Schim van der Loeff MF, Aaby P, Ariyoshi K, Vincent T, Aveika AA, Da Costa C, et al. HIV-22 does not protect against HIV-1 infection in a rural community in Guinea-Bissau. AIDSAIDS 2001; 15: 2303-2310. 33.. Aaby P, Ariyoshi K, Buckner M, Jensen H, Berry N, Wilkins A, et al. Age of wife as a majorr determinant of male-to-female transmission of HIV-2 infection: a community study fromm rural West Africa. AIDS 1996; 10: 1585-1590. 34.. Holmgren B, Aaby P, Jensen H, Larsen O, da Silva Z, Lisse IM. Increased prevalence of retroviruss infections among older women in Africa. Scand J Infect Dis 1999; 31:459-66. 35.. Djomand G, Greenberg AE, Sassan-Morokro M, Tossou O, Diallo MO, Ekpini E, et al. The epidemicc of HIV/AIDS in Abidjan, Cöte d'lvoire: a review of data collected by Projet RETRO-CII from 1987 to 1993. J Acquir Immune Defic Syndr 1995, 10: 358-365. 36.. Schim van der Loeff MF, Sarge-Njie R, Ceesay S, Aveika AA, Jaye P, Sam O, et al. Regionall differences in HIV trends in The Gambia:results from sentinel surveillance amongg pregnant women. AIDS Accepted.
87 7 HIV-21\\ WEST AFRICA
88 8 5 5
HIV-22 does not protect against HIV-1 infection inn a rural community in Guinea-Bissau
AIDSAIDS 2001; 15 (17): 2303-10
Maartenn F Schim van der Loeff ', Peter Aaby 2, Koya Aryioshi ', Timothy Vincent ,
Akumm Aveika Awasana , Carlos Da Costa ~, Lucy Pembrey ', Francisco Dias ,
Elizabethh Harding ', Helen A Weiss \ Hilton C Whittle '
(1)(1) the MRC laboratories, Banjul, The Gambia, (2) Projecto de Saüde de Bandim, Bissau, Guinea-
Bissau,Bissau, (3) Ministério de Saüde Publico, Bissau, Guinea-Bissau, and (4) London School of Hygiene
andand Tropical Medicine, lôndon, UK.
Thiss research was funded by the Medical Research Council (UK).
HIV-22 does not protect against HIV-1 infection inn a rural community in Guinea-Bissau
ABSTRACT T
Objective:: To examine the putative protective effect of HIV-2 infection againstt subsequent HIV-1 infection. Design:: Retrospective analysis of data from two cross-sectional surveys in thee same community. Methods:: Two surveys between 1989 and 1998 in a rural area in North- westernn Guinea-Bissau. Data from residents aged 15-59 years were analysedd for the study. HIV testing was done in the first survey, and in the secondd survey tests for both HIV and syphilis were done, and data on socio- demographicc factors and sexual behaviour, including commercial sex work, weree gathered. Qualitative PCR amplification of HIV-1 and HIV-2 viral DNA wass performed on serologically dually reactive samples. Results:: Of the 2276 eligible adult villagers initially tested 60% (1360) providedd a second sample. Of 110 HIV-2 infected subjects, 17 became additionallyy infected with HIV-1 (incidence rate (IR) = 26.3/1000 person-years off observation (pyo)). Of the 1250 HIV seronegative subjects, 24 became infectedd with HIV-1 (IR = 2.8/1000 pyo). The incidence rate ratio (IRR), comparingg the incidence rate in HIV-2 infected people with the rate in HIV- seronegativee subjects, was >1 in all three 'risk groups': men, female commerciall sex workers, and other women. The overall estimate of the IRR, adjustedd for age group and risk group, was 3.24 (CI: 1.5 - 7.1). Conclusions:: We found no protective effect of HIV-2 in this population. HIV- 22 cannot be regarded as a vaccine, but instead, may be a risk factor for HIV-1 infection. . HIV-22 IN WEST AFRICA
Introduction n
Thee Human Immunodeficiency Virus Type 2 (HIV-2) is a retrovirus with a genetic structuree very similar to HIV-1, but a nucleotide homology of only 60% [1]. It can cause
AIDSS [2,3], though the progression to disease and to immune deficiency appears slower thann that of HIV-1 [4], and HIV-2 associated mortality is less than that of HIV-1 [5]. In
19955 Travers et al. reported that the incidence of HIV-1 in commercial sex workers
(CSWs)) infected with HIV-2 was lower than the incidence of HIV-1 in seronegative
CSWs.. The "protective effect" of HIV-2 was estimated at about 70% (95% confidence intervall (CI) 41 - 80%) [6]. However, this study has been criticised for a serious selection bias,, which may be responsible for the finding [7]; the authors of the original paper have donee additional analyses rejecting this possibility [8].
Fourr other cohort studies, all conducted in West Africa, have examined this putative protectivee effect, but none of them confirmed the finding (Larsen, unpublished data) [9,
10,, 11, 12]. In The Gambia among sex workers the HIV-1 incidence rate ratio (IRR), comparingg HIV-2 infected subjects to seronegative subjects, was 1.7 (CI 0.7 — 3.9); howeverr in this study the loss to follow-up was substantial at 59% [9]. In Abidjan, Cöte d'lvoire,, a study among a cohort of women attending antenatal clinics, estimated an IRR off 2.7 (CI 0.7 — 11.2), but this study could not adjust for sexual behaviour, and dual infectionn was not defined by PCR [12]. In a peri-urban area in Guinea-Bissau, an IRR of
7.55 (CI 1.95 — 27.8) was found in the general population, adjusted for symptoms of sexuallyy transmitted diseases (STD's) (Larsen, unpublished data). Finally, in Guinea-
Bissau,, Norrgren et al. found an IRR of 1.65 (CI 0.7-3.7) in a cohort of police officers, afterr adjusting for various confounding variables, including sexual behaviour [11]. In all off these studies, including the paper by Travers et al, the number of incident HIV-1 infectionss among HIV-2 infected subjects was small, being between 6 and 8.
Inn this paper we investigate the possible protective effect of HIV-2 in preventing HIV-1 infectionn in a retrospective analysis of two cross-sectional surveys in the same rural area off North-western Guinea-Bissau.
92 2 HIV-22 IN WEST AFRICA
Methods s
Subjects Subjects Thee study was set in a string of villages in north-west Guinea-Bissau. From 1989 to 1991 ann HIV sero-survey was conducted among 2770 inhabitants targeted at diose aged 15 yearss or older [13]. In 1991 and 1992 additional small surveys were carried out in order too recruit subjects who had been absent during the first survey (n = 294) [14]. In our analysiss these surveys have been grouped together as "first survey". In the second survey fromm 1996 to 1998 all adults in the same area were re-surveyed [15].
AA yearly census by a team of local field workers records births, deaths, immigration into, andd emigration from die area. Many male villagers work and live in large towns in the sub-region,, and in Portugal and France, and many women work or used to work as prostitutess in the capital (Bissau) or in neighbouring countries (The Gambia, Senegal) [16].. However, due to strict observation of death rituals, which also apply to villagers dyingg abroad, few adult deaths of subjects native to the area go unnoticed. Thus we are confidentt that most deaths are recorded accurately in the census.
Inn the second survey (1996-8) a questionnaire concerning risk factors for HIV, including agee of first sexual intercourse, number of sexual partners within last 12 months, history off sexually transmitted diseases (STD's), frequency of sexual intercourse in the last month,, history of sex with a CSW, and a history of sex work, was administered.
HIVV infected subjects and a similar number of seronegative controls identified in 1989- 922 were followed up more regularly than the other participants, and had free access to a clinicc run by a research physician[17]. Whether subjects in this group were involved in commerciall sex work was established by field assistants [14, 16], as well as by the questionnairee used in the second survey.
93 3 HIV-22 IN WEST AFRICA
LaboratoryLaboratory methods
Capillaryy blood samples (1989-91) or venous blood samples (1991 onwards) were collected.. Serum and cells were separated in the field laboratory in the study area, and transportedd to the main laborator)' in Fajara, The Gambia, where they were stored, and wheree all tests were performed. Initially, sera were screened separately for HIV-1 and
HIV-2,, using respectively a competitive HIV-1 ELISA (Wellcozyme, Wellcome
Diagnostics,, Dartford, UK), and a previously described [18] competitive test specific for
HIV-2.. From October 1990 sera were screened using the Wellcozyme combined test for
HIV-11 and HIV-2 (Murex Diagnostics, Dartford, England, UK), and since August 1996 usingg the ICEHIV-l.O.2 (Murex). Reactive sera were further tested using mono-specific
ELISA'ss (Wellcozyme 1 for HIV-1 throughout, and initially Wellcozyme II, and from
Marchh 1996 ICEHIV-2 for HIV-2[Murex]). Samples that were dually reactive were furtherr tested with a line immuno-assay based on synthetic peptides (Pepti-Lav, Pasteur,
Marnes-la-Coquette,, France).
Sampless that were dually reactive by both ELISA and Pepti-Lav, were considered as potentiallyy dual infections. These samples, and those where a firm serological diagnosis couldd not be achieved, were further examined by both HIV-1 specific and HIV-2 specific
PCR,, performed on DNA obtained from Peripheral Blood Mononuclear Cells, using nestedd PCR primers based on the HIV-type specific LTR region [19, 20, 21]. Samples in whichh both HIV-1 and HIV-2 LTR signals were detected by nested PCR, were defined as duall infection (HIV-D).
Sampless from the second survey were screened for syphilis antibodies by Treponema pallidumpallidum haemagglutination assay (TPHA) (Microsyph TP, Porton Cambridge,
Newmarket,, UK). All subjects who had a positive TPHA and a positive RPR result, were treatedd with a single benzathine penicillin (2.4 MU) IM injection.
StatisticalStatistical Methods
Dataa were double entered and analysed with Epilnfo version 6, dBase V, FoxPro version
2.6,, and Stata version 6.0. Differences between groups were compared using a t-test if
94 4 HIV-22 IN WEST AFRICA dataa were approximately normally distributed, otherwise with the ranksum test. Comparisonn of two proportions was made with the chi-squared test. The observation timee per subject was calculated as die time between the first and the last blood sample. HIVV infections were assumed to have occurred midway between the dates of the last seronegativee and the first seropositive sample. Incidence rates were calculated as the numberr of HIV-1 infections per 1000 person-years of observation (pyo), and 95% confidencee intervals (CI) for rates were calculated. Poisson regression was used to obtain incidencee rate ratios and 95% confidence intervals, adjusting for age and other confoundingg variables where appropriate. In the stratified analysis (Table 3) different age groupss were chosen for the three risk groups, because die age distribution in the three groupss was very different
Thee study was approved by the Ethics Committee of the MRC, The Gambia, and the Nationall AIDS Control Programme in Guinea-Bissau. All participants gave informed consent,, and HIV test results were made available upon request by a village-based counsellor. .
Results s
Subjects Subjects Inn the first survey 3064 out of 4324 eligible subjects (71%) provided a blood sample. Onee hundred and nineteen subjects younger than 15 years , and 652 subjects aged 60 yearss or older were excluded from the analysis. Of the remaining 2293 individuals, 174 (7.6%)) were HIV-2 seropositive, 5 (0.2%) HIV-1 seropositive, and 9 (0.4%) were dually infectedd (HIV-D). These last two categories were also excluded from the analysis, as theyy were not at risk for HIV-1 infection. Three subjects could not with certainty be re- identifiedd later and were excluded as well.
AA second sample was available from 1360 of the 2276 remaining subjects (59.8%). The reasonss that no second sample was obtained were as follows: 102 died (4.5%), 209 had
95 5 HIV-22 IN WEST AFRICA movedd away permanently (9.2%), 297 were residing long-term elsewhere, though expectedd back (13.0%), 94 refused to provide a sample in the second survey (4.1%), and
2144 were repeatedly not at home when visited (9.4%).
Tablee 1: Comparison of subjects with one vs. those with two samples
Twoo samples One sample
N N 1360 0 916* *
Mediann age 30.3 3 26.5 5 <0.0001 1 (21.8-44.3) ) (20.5-37.0) ) (25thh and 75,h percentiles) 9311 (68.5%) 4977 (54.3%) <0.001 1 Sex,, number of women (%) 633 (6.8%) 18(4.1%)** * 0.06 6 CSWss (% of women) 9311 (68.5%) 6222 (69.9%) 0.47 7 Area:: central (%) 110(8.1%) ) 611 (6.7%) 0.21 1 HIV-22 positive
CSWW = commercial sex worker. * Reasons for loss to follow-up were: died before second surveyy (102); moved away permanently before second survey (209); long-term resident abroadd (297); refused to provide blood sample in 2nd round (94); repeatedly not at home when visitedd (214). ** Data missing for 62 women with one sample.
BaselineBaseline Characteristics
AA comparison of baseline characteristics shows that the subjects with a second sample weree more likely to be female (68.5°'o vs. 54.3%; p<0.001), and were older on average
(mediann 30.3 years vs. 26.5 years; p<0.0()01) (Table 1). There was also some evidence thatt the women who provided two samples were more likely to be sex workers (6.8% vs.
4.1%;; p=0.06). There was no difference in the proportion HIV-2 infected (8.1% vs.
6.7%;; p=0.21), or the proportion living in the central area of the village (68.5% vs.
69.9%;; p=0.47).
Amongg the 1360 persons who provided two samples 1250 were seronegative at baseline andd 110 infected with HIV-2. HIV-2 infection was significandy associated with older
96 6 HIV-22 IN WEST AFRICA age,, being a sex worker, TPHA seropositivity, an older age at time of first sexual intercourse,, and living in the central part of the study area (Table 2).
Tablee 2: Characteristics of 1360 subjects with 2 biood samples
HIVV HIV-2 seronegativee seropositive p-value
1250 0 110 0
Mediann age 28.5 5 42.4 4 <0.000' ' (21.2-43.5) ) (35.0-50.7) ) (25thh and 75th centiles) 8577 (68.6%) 7474 (67.3%) 0.78 8 Numberr of women (%) 288 (3.3%) 355 (47.3%) <0.001 1 CSWss (% of all women) 8466 (67.7%) 855 (77.3%) 0.04 4 Area:: central (%) 188(15.2%) ) 34(31.8%) ) <0.001 1 TPHA+vee (%)# Mediann age of 1st sex 18(16-20) ) 19(18-20) ) 0.01 1 (25thh and 75th centiles) *
Historyy of STD*: none 8388 (67.7%) 522 (59.1%) 0.08 8 onee only 267(21.6%) ) 200 (22.7%) twoo or more 133(10.7%) ) 16(18.2%) )
Mediann number of sex partners in last 12 mo 1(1-1) ) 11 (1-2) 0.06 6 (25thh and 75th centiles)**
Freqq (%) of sex in preceding month: 0.65 5 nil l 6777 (54.9%) 444 (50.0%) 1-44 times 3944 (32.0%) 322 (36.4%) >> 5 times 162(13.1%) ) 12(13.6%) )
Everr had sex with CSW (%)[men only] **** 138(35.4%) ) 15(48.4%) ) 0.15 5
Numberr of HIV-1 infections 24 4 17 7 person-yearss of observation 8594.2 2 646.6 6
Mediann observation time in years (range) 7.2 2 6.2 2 <0.0001 1 (2.7-8.6) ) (2.6-8.1) ) Incidencee Rate (CI) of HIV-1 per 1000 pyo 2.8 8 26.3 3 <0.001 1 (1.9-4.2) ) (16.3-42.3) )
Abbreviations:: CSW = commercial sex worker; pyo = person-years of observation; TPHA = TreponemaTreponema pallidum haemagglutination assay; CI = 95% Confidence Interval; STD = sexually transmittedd disease. # [data lacking for 11, respectively 3]; * [data lacking for 15, respectively 22];22]; ** [data lacking for 36, respectively 23]; *** [data lacking for 17, respectively 22]; **** [data lackinglacking for 3, respectively 5].
97 7 HIV-22 IN WEST AFRICA
Tablee 3: HIV-1 Incidence Rates and HIV-1 Incidence Rate Ratios, in three risk groups (unadjusted) )
Tablee 3.a. commercial sex workers
N N HIV-1 1 Person n Incidencee rate Incidence e inci-- yearss of perr 1000 pyo Ratee ratio P P dent t obser-- (CI) ) (CI) ) value e cases s vation n Total l 63 3 15 5 351.7 7 42.66 (25.7-70.7) HIV-22 status Negative e 28 8 3 3 173.2 2 17.3(5.6-53.7) ) 1.00 0 Positive e 35 5 12 2 178.6 6 67.2(38.2-118.3) ) 3.88(1.1-13.7) ) 0.02 2 Agee group <400 years 22 2 3 3 125.5 5 23.9(7.7-74.1) ) 1.00 0 40-499 years 28 8 7 7 159.5 5 43.9(20.9-92.1) ) 1.84(0.5-7.1) ) 0.11 1 50-599 years 13 3 5 5 66.8 8 74.9(31.2-179.8) ) 3.13(0.7-13.1) ) (trend) )
Area a Central l 50 0 12 2 285.6 6 42.00 (23.9-74.0) 1.00 0 Periphery y 13 3 3 3 66.1 1 45.4(14.6-140.7) ) 1.08(0.3-3.8) ) 0.9 9 TPHAA status Negative e 39 9 8 8 231.1 1 34.6(17.3-69.2) ) 1.00 0 Positive e 24 4 7 7 120.6 6 58.0(27.7-121.8) ) 1.68(0.6-4.6) ) 0.3 3 Historyy of STD's none e 32 2 8 8 180.6 6 44.33 (22.2-88.6) 1.00 0 one e 14 4 1 1 88.0 0 11.4(1.6-80.7) ) 0.26(0.03-2.1) ) 0.3 3 22 or more 7 7 1 1 37.7 7 26.5(3.8-188.4) ) 0.600 (0.07-^.8) (trend) ) Agee at first sex >> 19 years 33 3 6 6 197.2 2 30.4(13.7-67.7) ) 1.00 0 << 18 years 19 9 4 4 102.0 0 39.2(14.7-104.5) ) 1.29(0.4-4.6) ) 0.7 7 Noo of partners inlas t t 122 months nonee or one 41 1 8 8 239.3 3 33.4(16.7-66.9) ) 1.00 0 twoo or more 11 1 2 2 60.6 6 33.0(8.2-131.9) ) 1.00(0.2^.6) ) 1.0 0 Frequencyy of sex in lastt month nil l 29 9 7 7 161.2 2 43.4(20.7-91.1) ) 1.00 0 1-44 times 13 3 1 1 82.2 2 12.2(1.7-86.4) ) 0.288 (0.03-2.3) 0.5 5 >> 5 times 11 1 2 2 62.9 9 31.8(8.0-127.2) ) 0.73(0.15-3.5) ) (trend) )
Incidencee rate ratio for HIV-1 incidence with HIV-2 status, adjusted for age as a continuous variablee = 3.05 (CI: 0.84 - 11.1). Abbreviations: pyo = person-years of observation; CI = 95%% Confidence Interval; TPHA = Treponema pallidum haemagglutination assay; STD = sexuallyy transmitted disease; CSW = female commercial sex worker; IRR = incidence rate ratio o
HIV-1HIV-1 incidence Amongg the HIV-2 infected, 17 HIV-1 infections occurred, and among the seronegative persons,, 77 new HIV infections occurred: 24 HIV-1, 33 HIV-2, 17 HIV-D, and 3 HIV-
98 8 HIV-22 IN WEST AFRICA
Tablee 3.b. women who are not sex workers
N N HIV-1 1 Person n Incidencee rate Incidence e inci-- yearss of perr 1000 pyo Ratee ratio PP value dent t obser-- (CI) ) (CI) ) cases s vation n Total l 868 8 16 6 6001.1 1 2.7(1.6-4.4) ) -- HIV-22 status Negative e 829 9 13 3 5753.9 9 2.3(1.3-3.9) ) 1.00 0 Positive e 39 9 3 3 247.2 2 12.11 (3.9-37.6) 5.37(1.5-18.8) ) 0.003 3 Agee group <255 years 293 3 4 4 1990.1 1 2.00 (0.8-5.4) 1.00 0 25-344 years 206 6 3 3 1462.1 1 2.11 (0.6-6.4) 1.02(0.2^.6) ) 35-444 years 158 8 6 6 1094.8 8 5.55 (2.5-12.2) 2.733 (0.8-9.7) 0.6 6 45-599 years 211 1 3 3 1454.1 1 2.11 (0.7-6.4) 1.03(0.2-4.6) ) (trend) ) Area a Central l 585 5 12 2 4034.3 3 3.0(1.7-5.2) ) 1.00 0 Periphery y 283 3 4 4 1966.8 8 2.00 (0.8-5.4) 0.68(0.2-2.1) ) 0.5 5 TPHAA status Negative e 728 8 14 4 5038.0 0 2.8(1.6-4.7) ) 1.00 0 Positive e 131 1 1 1 909.5 5 1.11 (0.2-7.8) 0.400 (0.05-3.0) 0.4 4 Historyy of STD's none e 624 4 10 0 4365.5 5 2.3(1.2-4.3) ) 1.00 0 one e 188 8 2 2 1281.6 6 1.6(0.4-6.2) ) 0.7(0.1-3.1) ) 0.11 1 22 or more 40 0 3 3 261.4 4 11.5(3.7-35.6) ) 5.0(1.4-18.2) ) (trend) ) Agee at first sex >> 19 years 402 2 6 6 2807.2 2 2.11 (1.0-4.8) 1.00 0 << 18 years 434 4 9 9 2983.5 5 3.0(1.6-5.8) ) 1.411 (0.5-^.0) 0.5 5 Noo of partners in last 122 months none e 273 3 4 4 1923.4 4 2.11 (0.8-5.5) 1.00 0 onee or more 579 9 11 1 3985.2 2 2.8(1.5-5.0) ) 1.33(0.4-4.2) ) 0.6 6 Frequencyy of sex in lastlast month nil l 532 2 8 8 3691.9 9 2.2(1.1-4.3) ) 1.00 0 1-44 times 237 7 4 4 1636.9 9 2.4(0.9-6.5) ) 1.13(0.3-3.7) ) 0.2 2 >> 5 times 80 0 3 3 556.6 6 5.4(1.7-16.7) ) 2.499 (0.7-9.4) (trend) )
Incidencee rate ratio for HIV-1 incidence with HIV-2 status, adjusted for age category = 4.70 (CI:: 1.3-17.0). Abbreviations: pyo = person-years of observation; CI = 95% Confidence Interval;; TPHA = Treponema pallidum haemagglutination assay; STD = sexually transmitted disease;; IRR = incidence rate ratio
indeterminate.. In none of the dual infections could it be established which infection camee first, and these have been excluded from the main analysis. The HIV-1 incidence ratee (IR) in the HIV-2 infected (26.3 per 1000 pyo, CI 16.4 - 42.3) was significandy higherr than in the HIV uninfected gtoup (2.8; CI: 1.9 — 4.2), crude incidence rate ratio =
9.422 (CI: 5.1 - 17.5) (Table 2). The overall IR for both groups was 4.4 per 1000 pyo (CI:
3.33 - 6.0).
99 9 HIV-22 IN WEST AFRICA
Tablee 3.c. Men
N N HIV-1 1 Person n Incidencee rate Incidence e inci-- yearss of perr 1000 pyo Ratee ratio PP value dent t obser-- (CI) ) (CI) ) cases s vation n Total l 429 9 10 0 2887.9 9 3.5(1.9-6.4) ) HIV-22 status Negative e 393 3 8 8 2667.1 1 3.0(1.5-6.0) ) 1.00 0 Positive e 36 6 2 2 220.8 8 9.11 (2.3-36.2) 3.022 (0.6-14.2) 0.14 4 Agee group << 29 years 241 1 2 2 1608.6 6 1.2(0.3-5.0) ) 1.00 0 30-399 years 70 0 3 3 475.3 3 6.3(2.0-19.6) ) 5.088 (0.8-30.4) 0.04 4 40-599 years 118 8 5 5 804.0 0 6.22 (2.6-14.9) 5.00(1.0-25.8) ) (trend) ) Area a Central l 296 6 6 6 2007.9 9 3.000 (1.3-6.6) 1.00 0 Periphery y 133 3 4 4 880.0 0 4.5(1.7-12.1) ) 1.52(0.4-5.4) ) 0.5 5 TPHAA status Negative e 357 7 9 9 2410.3 3 3.7(1.9-7.2) ) 1.00 0 Positive e 67 7 1 1 448.7 7 2.22 (0.3-15.8) 0.600 (0.08-4.7) 0.6 6 Historyy of STDs' s none e 234 4 4 4 1580.6 6 2.5(1.0-6.7) ) 1.00 0 one e 85 5 2 2 568.5 5 3.5(0.9-14.1) ) 1.39(0.3-7.6) ) 0.5 5 more e 102 2 3 3 692.8 8 4.3(1.4-13.4) ) 1.711 (0.4-7.6) (trend) ) Agee at first sex >> 19 years 163 3 4 4 1117.2 2 3.6(1.3-9.5) ) 1.00 0 << 18 years 252 2 5 5 1683.8 8 3.0(1.2-7.1) ) 0.83(0.2-3.1) ) 0.8 8 Noo of partnersnn last 122 months nonee or one 190 0 6 6 1280.0 0 4.7(2.1-10.4) ) 1.00 0 twoo or more 229 9 3 3 1546.8 8 1.9(0.6-6.0) ) 0.4(0.1-1.7) ) 0.2 2 Frequencyy of sex in lastt month nil l 160 0 3 3 1072.6 6 2.88 (0.9-8.7) 1.00 0 1-44 times 176 6 2 2 1187.6 6 1.7(0.4-6.7) ) 0.60(0.1-3.6) ) 0.2 2 >> 5 times 83 3 4 4 566.2 2 7.11 (2.7-18.8) 2.53(0.6-11.3) ) (trend) ) Everr had sex with CSW W no o 268 8 5 5 1804.7 7 2.8(1.2-6.7) ) 1.00 0 yes s 153 3 4 4 1037.2 2 3.9(1.4-10.3) ) 1.39(0.4-5.2) ) 0.6 6
Incidencee rate ratio for HIV-1 incidence with HIV-2 status, adjusted for age = 1.91 (CI: 0.4- 9.3);; IRR adjusted for age category and history of STD's: 1.01 (CI: 0.12-8.4). Abbreviations:: pyo = person-years of observation; CI = 95% Confidence Interval; TPHA = TreponemaTreponema pallidum haemagglutination assay; STD = sexually transmitted disease; CSW = femalee commercial sex worker; IRR = incidence rate ratio
Thee incidence rate in CSWs (42.6, CI 25.7 - 70.7) was very high compared to other womenn (IR = 2.7, CI 1.6 - 4.4), or men (IR = 3.5, CI 1.9 - 6.4). Therefore subsequent univariatee analyses were stratified by 'risk group': CSWs, all other women, and men, and
100 0 HII V-2 IN WEST AFRICA
factorss associated with HIV-1 infection are shown in Tables 3.a — c. Despite the small numberr of seroconversions in each group (less than 20), HIV-2 infected subjects had a significantlyy higher HIV-1 incidence in CSWs (IRR = 3.88, CI 1.1-13.7) and other womenn (IRR = 5.37, CI 1.5-18.8), and a non-significandy increased risk in men (IRR = 3.02,, CI 0.6-14.2). Age was associated with increased risk of seroconversion among CSWss and men. CSWs aged over 50 were at three times die risk of those aged under 40 (IRR=3.13,, CI 0.7-13.1; Table 3a), and men aged 30 or over were at five times the risk of thosee aged less than 30 (IRR=5.03, CI 1.1-23.7; Table 3c). The incidence rate did not varyy significandy with area, TPHA status, history of STD's, age at first sex, number of sexx partners in last 12 months, or frequency of sex in last month, for any risk group, althoughh power to detect difference was low due to small numbers. A Poisson model incorporatingg age was fitted for each risk group, resulting in an IRR (comparing HIV-1 infectionn in HIV-2 seropositive to seronegative subjects) of 3.05 (CI 0.84 - 11.10) for CSWs,, IRR = 4.70 (CI 1.3 - 17.0) for all other women, and an IRR = 1.91 (CI 0.4 - 9.3) forr men. An overall estimate of the IRR in a Poisson model adjusting for age group and riskk group was 3.24 (CI 1.5 - 7.1).
Inn order to examine whether the midpoint assumption of HIV-1 infection had unduly influencedd the results, we repeated the above analysis assuming HIV-1 infection had occurredd at the end of the observation period. This did not substantially change the IRR inn either of the three risk groups (data not shown). The overall IRR estimate, adjusted forr age group and risk group, was 3.07 (CI 1.4 — 6.8).
Discussion n
Wee found a three-fold increased incidence of HIV-1 among HIV-2 infected subjects, whichh was statistically significant after adjusting for confounding variables. The effect of increasedd incidence was not only found overall, but also in the three subsets of the study population:: sex workers, all other women, and men. Our study, which was strengthened byy the diagnosis of dual infection by PCR, had die largest power of all five studies done
101 1 HIV-22 IN WEST AFRICA soo far, for it contained many more incident HIV-1 infections among HIV-2 infected subjects. .
Fortyy percent of enrolled subjects did not provide a blood sample in the second survey. Thiss was largely due to the long follow-up period: many people had died (4.5%), and manyy people had moved temporarily (13.0%) or permanendy (9.2%). So only 13.5% weree still resident in the study area and not included in the second survey; they either refusedd (4.1%) or were repeatedly not found at home during the survey (9.4%). The loss too follow-up was similar among HIV-2 positive and HIV-negative subjects (36% vs. 41%,, p=0.2), so does not seem to be related to the exposure. Hence, the increased rate ratioo is unlikely to be due to bias in follow-up. This would have occurred only if the HIV-negativee subjects who were lost, would have had a substantially higher HIV-1 incidencee (five-fold or more) compared to those who remained in the village. A smaller differencee in incidence would have lead to an over-estimate of the increased risk with HIV-22 infection; this possibility cannot be fully excluded.
Historyy of sex work was the single most important risk factor for incident HIV-1 infection,, and controlling for this was therefore essential. Some misclassification of CSW statuss is possible, but the findings that the incidence rate of HIV-1 among CSWs was moree than ten-fold higher than the rate among other women or men, and that the rates amongg other women and men were similar, suggests that misclassification was limited. It iss possible that those with HIV-2 infection engaged more in high risk sexual activities, andd this could cause an increased incidence of HIV-1 among them. However, analysis of incidencee according to TPHA status and sexual behaviour, did not confirm this suspicion. .
Manyy sex workers now living in the village would have given up their trade and were thereforee no longer at increased risk. It is therefore likely that the group of CSW includess a subgroup of still practising women with an even higher HIV-1 incidence rate. However,, the fact that most women who have never been a CSW showed a similar
102 2 HIV-22 IN WEST AFRICA incidencee rate ratio, suggest that better information on CSW status would not alter the conclusionn that previously HIV-2 infected individuals had higher HIV-1 incidence.
Thee median interval between the two tests was long (7.2 years), and therefore the date of seroconversion,, which was taken as the midpoint between the last seronegative and the firstt seropositive sample, was not precise. However, for this analysis the exact date of HIV-11 infection is not essential, and our sensitivity analysis using an end point assumption,, which reduced die rate ratio slightly, from 3.24 to 3.07, assured us that the midpointt assumption has not substantially influenced die results.
Seventeenn seronegative subjects became infected with both HIV-1 and HIV-2 in die studyy period. We excluded these from the analysis, as we did not know the order in whichh tiiey were infected. Three subjects became HIV-infected, but we were unable to determinee the type of infection (HIV-1, HIV-2, or bodi); diese were also excluded from thee main analysis. In order to examine whether these exclusions may have influenced ourr results substantially, we did a two-way sensitivity analysis. In the first extreme scenarioo we assumed diat all 20 subjects became infected first widi HIV-2, and dien with HIV-1.. This increased the incidence rate among HIV-2 infected subjects, resulting in an overalll rate ratio, adjusted for age group and risk group, of 15.65 (CI: 8.5 - 28.9). In the secondd extreme scenario we assumed that all had become infected with HIV-1 first, and thenn with HIV-2. This decreased the overall rate ratio, adjusted for age group and risk group,, to 2.03 (CI 1.0—4.1). Additional adjustment for history of STD's led to a rate ratio off 1.93 (0.9-4.2). As the total incidence among seronegatives of HIV-2 in the period (n=33)) is larger than the incidence of HIV-1 (n=24), and the prevalence of HIV-2 was muchh higher when the study started [15], it is likely diat the majority of dually infected individualss were first infected with HIV-2, and subsequendy witii HIV-1. This would meann diat the adjusted IRR of 3.24 is likely to be a substantial underestimate.
Whatt could be the possible explanation for die finding of an increased risk of HIV-1 infectionn in HIV-2 infected subjects, which is in stark contrast to the Senegalese sex workerr study? Both in Dakar [22] and in Guinea-Bissau [23, 24] the HIV-2 subtype is A,
103 3 HIV-22 IN WEST AFRICA soo differences in circulating subtype of HIV-2 cannot explain this. The Dakar study was carriedd out in a group of sex workers only, but so was the earlier Gambian study [9], whichh did not find a protective effect. Our study, like the other four studies that were unablee to replicate the original finding, relied on an analysis including all subjects enrolled inn the cohort, in contrast to the Senegalese study, which selected a group from within a largerr study.
Thee incidence rate ratio of 3.24 could be interpreted in several ways. Residual confoundingg could have partly caused this, as unsafe sexual behaviour is a strong risk factorr for HIV-1 infection, and the analysis may not have controlled completely for that. Otherr possible explanations are that some subjects are genetically susceptible to HIV infection,, be it HIV-1 or HIV-2, or that HIV-2 infection increases the biological susceptibilityy to HIV-1, e.g. through enhancing antibodies. Since antibodies against env off HIV-2 cross-react with the HIV-1 envelope [18], this may lead to an increased susceptibilityy to HIV-1 infection; this could have important repercussions for HIV vaccinee research.
Forty-onee percent of incident HIV-1 infections occurred in HIV-2 infected persons, so fromm a public health perspective HIV-2 is a strong risk indicator for HIV-1 infection.
Inn conclusion, this large study did not confirm a protective effect of HIV-2 against subsequentt HIV-1 infection. Instead, subjects who have HIV-2 infection are at a substantiallyy higher risk of acquiring HIV-1 infection, whether due to behavioural or biologicall reasons.
Acknowledgements s
Wee are grateful to the population of Caio for participating in the studies over the years. Wee thank Dr Andrew Wilkins, Dr Dominic Ricard, and Margaret Buckner for the initial workk on which this study is built; Alhajie Bayang, Tom Blanchard, Assan Jaye, Pa Tamba
104 4 HIV-22 IN WEST AFRICA
N'gom,, Abraham Alabi, Ramu Sarge-Njie, Mamady Njie, and Bakary Sanneh for tfieir lab work;; the team of Caio field assistants for the field work; Bryan Savage, Glyn Taylor,
Chadd Anderson, Stephanie Constantine, and Margaret Buckner for supervising the field work;; Dr. Mamae Barbosa (Ministério de Saüde Püblica, Bissau) for her support; Prof.
Richardd Hayes, Prof. Roel Coutinho and Dr Birgit van Benthem for advice; and Peter
Langfieldd and M'bemba Ceesay for transport and transit.
References s
1.. Guyader M, Emerman M, Sonigo P, Clavel F, Montagnier L, Alizon M. Genome organizationn and transactivation of the human immunodeficiency virus type 2. NatureNature 1987; 326: 662-669. 2.. Clavel F, Guetard D, Brun-Vezinet F, et al. Isolation of a new human retrovirus fromm West African patients with AIDS. Science 1986; 233: 343-346. 3.. Brun-Vezinet F, Rey MA, Katlama C, et al. Lymphadenopathy-associated virus typee 2 in AIDS and AIDS-related complex. Clinical and virological features in four patients.. Lancet 1987; i: 128-132. 4.. Marlink R, Kanki P, Thior I, et al. Reduced rate of disease development after HIV- 22 infection as compared to HIV-1. Science 1994; 265: 1587-1590. 5.. Poulsen AG, Aaby P, Larsen O, et al. 9-year HIV-2-associated mortality in an urbann community in Bissau, west Africa. Lancet 1997; 349: 911-914. 6.. Travers K, Mboup S, Marlink R, et al. Natural protection against HIV-1 infection providedd by HIV-2 [published erratum appears in Science 1995 Jun 30; 268(5219):: 1833]. Science 1995; 268: 1612-1615. 7.. Greenberg AE, Wiktor SZ, DeCock KM, Smith P, Jaffe HW, Dondero TJ, Jr. HIV- 22 and natural protection against HIV-1 infection [letter; comment]. Science 1996; 272:: 1959-1960. 8.. Kanki PJ, Eisen G, Travers KU et al.: HIV-2 and natural protection against HIV-1 infectionn [letter]. Science 1996, 272:1959-1960. 9.. Ariyoshi K, Schim van der Loeff M, Sabally S, Cham F, Corrah T, Whittle H. Does HIV-22 infection provide cross-protection against HIV-1 infection? [letter]. AIDS 1997;; 11 : 1053-1054. 10.. Aaby P, Poulsen AG, Larsen O, et al. Does HIV-2 protect against HIV-1 infection?? [letter], AIDS 1997; 11: 939-940. 11.. Norrgren H, Andersson S, Biague AJ, et al. Trends and interaction of HIV-1 and HIV-22 in Guinea-Bissau, west Africa: no protection of HIV-2 against HIV-1 infection.. AIDS 1999; 13: 701-707. 12.. Wiktor SZ, Nkengasong JN, Ekpini ER, et al. Lack of protection against HIV-1 infectionn among women with HIV-2 infection. AIDS 1999; 13: 695-699. 13.. Wilkins A, Ricard D, Todd J, Whittle H, Dias F, Paulo Da Silva A. The epidemiologyy of HIV infection in a rural area of Guinea- Bissau. AIDS 1993; 7: 1119-1122. . 14.. Aaby P, Ariyoshi K, Buckner M, et al. Age of wife as a major determinant of male- to-femalee transmission of HIV-2 infection: a community study from rural West Africa.. AIDS 1996; 10: 1585-1590.
105 5 HIV-22 IN WEST AFRICA
15.. Schim van der Loeff MF, Aaby P, Dias F, Pembrey L, da Costa C, Ariyoshi K, and Whittlee H. Trends in HIV-1 and HIV-2 prevalence in a rural area of Guinea- Bissau.. Xlth International Conference on AIDS and STDs in Africa. Lusaka, September19999 (abstract 13PT31-6). 16.. Buckner M. Village women as town prostitutes: cultural factors relevant to prostitutionn and HIV epidemiology in Guinea-Bissau. In: Vivre et penser le sida en Afrique.. Edited by Becker C, Dozon J-P, Obbo, Toure M. Paris: Editions Karthala;; 1999: 391-404 17.. Ricard D, Wilkins A, N'Gum PT, et al. The effects of HIV-2 infection in a rural area off Guinea-Bissau. AIDS 1994; 8: 977-982. 18.. Tedder RS, O'Connor T, Hughes A, N'jie H, Corrah T, Whittle H. Envelope cross- reactivityy in Western blot for HIV-1 and HIV-2 may not indicate dual infection. /_anceM988;ii:: 927-930. 19.. Berry N, Ariyoshi K, Jobe O, et al. HIV type 2 proviral load measured by quantitativee polymerase chain reaction correlates with CD4+ lymphopenia in HIV typee 2- infected individuals. AIDS Research & Human Retroviruses 1994; 10: 1031-1037. . 20.. Ariyoshi K, Jaffar S, Alabi AS, et at. Plasma RNA viral load predicts the rate of CD44 T cell decline and death in HIV-2-infected patients in West Africa. AIDS 2000;; 14: 339-344. 21.. Berry N, Ariyoshi K, Jaffar S, et al. Low peripheral blood viral HIV-2 RNA in individualss with high CD4% differentiates HIV-2 from HIV-1 infections. Journal of HumanHuman Virology 1998; 1:457-68 22.. Sarr AD, Sankale JL, Gueye-Ndiaye A, Essex M, Mboup S, Kanki PJ. Genetic analysiss of HIV type 2 in monotypic and dual HIV infections. AIDS Res Hum RetrovirusesRetroviruses 2000; 16 : 295-298. 23.. Norrgren H, Marquina S, Leitner T, et al. HIV-2 genetic variation and DNA load in asymptomaticc carriers and AIDS cases in Guinea-Bissau. Journal of Acquired ImmuneImmune Deficiency Syndromes & Human Retrovirology 1997; 16: 31-38. 24.. Xiang Z, Ariyoshi K, Wilkins A, Dias F, Whittle H, Breuer J. HIV type 2 pathogenicityy is not related to subtype in rural Guinea Bissau. AIDS Research & HumanHuman Retroviruses 1997; 13: 501-505.
106 6 6 6
Mortalityy of HIV-1, HIV-2 and HIV-1 / HIV-2 duallyy infected patients in a clinic-based cohort inn The Gambia, West Africa
AIDSAIDS 2002; 16 (13): 1775-83
Maartenn F Schim van der Loeff12, Shabbar Jaffar 2, Akum A Aveika ', Saihou Sabally ', l Tumanii Corrah y Elizabeth Harding ', Abraham Alabi , Alhajie Bayang , Koyaa Ariyoshi', and Hilton C Whittle '
(1)(1) the MRC Laboratories, Banjul, The Gambia and (2)(2) the London School of Hygiene and Tropical Medicine, London, UK
Thiss study was financedb y grants from the Medical Research Council (UK), the Japanese Foundationn for AIDS Prevendon, and the Health Sciences Foundation (Japan).
Mortalityy of HIV-1, HIV-2 and HIV-1 / HIV-2 duallyy infected patients in a clinic-based cohort inn The Gambia, West Africa
ABSTRACT T
Objective:: To assess and compare the mortality rate of patients with HIV-1, HIV-22 or both infections (HIV-D) in the same population. Design:: Clinic-based cohort study Methods:: HIV-seropositive subjects aged 15 years and above who attended thee MRC clinics in Fajara between May 1986 and September 1997 were recruited.. Clinical assessment using the Kamofsky score, CDC staging, and WHOO staging, as well as CD4 counts, were done at baseline. Patients were invitedd to attend the clinic every 3 months; if they did not attend, they were visitedd at home by field workers to ascertain survival status. No patient was onn anti retroviral therapy during the study period. Results:: Data from 1519 HIV-positive adult patients were analysed. Seven hundredd and forty-six patients had HIV-1, 666 HIV-2, and 107 patients had HIV-DD infection. Eight hundred and twenty-eight (55%) subjects died, and 161 (11%)) were lost to follow up. The median follow-up time was 12 months (rangee 0 to 128). CD4 counts were available for 894 patients. Compared to HIV-1,, the adjusted Hazards Ratio (HR) for mortality in the CD4 category > 5000 /îL was 0.50 for HIV-2 (95% Confidence Interval (CI): 0.28-0.88), and 1.277 (95% CI 0.51-3.7) for HIV-D. Among those with CD4 <500 /^L the mortalityy rates in HIV-2 and HIV-D infection were similar to that in HIV-1. Discussion:: HIV-2 infected patients with CD4 counts > 500 /|aL have a significantlyy lower mortality rate than HIV-1 infected patients. HIV-2 infected patientss with advanced disease have the same poor prognosis as patients withh HIV-1. Dually infected patients have mortality rates similar to HIV-1. HIV-22 IN WEST AFRICA
Introduction n
Survivall with the Human Immunodeficiency Virus type 1 (HIV-1) varies with age at infection,, with an overall median survival of 11 years in developed countries [1]. Recent dataa indicate that the survival in Africa may be similar [2]. The other human immunodeficiencyy virus, HIV-2, is less well studied, but the available evidence indicates a longerr survival for patients with HIV-2 infection [3,4].
Twoo community-based studies and one occupational cohort study from Guinea-Bissau foundd that the mortality rate of HIV-2 infected subjects was between 2.3 and 6.6 times higherr than that of seronegative subjects [5,6,7]. This compares to a mortality rate ratio off 11.9 in men and 13.9 in women, comparing HIV-1 to seronegative subjects in a community-basedd study in rural Uganda [8].
Inn West Africa, where both viruses circulate, it is not uncommon for one patient to be infectedd with both [9]. The clinical course and the associated mortality rate of dual infectionn with HIV-1 and HIV-2 (HIV-D) has not been described so far, apart from a smalll study in Cote d'lvoire in tuberculosis (TB) patients [10], which reported no differencee in mortality rates between HIV-1 and HIV-D.
Wee compared the mortality rates and survival of patients with HIV-D infection with diosee who had HIV-1 or HIV-2 single infections, in a hospital-based cohort study. This representss a continuation of an earlier study of the same cohort [4].
Patientss and Methods
Patients Patients
Thee first case of HIV in The Gambia was identified in 1986. In 1995, the prevalence of
HIV-11 among pregnant women was 0.5% (95% CI 0.4-0.6%), of HIV-2 1.1% (95% CI
1.0-1.2%),, and of HIV-D 0.05% (95% CI 0.03-0.09%) [11]. At the clinic and hospital of
110 0 HIV-22 IN WEST AFRICA thee Medical Research Council (MRC) Laboratories in Fajara, The Gambia (West Africa), subjectss are tested for HIV antibodies for a variety of reasons. All blood donors and self- identifiedd female commercial sex workers (CSWs) are routinely tested. Most patients withh TB or a suspected sexually transmitted disease (STD), and partners of HIV and STDD cases are tested as well. Finally, patients with a clinical presentation suggestive of HIVV disease are tested. The study population therefore comprises both healthy and symptomaticc patients. The clinic is situated in an urban area near the capital, but attracts patientss from the rural areas as well. It is the national referral centre for HIV patients in Thee Gambia. Patients aged 15 years or older attending the clinic who were HIV-positive uponn screening, and who gave informed consent, were included in the study. All patients weree counselled before and after HIV testing.
Thee date of seroconversion was unknown for most patients. The date of the first positive HIVV test was taken as the date of enrolment. The clinic started recruiting patients in May 1986,, and enrolment for this study closed on 30th September 1997.
Patientss who were tested in outside laboratories and not re-tested at the MRC (due to earlyy death or loss to follow-up) were excluded. None of the patients is known to have beenn on anti-retroviral therapy during the study period. Patients with TB were treated withh multidrug therapy according to the national guidelines. Prophylaxis against opportunisticc infections or TB was not provided during the period of the study.
SerologicalSerological diagnosis ofYilV infections Serumm was screened by the Wellcozyme HIV 1+2 (Murex Diagnostics Ltd, Dartford, UK)) until August 1996, and after that date by the ICEHIV-l.O.2 (Murex). If reactive, sampless were re-tested by type-specific ELISA's. For HIV-1 this was die Wellcozyme HIVV recombinant-1 (Murex), and for HIV-2 the Wellcozyme HIV-2 (Murex Diagnostics) fromm the start till April 1996, and after that the ICE*-HIV-2 test (Murex). Samples that weree clearly reactive in only one type-specific ELISA were assigned a serological diagnosiss accordingly. Samples positive in both ELISA's were further tested by a syntheticc peptide-based strip method, Pepti-Lav 1-2 (Sanofi Diagnostics Pasteur, Marne
111 1 HIV-22 IN WEST AFRICA lala Coquette, France). We interpreted the appearance of a clear band (++) or a very clear bandd (+ + +) as evidence of infection with the relevant HIV type; samples with clear or veryy clear lines for both virus types were considered as dually infected. A second, confirmatoryy serum sample, usually taken 2 to 8 weeks later, was tested in the same way.
Sampless with at least one positive test result, but inconclusive ELISA and Pepti-Lav results,, or with insufficient serum to do all required tests, and patients with two samples havingg incompatible results were classified as indeterminate. A serological diagnosis was assignedd by investigators unaware of the clinical condition of the patient.
PCRPCR confirmation of dual infection
Peripherall blood mononuclear cells (PBMC's) stored at - 70 "C from patients who were serologicallyy dually reactive, were tested by qualitative polymerase chain reaction (PCR), usingg methods that have been described elsewhere [12,13]. If both the HIV-1 specific
PCRR and the HIV-2 specific PCR signals were positive, the patient was considered as havingg dual infection. Cases from whom PBMC's were not available but in whom the serologicall pattern was unequivocally suggesting dual infection, were also considered as
HIV-D.. Finally, some cases positive by PCR for HIV-1 but not for HIV-2, with unequivocall serological evidence of dual infection were considered to be HIV-D, as in
HIV-DD patients with advanced immunodeficiency the PCR signal for HIV-2 may disappearr [14,15]. The HIV status at recruitment was used to classify patients as HIV-1,
HIV-2,, or HIV-D.
ClinicalClinical stage
Att the first visit after a positive test result a research clinician took a history7 and conductedd a full physical examination. All patients were given a score on the Karnofsky performancee scale, ranging from 10 (moribund) to 100 (asymptomatic and well) [16].
Patientss were staged according to the CDC 1993 system [17]. From 1st January 1993 patientss were also categorised according to the WHO clinical staging system, ranging fromm 1 (asymptomatic infection, persistent generalised lymphadenopathy or acute retrovirall syndrome), to 4 (AIDS) [18]. Due to limited options for investigations, most diagnosess and staging were clinical rather than laboratory-con firmed.
112 2 HIV-22 IN WEST AFRICA
CD4CD4 measurement Thee CD4 count was estimated by FACScan (Becton-Dickinson, Oxford, UK). Lymphocytee subset measurements were started routinely in November 1988. These were performedd as soon as possible after patient enrolment. In some patients a CD4 count wass not available for the first visit, but for a subsequent visit. These were used as baselinee CD4 count if the measurement was performed within 3 months of recruitment.
FollowFollow up Patientss were invited to attend the clinic at least every three months, regardless of symptoms.. Those who failed to do so were visited at home by a fieldworker to ascertain thee vital status. A patient was considered lost to follow-up if the study team had no informationn on his or her vital status at the close of the study. The observation time of patientss who were lost to follow-up was censored at the last date they were known with certaintyy to be alive. Observation time of patients who refused further contact with the studyy team was censored at the date of their refusal. The observation period closed at 31stt December 1997, and we aimed to establish the vital status of all originally recruited patientss at that date.
Outcome Outcome Thee cause of death is unknown in the majority of cases, since most patients died at home.. For patients who died in the MRC hospital the date of death was extracted from hospitall records. For patients dying at home a field worker obtained a date of death by interviewingg relatives.
DataData management and statistical analysis Dataa were entered initially in dBASE III PLUS (Ashton-Tate, CA, USA), and later in FoxProo 2.6a for DOS (Microsoft Corporation, WA, USA). Statistical analysis was done usingg The SAS System (SAS Institute Ine, NC, USA), and Stata version 6.0 (Stata Corporation,, College Station, TX, USA).
113 3 HII V-2 IN WEST AFRICA
Continuouss data were compared by Wikoxon test, or by the t-test if approximately normallyy distributed. Proportions were compared with the %2 test or Fisher's exact test, ass appropriate. Differences in survival were examined with Kaplan-Meier survival curves,, log-rank test, and multivariate analysis by Cox regression. Significance was assessedd using the likelihood ratio test.
Thee study was approved by the Gambian Government/MRC Joint Ethics Committee.
Results s
BaselineBaseline characteristics AA total of 1534 adult HIV-infected patients were recruited; from 642 patients (42%) of thesee only one sample was available for serology. There were 746 HIV-1 infected patients,, 666 HI V-2 infected and 107 patients with dual infection (HIV-D); 15 patients withh indeterminate HIV infection were excluded from the analysis. PBMC's were availablee for PCR from 82 of the 107 patients diagnosed as dually infected by serology, andd in 73 of these, dual infection was confirmed by PCR. The other nine, who had advancedd disease with a median CD4 count of 40/fJ.L (interquartile range (IQR) 30 - 250),, tested positive by PCR only for one of the two HIV types.
Thee characteristics of the patients are shown in Table 1. The annual number of recruited HIV-11 cases increased over the yfears, but the number of HIV-2 cases stabilised after 1993.. HIV-1 infected patients were significantiy younger than HIV-2 and HIV-D infectedd patients on first presentation (p < 0.0001 and p = 0.04, respectively). A higher proportionn of HIV-2 infected patients than of HIV-1 patients were female (p<0.001). Onn average, women were 6 (95% CI 5 - 7) years younger than men at first presentation.
Theree were significant differences between the 3 types of infection at presentation accordingg to the distribution of the clinical stage by Karnofsky score, by CDC and by WHOO classifications (p = 0.0007, p = 0.O3, and p < 0.0001 respectively). HIV-1 and
114 4 HIV-22 IN WEST AFRICA
Tablee 1. Clinical and immunological characteristics of patients at enrolment
HIV-1 1 HIV-2 2 HIV-D D
Numberr of patients 746 6 666 6 107 7
Numberr recruited by year (%) Mayy 1986-Dec 1989 366 (5%) 1233 (18%) 14(13%) ) Jann 1990-Dec 1993 2655 (36%) 2988 (45%) 411 (38%) Jann 1994-Sept 1997 4455 (60%) 2455 (37%) 522 (49%)
Age,, median (range) 30(15-68) ) 35(16-70) ) 32(19-72) )
Sex,, number female (%) 3455 (46%) 4022 (60%) 599 (55%)
Femalee commercial sex workers, 58/3455 (17%) 80/4022 (20%) 16/599 (27%) numberr (%)
Karnofsky,, median (range) 70(10-100) ) 80(10-100) ) 70(10-100) )
Karnofskyy < 60 (%) 2400 (32%) 1899 (28%) 344 (32%) Karnofskyy 70-80 (%) 2633 (35%) 1966 (29%) 44(41%) ) Karnofskyy 90-100 (%) 2433 (33%) 2811 (42%) 299 (27%)
CDCC 1 classification A 2399 (32%) 2866 (43%) 311 (29%) B B 3266 (44%) 2400 (36%) 500 (47%) C C 179(24%) ) 139(21%) ) 255 (24%)
WHOO 2 Stage I 179(32%) ) 138(39%) ) 133 (20%) II I 233 (4%) 13(4%) ) 0 0 III I 2300 (42%) 142(41%) ) 377 (57%) IV V 1222 (22%) 577 (16%) 155 (23%)
Mediann CD4 cell count3 pern LL (IQR) 210 210 325 5 230 0 (70,410) ) (130,590) ) (110,510) )
Numberr (%) with <200 2211 (46%) 128(37%) ) 299 (46%) CD44 cell count3 200-499 175(36%) ) 110(31%) ) 18(29%) ) perr \xL >500 855 (18%) 112(32%) ) 16(25%) )
Notes:: 1. Missing for four patients; 2. Collected routinely since 1st January 1993 and availablee from 969 patients; 3. Available for 894 patients within 3 months of recruitment. HIV-DD = coinfected with both HIV-1 and HIV-2. CDC = Centers for Disease Control and Prevention.. IQR = inter-quartile range.
HIV-DD infected patients tended to present at a similar, more advanced stage of disease
thann HIV-2 infected patients. CD4 counts were measured in 1126 (74%) patients, and
weree available within 3 months of their first presentation from 894 (59%) patients. HIV-
115 5 HIV-22 IN WEST AFRICA
22 infected patients presented with a significantly higher median CD4 count than HIV-1 infectedd patients (p < 0.001); there was no significant difference in CD4 count between
HIV-DD and HIV-1 infected patients (p = 0.3, Wilcoxonn test). The median CD4 count at presentationn was 330 (IQR 160-570) for women and 180 (IQR 60-370) for men
(p<0.0001).. The median Karnofsky score at presentation was 80 (IQR 70-90) for womenn and 70 (IQR 50-80) for men (p<0.0001).
Outcome Outcome
Overalll 161 (11%) of patients were lost to follow-up either because thev had moved or emigrated,, or had refused further contact with the clinic (Table 2). There were no
significantt differences in the proportions lost to follow-up between the three types, of infection.. Those lost to follow-up were younger, more often female, more often sex workers,, and had a higher Karnofsky score, better WHO and CDC stages, and a higher
CD44 count (data not shown). These differences were highly significant (p < 0.001 in eachh case).
Thee median follow-up was 12 months (range 0 — 128). Four hundred and twenty three
(57%)) HIV-1 infected, 342 (51%) HIV-2 infected and 63 (59%) HIV-D infected patients diedd (Table 2). The crude mortality rate in HIV-1 was higher than in HIV-2, but similar too that in HIV-D infected patients. The survival of HIV-2 patients was significantly longerr than that of HIV-1 (log rank test, p — 0.006). There was no significant difference inn survival of HIV-D compared to HIV-1 (p = 0.8) or to HIV-2 (p = 0.13).
MortalityMortality rates stratified by age and sex
Whenn subjects were divided into four age groups, the mortality rate increased with increasingg age (Table 3). This effect remained significant after adjusting for sex, CD4 countt category, and HIV type: hazards ratio (HR) for patients 25 - 34 years (95% CI) wass 1.44 (1.01 - 2.07; p = 0.04), for patients 35 - 44 years 1.47 (1.00 - 2.15; p = 0.05), andd for patients aged 45 years and above 1.93 (1.30 - 2.88; p = 0.0012) relative to those agedd 15 - 24 years.
116 6 HIV-22 IN WEST AFRICA
Tablee 2. Follow-up, mortality rates, and survival of HIV-1, HIV-2, and dually infected patients s
HIV-11 HIV-2 HIV-D
Number r 746 6 666 6 107 7 Lostt to follow-up 677 (9.0%) 80(12.0%) ) 14(13.1%) ) Died d 4233 (56.7%) 342(51.4%) ) 633 (58.9%) Mediann number of consultations (range) 3(1,94) ) 3(1,74) ) 3(1,68) ) Mediann follow-up time in months (range) 10.6(0,, 117) 15.7(0,, 128) 8.66 (0, 106) Totall observation time in years 1197 7 1556 6 151 1 Mortalityy rate per 100 person years 35.3 3 22.0 0 41.7 7 (95%% CI) (32.0-38.7) ) (19.7-24.3) ) (31.4-52.0) )
Survivall probability (%) (standard error) 11 year 58.4(1.9) ) 66.11 (1.9) 57.99 (5.2) 22 years 48.00 (2.0) 57.22 (2.0) 40.33 (5.6) 55 years 29.44 (2.3) 40.00 (2.3) 19.8(5.3) ) 7.55 years 15.3(3.7) ) 32.99 (2.7)
Mediann survival time in years (95% CI) 1.82 2 3.15 5 1.53 3 (1.39-2.26) ) (2.38-4.19) ) (0.89-2.02) ) Notes:: HIV-D = co-infected with HIV-1 and HIV-2. 95% CI = 95% Confidence Interval.
Menn had a higher mortality rate than women. This effect was significant in HIV-1 and HIV-22 (p < 0.0001 for both; p = 0.06 in HIV-D). After adjusting for age, CD4 count category,, and HIV type, the HR (95% CI) for men was 1.63 (1.3 - 2.0; p < 0.0001) comparedd to women.
MortalityMortality rates by CD4 category Absolutee CD4 cell counts were available within three months of the first visit for 894 patients.. There were no significant differences between this subgroup and die total patientt group in age, sex and mortality rates (data not shown). The mortality rate was inverselyinversely related to CD4 count for all three infection types. Figure 1 shows the Kaplan- Meierr survival curves for patients in the high (> 500 cells/u.L), intermediate (200 — 499 cells/u.L),, and low (< 200 cells/uL) CD4 categories, comparing the three types of infection.. The median survival of HIV-1 subjects in the highest CD4 category was 4.9 yearss (95% CI 3.7 - 8.7); for HIV-2 and HIV-D the median survival could not be calculatedd due to insufficient data. The median survival for subjects in the lowest CD4 categoryy was 6 montiis for HIV-1, 8 months for HIV-2, and 6 months for HIV-D.
117 7 ^lOO 00 co ^^ i^ s ww oó c\i CM' r^.. co r^ oo CUU 3? "77 a> T- T- oo CM coC O O O^CTÔOÔJ. . ^ff co ïi CM M rara;; >lo Oll oo aWW ö iri O O Sss 5 CMM Tf «3- E E
O O môo^ ^ o o ebb J£ * co ö ö O O
il l Ö Ö
SCLÜ L Ü ~~ 00 o co «?? ^ ö co

co«
^r-nm m coo r-- co iq oo o fjl l E r^^ oS co ui > > CMM co o o c c X X ca a Q. . 3 3 c\ii ai Tt CM O O ^^ CM co ^r irii irj ö Is i_ _ coo m *- oo O) ) COO T— T— T— OO i- CM CM 0) ) CMinCMT-- O) ) COO O CM CO >co. o . O* * I I
> > j^paii CO Tj" I I „-;; cd ir> co' ö ö c c AA CM 00 O) ra ra rara £ > > S2-™ ™
> > X X
COO co ^ e- CMM co S £2 N ra ra coo s ^ t t o o S S co o ^^ ^ * ^^ ^ 3 «MM CO ^ II I l to CMM CO ^- U»» Ift v> "* II I t m T-- CM CO Al IAA IO IO ^ *-- CM CO Al ... o ££ to .§§ s
O) ) 0. . ö ö coo 2 5 5 CX>> CV w w i-- Q. O O f-- 33 ii X X 0)) O 5 5 o o Tï ï J2 2 C C £^ ^ o o 00 0 co o OÏ5)) >— > 1^ ^ lO O —'X X co o 0 0 tttt « CMM *- CM CM' r>- , Q o oC M Mo o co o COO *~ CJ> CM O) ) »» | T- | N I o o 1 1 I I o o 1 1 99 I « I xg g c57°° ' - ' T CM M 5 5 '' Ti - m ~ C~O O « « oo o co o 000 Tf 00 0 co o COO CO o o ö ö ÖÖ oo ü o o o o o o o öö ö-« tt a. Q Q #> > O O Sx x c c o o __ CD £ £ < < .22 0 o o COO -t u u co o £ £ Q. . mm a) 2 2 -- « l__ «J o o oo ® lc_ _c **== <2 co o O O ÜÏÏ Q i55 o 3 3S 5 5 co o 00 0 O)) OJ ^-, O O o o CDD 2 CM M O)) O) p co o KKenen . £ OO^fflôO O (OO O in O T— 1- SOî-iflO O co o oO O CM MC OO g) > > s 00 0 co o n n COO | CO | Ifi | :: I ^ I ^ I CDD | 00 | I -; | 1 1 1 1 1 1c oo O —X£x £x OO (s. O ^ O O O oo co o o C3 co OsOsOio o o o o oC O O II O O "^CM M f-,, to CM ^rr co IO CM M CO O odd d ">> ^ ö O O o o o oC OO Q > > ü> > w öö ö ö X X §ï ï oo o iss 8 ö "O O ^ ^ CC 'A c c < < |NE E CM M COO < CO O)) I (Ö > > CMM C~ —, *~~ "O N X X co"" P co ">"> CM inn I] ™ too ^ CD CMM CM ^" " Q Q CMM to jo in > > ^> > o>> i S! CDD I CM'' I 5 I X X ül l artt w (o '-s o oi nn to *- r^ ** T- " T- CMM r-^ CO rrii CO CMM ^* M M o o 3? ? in n 88 S si J2.. co (0 0 o> > joo 5 II o o o o CÖÖ T3 GC n n O>». . » (^0 0^ e e f-.. ST ~ c^^ 'T CD CM M 1^^ (B N «« « -* ">> 1^ CM = = "E E N N COO c > <°.. e» "* CM T- r: tr^N,.» » U5C»CoO -- . o^îST T
2 -oo — c uu > --
3 3 -- 4 I CO I CM I ^^ t oi I r-. I 0)) — 7 .. o ó 10 CM COO CO ^p c c 0 0 O) ) CM M CO O CMM CO flj flj 55-- ^ 5CO 00* o>*o«o o ST T .. , m "t ^ -ii- S3i rfi 'T ^^ i s l o l > > ££ T- CO^- ^ ^ öö i r^ I I ^^ ' CO^CM^. CO O II I 2 2 X X w r CMM m « « oo ~~ co «. i-- ' CO CM ^ CD * * ooo C- ^rr CM t 55 a £ iü-- co -L- § § (00 <]) o o > > O O S S o o .3. . je e O) ) 0» » Cft t £ £ 3 3 -f™ ™ oo -5- Ol l .3. . o o 10 0 o o o o O O ^^ o3 « « o o co o o o O O ol ol m m coo co en n O O X X p pCM M A A XX VI r-. . Al l ü< < mm ü Zinn o OO v 5_ _ HIV-22 IN WEST AFRICA
Figg 1a. Kaplan-Meier survival graph comparing HIV-1, HIV-2, and HIV-D in patients with CD44 > 500/ nL
i i 11 1 1.00 0
VV 4 *—ii ' 1
II L _ -i i
i i —— L 0.75 5 1 1 1-1 2 2 L L -L L L L 1 1 1 1 i i L L 0.50 0 D D
0.25 5 -- 1 1
0.00 0 -- I I i i i i II I timee in years Numberr at nsK HIV-1 1 85 5 35 5 19 9 3 3 1 1 HIV-2 2 112 2 60 0 37 7 12 2 6 6 Duallyy infected 16 6 7 7 3 3 0 0 0 0
Whenn adjusting for age and sex in a Cox proportional hazards regression there was littlee difference in mortality rates between HIV-1 and HIV-2 in the low CD4 (< 200/u.L)) and the intermediate (200 - 499/u.L) categories (see Table 4). In the high CD44 category (> 500/u.L) the HIV-2 mortality rate was significandy lower than that in HIV-11 (HR = 0.50, 95% CI 0.28 - 0.88; p = 0.02). The mortality rate in HIV-D was nott significantly different from that in HIV-1 in any of the categories, but was significantlyy higher than that of HIV-2 in the high CD4 category (p = 0.04).
MortalityMortality rates by Kamofsky score
Karnofskyy scores were divided in three categories (< 60, 70 - 80, and 90 - 100), accordingg to tertiles. Mortality rates were inversely related to Karnofsky category in all
120 0 HIV-22 IN WEST AFRICA
Figg 1b. Kaplan-Meier survival graph comparing HIV-1, HIV-2, and HIV-D in patients withh CD4 200 - 499/ uL
1.00 0
0.75 5
0.500 -
o o
? ? 0.25 5
0.00 0 TT r 44 6 timee in years NumberNumber at risk HIV-11 175 68 8 27 7 HIV-22 110 48 8 25 5 Duallyy infected 18 4 4 1 1
threee infection types (Table 4). They were lower for HIV-2 than for HIV-1 in all three Karnofskyy categories; the overall hazards ratio adjusted for age, sex, and Karnofsky scoree was 0.75 (95% CI 0.64 - 0.87; p < 0.0001). The HIV-D mortality rate was similarr to that of HIV-1 in all strata, and the overall hazards ratio of HIV-D relative to HIV-11 was 1.09 (95% CI 0.83 - 1.42; p = 0.5).
MortalityMortality rates by CDC clinical stage CDCC clinical stage was associated significandy with mortality rates in all three infection typess (Table 4). The overall hazards ratio, adjusted for sex, age, and CDC stage, of HIV-22 relative to HIV-1 was 0.76 (95% CI 0.65 - 0.88; p = 0.0003). HIV-D mortality ratess were similar to those of HIV-1 in each stratum; the overall hazards ratio adjusted
121 1 HIV-22 IN WEST AFRICA
Figg 1c, Kaplan-Meier survival graph comparing HIV-1, HIV-2, and HIV-D in patients with CD44 < 200/ nL
1.00 0
0.75 5
o o a-a- 0.50 0
0.25 5
0.00 0
timee in years Numberr at risK
HIV-11 221 24 4 HIV-22 128 18 8 Duallyy infected 29 2 2
forfor sex, age, and CDC stage, of HIV-D relative to HIV-1 was 1.19 (95% CI 0.91 - 1.55; pp = 0.2).
MortalityMortality rates by WHO clinical stage WHOO clinical stage at baseline was available for only 930 patients (61%), as this informationn was only recorded since 1993. Stage was significantly associated with mortalityy rates in HIV-1 and HIV-2; for HIV-D not enough data were available (Table 4).. The overall hazards ratio of mortality rate of HIV-2 relative to HIV-1 was 0.72 (95%% CI 0.61 - 0.84; p < 0.0001); and for HIV-D relative to HIV-1: 1.07 (95% CI 0.87 -1.40). .
122 2 HIV-22 IN WEST AFRICA
RestrictingRestricting analysis to PCKproven HIV-D Wee repeated the analyses restricting the group of HIV-D to those with PCR confirmed duall infection status (n — 73); the results were similar to those in the main analysis (dataa not shown).
Discussion n
Thiss is the largest comparative survival study of HIV-1, HIV-2 and HIV-1 /HIV-2 dual infection.. We found that among patients widi CD4 count of 500 cells/u,L or higher, thosee infected with HIV-2 have a significantly lower mortality rate than HIV-1 or HIV- DD infected subjects. In the low CD4 range (< 200/|U.L) there appears to be no differencee in mortality rates between HIV-1, HIV-2, or HIV-D. Mortality rates appear too be similar between HIV-1 and HIV-D in all CD4 categories. Mortality rates were significandyy higher among older patients (after adjusting for sex, CD4 count, and HIV type),, and among males (after adjusting for age, CD4 count, and HIV type). The Karnofskyy performance status was a very good predictor of mortality.
Too overcome the problem of unknown time since infection, we decided to stratify HIV patientss by CD4 count. This is not ideal, as some patients are known to progress to low CD44 counts rapidly and others appear not to progress for > 10 years, both in HIV-1 [19],, and in HIV-2 [20,21]. One cannot assume that the time since infection was similarr for HIV-1 and HIV-2 patients. The resulting bias will lead to an underestimationn of die difference in mortality rates between the two infections [22]. In thee same way, a real difference between HIV-1 and HIV-D could exist, although we weree unable to detect it. In seroprevalent cohort studies subjects can be categorised intoo CD4 strata or clinical stages [23,24]. Although the CD4 count does not indicate timetime since infection, it indicates disease progression, and is a more reliable predictor of AIDSS and death in HIV-1 than time since infection [25]. The high CD4 category
123 3 HIV-22 IN WEST AFRICA comprisess Long-Term Non-Progressors (LTNPs) and recently infected subjects in an unknownn case-mix.
Wee used the HIV serology at enrolment to assign patients to the HIV-1, HIV-2, or
HIV-DD group. In some padents seroconversion from single infection to HIV-D may havee occurred. As HIV-2 has been prevalent for longer than die more recently introducedd HIV-1 [26j we assume most seroconversions will have been from HIV-2 -^
HIV-D.. We decided to ignore these seroconversions. This may have lead to a dilution off the differences between HIV-2 and the other two types of infection, and therefore thee differences that we found may be underestimates.
Routinee CD4 counts were started in November 1988, 2.5 years after the beginning of thee study. Reasons for lacking CD4 counts of patients recruited after that date were rapidd clinical deterioration and death, early loss to follow-up, non-attendance, or the
FACScann being temporarily out-of-service. CD4 counts are lacking from a larger proportionn of HIV-2 than HIV-1 patients (p = 0.00005). Those missing data have reducedd the power of the study and led to a different distribution of patients among thee HIV infection types. However, as CD4 count was the key stratifying factor, it is unlikelyy to have caused bias.
Thee losses to follow-up were limited to 11%. Those lost to follow-up were less ill and hadd a higher CD4 count than those who remained in die study. As the main predictor off death was CD4 count, and the analysis was stratified by CD4 count, the loss to follow-upp has reduced the power of the study marginally in the high CD4 category, but iss unlikely to have introduced much bias.
Theree is broad agreement tJiat in some patients HIV-2 can be pathogenic and can causee immunodeficiency and AIDS. However, the excess mortality risk of HIV-2 infectedd subjects compared to HIV negative subjects appears to be limited [5,6,7].
Becausee the proportion of HIV-2 subjects in the high CD4 category is much larger in thee population at large than in our study population, the contrast between HIV-1 and
124 4 HIV-22 IN WEST AFRICA
HIV-22 is most appropriately characterised by the differences found in the high CD4 category.. The significantly lower mortality rate in HIV-2 in that category (hazards ratio == 0.5, p = 0.02) is compatible with findings of limited excess mortality among HIV-2 infectedd people [5,6,7]. Our finding that people in die advanced stage of infection (CD44 < 200/(iL) have the same high mortality rate, irrespective of HIV type, suggests thatt HIV-1 and HIV-2 run the same course once the immune system is severely affected.. The finding that the prognosis for subjects with CD4 counts > 500/|iL is betterr for HIV-2, coupled with the identification in community-based studies of many oldd and healthy HIV-2 infected subjects with normal CD4 counts [5,27], suggests that a substantiall proportion of HIV-2 infected subjects is not harmed by the infection; otherss may experience a disease course which is indistinguishable from that of HIV-1.
HIV-DD infected patients had a mortality rate similar to HIV-1 in all CD4 categories, althoughh a non-significant trend to worse survival was seen in the category of CD4 > 500// fiL. The power to detect significant differences in survival between HIV-1 and HIV-DD was limited: given die number of deaths, the study had 90% power to detect as statisticallyy significant a hazards ratio (comparing HIV-D to HIV-1) of 1.56 or greater. Thereforee smaller differences cannot be excluded. The poor survival of HIV-D suggestss that preceding HIV-2 does not act as a 'vaccine' mitigating the disease course off subsequent HIV-1 infection.
Thee CASCADE study showed that in HIV-1 age at infection is strongly predictive of mortalityy [1]. In our sero-prevalent study age at enrolment was also significantly associatedd with mortality. The CASCADE study did not find a significant difference in survivall between men and women, nor did several other cohort studies of HIV-1 in Africaa [8,28,29], but in our study men had a significandy higher mortality rate than womenn in both HIV types. Even after adjustment for CD4 count category, HIV type, andd age, the mortality' hazards ratio for men relative to women was 1.63 (95% CI 1.3 — 2.0).. In an occupational cohort study in Tanzania men with HIV-1 had a mortality rate almostt twice that of women [30]. In our study men were at a much more advanced stagee of infection than women, and perhaps adjusting for CD4 count did not capture
125 5 HIV-22 IN WEST AFRICA thatt difference fully. Other potential reasons for this higher mortality rate among males,, such as higher viral loads, should be further examined.
Thee WHO clinical classification does not require sophisticated laborator}' support; it is aa scoring system taking into account a wide range of clinical conditions. It has been usedd with success in research settings [23,24], but it is unclear how widespread its use is inn general hospitals or health centres. The Karnofsky index is a simple clinical assessmentt that is made without the need for a laborator)', and which was originally createdd to assess the prognosis of patients with cancer [16]. It correlated remarkably welll with mortality rates, and may be a simple and useful tool for routine clinic settings andd home care to classify HIV positive patients.
Inn conclusion, this study confirms that HIV-2 infection is associated with a lower mortalityy rate than HIV-1, but shows for the first time that this difference is limited to patientss with CD4 counts above 500 /|xL. Earlier studies demonstrated the lower heterosexuall [31,32] and the lower perinatal transmission [11,33]. Asymptomatic HIV-
22 patients should be counselled that they have a better prognosis, that they not necessarilyy get AIDS, and that the risk of transmitting the infection to an infant is muchh smaller than in HIV-1. Infection with both HIV-1 and HIV-2 carries the same prognosiss as single HIV-1 infection.
Acknowledgements s
Wee thank all the patients in the study for their participation. We thank Tom Blanchard,
Paa Tamba Ngom, Ramu Sarge-Njie, Bakary Sanneh, Mamady Njie, and Andrew N orris forr skillful laboratory work; Ken J oof, Ramu Jaigne, Alieu Jatta, Musukebba Sanyang, andd Babucarr Jawneh for counselling in the clinic and follow-up in the field; Amie
Sarjo,, Awa Kendah and Fatou Njie for data entry; and Roel Coutinho for comments onn drafts and helpful discussions. We thank two anonymous referees for helpful
126 6 HIV-22 IN WEST AFRICA commentss and suggestions. We appreciate the continued support of Saihou Ceesay, directorr of the Gambian National AIDS Secretariat. This study built on earlier work by Drs.. Andrew Wilkins, Jacques Pepin, Arinze Egboga, and Sarah Hawkes.
References s
1.. Collaborative Group on AIDS Incubation and HIV Survival including the CASCADEE EU Concerted Action. Time from HIV-1 seroconversion to AIDS and deathh before widespread use of highly-active antiretroviral therapy. Lancet 2000,355:1131-1137. . 2.. Morgan D, Mahe C, Mayanja B, Okongo JM, Lubega R, Whitworth J. HIV-1 infectionn in rural Africa: is there a difference in median time to AIDS and survivall compared with that in industrialized countries? AIDS 2002, 16:597- 603. . 3.. Matheron S, Mendoza-Sassi G, Simon F, Olivares R, Coulaud JP, Brun- Vezinett F. HIV-1 and HIV-2 AIDS in African patients living in Paris [letter]. AIDS 1997,, 11:934-936. 4.. Whittle H, Morris J, Todd J, et al. HIV-2-infected patients survive longer than HIV-1-infectedd patients. AIDS 1994, 8:1617-1620. 5.. Ricard D, Wilkins A, N'Gum PT, et al. The effects of HIV-2 infection in a rural areaa of Guinea-Bissau. AIDS 1994, 8:977-982. 6.. Poulsen AG, Aaby P, Larsen O, et al. 9-year HIV-2-associated mortality in an urbann community in Bissau, west Africa. Lancet 1997, 349:911-914. 7.. Norrgren H, Da Silva ZJ, Andersson S, et al. Clinical Features, Immunological Changess and Mortality in a Cohort of HIV-2-infected Individuals in Bissau, Guinea-Bissau.. Scan J Infect Dis 1998, 30:323-329. 8.. Nunn AJ, Mulder DW, Kamali A, Ruberantwari A, Kengeya-Kayondo JF, Whitworthh J. Mortality associated with HIV-1 infection over five years in a rural Ugandann population: cohort study. BMJ 1997, 315:767-771. 9.. Nkengasong JN, Kestends L, Ghys PD, et al. Dual infection with Human Immunodeficiencyy Virus Type 1 and Type: impact on HIV Type 1 viral load and immunee activation markjers in HIV-seropositive female sex workers in Abidjan, Ivoryy Coast. AIDS Res Hum Retroviruses 2000, 16:1371-1378. 10.. Kassim S, Ackah A, Abouya L, et al. Prospective cohort study of HIV-1/HIV-2 duall infection among HIV-infected tuberculosis patients in Abidjan, Cote d'lvoire.. 13th World AIDS Conference. Durban, 2000 [Abstract TuPeC3355]. 11.. O'Donovan D, Ariyoshi K, Milligan P, et al. Maternal plasma viral RNA levels determinee marked differences in mother-to-child transmission rates of HIV-1 andd HIV-2 in The Gambia. AIDS 2000,14:441-448. 12.. Berry N, Ariyoshi K, Jobe O, et al. HIV type 2 proviral load measured by quantitativee polymerase chain reaction correlates with CD4+ lymphopenia in HIVV type 2- infected individuals. AIDS Research & Human Retroviruses 1994, 10:1031-1037. . 13.. Berry N, Ariyoshi K, Jaffar S, et al. Low peripheral blood viral HIV-2 RNA in individualss with high CD4 percentage differentiates HIV-2 from HIV-1 infection. Journall of Human Virology 1998, 1:457-468.
127 7 HIV-22 IN WEST AFRICA
14.. Dieng Sarr A, Hamel DJ, Thior I, et al. HIV-1 and HIV-2 dual infection: lack of HIV-22 provirus correlates with low CD4+ lymphocyte counts. AIDS 1998, 12:131-137. . 15.. Ariyoshi K, Cham F, Harding E, et al. Diagnosis of HIV-1/2 dual infection using dilutionall analysis of type specific antibody [letter]. AIDS 1998, 12:2504-2505. 16.. Karnofsky DA, Abelmann WH, Craver LF, Burchenal JH. The use of the nitrogenn mustards in the palliative treatment of carcinoma. Cancer 1948, 2:634- 656. . 17.. Centers for Disease Control and Prevention. Revised Classification System for HIVV Infection and expanded surveillance case definition for AIDS among adolescentss and adults. MMWR 1992, 41/RR-17:1-19. 18.. World Health Organization. Acquired immune deficiency syndrome (AIDS): interimm proposal for a WHO staging system for HIV infection and disease. Wkly Epidemoll Rec 1990, 65:221-228. 19.. Keet IPM, Krijnen P, Koot M, et al. Predictors of rapid progression to AIDS in HIV-11 seroconverters. AIDS 1993, 7:51-57. 20.. van der Ende ME, Schutten M, Ly TD, Gruters RA, Osterhaus AD. HIV-2 infectionn in 12 European residents: virus characteristics and disease progression.. AIDS 1996,10:1649-1655. 21.. Ancelle R, Bletry O, Baglin AC, Brun-Vezinet F, Rey MA, Godeau P. Long incubationn period for HIV-2 infection [letter]. Lancet 1987, 1:688-689. 22.. Brookmeyer R, Gail MH. Biases in prevalent cohorts. Biometrics 1987, 43:739-749. . 23.. Morgan D, Maude GH, Malamba SS, et al. HIV-1 disease progression and AIDS-definingg disorders in rural Uganda. Lancet 1997, 350:245-250. 24.. Kassa E, Rinke De Wit TF, Hailu E, et al. Evaluation of the World Health Organizationn system for HIV infection and disease in Ethiopia: association betweenn clinical stages and laboratory markers. AIDS 1999,13:381-389. 25.. Philips AN, Pezzotti P, Lepri AC, Rezza G, and the Italian Seroconversion Study,, CD4 lymphocyte count as a determinant of the time from HIV seroconversionn to AIDS and death from AIDS: evidence from the Italian Serocconversionn Study. AIDS 1994, 8:1299-1305. 26.. Wilkins A, Oelman B, Pepin J, et al. Trends in HIV-1 and HIV-2 infection in The Gambiaa [letter]. AIDS 1991, 5:1529-1530. 27.. Larsen O, da Silva Z, Sandstrom A, et al. Declining HIV-2 prevalence and incidencee among men in a community study from Guinea-Bissau. AIDS 1998, 12:1707-1714. . 28.. Todd J, Balra R, Grosskurth H, et al. HIV-associated mortality in a rural Tanzaniann population. AIDS 1997, 11:801-807. 29.. Sewankambo NK, Wawer MJ, Gray RH, et al. Demographic impact of HIV infectionn in rural Rakai District, Uganda: results of a population-based cohort study.. AIDS 1994, 8:1707-1713. 30.. Senkoro KP, Boerma JT, Klokke AH, et al. HIV incidence and HIV-associated mortalityy in a cohort of factory workers and their spouses in Tanzania, 1991 throughh 1996. J Acquired Immune Deficiency Syndromes 2000, 23: 194-202. 31.. Kanki PJ, Travers KU, Mboup S, et al. Slower heterosexual spread of HIV-2 thann HIV-1. Lancet 1994, 343:943-946. 32.. Donnelly C, Leisenring W, Kanki P, Awerbuch T, Sandberg S. Comparison of transmissionn rates of HIV-1 and HIV-2 in a cohort of prostitutes in Senegal. Bulletinn of Mathematical Biology 1993, 55:731-743. 33.. Adjorlolo-Johnson G, De Cock KM, Ekpini E, et al. Prospective comparison of mother-to-childd transmission of HIV-1 and HIV-2 in Abidjan, Ivory Coast. JAMA 1994,, 272:462-466.
128 8 7 7
Contrastss in plasma viral load, CD4% and survivall in a community-based cohort of HIV-1 andd HIV-2 infected women in The Gambia
Submitted Submitted
Andreass Hansmann ',2, Maarten Schim van der Loeff u, Steve Kaye ',
Akumm Aveika Awasana ', Ramu Sarge-Njie ', Diarmuid O'Donovan 4, Koya Ariyoshi \
Abrahamm Alabi ', Paul Milligan ', Hüton Whitde '
(1)(1) MRC Laboratories, Fajara, The Gambia; (2) Westfalische Wilbelms-Universitat Munster,
Germany;Germany; (3)London School of Hygiene and Tropical Medicine, London, UK;
(4)(4) Department of Public Health, University College, Galway, Ireland;
(5)(5) National Institute of Infectious Diseases, Tokyo, Japan
Thiss study was funded by the Medical Research Council (UK), and the Japanese Foundation for AIDSS prevention. Andreas Hansmann received a student travel award from the Deutscher Akadcmischerr Austausch Dienst.
Contrastss in plasma viral load, CD4% and survivall in a community-based cohort of HIV-1 andd HIV-2 infected women in The Gambia
ABSTRACT T
Objectives:: To estimate and compare all-cause mortality rates of HIV-1, HIV- 22 and uninfected women; to assess the predictive value of baseline plasma virall load (PVL) and CD4% for mortality. Methods:: On presentation to antenatal clinics in The Gambia in 1993-5, pregnantt women were screened for HIV-1 and HIV-2 antibodies. Seropositive subjectss and a similar number of seronegative controls were enrolled and baselinee PVL and CD4% measured. Participants were visited regularly by field workerss until 18 months post-delivery and again 4-7 years later. Findings:: Thirty-two of 101 women infected with HIV-1, 23/243 with HIV-2 andd 9/468 seronegative women died during a median follow-up period of 6.9 years.. The mortality rate was 55 per 1000 person years-of-observation (pyo) forr HIV-1 infected, 16 for HIV-2 infected, and 3.1 for HIV uninfected women. Inn a multivariate analysis, one log increase of PVL was associated with a 1.8 foldd higher mortality rate in HIV-1 (95% Confidence Interval (CI) 0.9-3.4). In HIV-22 infection women with a high PVL (>10,000 copies/ml) had an 8.7 (95%CII 2.8-28) higher mortality than those with a low PVL (<1,000 copies/ml). AA 10% decrease in CD4% was associated with higher mortality rates in HIV-1 (1.6-fold;; 95%CI 1.1-2.3) and HIV-2 infected subjects (1.5-fold; 95%CI 1.0- 2.3). . interpretation:: Survival of HIV-1 infected women in The Gambia is similar to thatt in industrialised countries prior to the introduction of anti-retroviral treatment.. Survival of HIV-2 infected women is much better. However, those withh high PVL die as quickly as HIV-1 infected women. HIV-22 IN WEST AFRICA
Introduction n
Inn Sub-Saharan Africa two types of HIV are prevalent and cause disease. HIV-1 is epidemic acrosss the continent with 29 million Africans infected,1 but HIV-2 is mainly limited to West Africaa with an estimated 1 million people infected. In general, HIV-2 infection has been associatedd with substantially lower rates of transmission, disease progression, and mortality comparedd to HIV-1.2"5
Inn industrialised countries there are good data on natural history and survival of HIV-1 infection inn men prior to the use of highly active antiretroviral treatment (HAART). Median survival is betweenn 9 and 13 years depending on age of infection.'
Longitudinall data on survival with HIV-1 infection in Africa are rare, and there are very few incidentt cohort studies. In a sero-incident cohort in rural Uganda, Morgan et al. estimated a mediann survival time from seroconversion of 9.8 years. 7 A community-based seroprevalent cohortt study in Malawi recenüy reported a median survival time of 8 years.8
Althoughh the natural history in African adults infected with HIV-2 is less well described, studies too date indicate disease progression is much lower compared to those infected with HIV-1. Two community-basedd studies and one occupational cohort study from Guinea-Bissau, have reported 2.33 to 6.6 times higher mortality rates in HIV-2 infected adults compared to uninfected adults.9 u Thesee rates contrast with an 11-fold higher mortality rate in HIV-1 infected compared to seronegativee adults in a rural community-based study in East Africa l and a 9-fold higher mortalityy in a community-based study in women in Rwanda.
Plasmaa viral load (PVL) and CD4 count are independent predictors of mortality in HIV-1 infection,, 14 and a reduction of PVL through HAART is strongly associated with improved survival.15188 There are no community-based studies of HIV-1 in sub-Saharan Africa that have assessedd the predictive value of PVL and CD4 count. A community-based study of HIV-2 in Guinea-Bissauu showed that PVL and CD4 count are clinically important independent predictors off mortality in HIV-2 infection.19
132 2 HIV-22 IN WEST AFRICA
Thiss study compares the mortality rates in HIV-1 infected, HIV-2 infected and uninfected adult womenn in The Gambia and assesses the predictive value of PVL and CD4% for mortality in bothh infections.
Methods s
Subjects Subjects Thee women in this cohort were recruited between 1993 and 1995 during a study estimating the mother-to-child-transmissionn (MTCT) of HIV-1 and HIV-2 in The Gambia (O'Donovan 2000). Afterr delivery, mothers and children were visited at 2, 6, 9, 12, 15, and 18 months by a field workerr or clinician.4'20
Follow-up Follow-up Betweenn July 9th and November 1st 2001 three experienced and well trained counsellors / field workerss from the genito-urinary clinic (GU) at the MRC in Fajara, the national referral centre for HIVV care, revisited all women in their homes. The counsellors were blinded to the HIV status of thee women. Study participants were asked for verbal informed consent, which was documented. Iff a woman was known to be alive but not seen at her compound, the date she was last seen alive byy a close relative was recorded. If the woman was temporarily absent, she was revisited. Women whoo had moved within The Gambia, were visited at their new address. For women who could nott be traced, or on whom no information could be obtained in the follow-up in 2001, the last datee she was known to be alive was obtained from earlier records. Where available, data from thesee follow-up and clinic visits were included in our analysis. To check the quality of the collectedd data, 5.3% of the women were visited twice in 2001 by two different field workers.
Ethics Ethics Inn the initial screening 1993-5, all women were counselled and only those agreeing to an HIV test,, were enrolled. The test results were available two weeks later at the same clinic from an HIV counsellorr for women who wanted to know their HIV status, in accordance with the National
133 3 HIV-22 IN WEST AFRICA
AIDSS Policy.21 Very few women took advantage of this opportunity. In the follow-up in 2001 thee results of the old HIV test were not provided, since the woman's sero-status might have changedd in the meantime. Instead, a new test was offered to all women. The study team offered freee medical treatment and fares to the next health centre to any women or other person found too be sick in the household. Women who wanted to be re-tested and who were found to be HIV infectedd were offered free care at the MRC GU clinic. HIV-1 infected women who were pregnantt were offered short course nevirapine to prevent mother-to-child transmission of HIV- 1.. In The Gambia anti-retroviral treatment is not yet available. The study was approved by the jointt Gambian Government-MRC Ethics Committee (Project Number 868/825). laboratorylaboratory methods Serology,, CD4 cell counting and plasma viral load assays were performed as described previously.. 4 Briefly, screening for HIV infection was done in pools of ten, using a combined enzymee immunoassay (EIA) for HIV-1 and HIV-2 (Wellcozyme HIV Recombinant, Murex Diagnosticss Ltd, Dartford, UK). Individual sera from a positive pool were re-tested in the same assay.. Those found to be positive were tested with the Wellcozyme HIV-1 and HIV-2 specific EIA'ss (Murex Diagnostics Ltd). Seropositivity was confirmed using a combined HIV-1 and HIV-22 peptide-based EIA (Pepti-LAV 1-2, Sanofi Diagnostic Pasteur SA, Marnes la Coquette, France).. To determine CD4 cell percentages, whole blood fluorescent antibody cell sorting (FACS)) analysis (FACScan, Becton Dickinson, Oxford, UK) was used, adapted for field use.9 Viruss loads were assayed in heparin-plasma samples stored at -80 C. RNA was extracted by the methodd of Boom et al., 22 reverse-transcribed and PCR-amplified using primers targeted to the LTRR region of HIV-1 or HIV-2, and PCR products quantified in a microtitre-format hybridisationn assay.23 The lower limit of detection of these assays was 500 copies per ml.
StatisticalStatistical methods Thee data were double entered. Analysis was done with Stata version 6 (Stata Corporation, Collegee Station, TX, USA). Person years of observation (pyo) were calculated from the time peoplee were enrolled until 1st July 2001 or until the date of death or until the last date known to bee alive, whichever came first. Mortality rates were calculated as deaths per 1000 pyo, with 95%
134 4 HIV-22 IN WEST AFRICA
Confidencee Intervals (CI). Data on 10 HIV-1 and HIV-2 dually infected women were excluded fromm this analysis.
Timee to death was examined by Kaplan-Meier graphs. Mortality rates were compared by strata off CD4% and of PVL, and Mantel-Haenszel mortality rate ratios calculated to adjust for these. Poissonn regression analysis was used for multivariate analysis of mortality rates. Because mortalityy rates increased witfi increasing time since recruitment, the observation period of 8.4 yearss was split in three time periods (0-2.99 years; 3-5.99 years; and 6-8.4 years) and time period waswas one of the factors adjusted for in the Poisson regression.
RoleRole of the funding sources Thee funding sources had no role in the data collection, analysis, or interpretation, nor in the writingg of the report or the decision to publish it.
Results s
BaselineBaseline characteristics Thee baseline characteristics of the 812 women at the time of recruitment (Table 1) show that HIV-11 positive women were younger than HIV-2 infected women (mean age 23.4 years vs. 27.0 yearss (p<10'4)) and younger than HIV-negative women (mean age 25.8 years (p~0.0002)). HIV- 11 infected women had lower median baseline CD4% (33%) compared to HIV-2 infected women (42%,, p<104) and uninfected women (48%, p<104). The median PVL, assayed in 94 HIV-1 infectedd women, was 13,600 copies/ml (range 250-265,000). The median PVL of 228 HIV-2 infectedd women was 500 copies/ml (range 250-63,000), which was 27 fold lower (p<104). Comparisonn of PVL distributions showed 6% and 61% of HIV-1 and HIV-2 infected subjects respectivelyy had loads below 1000 RNA copies/ml and 9% and 0% respectively had loads above 100,0000 copies/ml (Figure 1).
135 5 HIV-22 IN WEST AFRICA
Tablee 1. Baseline characteristics of 812 women at time of recruitment
HIVV status of woman
HIV-negative e HIV-1 1 HIV-2 2 nn = 468 nn = 101 nn = 243
Meann age (SD) 25.88 (5.7) 23.44 (5.5)a 27.00 (5.5)
c d Mediann CD4% (range) 488 (8-68) b 333 (1-57) 422 (5-66)
Numberr {%) of women with: CD44 < 14% 33 (3%) 66 (7%) 10(5%) )
CD44 14-28% 44 (3%) 266 (30%) 18(8%) )
CD44 > 28% 112(94%) ) 566 (64%) 188(87%) )
s f Plasmaa viral load: — — 13,6000 500500 mediann RNA copies / ml (range) (250-265,000) ) (250-63,000) )
Numberr (%) of women with: << 1,000 RNA copies/ml — — 66 (6%) 140(61%) )
1,0000 - 9,999 RNA copies/ml — — 266 (28%) 600 (26%)
10,0000 - 99,999 RNA copies/ml — — 544 (57%) 29(13%) )
>> 100,000 RNA copies/ml — — 88 (9%) 00 (0%)
Notes:: a. age not available for one subject; b. available for 119 subjects; c. available for 88 subjects;; d. available for 216 subjects; e. available for 94 subjects; f. available for 229 subjects. SDD = standard deviation. Seronegative women were selected matched for age to HIV-infected women.. Percentages do not add always up to 100% due to rounding. All differences between the threee groups P < 0.01.
Follow-up Follow-up
Duringg the follow-up exercise between July and November 2001, 655 of the 812 women (81%) weree found to be alive, 64 (8%) had died, 8 (1 %) had refused further follow-up prior to 2001 and
855 (10%) were lost during follow-up (Table 2). The rate of loss during follow-up was 17 (95%CI
14-21)) per 1000 pyo. The median follow-up time was 6.9 years (range: 3 months to 8.4 years).
Noo significant differences were found in HIV status, age, CD4%, PVL, or parity between womenn who refused or who were lost during follow-up and women with known survival status,
136 6 HIV-22 IN WEST AFRICA
Figuree 1: Plasma viral load in relation to HIV status and survival (horizontal lines show the median virall load).
100000 0
§§ 10000
1000 0 s s ft, ft, 500 0
O O -0-- î> > o o HIV-1HIV-1 alive died died A A HIV-2HIV-2 alive died died
butt those lost had had less education, were more often of the Jola ethnic group, were more often fromm the coastal area, and tended to have a poorer water supply (data available upon request).
Forr quality assurance purposes 43 women (5.3%) were visited on two occasions in 2001 by differentt fieldworkers. No discrepancies in recorded survival status occurred. There were four discrepanciess in the date the woman was last known to be alive (median difference 10 months); thiss was caused by interviewing different informants. Despite intensive counselling by the fieldworkers,fieldworkers, only twelve women opted to have a new HIV test.
137 7 3^^ en o ^^ ^ 5 enn r>- oo
r-- CO T- HH có 4 t«« £ cd""" M
5«« 2 ~~ O ^ £$3 3
coo o CMM C\i
cc £ g^ ^ >-- c Q. . OO qj 3 3 O O 33 o> Oï ï Cöö ^> CD D OO <£ O) ) ©© C Cö ö SSo oo .O O 3ÊÊ II oo _ ü ü oo o O O 5§ § IT) ) -oo 5 2-- 21 1 roro £ coo a> lO O o o V) V) Q. . *-^ ^ a a (00 -£3 co o c c cö ö O O m m a> > CD D o o O O CO O 55 en n £ £ o o O O Cöö 0 c c o o ro ro c c o o g g2 2 (0 0 CO O 10 0 00 e CD D ro ro o o 0 0 0} } . . o o ü ü ro ro 1 1£ £ ir>> i o o COO S 2 2 c c "ro o co o c c c c o o "ro o a> > COO -£ £ £ D D f3 3 >s s V VC M MA l l roro o) CM M c c >, a,> > CD D cc Q. 0) ) caa o^ co o 2 2 2 2 "CD D O) ) CÖ ÖC S S .Sï ï a) ) co o CÖ Ö CUU H r-- .?? "O to o > > «0 0 e e ro ro o o o o o o »-- 0 %% .2 o o .Q Q
Mortality Mortality Thirty-twoo out of 101 HIV-1, 23 of 243 HIV-2 and 9 of 468 seronegative women died. The mortalityy rates were 56 per 1000 pyo in HIV-1 (95%CI: 39-79), 16 in HIV-2 (95%CI: 10-23) and 3.11 in sero-negative women (95%CI: 1.6-6.0). Table 2 and Figure 2 compare the mortality rate andd survival probability of the three groups. The mortality rate of women with HIV-1 infection waswas 18-fold higher (95%CI 8.5-37; p<104) than that of HIV uninfected women. The mortality ratee of women with HIV-2 was 5.0 fold higher (95%CI 2.3-11; p<104) than that of HIV uninfectedd women and 3.6 fold lower (95%CI 2.1-6.1; p<10~4) than of HIV-1 infected women. Theree were no significant differences in mortality rate by age group in HIV-1 infected (pÔ.ó), HIV-22 infected (p—0.11) or seronegative women (p=0.5; see Table 2). Thee mortality rate increased over time. This trend was significant overall (p=0.001), in HIV-1 infectedd women (p=0.02), in HIV-2 infected women (p=0.005), but not in HIV-negative women (p=0.5). .
MortalityMortality by PVL and CD4 % Amongg HIV-1 infected study subjects the mortality rate increased with the log of PVL, though nott significantly so (p=0.09) (Table 3); among HIV-2 infected women the mortality rate increasedd significantly with PVL (p=0.002) (Table 3). When comparing the mortality rates betweenn HIV-1 and HIV-2 infected women within the same PVL category, the rates of HIV-2 weree lower, but mis was only significant in one category. A Mantel-Haenszel rate ratio, adjusted forr PVL category showed no significant difference between the two infections (p=0.4; table 3). Amongg HIV-1 infected subjects the mortality rate increased with decreasing CD4 % (p=0.008) (Tablee 3). In HIV-2 infected women the mortality rates also increased with decreasing CD4% (p<0.0005).. Among women with CD4% > 28%, the mortality rate was significantly lower in HIV-22 dian in HIV-1 infected subjects (p<104), but in the two lower CD4% bands there were noo significant differences in mortality rates between the two infections (p=0.4 each).
MultivariateMultivariate analysis Inn the multivariate analysis only PVL and CD4% were independent predictors of mortality. Age waswas a priori in the model and time period was a confounder and kept in the model as well. After adjustingg for the other variables, each 10 percent decrease in CD4% was associated with a 1.6
139 9 't t ö ö ö ö (11 1 N N Cf> 00 0 r*-. . CN N CD D ra ram m >»» o ^ ^ > CD D 1— — T— — CN N "55 5 O O C C o o O O O O CN N c c O, , i i o, , == "* o o ^—' ' "—' ' O O ra ra© © ~~ ... ^ O O a> > > ^-^00 0 ^-^ o o CN N ip p -> > 't t CO O
rara o 322 o [OO ac >.oO O ^^ > CO-- V^^ co ra rra r *?£ £ CO O fc ofc o 1 (O O w w 22 fc— tb b -- r--- ra rrao o n nt t (a0> 0 > tni i ni } } CN N CN N CD Do o 00 0 CO O r r u u o ot/ > > > > co o I I ^ ^ c c a> > »
..
»--o o oo o >i ,i L>,L » L » ra ra-o o oo S" -e ero ro o o. o o E E n n ro ro Cl) ) c c> > (1) ) o o o oro ro F Fb b ra raQ. .^^ ^^ o ow w >,o oO O r-- ™ o o( 0 0 O O CD DO O ï^^ £i co üü Tj- !£> O O E EO . . rara D. II . ——CUU o's-? OO co coo o i-- co ra Ora O -OO £
2^^ ra > >o o .. © T T. p p oo -o EE - S-S-or ro .*ti i o o < < CN No " " > >ro ro oo _5^ T T CU U O) ) ro ro cc c c c T3 3 Cö ö— — CO O en n U. . co o O O CD D F Fe e O) ) O) ) O O o o ro ro o o O) ) co o o o 5* * O Ob b CR R o o 00 0 5? ? co o § § ó ó o" " 3? ? CN N CU U ó ó o o o o co o 1 1 C_ _ i o o o o CN N co o c/>> - o o A A V V EE o o o Al l F F 122 v o oo o co o O O CL L O O co o * * co o ir> > co o
CD D o o CD D m m co o co o
M-- m m ro o
'S S > > X X
CD D CN N co o co o o o E E o o co o o o CN N
CN N
O O 00 0 co o o o co o CM M 03 3 .O O 2 2 O O lO O o o o o o. . in n CM M o o o1-- o o o o o o o o o CD D E E '5 5 i_ _ o o Q. . CU U T3 3 0! ! CD D CD D o o O O CD D CM M (I) ) c c C C '(/) ) > > > > t t 'E E CO O X X ro ro 3 3 o. . >o ,o , :c*c c* oi i o> >
O) ) HIV-22 IN WEST AFRICA foldd increase in mortality rate (95%CI 1.1-2.3, p=0.02) in HIV-1 and a 1.5-fold increase (95%CI 1.0-2.3,, p=0.046) in HIV-2. After adjusting for die odier variables, an increase of one log in PVLL was associated widi a 1.8 fold increase in mortality rate (95%CI 0.9-3.4; p=0.08) in HIV-1. Ass 41% of HIV-2 infected women had a PVL below the detection limit, we could not calculate thee effect of a one log increase in PVL in HIV-2. In a multivariate analysis using categories, the mortalityy rate was 1.3 fold (95%CI 0.36-4.5) higher in women with PVL between 1,000 and 9,9999 copies per ml, and 8.7 fold (95%CI 2.8-28) higher in women with PVL > 10,000 copies per ml,, compared to women widi PVL < 1,000 copies per ml. The mortality rate in HIV-2 infected womenn witii normal CD4% and undetectable PVL was not significandy different from that amongg seronegative women (p—0.6).
Inn an analysis combining HIV-1 and HIV-2, each 10% decrease in CD4% was associated with a 1.77 fold increased risk of mortality (95%CI: 1.3-2.2, p<0.0005). PVL was also associated with mortality:: rate ratios were 2.1 (95%CI 0.71-6.1) for those with PVL between 1,000 and 9,999 copiess per ml and 3.9 (95%CI 1.3-12) for those with PVL > 10,000 copies per ml, when comparedd to die baseline group (< 1,000 copies per ml). When adjusted for CD4%, PVL, age groupp and time period, there was no significant difference in mortality rates between the two infectionss (p=0.7).
Discussion n
MainMain findings Thiss is the first study that has been able to compare the mortality of both infections in persons fromm the same population, and it confirms that the excess mortality of HIV-2 is much less than thatt of HIV-1.7'50'12
Thee study confirms data from other community-based studies that showed that mortality rates in Africann HIV-1 patients are comparable to those of HIV-1 infected patients in industrialised countriess before the introduction of HAART. 7> 8 Eight years after recruitment into this seroprevalentt cohort, more than 50% of HIV-1 women were still alive.
142 2 HIV-22 IN WEST AFRICA
Finally,, this study shows that PVL and CD4% are independent and significant predictors for mortalityy in HIV-2 in West Africa, confirming the results of an earlier community-based study in Guinea-Bissauu 19 and a clinic-based study in The Gambia.24 HIV-2 infected women with normal CD4%% had a significantly lower mortality than HIV-1 infected women with normal CD4%, but amongg women with low CD4% there were no significant differences in mortality rate between thee two HIV types. These findings are in agreement with findings from a clinical cohort in The Gambia.5 5
PossiblePossible biases AA possible source of selection bias is that women were pregnant when they were recruited, and fertilityy is reduced among HIV-infected women. 25~27 As HIV-associated sub fertility affects recentlyy and long infected women to a similar degree, 21 this is unlikely to have lead to underestimatess of mortality rates, or overestimates of survival probabilities.
Upp to 18 months after delivery, the women were visited quarterly and the loss to follow-up was limited.. The next visit was 4-7 years later, and this long time gap and the associated loss to follow-upp could be a source of bias. However, there were no significant differences in the key baselinee variables age, CD4%, PVL, and parity between those lost to follow-up and those who weree not.
Wee had anticipated that many women would wish to obtain their HIV status after being counselledd a second time, but this did not transpire. Potential reasons for the poor uptake could bee the stigma attached to the disease, lack of effective anti-retroviral treatment on offer, and counsellingg by male field workers. Voluntary counselling and testing is considered to be a valuablevaluable tool in the fight against HIV and AIDS, 2S but this low acceptance ratio of HIV tests suggestss that the uptake may be very low in societies where HIV disease is stigmatised and effectivee antiretroviral therapy unavailable.
143 3 HIV-22 IN WEST AFRICA
Pathogenesis Pathogenesis Thiss study shows that, like in HIV-1, '4 PVL is also a key predictor of mortality in HIV-2 infection.. Although in univariate analysis the mortality of HIV-2 was much lower than in HIV- 1,, HIV type was not an independent predictor of mortality after adjusting for PVL, CD4%, age groupp and time period. This suggests that the lower mortality of HIV-2 can be explained by the generallyy lower PVL and higher CD4% in HIV-2,2i'29 and indicates a lower virulence of HIV-2.
Ann intriguing question remains why some people infected with HIV-2 develop a high PVL and rapidd decay of the immune system, and others not. This is not due to subtype (all HIV-2 patientss in The Gambia have subtype A), ° but could be due to intra-subtype variations. The transmissionn route, infecting dose, genetic factors such as HLA type, and the HLA type of the infectingg partner may be important as well. Further studies to elucidate this interaction between agent,, host and environment are needed.
Implications Implications Thiss study indicates that PVL and CD4% are of great value in predicting the outcome of infection:: women with a normal baseline CD4% and a baseline PVL < 500 copies/mL had a mortalityy rate that was not significandy different from that of seronegative women (p=0.6). Sincee clinical markers alone do not reliably predict mortality, 32 3i baseline CD4% or PVL measurementt may be important to guide decisions to start ART. Currently there are no guideliness for the treatment of HIV-2. Our data suggest that the same clinical, virus load, and CD44 criteria should be applied in cases of HIV-2 infection as in HIV-1 when considering the use off ART. Most HIV-2 infected people live in Africa, but unfortunately few HIV-2 infected patientss in Africa are receiving ART.
Acknowledgements s
Wee thank all women and their families participating in the study; Alieu Jatta, Baba Danso, Saikou Badjiee (counsellors / field workers); Aji Fatou Jallow, Awa Kenda, and Majulla Saidy for data entry;; the Divisional Health Teams in The Gambia; Dr. Martin Weber (WHO, Geneva), Dr.
144 4 HIV-22 IN WEST AFRICA
Tumanii Corrah, and Prof. Keith McAdam (MRC Laboratories, The Gambia), Saihou Ceesay
(Nationall AIDS Secretariat The Gambia), Dr. Gisela Schneider (WEC International), Dr.
Shabbarr Jaffar (LSHTM), Prof. Roel Coutinho (GG&GD Amsterdam) for their advice and support. .
Thiss study was funded by the Medical Research Council (UK), and the Japanese Foundation for
AIDSS prevention. Maarten Schim van der Loeff has an MRC-funded Linked Fellowship.
Andreass Hansmann received a student travel award from the Deutscher Akademischer
Austauschh Dienst.
References s
1.. UNAIDS. Epidemic update, www.unaids.orq/ worldaidsday/ 2002/press/ update/ epiupdate20022 en.doc. Accessed 17 December 2002. 2.. Marlink R, Kanki P, Thior I, et al. Reduced rate of disease development after HIV-2 infectionn as compared to HlV-1. Science 1994; 265: 1587-1590. 3.. Kanki PJ, Travers KU, MBoup S, et al. Slower heterosexual spread of HIV-2 than HIV-1. LanceMM 994; 343: 943-6. 4.. O'Donovan D, Ariyoshi K, Milligan P, et al. Maternal plasma viral RNA levels determine markedd differences in mother -to-child transmission rates of HIV-1 and HIV-2 in The Gambia.. AIDS 2000; 14: 441-448. 5.. Schim van der Loeff MF, Jaffar S, Aveika AA, et al. Mortality of HIV-1, HIV-2 and HIV-1 / HIV-22 dually infected patients in a clinic-based cohort in The Gambia, West Africa. AIDS 2002;; 16: 1775-83. 6.. Collaborative Group on AIDS Incubation and HIV Survival. Time from HIV-1 seroconversionn to AIDS and death before widespread use of highly-active antiretroviral Therapy:: a collaborative re-analysis. Lancet 2000; 255: 1131-1137. 7.. Morgan D, Muwonge R, Mahe C, et al. HIV-1 infection in rural Africa: is there a difference inn median time to AIDS and survival compared with that in industrialised countries? AIDS 2002;; 16:597-603. 8.. Crampin AC, Floyd S, Glynn JR, et al. Long-term follow-up of HIV-positive and HIV- negativee individuals in rural Malawi. AIDS 2002; 16: 1545-1550 9.. Ricard D, Wilkins A, N'Gum PT, et al. The effects of HIV-2 infection in a rural area of Guinea-Bissau.. AIDS 1994; 8: 977-82. 10.. Poulsen AG, Aaby P, Larsen O, et al. 9-year HIV-2 associated mortality in an urban communityy in Bissau, West Africa. Lancet 1997; 349: 911-914. 11.. Norrgren H, da Silva ZJ, Andersson S, et al. Clinical features, immunological changes andd mortality in a cohort of HIV-2 infected individuals in Bissau, Guinea-Bissau. Scand J InfectInfect Dis 1998; 30: 323-329. 12.. Nunn AJ, Mulder DW, Kamali A, Ruberantwari A, Kengeya-Kayondo JF, Withworth J. Mortalityy associated with HIV-1 infection over five years in a rural Ugandan population: cohortt study. BMJ 1997; 315: 767-771. 13.. Leroy V, Msellati P, Lepage P, et al. Four years of natural history of HIV-1 infection in Africann women: a prospective cohort study in Kigali (Rwanda), 1988-1993. J Acq Imm DefSyndrDefSyndr 1995; 9:415-21.
145 5 HIV-22 IN WEST AFRICA
14.. Mellors JW, Rinaldo Jr. CR, Gupta, et al. Prognosis in HIV-1 infection predicted by the quantityy of virus in plasma. Science 1996; 272: 1167-1170. 15.. Katzenstein DA, Hammer SM, Hughes MD, et al. The relation of virological and immunologicall markers to clinical outcomes after nucleoside therapy in HIV infected adultss with 200 to 500 CD4 cells per cubic millimeter. N EnglJ Med 1996; 335: 1091-8. 16.. O'Brien WA, Hartigan PM, Martin D, et al. Changes in plasma HIV-1 RNA and CD4+ lymphocytee counts and the risk of progression to AIDS. N Engl J Med 1996; 334: 426- 431. . 17.. O'Brien WA, Hartigan PM, Daar ES, Simberkoff MS, Hamilton JD. Changes in plasma HIVV RNA levels and CD4+ lymphocyte counts predict both response to anti retroviral therapyy and therapeutic failure. Ann Intern Med 1997; 126: 939-45. 18.. Ho DD. Viral counts in HIV infection. Science 1996; 272: 1124-5. 19.. Berry N, Jaffar S, Schim van der Loeff MF, et al. Low level viremia and high CD4% predictt normal survival in a cohort of HIV-2 infected villagers. AIDS Res Hum RetrovirusesRetroviruses 2002; 18: 1167-73. 20.. Ota M, O'Donovan D, Alabi AS, et al. Maternal HIV-1 and HIV-2 infection and child survivall in The Gambia. AIDS 2000, 14:435-439. 21.. Ministry of Health, National AIDS Control Programme, The Gambia. Policies and GuidelinesGuidelines on HIV and AIDS. Banjul: Ministry of Health; 1995. 22.. Boom R, Sol CJ, Salimans M, Jansen C, Wertheim-van Dillen P, van der Noorda J. Rapidd and simple methods for purification of nucleic acids. J Clin Micro 1990; 28: 495- 503. . 23.. Berry N, Ariyoshi A, Shabbar J, et al. Low peripheral blood viral HIV-2 RNA in individuals withh high CD4 percentage differentiates HIV-2 from HIV-1 infection. Journal of Human VirologyVirology 1998; 1:457-468. 24.. Ariyoshi K, Jaffar S, Alabi A, et al. Plasma RNA viral load predicts the rate of CD4 T cell declinee and death in HIV-2-infected patients in West Africa, AIDS 2000; 14: 339-344. 25.. Ryder RW, Batter VL, Nsuami M, et al. Fertility rates in 238 HIV-1 seropositive women in Zairee followed for 3 years post-partum. AIDS 1991; 5:1521 -7. 26.. Gray RH, Wawer MJ, Serwadda D, et al. Population-based study of fertility in women withh HIV-1 infection in Uganda. Lancet 1998; 351: 98-103. 27.. Glynn JR, Buvé A, Caraël M, et al. Decreased fertility among HIV-1 infected women attendingg antenatal clinics in three African cities. JAIDS 2000; 25: 345-52. 28.. UNAIDS. Report on the global HIV/AIDS epidemic. Geneva: UNAIDS; 2002. 29.. Atabi AA, Jaffar S, Ariyoshi K, et al. Plasma viral load, CD4 Lymphocyte counts and survivall of HIV-1, HIV-2 and dually infected Gambian patients. AIDS Accepted. 30.. Berry N, Ariyoshi K, Balfe P, Tedder R, Whittle H. Sequence specificity of the human immunodeficiencyy virus type 2 (HIV-2) long terminal repeat u3 region in vivo allows subtypingg of the principal HIV-2 viral subtypes A and B. AIDS Res Hum Retroviruses 2001;; 17:263-7. 31.. Grassly NC, Xiang Z, Ariyoshi K, et al. Mortality among human immunodeficiency virus typee 2-positive villagers in rural Guinea-Bissau is correlated with viral genotype. J Virol 1998;; 72: 7895-9. 32.. Kilmarx PH. Commentary: Virus, host, or environment? BMJ 2002; 324: 197. 33.. Morgan D, Mahe C, Mayahja B, et al. Progression to symptomatic disease in people infectedd with HIV-1 in rural Uganda: prospective cohort study. BMJ 2002; 324: 193-196.
146 6 8 8
Survivall of HIV-1 and HIV-2 perinatallyy infected children inn The Gambia
AIDSAIDS 2003; in press
Maartenn F Schim van der Loeff 12, Andreas Hansmann u, Akum Aveika Awasana , Martinn O. Ota ', Diarmuid O'Donovan ', Ramu Sarge-Njie XA, Koya Ariyoshi ll5, Paull MiUigan ', Hilton Whitde '
(1)(1) the Medical Research Council Laboratories, Fajara, The Gambia; (2) London School of Hygiene <&<& Tropical Medicine, London, UK; (3) Westfdlische Wilhelms-Universitdt, Munster, Germany; (4) NationalNational Health Laboratories Services, Banjul, The Gambia; (5) AIDS Research Center, National InstituteInstitute of Infectious Diseases, Tokyo, japan
Thiss study was funded by the Medical Research Council (UK), andd the Japanese Foundation for AIDS prevention.
Survivall of HIV-1 and HIV-2 perinatallyy infected children inn The Gambia
ABSTRACT T
Background:: The risk of mother-to-child transmission (MTCT) of HIV-2 is muchh lower than that of HIV-1, but the long-term prognosis of perinatally infectedd HIV-2 children is unknown. We re-visited children who were part of a largee MTCT study in The Gambia (conducted 1993 - 97), in order to compare thee long-term survival of children perinatally infected with HIV-2 with that of seronegativee and of HIV-1 infected children. Methods:: Five to eight years follow-up of a cohort of children born to HIV- negative,, HIV-1 positive, and HIV-2 positive mothers. Results:: Seven hundred and seventy-four children were followed up for a mediann of 6.6 years. Of 17 perinatally HIV-1 infected children, 3 were still alivee on 1st July 2001, 2 had been lost to follow up, and 12 had died. The mediann survival was 2.5 years. Of 8 HIV-2 infected children 5 were still alive, nonee were lost to follow up and 3 had died. The mortality hazards ratio of bothh HIV-1 (9.9; 95% Confidence Interval (CI) 5.2-19), and of HIV-2 infected childrenn (3.9; CI 1.2-12) was significantly increased compared to children of seronegativee mothers. The mortality hazards ratio of HIV uninfected children off HIV-1 or HIV-2 infected mothers was not significantly increased compared too children of seronegative mothers (P = 0.17 and P = 0.5 respectively). Conclusions:: Children with perinatally acquired HIV-2 infection have a higherr mortality than children of seronegative mothers. Guidelines for treatmentt of HIV-1 infected children should be used for treatment of HIV-2 infectedd children. HIV-22 IN WEST AFRICA
Introduction n
Thee majority of the 3 million HIV infections in children world-wide results from mother-to-childd transmission (MTCT) [1-3]. The MTCT rate in HIV-1 is between
211 % and 43% in breast feeding populations without access to antiretrovirals [4], but cann be as low as 1% in women using antiretro viral therapy and delivered by elective caesareann section [5]. The mortality of perinatally infected children in affluent countriess has improved dramatically in recent years due to better treatment, notably anti-retrovirall treatment [6-9]. Even the median survival in children not on antiretrovirall therapy (ART) is longer than 8 years [6,7,10]. The 5-year probability of survivall ranged from 0.60 to 0.75 [6,10,11].
Almostt 90% of children with HIV-1 live in sub-Saharan Africa [1], where the survival off children with HIV-1 is much worse. In studies conducted in Africa the survival probabilityy of HIV-1 infected children ranged from 0.66 to 0.80 at 1 year [12-14], and fromm 0.25 to 0.43 at 5 years [12,15,16]. Reported median survival ranged from 12.4 monthss to 21 months [13,15,17]. An estimated 8% of all under-5 mortality in sub-
Saharann Africa is caused by HIV. [18]
Thee mother-to-child transmission of HIV-2 is estimated to be 4%, one-sixth that of
HIV-11 [19]. Survival of adults with HIV-2 is better than in HIV-1 infection [20 22], butt few data exist on the survival of children with HIV-2 infection. There are several casee reports of presumably perinatally infected children who reached teen [23-25], To ourr knowledge, there are no prospective cohort studies of perinatally infected HIV-2 children. .
Inn order to estimate the long-term survival of children with HIV-2 infection, and to comparee this with survival in HIV-1 infected and in uninfected children in the same population,, we re-visited mother and child pairs four years after the completion of a largee observational MTCT study that was conducted between 1993 and 1997 [19,26].
150 0 HIV-22 IN WEST AFRICA
Methods s
PerinatalPerinatal study Betweenn January 1993 and March 1995 HIV-negative, HIV-1 infected and HIV-2 infectedd pregnant women were recruited at eight health centres throughout The Gambia.. They were followed up at regular intervals up to 18 months postpartum. HIVV infection in infants was diagnosed by PCR at 2, 9 and 18 months and by serology att 18 months. Babies were considered infected if they tested HIV seropositive at 18 monthss of age, or, if this specimen was not available (through death or loss to follow- up)) if they tested positive by PCR at 2 or 9 months of age [19]. Infants who died or weree lost to follow-up before a first post-natal sample was taken (typically at 2 months),, could not be assigned an HIV diagnosis. The survival probability at 18 monthss was 0.64 (95% confidence interval (CI) 0.37 - 0.82) for the seventeen HIV-1 infectedd and 1.0 (95%CI 0.63-1.0) for the eight HIV-2 infected children [26].
FieldField follow-up Fromm July to November 2001 an attempt was made to visit all 819 women and their childrenn who were seen at least once since their delivery in 1993-5. No visit was plannedd when the woman had refused participation in the earlier study or when it was knownn diat both mother and child had died. In the latter case the mortality data from thee earlier study were used in the analysis. Women were visited by one of three male Gambiann field workers, who were unaware of the sero-status of mother and child. The fieldfield workers were experienced counsellors in the genito-urinary medicine (GUM) clinicc of the Medical Research Council (MRC) Laboratories in Fajara, and were trained too explain the study, and discuss the purpose and meaning of an HIV test. Interviews weree conducted in the language of the study participants. A short questionnaire was completedd with some key demographic details. The field workers aimed to speak to thee study woman, or if she had died, to a close relative, and relied on their information regardingg the vital status of the study child. All women were offered a new HIV test.
StatisticalStatistical analysis
151 1 HIV-22 IN WEST AFRICA
Losss to follow-up was defined as lack of information concerning the survival status of thee child as per Is'July 2001. Observation time started at the date of birth, and ended at thee date of death, or the date the child was last known to be alive, whichever came first.first. Children of a twin deliver)7 were both included in the analysis. Time to death was examinedd using Kaplan-Meier graphs. Cox proportional hazards analysis and log rank testss were performed to compare survival between groups.
Htbics Htbics
Thee study was approved by the Gambia Government / MRC Laboratories Joint Ethics
Committee.. Participants were asked for verbal informed consent, which was documentedd by the field worker. In the original study the results of HIV testing had beenn available from a counsellor based in government health centres. Ver)7 few women camee to obtain their test results (< 1%), and consequently the current study followed womenn who were mostly unaware of their HIV status. Instead of offering the result of thee test done 6 to 8 years previously, all participants were offered a new HIV test, irrespectivee of their original HIV status. All newly tested positive subjects were referredd for free clinical care to the MRC GUM clinic. Perinatal antiretroviral prophylaxiss was not available in The Gambia at the time of birth of the study children.
Nonee of the children received antiretroviral therapy or prophylaxis against opportunisticc infections during the study period.
Results s
Eightt hundred and nineteen women had 832 children (including 10 sets of twins). The tenn children (all of them HIV-negative) born to HIV-1 and HIV-2 dually infected
(HIV-D)) women, and 48 stillborn children were excluded, leaving 774 children for analysis.. As of Is' July 2001, 104 chÜdren (131'o) had died, 182 (24%) were lost to followw up, and 488 were alive. The median observation time was 6.6 years
(interquartilee range (IQR) 1.5-7.5). The rate of loss to follow up was 4.6 (95% CI 4.0-
5.3)) per 100 person-years of observation, and did not differ by HIV status of mother orr child.
152 2 ^T-^T- s? ? atat m co co o •t t s-- ô co ô a> ^ c-- er- « « co o 22. . ES S O) ) ra ra't t mm O) o o oo v o o ai i o o 00 0 -—-- >—- c c c c •* * o o coo o> oo^ CL L £ £ co o Ui Ui -o o o o .o o XI I E E S S £ £ ^ ^ O OC D D CM M c c CM M 2, , co o J£ £ -- C C 'T 00 0 3 3
0) ) s~*s~* ^p vp o "•5" " CDD ^ o^ ^.; co"" ^ CDD vP 00 .o CO ^ TT T OO CO CM ^ ,-,S*-COOÔ>CMM i»n Q QC D D CM M CMM -A CM T- O) ) c c ^J-ïï^J-ïï l S l -po 4) ) CD D ,-- co co ^ C C ^CM ^M t t •—•• CO CO o
(A A
E E 3 3 ) ) co o flj flj % ) ) O O % (N N •5" " 2
CU U 369 > > 00 0 O, , % 00 0 0 - -
> > 9. 6. > > 3. 0.0 x x «o, , 63 100 o o 33
X X X X O O (2.7-32 00 •3-100 : 1.2-12 co o (1.9-21 co o ^^ CD CM ^_^ T3 3 co o CD D CD D *o *o T3 3 £ £ O O u u E E C C # # 5* * o o r-*-- ff-ff-CO O c c T3 3 JC C ^ ^ Si i C C c c h-- in n 3 3
^^ 3" 3" co cp'' ^ _> > Tff o"- c?-- - co o CD D coo i oôjôico i rr r».. iri co oo ' oo ' r: o V V o o O O QQ CO co oq c c v-- T— — üü co, r-^ o •^^•^^ ^^ ) ) ) ) ) ) ^5" " 5 499 5 82% 9 % % 2 9
r-- CM M 44% - 1 - > > 2 9. - 23 59 21 9.5-53 9.3-38
X X 88
CM M 33 <0.000 CM M 55 (5.2-19 ( (6 (3
OO c OO "£• '7. 00 0 ^-s s 5?? in g**- ! ! oo o 0s s ISS 91 eö CU U O. . XXCftt 3 3 .2 iif-f o o .22 > .2 > 3 3 ••OO55 mo^ ion^ *5- T-«CJT J J i c c i i 22 "E 2 £ 5 5 coo r CO co . CD D __ 3 =3 ( ( d d d s 0 0 s ll ll
fo CO O yy ) ) n CL L n tt o E E o 0 0 T) ) ll o o O) ) ll tes rtalit rviva rviva hazar
hazar o o hazar (J J a a <11 1 95% 95% OO 3 3 ,- _*: o o Wi i f-- c c C.. « CO ^ c 5^ ^ 3 3 •ê\22 w.$\o to O O o o i "cö ö .O O T3 3 •t;; m m (ö 2 CLL co ~ £ co ~ £ O OB B ; t —' ' r > > 5 5 •o o II I o o g 0) ) ye £ £oo o £ o o £ o (0 0 > > O O 3 3 n> > Lo
mor __ _ Infa 55 2ye OTEOTE 8 E E 8 E u u xi i LJ J b b _l l CA A HIV-22 IN WEST AFRICA
Figuree 1. Kaplan-Meier graph comparing survival of HIV-1 infected and HIV-2 infected children, and off children of HIV uninfected mothers. (The Gambia 1993-2001)
I I 1.00 0 HIVV uninfected
0.75 5 u u HIV-2 2 1 1 0.50 0 'l l
HIV-1 1 0.25 5
0.00 0 I I 1 1 ii i analysis stim e e
0 0 1 1 2 2 3 3 4 4 5 5 6 6
Childrenn at risk: HIV-uninfected d 448 8 411 1 312 2 307 7 307 7 304 4 301 1 HIV-1 1 17 7 13 3 8 8 6 6 5 5 3 3 3 3 HIV-2 2 8 8 8 8 8 8 5 5 5 5 5 5 5 5
Twelvee out of 17 HIV-1 infected children died (71%), the median survival being 2.5 years (IQR 1.0-4.5)) (see Table 1). The mortality hazards ratio of HIV-1 infected children was 9.9 (CI 5.2- 19)) compared to children of HIV uninfected mothers (P<0.0005). Three out of 8 HIV-2 infectedd children died (38%). The median survival could not be calculated, but after 6 years 5 (63%)) were still alive (CI 23-86%). The mortality hazard ratio of HIV-2 infected children was 3.99 (CI 1.2-12) compared to children of HIV uninfected mothers (P=0.02). The mortality hazardd ratio of HIV-1 infected children was 3.1 (0.87-11) times that of HIV-2 infected children, butt this was not significant (P=0.08). Figure 1 compares the survival of HIV-1 and HIV-2 infectedd children with children born to HIV seronegative women. The characteristics of the eightt HIV-2 infected children are listed in Table 2.
154 4 o o 8* * T-- CU i-- -C CDD ** Q-cü ü coo £f .922 co Q.. tl| | 8f f rr2 to e*. >-- >- >- > > << O'S ÜÜ 0) ••*= CO -cc -Si JS ZZ E o o-oo S ... .2 > CUU "O cö
COO O ,_ .>> co CD lil l £-« « << < < O < < < < 33 (0 O «i-s s WW m E J33 " CL OO ^ II (00 OT3 oo gj "- 55 TÏ O o o !>! ! h-^-CDcsioqcN-*Tt t *-.. "O o o CNoiTtTtTtTteócvi i «« CO II CM M üü CL "5 -^z z i i cc co efc CDD 2 CO O << o o o CO O roZroZ '" co o T33 o ojgg ïi' j.. E °cü ü ££ t: > COO CD » CÖÖ « .25 "OO CO J2 'ss e S iBB S^ JSS o •- Cöö >N £- CUU £ cc o^ £ " ^ .2 cc O E "55 -p "o •o•oCDD £C = T3 3 .. o CU U CD D ^22 ca t3 3 ."QQÊÊ cu• E< u 22 2 rou ^^ . co 2 , O ü.. "> J3 CDD •? II 'S-Euroro In•• .. ^.9>> © ca a o.. co Tt'-a-r^.coocNirôp p o oOO E cvicvjcvicor^r^r^r^ ^ CDD < n n CD D O) ) gaf. . < < >** CO CD Cöö O T3 >> co '?? — 5 QQQ<<<<< < =33 <0 (0 « «c cc = CD +-- T3 C .22 o = COO CD ™ "OO i_ «J CUU "O £ jjOOOONÔO'r r CCMCO'CMCOCOCO^ ^ EE S^ ,,mm,, ° CD Tc33o o C CO O UJJ cu 5 §3o 3o ww ö „ oo e •> ®® x) £ «o-coo 2S .So 55 J3 < "PP "5 -^ CRR © CM T- o m CMC M M 0 0 COO CD CD (incoiotoocoocOO (O O O) (O ifl CMo Oo COO lf> CM CM CM i- T- S'CTJ J o o .. CD CO OO to c coo ^j- co ^t •* CM o COO CL-C Hlv-22 IN WEST AFRICA
Survivall to the time of the first blood sample was essential for making an HIV diagnosis. A bloodd sample was not available for 17 (17%) children of HIV-1 infected mothers, and for 26 (11%)) children of HIV-2 infected mothers. Therefore we repeated the proportional hazards calculations,, conditional on survival up to 4 months. Also in this approach the mortality hazards off HIV-1 and HIV-2 infected children were significantly higher than in children of uninfected motherss (P<0.0005 and P=0.003 respectively).
Thee mortality hazard ratio of children according to their mothers' HIV status (and ignoring their ownn status) was 3.6 (CI 2.2-5.7 ; P
Provirall load was available for 11 of 17 HIV-1 infected children; each logu, increase of proviral loadd in the infant was associated with a 1.5 rise (CI 0.89-2.5) in mortality hazard ratio (P=0.13). Inn HIV-2 children there were only 4 proviral loads done, and no associations were estimated.
Nonee of the following variables were associated with mortality hazard ratios in HIV-1 or HIV-2 infectedd children: maternal antenatal or postnatal CD4%, maternal plasma viral load, age, parity, orr ethnic group, and baby's sex, but power was low due to small numbers.
Discussion n
Thiss study shows that the survival rate of children with perinatally acquired HIV-2 is worse than thatt of uninfected children, but may be better than in HIV-1 infected children. The mortality of 156 6 HTV-22 IN WEST AFRICA childrenn with perinatally acquired HIV-t was similar to that found in other studies [12,13,15,17]: thee median survival time was 2.5 years, and the survival probability at 5 years was 0.21. The mediann survival in HIV-2 could not be estimated, but at age 7 years 5/8 of the children were still alive. .
Thee rate of loss to follow-up was limited with 4.6 per 100 pyo. The observed 4-fold higher mortalityy of HIV-2 infected children compared to children of HIV uninfected mothers is likely too be an underestimate, as diagnosis of HIV-2 infection was conditional on surviving up to the firstt sample (typically at 2 months). Twelve children born to HIV-1 infected, and 19 children bornn to HIV-2 mothers died or where lost to follow-up before the HIV status could be established,, but some of these may have been infected. Therefore the proportion of infected childrenn surviving may in fact be lower and the mortality hazard ratio higher than we estimated. Inn order to obtain unbiased estimates, hazard ratios were calculated conditional on survival to 4 monfhs.. The mortality HR appeared higher than in the overall analysis, confirming an increased mortalityy rate in both HIV-1 and HIV-2 infected children.
Perinatallyy acquired HIV-2 infection may have a better prognosis than perinatally acquired HIV- 11 infection. Although the plasma viral load of the mothers who infected their children widi HIV-22 was high [19], it appears that this does not necessarily lead to a rapid disease progression inn the children. This underlines the importance of host factors in the natural course of HIV-2 infection. .
Motherss with HIV-2 must be counselled that they have a small chance of transmitting the infectionn to their baby, but if they do, their babies are at increased risk of early death. HIV-2 infectedd children need antiretroviral treatment and may need prophylaxis against opportunistic infections.. There are few studies on treatment of HIV-2 in adults, [27-29] and none of treatment off HIV-2 in children. In the absence of HIV-2 specific guidelines and lack of data on PVL, CD4 countt and morbidity, we advise that guidelines in use for HIV-1 [30,31] should be followed, with diee exclusion of non-nucleoside analogue reverse transcriptase inhibitors (e.g. nevirapine), which aree not effective against HIV-2. [32]
157 7 H1V-22 IN WEST AFRICA
Acknowledgements s
Wee thank all women and children who participated in the study; Alieu Jatta, Saikou Badgie, Baba
Dansoo (field workers); Awa Kendah, AjiFatou Jallow, Sulayman Sowe, Majulla Saidy (data entry staff);; Pa Tamba N'gom, Elizabeth Harding, Dr Abraham Alabi (laboratory'); Lawrence K
Yamuah,, Dr Andrew Wilkins (investigators of the original study); Saihou Ceesay, (National
AIDSS Secretariat), Dr Tumani Corrah and Prof. Keith McAdam (MRC Laboratories The
Gambia),, Dr Gisela Schneider (WEC International), PD Dr Martin Weber (University of
Hannover,, Germany) and Dr Shabbar Jaffar (London School of Hygiene & Tropical Medicine,
UK)) for support and advice; and Prof Roe! Coutinho and Prof. Richard Hayes for feedback on earlierr drafts.
References s
1.. UNAIDS. Report on the global HIV/AIDS epidemic. Geneva: UNAIDS; 2002. 2.. Mofenson LM, Mclntyre J. Advances and research directions in the prevention of mother-to- childd HIV-1 transmission. Lancet 2000; 355: 2237-44. 3.. Dabis F, Ekpini ER. HIV-1/AIDS and maternal and child health in Africa. Lancet 2002; 359: 2097-104. . 4.. Working Group on Mother-To-Child Transmission of HIV. Rates of mother-to-child transmissionn of HIV-1 in Africa, America, and Europe: results from 13 perinatal studies. Journall of Acquired Imm Defic Syndr and Hum Retrovir. 1995; 8: 506-10. 5.. International Perinatal Group. The mode of delivery and the risk of vertical transmission of Humann Immunodeficiency Virus type 1. NEJM 1999; 340: 977-87. 6.. European Collaborative Study. Fluctuations in symptoms in human immunodeficiency virus- infectedd children: the first 10 years of life. Pediatrics 2001; 108: 116-22. 7.. De Martino M, Tovo P-A, Balducci M, et al. Reduction in mortality with availability of antiretrovirall therapy for children with perinatal HIV-1 infection. JAMA 2000; 284: 190-7. 8.. McKinney RE, Johnson GM, Stanley K, et al. A randomized study of combined zidovudine- lamuvidinee versus didanosine monotherapy in children with symptomatic therapy-naive HIV-1 infection.. J Pediatr 1998; 133: 500-8. 9.. Gortmaker SL, Hughes M, Cervia J, et al. Effect of combination therapy including protease inhibitorss on mortality among children and adolescents infected with HIV-1. N Engl J Med. 20011 Nov 22;345:1522-8. 10.. Barnhart HX, Caldwell MB, Tomas P, et al. Natural history of human immunodeficiency virus diseasee in perinatally infected children: an analysis from the Paediatric Spectrum of Disease Project.. Pediatrics 1996; 97: 710-16. 11.. Langston C, Cooper RE, Goldfarb J, et al. Human immunodeficiency virus-related mortality in infantss and children: data from the pediatric pulmonary and cardiovascular complications of verticallyy transmitted HIV (P2C2) study. Pediatrics 2001; 107: 328-38. 12.. Lepage P, Spira R, Kalibala S, etal. Care of human immunodeficiency virus-infected children inn developing countries. Pediatr Infect Dis J 1998; 17: 581-6.
158 8 HIV-22 IN WEST AFRICA
13.. Marum LH, Bagenda D, Guay L, et al. Three-year mortality in a cohort of HtV-1 infected and uninfectedd Ugandan children. XI International Conference on AIDS. Vancouver. July 1996 [abstractt We.B.312]. 14.. Ryder RW, Nsuami M, Nsa W, et al. Mortality in HIV-1 seropositive women, their spouses andd their newly born children during 36 months of follow-up in Kinshasa, Zaire. AIDS 1994; 8: 667-72. . 15.. Sptra R, Lepage P, Msellati P, et al. Natural history of human immunodeficiency virus type 1 infectionn in children: a five-year prospective study in Rwanda. Mother-to-Child HIV-1 Transmissionn Study Group. Pediatrics 1999; 104: e56. 16.. Marum LH, Tindyebwa D, Gibb D. Care of children with HIV infection and AIDS in Africa. AIDSS 1997; 11 (suppl B): S125-S134. 17.. Dabis F, Elenga N, Meda N, et al. 18-Month mortality and perinatal exposure to zidovudine in Westt Africa. AIDS 2001; 15: 771-9. 18.. Walker N, Schwartlaender B, Bryce J. Meeting international goals in child survival and HIV/AIDS.. Lancet 2002; 360; 284-9. 19.. O'Donovan D, Ariyoshi K, Milligan P, et al. Maternal plasma viral RNA levels determine markedd differences in mother-to-child transmission rates of HIV-1 and HIV-2 in The Gambia. AIDSS 2000, 14:441-448. 20.. Poulsen AG, Aaby P, Larsen O, et al. 9-year HIV-2 associated mortality in an urban communityy in Bissau, West Africa. Lancet 1997; 349: 911-914. 21.. Morgan D, Muwonge R, Mahe C, et al. HIV-1 infection in rural Africa: is there a difference in mediann time to AIDS and survival compared with that in industrialised countries? AIDS 2002; 16:: 597-603. 22.. Schim van der Loeff MF, Jaffar S, Aveika AA, et al. Mortality of HIV-1, HIV-2 and HIV-1 / HIV- 22 dually infected patients in a clinic-based cohort in The Gambia, West Africa. AIDS 2002; 16: 1775-83. . 23.. Rosado L, Neves C, Bessa A, et al. Survival among HIV-2 perinatally infected infants. XI Internationall Conference on AIDS. Vancouver. July 1996 [abstract We.C.3436] 24.. Faye A, Burgard M, Crosnier H, Retbi JM, Blanche S. Human Immunodeficiency virus type 2 infectionn in children. J Pediatr 1997,130:994-997. 25.. Mota-Miranda A, Gomes H, Lima-Alves C, Araujo F, Cunha-Ribeiro LM, Taveira N. Perinatally acquiredd HIV-2 infection diagnosed at 15 and 24 years of age. AIDS 2001; 15: 2460-1. 26.. Ota M, O'Donovan D, Alabi AS, et al. Maternal HIV-1 and HIV-2 infection and child survival in Thee Gambia. AIDS 2000, 14:435-439. 27.. Smith NA, Shaw T, Berry N, Vella C, Okorafor L, Taylor D, et al. Antiretroviral therapy for HIV-22 infected patients. J Infect 2001;42:126-33. 28.. Bissagnene E, Eholie S, Tanon A. Evaluation du traitement antiretroviral cheze les patients VIH-22 positifs è Abidjan. Xllth International Conference on AIDS and STDs in Africa. Ouagadougou,, December 2001 [abstract 10DT3-6]. 29.. Adjé-Touré CA, Cheignsoni R, Garcia-Lerma G, Adams DR, Konan-Koko R, Otten R, et al. Virologiee and immunologic responses and drug resistance profiles of HIV-2 infected patients receivingg antiretroviral treatment in Abidjan. Xllth International Conference on AIDS and STDs inn Africa. Ouagadougou, December 2001 [abstract 11DT4-5]. 30.. Sharland M, di Zub GC, Ramos JT, Blanche S, Gibb DM; PENTA Steering Committee. PENTAA guidelines for the use of antiretroviral therapy in paediatric HIV infection. Pediatric Europeann Network for Treatment of AIDS. HIV Med 2002 ;3:215-26. 31.. Anonymous. Guidelines for the use of antiretroviral agents in pediatric HIV infection. http://vww.aidsinfo.nih.gov/guidelines/pediatric/pediatric.pdf.. Accessed 28 February 2003. 32.. Bardsley-Elliot A, Perry CM. Nevirapine: a review of its use in the prevention and treatment of paediatricc HIV infection. Paediatr Drugs 2000; 2: 373-407.
159 9 H1V-22 IN WEST AFRICA
160 0 9 9
Conclusions s
HIV-22 IN WEST AFRICA
Introduction n
Thiss thesis investigates the epidemiology of HIV-2 infection in The Gambia and Guinea- Bissau.. In this last chapter die key findings of tlie preceding chapters are summarised, putt into context and some conclusions will be drawn. The themes are: trends in HIV-2 incidencee and prevalence, the cause and maintenance of the HIV-2 epidemic, risk factors forr infection, interactions with HIV-1, mortality, die role of plasma viral load, host versus virall factors to explain pathogenesis, HIV-2 infection in children, treatment and HIV-2 vaccines.. Finally, four questions that are still open and deserve further research will be identified. .
Prevalencee and incidence of HIV-2
Alll studies from West African countries have shown stable or declining prevalences of HIV-22 in diverse populations: female commercial sex workers [1-7], an occupational cohortt [8], pregnant women [6, 8, 9, 10], STD patients [11, 12], and the general populationn [13]. Three cohort studies reported incidence rates over different time periods;; among commercial sex workers in Dakar the incidence was stable [4] and in an occupationall cohort and in a peri-urban community in Bissau the rates were falling over timee in men and stable in women [8, 13].
Chapterr 3 compared the HIV-2 prevalence among pregnant women attending antenatal clinicss in The Gambia in 2000-1 with the prevalence in a similarly defined population in 1993-5.. A non-significant decline of 0.2% (from 1.0% to 0.8%) was observed, in line withh the trends seen elsewhere. Chapter 4 estimated the HIV-2 incidence in a rural communityy in Guinea-Bissau at 4.8 (95% confidence interval (CI) 3.7-6.4) per 1000 person-yearss of observation (pyo). This is a low incidence in view of the high prevalence off around 8.0% in adults. Although the overall prevalence did not change, the prevalencee among young people (< 45 years) had decreased from 5.9% to 4.8% (p = 0.10),, and the prevalence among people aged 45 years or above had increased from
163 3 H1V-22 IN WEST AFRICA
11.8%% to 14.8% (p = 0.04) between two rounds of cross-sectional studies (1989-91 and 1996-8),, suggesting a cohort effect [unpublished data]. Mathematical modelling is underwayy to explore this phenomenon in more detail. Because all these studies report stablee or declining prevalences in West Africa, and because no new HIV-2 epidemics havee been observed outside West Africa, [chapter 2] we can conclude that HIV-2 is not ann "emergent epidemic", but an epidemic in decline.
Causee and maintenance of the epidemic
Thiss raises one of the fundamental, still unanswered questions about HIV-2: which eventss created the epidemic, and what has changed so that the epidemic is no longer sustained?? HIV-2 is genetically close to die Simian Immunodeficiency Virus of die sooty mangabeyy (SIVsm) [14], a monkey. The natural habitat of the sooty mangabey (Cercocebm torquatmtorquatm atys) is West Africa. Based on the large degree of genetic homology, the geographicc overlap, and the fact that human-monkey contacts are common in West
Africa,, most researchers maintain that SIVsm is the source of HIV-2 [15]. In a phylogeneticc tree SIVsm and HIV-2 are not separate branches; most subtypes of HIV-2 clusterr closer to specific strains of SIVsm than to each other [16, 17]. Therefore it is assumedd that the 7 different clades of HIV-2 [18, 19] represent at least 7 different zoonoticc events. All subtype A strains are thought to be derived from one such zoonotic event. .
HIV-11 is genetically related to SIV a virus of chimpanzees. The habitat of chimpanzeess in West-Central Africa overlaps with the area of the greatest diversity of HIV-1,, and there appears to be increasing human-primate contact, partly by hunting. In phylogeneticc analyses some strains of SIV are closely related to HIV-1 [20]. Because of thesee reasons it is generally assumed that HIV-1 has originated from SIV [20, 21]. The exactt details of this zoonotic event are unknown and are controversial [22, 23]. By using mutationn rates of the viral genome as a molecular clock, back-calculations estimated the timingg of the original transmission from chimpanzee to human at about 1930 [24, 25].
164 4 HIV-22 IN WEST AFRICA
Recentlyy this has been done for HIV-2 as well. Based on partial sequences of 33 samples thee most likely date of the zoonotic event giving rise to the HIV-2 subtype A epidemic waswas estimated to be 1940 6 years) [26].
Assumingg that one person or a few persons became infected with HIV-2 clade A throughh contact with a sooty mangabey, it is still unclear how this led to an epidemic. It iss theoretically possible that the virus used to be more virulent than it is now, but this is unlikely:: there is no record of an epidemic of slim disease in West Africa prior to 1985. So,, if we assume that the virulence of the virus has not substantially changed over time, theree must have been an increase in transmission that amplified a small outbreak into a largee epidemic. This could have been due to an increase in unscreened blood transfusionss on a large scale, high rates of sexual partner change, many concurrent sexual partnershipss or presence of co-factors enhancing sexual transmission like sexually transmittedd infections (STI's), or unsterile injections. Several of these factors may have beenn present at the same time in Guinea-Bissau.
Fromm 1963 to 1974 a bloody war of independence was fought, and a large colonial Portuguesee army was present in Guinea-Bissau. During the war many blood transfusions mayy have been given. Commercial sex may have flourished around the barracks, and STI prevalencee was probably high; this is usually seen in wars. In a cross-sectional study amongg older residents of Bissau it was found that sex with a white man was an independentt risk factor for women for HIV-2 infection [27]. As there were very few whitee men in the country after 1974, this was seen as a proxy for sex with a soldier. Thosee having served in the army and having had blood transfusions in the 1960s tended too have a higher prevalence of HIV-2 [27]. The authors concluded that the independencee war may have been critical for the propagation of HIV-2 through sex work andd blood transfusions, and that, once established, sexual relations maintained the epidemic.. This could also explain why HIV-2 is only really common in Guinea-Bissau (upp to 8% of adults) and nowhere else. In most of West Africa there was peace in die 1960'ss and 70's, with the notable exception of Nigeria in 1967-70. The hypothesis that thee war was responsible for the amplification of the epidemic fits also with the
165 5 HIV-22 IN WEST AFRICA observationss in Caio and Bissau that the highest prevalence is found in people who were youngg adults in the war years, and that the prevalence in later generations is lower. The transmissionn of the virus may not be efficient enough to maintain ongoing epidemic spreadd in the absence of important amplifiers like frequent commercial sex, high levels of STI's,, and unscreened blood transfusions.
Marxx hypothesised that reuse of unsterilised needles may have been responsible for both thee HIV-1 and HIV-2 epidemics [15]. In West Africa various mass vaccination and treatmentt campaigns against yaws, yellow fever, and small pox were conducted in the finall decades of the colonial era [28], and these may have been responsible for mass inoculationss with HIV-2. There is no proof for this and it does not explain why HIV-2 becamee epidemic in Guinea-Bissau and nowhere else.
Riskk factors
Chapterr 3 examined risk factors associated with prevalent HIV-2 infection among pregnantt women in The Gambia. This study was limited by the sentinel surveillance formatt and could examine very few risk factors. It found recruitment at one particular sitesite in The Gambia (Sibanor) and higher parity to be significantiy associated with HIV-2 infection.. Parity can be interpreted as a proxy for years of sexual experience, as use of contraceptivess is very low in The Gambia [29]. Chapter 4 examined risk factors for incidentt infection among adults in Caio. Even thought the study identified only 51 sero- incidentt HIV-2 infections, it is the world's largest study of incident HIV-2. It found that bloodd transfusions after 1989 were not a risk factor, and all persons who became infected hadd had sexual intercourse. Among women mobility, having been divorced, having had vaginall discharge, having serologically defined active syphilis were significant independentt risk factors, confirming earlier cross-sectional studies identifying these as riskrisk factors [1, 12, 30-38]. These risk factors are similar to those identified for HIV-1 infection.. Having drunk alcohol on the day of the interview was also significantiy associatedd with HIV-2 infection, as was a history of injections in the 12 months
166 6 HIV-22 IN WEST AFRICA precedingg the interview and sample. The latter may have been a cause or a consequence off HIV-2 infection, so caution is needed in interpreting this association. In men, age beloww 30 years and serological evidence of active syphilis were significant risk factors for HIV-2.. In men injections outside a clinic setting were also significandy associated with HIV-22 infection. The same caution as mentioned earlier for women is needed to interprett this finding.
HIV-11 and HIV-2 interactions
DualDual infection Afterr type-specific antibody tests became available, it became clear that both HIV-1 and HIV-22 circulated in West Africa. Samples of some people in West Africa showed dual serologicall reactivity, and it was not clear whether this was mainly due to antibody cross- reactivity,, dual infection, or an infection with a third, unknown virus [39]. Quite early it waswas demonstrated by PCR that dual infection wim bom HIV-1 and HIV-2 did occur [40, 41].. Later improvements in PCR techniques showed that a large proportion of people withh dually reactive samples (up to 86%) is truly dually infected [42, 43].
NoNo protection against HIV-1 infection Inn 1995 Travers et al. reported that HIV-2 seemed to offer protection against subsequent HIV-11 infection in a cohort of commercial sex workers in Dakar [44]; this caused excitementt and hope [45]. Several research groups in Guinea-Bissau [8, 46], Cote d'lvoiree [47, 48] and our group in The Gambia [49] examined this putative effect in other cohorts.. None of die seven analyses so far have been able to reproduce this finding. Onee of our own studies (Chapter 5) even found that HIV-2 was a significant risk factor forr HIV-1 infection. An editorial in AIDS concluded that the available epidemiological dataa do not support a protective effect of HIV-2 infection upon incident HIV-1 infection [50].. The investigators of the cohort in which the original finding was made, have providedd no updates of the effect in that cohort since 1999 [51-53], so it is unknown whetherr the effect in that cohort persisted, or declined over time, or even reversed.
167 7 HIV-22 IN WEST AFRICA
NoNo protection against disease progression Gambiann clinicians had the impression that HIV-1 and HIV-2 dually (HIV-D) infected patientss had a more rapid disease progression than HIV-1 singly infected patients. Others,, noting a lower plasma HIV-1 viral load in dually infected patients, [54] hypothesisedd a partially protective immune response of one virus against the other, presumablyy leading to a milder disease course. Some studies assessed survival of HIV patientss with TB, and found that the mortality rates among HIV-D infected patients weree significantly higher than among HIV-2 infected patients and similar to those among HIV-11 patients [55-57]. These studies were restricted to patients with TB, and the follow-upp periods were short. In a community-based study in Bissau, no significant differencee in mortality between HIV-1, HIV-2 and HIV-D infected subjects was found, butt the numbers with HIV-1 and HIV-D in this study were very small (26 and 20, respectively)) [58].
Chapterr 6 is the first published long-term study analysing the survival and mortality of subjectss with dual HIV infection, regardless of a TB diagnosis. Among patients of the genito-urinaryy (GU) clinic in Fajara, The Gambia, the mortality rate of dually infected patientss was similar to that of HIV-1 infected patients, and worse than that of HIV-2 infectedd patients. This was true overall, and after adjusting for baseline CD4 count. A studyy among a subset of this cohort found that the plasma viral load of HIV-D infected patientss was similar to that in HIV-1 singly infected patients [59]. These data do not supportt suggestions that HIV-2 infection could mitigate the course of HIV-1 infection.
HIV-1/HIV-2HIV-1/HIV-2 recombination Iff a person is infected with two or more subtypes of HIV-1, these can recombine their genomess to form new strains of HIV-1, and these can be transmitted [60-63]. Some of thee recombinant strains are successful in spreading, e.g. CRF01_AE in Thailand and CRF02_AGG in West Africa [64]. Recombinations of the genetically rather distant groups OO and M have been described [65-67], but so far no recombinations of HIV-1 and HIV- 22 have been reported. It is unclear whether this is biologically possible or plausible [68,
168 8 HIV-22 IN WEST AFRICA
69]] and if so, what the epidemic potential of such a recombinant would be. A search for possiblee HIV-1/HIV-2 recombinant should be done in populations where both viruses circulatee in relatively high frequencies, e.g. in the general population of Bissau [8] or Caio [Chapterr 4], or among commercial sex workers in Abidjan [48].
Mortalityy of HIV-2 infected subjects
Onee of the first epidemiological studies on HIV-2, conducted in Bissau in 1987-1988 [31],, showed that the mortality associated with HIV-2 infection was much lower than thatt usually found in HIV-1 infection. All subsequent studies have confirmed this [TO- TS].. Chapters 6 and 7 examined adult all-cause mortality in HIV-2 infected people and comparedd this with that found in HIV-1 infected people. The analysis in the seroprevalentt clinical cohort [Chapter 6] demonstrated a lower mortality in HIV-2 patientss than in HIV-1 patients, but found that this lower mortality was limited to those withh a normal CD4 count (> 500/jJ.l). Among those with a CD4 count < 200/^1 the mortalityy rate was similar between HIV-1 and HIV-2 infected subjects. This could be explainedd in two ways. The first possibility is that all HIV-2 infected subjects experience aa deterioration of their immune system, but this decline is slower than in HIV-1 infection.. Once the CD4 count has declined to < 200/(il, patients are at high risk of fatall opportunistic infections and diere is no difference in mortality; only the time to reachh this low CD4 is longer for HIV-2 infected than for HIV-1 infected subjects. The otherr possibility is that those with HIV-2 infection fall in either of two categories: those whosee immune system is not affected at all by the infection, and those whose immune systemm is damaged by the infection, at a rate similar to HIV-1.
Iff the first explanation is right we would expect that old people with HIV-2 would have lowerr CD4 counts than young people, as on average they will have been infected for longer.. There was some indication of lower CD4% in older HIV-2 infected individuals inn Caio (r = -.017; p=0.06) [73], but a study from Bissau did not find this (r = -.0038; p>0.05)) [74]. This could be examined in other cohorts in West Africa.
169 9 HIV-22 IN WEST AFRICA
Iff the second explanation is correct, then we would expect a relatively high excess mortalityy among recent seroconverters, and a limited excess mortality among people who seroconvertedd long ago. There are too few people with known date of seroconversion to analysee this, but if age is taken as a proxy for time since infection, then this can be examinedd in a seroprevalent cohort. Two studies so far found that the mortality rate ratioo of HIV-2 infected vs. HIV uninfected subjects was higher in the young (<45 years) thann in older people and that in older people the rate ratio was not significandy different fromm one [70, 71]. Due to a higher background mortality in older people the rate ratio is expectedd to decline even if the rate difference between infected and uninfected subjects is similarr to that observed in younger age groups. This question is unresolved, and we are currentlyy investigating this with a survival analysis over a 12-year period in the Caio population. .
Rolee of plasma viral load
Bothh sexual transmission [75] and mother-to-child transmission (MTCT) of HIV-1 are to aa large degree determined by plasma viral load [76]. Transmission rates of HIV-2 are lowerr than of HIV-1 [Chapter 2]. A large Gambian study showed that the lower transmissionn from mothers to babies can be explained by the difference in plasma viral loadd between the two infections [77]. It is likely that the same holds true for heterosexual transmission,, but no data have been published so far to support this.
Severall studies, both in sero-incident and in seroprevalent cohorts, have shown that baselinee plasma viral load is an important independent predictor for survival in HIV-1 [78-80].. Survival with HIV-2 infection is better than with HIV-1 [31, 70-72, 81-84, chapterr 6]. Chapter 7 shows that this can be explained by the lower plasma viral load in HIV-2:: plasma viral load at baseline was significandy associated with the mortality rate in HIV-22 infected women. In a multivanable model of all HIV infected women (both thosee with HIV-1 and those with HIV-2) that included plasma viral load, CD4%, HIV-
170 0 HIV-22 IN WEST AFRICA typee and age, virus type per se was not a significant predictor of mortality, but plasma viral loadd was. This is in agreement with findings of a clinic-based study from Senegal, which foundd that plasma viral load, regardless of HIV type, predicts the rate of CD4 cell decline [85].. These data indicate that plasma viral load is a more important marker for disease progressionn man HIV type.
Hostt and Virus
Thee lower transmission, morbidity, and mortality associated with HIV-2 can thus in part bee explained by the lower plasma viral load. HIV-2 infection can occasionally lead to
veryy high plasma viral loads in humans, but this is rare [59, 77]. SIVsm, the presumed recentt ancestor of HIV-2, grows to very high titres in its natural host without causing a
markedd CD4 cell depletion or disease [86-88]. Infection with SIVsm leads to both humorall and cellular immune responses in sooty mangabeys, but these do not reduce the plasmaa viral load to low or undetectable levels [87]. This indicates that a high plasma virall load alone is not enough to lead to immunodeficiency; differential host responses to SIV/HIVV infection are important as well [87, 89].
Itt has been hypothesised that the reason for the generally lower plasma viral load is that HIV-22 induces a more vigorous and effective immune response than HIV-1 [90]. Based onn this hypothesis our group has conducted detailed studies to map differences in proliferativee responses and cytotoxic T lymphocyte (CTL) responses between HIV-1 and HIV-22 infected people at similar, moderate levels of immunodeficiency; so far no obviouss differences have been found [unpublished studies, 91].
Integratedd viral DNA (provirus) is the source of all plasma virions. HIV-2 infected patientss have DNA viral loads similar to those in HIV-1 patients [85, 92-94], but the plasmaa concentration of virions is lower in HIV-2 infection [95]. This could be explainedd in several ways. Perhaps a larger proportion of HIV-1 proviruses is actively replicatingg [85]. Another possibility is that in HIV-1 more DNA is integrated and
171 1 HIV-22 IN WEST AFRICA replication-competentt compared to HIV-2 [85]. It is also possible that in HIV-1 infectionn increased proviral DNA levels exist in other compartments man blood [85]. Finally,, it may be that HIV-2 virions are cleared more efficiendy.
HIV-22 in children
Chapterr 8 presents the first long-term observational study of children with perinatally acquiredd HIV-2 infection. The study reports all-cause mortality but was not able to comparee morbidity patterns. The median follow-up time was 6.6 years. Three out of eightt HIV-2 infected children died (38%) compared to 12 out of 17 HIV-1 infected childrenn (71%) and 40 out of 448 children of HIV uninfected mothers (9%). The mortalityy rate of HIV-2 infected children was significandy higher than that of uninfected childrenn (p=0.02), but the difference with HIV-1 infected children did not reach statisticall significance (p=0.08). The numbers are small and it is hard to draw firm conclusions.. Survival among children with HIV-2 infection may be similar or better to thatt in HIV-1 infected children. Mortality among children with HIV-2 appears to be higherr than in seronegative children. These data suggest that children with HIV-2 infectionn need the same care as HIV-1 infected children. Unfortunately specific medical caree for HIV infected children in Africa, the part of the world with >90% of all cases [96],, is almost non-existent.
Treatment t
Nonee of the studies in this thesis have examined treatment of patients with HIV-2, and nonee of the participants in these studies were treated with antiretrovirals. The recent debatee and activism about access to treatment for HIV infection has brought this treatmentt closer to reality for patients in Africa. There are few data on treatment of HIV-22 with antiretrovirals from the developed world [97-102], and some studies from pilott projects of antiretroviral treatment in Africa [103, 104]. From pharmacological
172 2 HIV-22 IN WEST AFRICA studiess it is known that nevkapine and other non-nucleoside reverse transcriptase inhibitorss are not effective against HIV-2 [105,106]. HIV-2 may be more sensitive to somee protease inhibitors [107]. Resistance to antiretroviral drugs may develop [100, 108, 109],, sometimes due to the same mutations that confer resistance to HIV-1 strains [99, 108]. .
Theree are no guidelines for treatment of HIV-2. Current guidelines for the treatment of HIV-11 in adults advise to start highly active anti-retroviral therapy (HAART) if the patientt is symptomatic or if the CD4 count is < 200 cells/jil [109-111]. Asymptomatic patientss with a CD4 count of more than 350 cells/p.1, should defer treatment. There is a greyy zone for asymptomatic patients with CD4 counts between 200 and 350 cells/|il. In thatt range treatment decisions depend on die rate of decline of CD4 count, the potential forr adherence, and the plasma viral load. If the PVL is higher than 50,000 - 100,000 copies/mll treatment may be considered irrespective of the CD4 count or presence of symptomss [110].
Inn view of the similar mortality rates of HIV-1 or HIV-2 infected patients with low CD4 countss (<200 cells/jil), [chapter 6] and the predictive power of plasma viral load for CD4 declinee [85] and mortality in HIV-2 [chapter 7], it is proposed to use the same criteria for thee start of HAART in HIV-2 as in HIV-1. In practice this would mean diat only a minorityy of HIV-2 infected persons in the community would need treatment. Based on thee CD4 and PVL data from the perinatal study (chapter 7), and ignoring symptomatology,, 5% of HIV-2 infected women would qualify for treatment at recruitment.. Based on data of a community-based study in Caio, Guinea-Bissau [73], 7% off HIV-2 infected adults would qualify for treatment. (In a clinic setting, where more subjectss with advanced disease are found, these numbers are different. Data from the GUU clinic in The Gambia [59, Chapter 6] indicate that 43% of newly registered HIV-2 infectedd patients meet the CD4 and PVL criteria for treatment). As more trials and projectss on HAART are started in Africa, [112] studies should be done comparing the treatmentt response of HIV-2 infected patients with that of HIV-1 infected patients. This willl help formulating rational policies for management of patients wiüi HIV-2, and may
173 3 HIV-22 IN WEST AFRICA providee valuable insights into the mechanisms of immune reconstitution in HIV-2 infection. .
Vaccines s
Africaa needs an AIDS vaccine (113]. In spite of increased investment into vaccine research,, including clinical trials in Africa (Uganda and Kenya) an AIDS vaccine is still farr away [113, 114]. It is possible that vaccines need to be specific to the clade of the virus,, and perhaps even to the HLA type of the host. It is doubtful whether sterilising immunityy is feasible at all, so a vaccine preventing disease, not infection, may be the best achievable. .
Inn view of the above, it is unlikely that a vaccine against HIV-1 infection or disease wouldd prove effective against HIV-2. This means that HIV-2 specific vaccines are needed.. Large trials are needed to demonstrate the effectiveness of a preventative vaccinee [115]. Because the incidence of HIV-2 is low, even in high prevalence areas
[Chapterr 4], a preventative HIV-2 vaccine trial needs to be quite large. Assuming a high incidencee of 5 per 1000 pyo [Chapter 4], a trial that is able to show at least a 75% efficacy againstt HIV-2 infection needs to follow up over 3000 subjects for three years (see Table
1).. If a lower incidence rate is assumed, which is more realistic in view of the observed trends,, and a higher power and shorter follow-up period are chosen, the trial size increasess to nearly 10,000 subjects. That is a large and very expensive trial.
Remainingg questions and further research
Soonn after HIV-2 was first described it became clear that this virus was less pathogenic andd less transmissible than HIV-1, and that its spread and potential spread were more limited.. Since then attention for the HIV-2 epidemic has waned. Among the more than
8,0000 abstracts of the Barcelona AIDS conference in 2002, less than 1% contained the
174 4 HIV-22 IN WEST AFRICA
Tablee 1. Sample sizes for HIV-2 preventative vaccine trial
Expectedd Power HIV-2 incidence HIV-2 incidence in Length of Total vaccinee rate in control intervention group, follow-up number efficacyy group, per 1000 per 1000 pyo (years) needed pyo o
90% % 80% % 5 5 0.5 5 3 3 1950 0 80% % 80% % 5 5 1.0 0 3 3 2615 5 75% % 80% % 5 5 1.25 5 3 3 3058 8 70% % 80% % 5 5 1.5 5 3 3 4493 3
90% % 80% % 5 5 0.5 5 2 2 2791 1 80% % 80% % 5 5 1.0 0 2 2 3742 2 75% % 80% % 5 5 1.25 5 2 2 4376 6 70% % 80% % 5 5 1.5 5 2 2 5159 9
90% % 80% % 3 3 0.3 3 3 3 3249 9 80% % 80% % 3 3 0.6 6 3 3 4354 4 75% % 80% % 3 3 0.75 5 3 3 5092 2 70% % 80% % 3 3 0.9 9 3 3 7479 9
90% % 80% % 3 3 0.3 3 2 2 4655 5 80% % 80% % 3 3 0.6 6 2 2 6239 9 75% % 80% % 3 3 0.75 5 2 2 7295 5 70% % 80% % 3 3 0.9 9 2 2 8599 9
90% % 90% % 3 3 0.3 3 2 2 5976 6 80% % 90% % 3 3 0.6 6 2 2 8061 1 75% % 90% % 3 3 0.75 5 2 2 9453 3 70% % 90% % 3 3 0.9 9 2 2 11,175 5
Alll scenarios are based on 0.05 significance level, 10% loss to follow up per year, and equal numberss of subjects in intervention and control arms, pyo = person-years of observation.
wordd HIV-2. This does not mean that all relevant research has been done. I conclude withh four important questions that are worthwhile to be examined. The first one is of historicc public health relevance, two others impinge on retroviral pathogenesis in general, andd the fourth concerns the clinical management of HIV-2.
1.. Why did a zoonotic event lead to an HIV-2 epidemic in Guinea-Bissau and whyy is this epidemic in decline now?
Inn recent decades several animal pathogens jumped the species barrier and caused epidemicss in humans (among others: HIV-1, HIV-2, the corona virus causing
175 5 HIV-22 IN WEST AFRICA
SARS,, Ebola virus, prions). In the case of HIV-1 and HIV-2 some widely discussedd hypotheses have held medical interventions responsible for the epidemics.. It seems important to elucidate the origin of these epidemics, whether thatt means confirming or rejecting these hypotheses. Larger scale phylogenetic analysess than have been done so far, and epidemic modelling studies that try to fitfit the existing data can contribute to answering this question.
2.. Which proportion of people that are HIV-2 infected develop immunodeficiency,, or AIDS, and die prematurely?
Thiss proportion needs to be known in order to better understand the pathogenesiss of HIV-2 infection, to inform patients about their prognosis, and to helpp identify factors that may determine non-progression. Long-term follow-up off sero-converters is needed to answer mis question. There are few such cohorts andd all are small; only collaboration between research groups in West Africa couldd help to answer this question.
3.. Why does HIV-2 infection usually not lead to high plasma viral loads, in spite off proviral loads similar to HIV-1? Is this due to aspects inherent to the virus, or iss it due to a more efficient immune response? Thiss question could be examined by detailed comparative virological and immunologicall studies in populations where both viruses circulate.
4.. Is the virological, immunological, and clinical response of HIV-2 infected peoplee to highly active antiretroviral therapy similar to that in HIV-1 infected people? ?
Inn order to establish effective, evidence-based treatment regimens for HIV-2 disease,, clinical trials and cohort studies of antiretroviral therapy should be conductedd in West Africa, and in Portugal or France.
176 6 HIV-22 IN WEST AFRICA
References s
1.. Pepin J, Dunn D, Gaye I, et al. HIV-2 infection among prostitutes working in The Gambia:: association with serological evidence of genital ulcer diseases and with generalizedd lymphadenopathy. AIDS 1991;5:69-75. 2.. Pepin J, Morgan G, Dunn D, et al. HIV-2-induced immunosuppression among asymptomaticc West African prostitutes: evidence that HIV-2 is pathogenic, but less so thann HIV-1. AIDS 1991;5:1165-72. 3.. Wilkins A, Oelman B, Pepin J, et al. Trends in HIV-1 and HIV-2 infection in The Gambia.. AIDS 1991;5:1529-30. 4.. Kanki PJ, Travers KU, Mboup S et al. Slower heterosexual spread of HIV-2 than HIV- 1.. Lancet 1994, 343:943-946. 5.. Hawkes S, West B, Wilson S, et al. Asymptomatic carriage of Haemophilus ducreyi confirmedd by the polymerase chain reaction. Genitourinary Medicine 1995, 71:224- 227. . 6.. Greenberg A, Coulibaly IM, Kadio A et al. Trends in the HIV-1 and HIV-2 epidemics in Abidjan,, Cote d'lvoire: 11 years of HIV serosurveillance at Project Retro-CI. X Internationall Conference on AIDS and STD in Africa. Abidjan, December 1997 [Abstractt B.041]. 7.. Peeters M, Koumare B, Mulanga C et al. Genetic subtypes of HIV type 1 and HIV typee 2 strains in commercial sex workers from Bamako, Mali. AIDS Research & Humann Retroviruses 1998, 14:51-58. 8.. Norrgren H, Andersson S, Biague AJ, et al. Trends and interaction of HIV-1 and HIV- 22 in Guinea-Bissau, west Africa: no protection of HIV-2 against HIV-1 infection. AIDS 1999;; 13: 701-707. 9.. De Cock KM, Adjorlolo G, Ekpini E et al. Epidemiology and transmission of HIV-2. Whyy there is no HIV-2 pandemic. JAMA 1993, 270:2083-2086. 10.. Djomand G, Greenberg AE, Sassan-Morokro M et al. The epidemic of HIV/AIDS in Abidjan,, Cote d'lvoire: a review of data collected by Projet RETRO-CI from 1987 to 1993.. J Acquir Immune Defic Syndr Hum Retroviral 1995, 10:358-365. 11.. Mabey DC, Tedder RS, Hughes AS et al. Human retroviral infections in The Gambia: prevalencee and clinical features. British Medical Journal 1988, 296:83-86. 12.. Pepin J, Quigley M, Todd J, et al. Association between HIV-2 infection and genital ulcerr diseases among male sexually transmitted disease patients in The Gambia. AIDSS 1992;6:489-93. 13.. Larsen O, da Silva Z, Sandstrom A, et al. Declining HIV-2 prevalence and incidence amongg men in a community study from Guinea-Bissau. AIDS 1998, 12:1707-1714. 14.. Hirsch VM, Olmsted RA, Murphy-Corb M, et al. An African primate lentivirus (SIVsm) closelyy related to HIV-2. Nature 1989; 339; 389-92. 15.. Marx PA, Alcabes PG, Drucker E. Serial human passage of simian immunodeficiency viruss by unsterile injections and the mergence of epidemic human immunodeficiency viruss in Africa. Phil Trans R Soc London B 2001; 356: 911-920. 16.. Gao F, Yue L, White AT, et al. Human infection by genetically diverse SIVsm-related HIV-22 in West Africa. Nature 1992, 358:495^99. 17.. Chen Z, Telfer P, Gettie A et al. Genetic characterization of new West African simian immunodeficiencyy virus SIVsm: geographic clustering of household- derived SIV strainss with human immunodeficiency virus type 2 subtypes and genetically diverse virusess from a single feral sooty mangabey troop. J Virol 1996; 70:3617-3627. 18.. Chen Z, Luckay A, Sodora DL et al. Human immunodeficiency virus type 2 (HIV-2) seroprevalencee and characterization of a distinct HIV-2 genetic subtype from the naturall range of simian immunodeficiency virus-infected sooty mangabeys. J Virol 1997;71:3953-3960. .
177 7 HIV-22 IN WEST AFRICA
19.. Yamaguchi J, Devare SSG, Brennan CA. Identification of a new HIV-2 subtype based onn phylogenetic analysis of full-length genomic sequence. AIDS Research and Humann Retroviruses 2000; 16: 925-930. 20.. Gao F, Bailes E, Robertson DR, et al. Origin of HIV-1 in the chimpanzee Pan troglodytess troglodytes. Nature 1999, 397:436-441. 21.. Hahn BH, Shaw GM, De Cock KM, Sharp PM. AIDS as a zoonosis: scientific and publicc health implications. Science 2000;; 287: 607-614. 22.. Hooper E. The River: a journey back to the source of HIV and AIDS. London: Allen Lanee and Penguin Press, 1999. 23.. Weiss RA. Natural and iatrogenic factors in human immunodeficiency virus transmission.. Phil Trans R Soc London B 2001; 356: 947-953. 24.. Korber B, Muldoon M, Theiler J, et al. Timing the ancestor of the HIV-1 pandemic strains.. Science 2000; 288: 1789-96. 25.. Yusim K, Peeters M, Pybus 0, et al. Using human immunodeficiency virus type 1 sequencess to infer historical features of the acquired immunodeficiency syndrome epidemicc and human immunodeficiency virus evolution. 2001. Phil Trans R Soc Londonn B 2001; 356: 855-866. 26.. Lemey P, Pybus OG, Wang B, et al. Tracing the roigin and history of the HIV-2 epidemic.. PNAS 2003; 100; 6588-92. 27.. Poulsen AG, Aaby P, Jensen H, Dias F. Risk factors for HIV-2 seropositivity among olderr people in Guinea-Bissau. A search for the early history of HIV-2 infection. Scandd J Infect Dis 2000; 32:169-75. 28.. Antal GM, Causee G. The Control of endemic treponematoses. Reviews of Infectious Diseasee 1985; 7 (Suppl 2):S220-S226. 29.. United Nations Population Fund. Recommendation by the Executive Director; Assistancee to the Government of The Gambia, http:// www.unfpa.org/ africa/ gambia/ 1gam0206.pdf.. [accessed 29 May 2003], 30.. Dufoort G, Courouce AM, Ancelle-Park R, Bletry O: No clinical signs 14 years after HIV-22 transmission via blood transfusion [letter]. Lancet 1988, 2:510. 31.. Poulsen AG, Kvinesdal B, Aaby P, et al. Prevalence of and mortality from human immunodeficiencyy virus type 2 in Bissau, West Africa. Lancet 1989;1:827-31 32.. Wilkins A, Hayes R, Alonso P, et al. Risk factors for HIV-2 infection in The Gambia. AIDSS 1991;5:1127-32. 33.. Naucler A, Albino P, Da Silva AP, et al. HIV-2 infection in hospitalised patients in Bissua,, Guiena-Bissau. AIDS 1991; 5: 301-4. 34.. Kanki P, M'Boup S, Marlink R et al.: Prevalence and risk determinants of human immunodeficiencyy virus type 2 (HIV-2) and human immunodeficiency virus type 1 (HIV-1)) in west African female prostitutes. Am J Epidemiol 1992, 136:895-907. 35.. Naucler A, Albino P, Da Silva AP, Biberfeld G. Sexually transmitted disease and sexuall beahviour as risk factors for HIV-2 infection in Bissau, Guinea-Bissau. Int J STDD AIDS 1993; 4:217-21. 36.. Poulsen AG, Aaby P, Gottschau A, et al.: HIV-2 infection in Bissau, West Africa, 1987-1989:: incidence, prevalences, and routes of transmission. J Acquir Immune Deficc Syndr 1993, 6:941-948. 37.. Abbott RC, NDour-Sarr A, Diouf A, et al.: Risk factors for HIV-1 and HIV-2 infection in pregnantt women in Dakar, Senegal. J Acquir Immune Defic Syndr 1994, 7:711-717. 38.. Sassan-Morokro M, Greenberg AE, Coulibaly IM et al.: High rates of sexual contact withh female sex workers, sexually transmitted diseases, and condom neglect among HIV-infectedd and uninfected men with tuberculosis in Abidjan, Cote d'lvoire. J Acquir Immunee Defic Syndr Hum Retrovirology 1996,11:183-187. 39.. De Cock KM, Brun-Vezinet F, Soro B. HIV-1 and HIV-2 infections and AIDS in West Africa.. AIDS 1991; 5 (suppl 1): S21-S28. 40.. Evans LA, Moreau J, Odehouri K, et al. Simultaneous isolation of HIV-1 and HIV-2 fromm an AIDS patient. Lancet 1988; 2; 1389-91.
178 8 HIV-22 IN WEST AFRICA
41.. Rayfield M, De Cock K, Heyward W, et al. Mixed human immunodeficiency virus (HIV) infectionn in an individual: demonstration of both HIV type 1 and type 2 proviral sequencess by using polymerase chain reaction. J infect Dis 1988; 158:1170-6. 42.. Ishikawa K, Fransen K, Ariyoshi K et al. Improved detection of HIV-2 proviral DNA in duallyy seroreactive individuals by PCR. AIDS 1998, 12:1419-1425. 43.. Walther-Jallow L, Andersson S, Da Silva Z, Biberfeld G. High concordance between polymerasee chain reaction and anitbody testing of specimens from individuals dually infectedd with HIV types 1 and 2 in Guinea-Bissau. AIDS Res Hum Retroviruses 1999, 15:957-962. . 44.. Travers K, Mboup S, Marlink R et al.: Natural protection against HIV-1 infection providedd by HIV-2. Science 1995, 268:1612-1615. 45.. Cohen J. Can one type of HIV protect against another type? [news; comment]. Sciencee 1995; 268:1566-1566. 46.. Aaby P, Poulsen AG, Larsen O et al. Does HIV-2 protect against HIV-1 infection? [letter].. AIDS 1997, 11:939-940. 47.. Wiktor SZ, Nkengasong JN, Ekpini ER, et al. Lack of protection against HIV-1 infectionn among women with HIV-2 infection. AIDS 1999; 13: 695-699. 48.. Ghys PD, Diallo MO, Ettiegne-Traore V, et al. Effect of interventions to control sexuallyy transmitted disease on the incidence of HIV infection in female sex workers. AIDSS 2001;15:1421-31. 49.. Ariyoshi K, Schim van der Loeff M, Sabally S, et al. Does HIV-2 infection provide cross-protectionn against HIV-1 infection? [letter]. AIDS 1997,11:1053-1054. 50.. Greenberg AE. Possible protective effect of HIV-2 against incident HIV-1 infection: revieww of available epidemiological and in vitro data. AIDS 2001; 15: 2319-2321. 51.. Kanki PJ, Eisen G, Travers KU et al. HIV-2 and natural protection against HIV-1 infection.. Science 1996, 272:1959-1960. 52.. Travers K, Eisen G, Hsieh C et al. HIV-2 provides natural protection against HIV-1 infection.. Xth International Conference on AIDS and STDs in Africa. Abidjan, Decemberr 1997 [Abstract A.072]. 53.. Kanki PJ. Human Immunodeficiency Virus Type 2 (HIV-2). AIDS Rev 1999; 1: 101- 108. . 54.. Andersson S, Norrgren H, da Silva Z, et al. Plasma viral load in HIV-1 and HIV-2 singlyy and dually infected individuals in Guinea-Bissau, West Africa. Arch Intern Med 2000,, 160:3286-3293. 55.. Kassim S, Sassan-Morokro M, Ackah A, et al. Two-year follow-up of persons with HIV-11 and HIV-2 associated pulmonary tuberculosis treated with short-course chemotherapyy in West Africa. AIDS 1995; 9: 1185-91. 56.. Norrgren H, Bamba S, Da Silva ZJ, et al. High mortality and severe immunosuppressionn in hospitalized patients with pulmonary tuberculosis and HIV-2 infectionn in Guinea-Bissau. Scan J Infect Dis 2001; 33: 450-456. 57.. Seng R, Gustafson P, Gomes VF, et al. Community study of the relative impact of HIV-11 and HIV-2 on intrathoracic tuberculosis. AIDS 2002; 16: 1059-1066. 58.. Holmgren BG. Age and gender patterns of HIV-1, HIV-2, and HTLV-I in Guinea- Bissau.. Epidemiological and behavioural aspects. PhD thesis. Copenhagen: Universityy of Copenhagen, 2002. 59.. Alabi AS, Jaffar S, Ariyoshi K, et al. Plasma viral load, CD4 % and survival of HIV-1, HIV-2,, and dually infected Gambian patients. AIDS 2003; 17: 1513-1520. 60.. Blackard JT, Cohen DE, Mayer KH. Human immunodeficiency virus superinfection andd recombination: current state of knowledge and potential clinical consequences. Clinn Infect Dis 2002; 34:1108-14. 61.. Jost S, Bernard M-C, Kaiser L, et al. A patient with HIV-1 superinfection. N Engl J Medd 2002; 347: 731-6. 62.. Goulder PJR, Walker BD. HIV-1 superinfection - a word of caution. N Engl J Med 2002;; 347: 756-8. 63.. Levy JA. Is HIV superinfection worrisome? Lancet 2003; 361: 98-9.
179 9 HIV-22 IN WEST AFRICA
64.. McCutchan FE. Understanding the genetic diversity of HIV-1. AIDS 2000; 14 (suppl 3): S31-S44. . 65.. Takehisa J, Zekeng L, Miura T, et at. Triple HIV-1 infection with group O and group M off different clades in a single Cameroon patient. J AIDS Hum Retroviral 1997; 14: 81- 2 2 66.. Takehisa J, Zekeng L, Ido E, et al. Human immunodeficiency virus type 1 intergroup {M/O)) recombination in Cameroon. J Virol 1999; 73: 6810-20. 67.. Peeters M, Liegeois F, Torimori N, et al. Characterisation of a highly replicative intergroupp M/O human immunodeficiency virus type I recombinant isolated from a Camerooniann patient. J Virol 1999; 73: 7368-75. 68.. Kaye JF, Lever AML. Nonreciprocal packaging of human immunodeficiency virus type 11 and type 2 RNA: a possible role for the p2 domain of gag in RNA encapsidation. J Viroll 1998; 72: 5877-5885. 69.. Kaye JF, Lever AM. Human immunodeficiency virus types 1 and 2 differ in the predominantt mechanism used for selection of genomic RNA for encapsidation. J Virol 1999;73:3023-31. . 70.. Ricard D, Wilkins A, N'Gum PT, et al. The effects of HIV-2 infection in a rural area of Guinea-Bissau.. AIDS 1994;8:977-82. 71.. Poulsen AG, Aaby P, Larsen O et al. 9-year HIV-2-associated mortality in an urban communityy in Bissau, west Africa. Lancet 1997, 349:911-914. 72.. Norrgren H, Da Silva ZJ, Andersson S, et al. Clinical features, immunological changess and mortality in a cohort of HIV-2-infected individuals in Bissau, Guinea- Bissau.. Scan J Infect Dis 1998, 30:323-329. 73.. Berry N, Jaffar S, Schim van der Loeff M, et al. Low level viremia and high CD4% predictt normal survival in a cohort of HIV type-2-infected villagers. AIDS Res Hum Retrovirusess 2002; 18:1167-73. 74.. Lisse IM, Poulsen AG, Aaby P, et al. Serial CD4 and CD8 T-lymphocyte counts and associatedd mortality in an HIV-2 infected population in Guinea-Bissau. J Acquir Immunee Defic Syndr 1996; 13:355-362. 75.. Quinn TC, Wawer MJ, Sewankambo N, et al. Viral load and heterosexual transmissionn of human immunodeficiency virus type 1. N Engl J Med 2000;342:921-9. 76.. Dickover RE, Garatty EM, Hermann SA, et al. Identification of levels of maternal HIV- 11 RNA associated with risk of perinatal transmission. JAMA 1996; 275: 599-605. 77.. O'Donovan D, Ariyoshi K, Milligan P, et al. Maternal plasma viral RNA levels determinee marked differences in mother-to-child transmission rates of HIV-1 and HIV- 22 in The Gambia. MRC/Gambia Government / University College London Medical Schooll working group on mother-child transmission of HIV. AIDS 2000;14:441-8. 78.. Mellors JW, Kingsley LA, Rinaldo Jr. CR, et al. Quantitation of HIV-1 RNA in plasma predictss outcome after seroconversion. Annals of Internal Medicine 1995; 122: 573- 579. . 79.. Mellors JW, Rinaldo Jr. CR, Gupta, et al. Prognosis in HIV-1 infection predicted by thee quantity of virus in plasma. Science 1996; 272:1167-1170. 80.. Mellors, JW, Munoz A, Giorgi JV, et al. Plasma viral load and CD4+ lymphocytes as prognosticc markers of HIV-1 infection. Annals of Internal Medicine 1997; 126: 946- 954 4 81.. Borgdorff MW, Barongo LR, Klokke AH, et al. HIV-1 incidence and HIV-1 associated mortalityy in a cohort of urban factory workers in Tanzania. Genitourin Med 1995; 71: 212-5. . 82.. Nunn AJ, Mulder DW, Kamali A, et al. Mortality associated with HIV-1 infection over fivee years in a rural Ugandan population: cohort study. BMJ 1997, 315:767-771. 83.. Todd J, Balira R, Grosskurth H, et al. HIV-associated adult mortality in a rural Tanzaniann population. AIDS 1997;11:801-7. 84.. Senkoro KP, Boerma JT, Klokke AH, et al. HIV incidence and HIV-associated mortalityy in a cohort of factory workers and their spouses in Tanzania, 1991 through 1996.. J Acquir Immune Defic Syndr 2000, 23: 194-202.
180 0 HIV-22 IN WEST AFRICA
85.. Gottlieb GS, Salif Sow P, Hawes SE, et al. Equal plasma viral loads predict a similar ratee of CD4+ T cell decline in human immunodeficiency virus (HIV) type 1 - and type 2-infectedd individuals from Senegal, West Africa. JID 2002; 185: 905-914. 86.. Rey-Cuillé M-A, Berthier J-L, Bomsel-Demontoy M-C et al. Simian Immunodeficiency Viruss replicates to high levels in Sooty Mangabeys without inducing disease. J Virol 1998,, 72:3872-3886. 87.. Chakrabarti LA, Lewin SR, Zhang L, et al. Normal T-cell turnover in sooty mangabeyss harboring active simian immunodeficiemcy virus infection. J Virol 2000; 74:: 1209-1223. 88.. Ling B, Santiago M, Gormus B, et al. Non-invasive detection and new lineages of simiann immunodeficiency viruses from sooty mangabeys. XIV International AIDS Conference,, Barcelona, July 2002 [TuPeA4418] 89.. Broussard SR, Staprans SI, White R, et al. Simian immunodeficiency virus replicates too high levels in naturally infected African green monkeys without inducing immunolohicc of neurologic disease. J Virol 2001; 75; 2262-2275. 90.. Whittle HC, Ariyoshi K, Rowland-Jones S. HIV-2 and T cell recognition. Current Opinionn in Immunology 1998, 10:382-387. 91.. Jaye A, Sarge-Njie R, Schim van der Loeff M, et al. Comparisons of cellular immune responsess in HIV-1 and HIV-2 infections. XIV International AIDS Conference. Barcelona,, July 2002 [Abstract A10083]. 92.. Berry N, Ariyoshi K, Jobe O, et al. HIV type 2 proviral load measured by quantitative polymerasee chain reaction correlates with CD4+ lymphopenia in HIV type 2-infected individuals.. AIDS Res Hum Retroviruses 1994;10:1031-7. 93.. Ariyoshi K, Berry N, Wilkins A, et al. A community-based study of human immunodeficiencyy virus type 2 provirus load in a rural village in West Africa. JID 1996;173:245-248. . 94.. Popper SJ, Sarr AD, Gueye-Ndiaye A, et al. Low plasma human immunodeficiency viruss type 2 viral load is independent of proviral load: low virus production in vivo. J Viroll 2000;74:1554-7. 95.. Berry N, Ariyoshi K, Jaffar S, et al. Low peripheral blood viral HIV-2 RNA in individualss with high CD4 percentage differentiates HIV-2 from HIV-1 infection. J Hum Viroll 1998;1:457-68. 96.. UNAIDS/WHO. AIDS Epidemic Update, December 2001. Geneva, Switzerland: UNAIDS/WHO;2001. . 97.. van der Ende ME, Schutten M, Ly TD, et al. HIV-2 infection in 12 European residents: viruss characteristics and disease progression. AIDS 1996,10:1649-1655. 98.. Clark NM, Dieng-Sarr A, Sankale JL, et al. Immunologic and virologie response of HIV-22 infection to antiretroviral therapy. AIDS 1998; 12: 2506-2507. 99.. Rodes B, Holguin A, Soriano V, et al. Emergence of drug resistance mutations in humann immunodeficiency virus type 2- infected subjects undergoing antiretroviral therapy.. J Clin Microbiol 2000; 38: 1370-1374. 100.. Smith NA, Shaw T, Berry N, et al. Antiretroviral therapy for HIV-2 infected patients.. J Infect 2001; 42:126-33. 101.. Soriano V, Gomes P, Heneine W, et al. Human immunodeficiency virus type 22 (HIV-2) in Portugal: clinical spectrum, circulating subtypes, virus isolation, and plasmaa viral load. J Med Virol 2000; 61:111-116. 102.. Houston SC, Miedzinski LJ, Mashinter LD. Rapid progression of CD4 cell declinee and subsequent response to salvage therapy in HIV-2 infection. AIDS 2002; 16:: 1189-91. 103.. Adjé-Touré CA, Cheignsoni R, Garcia-Lerma G, et al. Virologie and immunologicc responses and drug resistance profiles of HIV-2 infected patients receivingg antiretroviral treatment in Abidjan. XII International Conference on AIDS and STDss in Africa. Ouagadougou, December 2001 [abstract 11DT4-5].
181 1 HIV-22 IN WEST AFRICA
104.. Bissagnene E, Eholie S, Tanon A. Evaluation du traitement antiretroviral chez less patients VIH-2 positifs a Abidjan. XII International Conference on AIDS and STDs inn Africa. Ouagadougou, December 2001 [abstract 10DT3-6]. 105.. Cox S, Aperia K, Albert J, Wahren B. Comparison of the sensitivities of primaryy isolates of HIV type 2 and HIV type 1 to antiretroviral drugs and drug combinations.. AIDS Res Hum Retrovir 1994; 12: 1725-28. 106.. Witvrouw M, Pannecouque C, Laethem KV, et al. Activity of non-nucleoside reversee transcriptase inhibitors against HIV-2 and SIV. AIDS 1999; 13: 1477-83. 107.. Tapia N, Clotet B, Martinez MA. A bacteriophage lambda-based genetic screeningg system for the characterization of the activity and phenotype of HIV-2 protease.. XIV International AIDS Conference. Barcelona, July 2002 [Abstract TuPeA4395]. .
108.. van der Ende ME, Guillon C, Boers PHt et al. Antiviral resistance of biologic HIV-22 clones obtained from individuals on nucleoside reverse transcriptase inhibitor therapy.. J Acquir Immune Defic Syndr 2000;25:11-8. 109.. Brandin E, Lindborg L, Brostrom C, et al. Antiretroviral resistance in Swedish HIV-22 patients. XIV International AIDS Conference, Barcelona, July 2002 [Abstract WePeB5979]. . 110.. British HIV Association Writing Committee. British HIV Association (BHIVA) guideliness for the treatment of HIV-infected adults with antiretrociral therapy. HIV Medicinee 2001; 2; 276-313. 111.. Yeni PG, Hammer SM, Carpenter CCJ, et al. Antiretroviral treatment for adultt HIV infection in 2002. Updated recommendations of the International AIDS Society-USAA panel. JAMA 2002; 288:222-235. 112.112. Rabkin M, El-Sadr W, Katzenstein D, et al. Antiretroviral treatment in resource-poorr settings: clinical research priorities. Lancet 2002; 360: 1503-5. 113.. World Health Organization. Scaling up antiretroviral therapy in resource- limitedd settings. Guidelines fora public health approach. WHO: Geneva, 2002. 114.. Weidle PJ, Mastro TD, Grant AD, et al. HIV/AIDS treatment and HIV vaccines forr Africa. Lancet 2002; 359: 2261-7. 115.. Schultz AM, Bradac JA. The HIV vaccine pipeline, from preclinical to phase III. AIDSS 2001; 15 (Suppl 5): S147-S158. 116.. Smith PG, Hayes RJ, Mulder DW. Epidemiological and public health considerationss in the design of HIV vaccine trials. AIDS 1991; 5 (suppl 2): S105- S111. .
182 2 HIV-22 IN WEST AFRICA
Listt of Abbreviations
AIDS S Acquiredd immunodeficiency syndrome ART T Antiretrovirall therapy BT T Bloodd transfusion CD4 4 Cellularr determinant no 4 CDC C Centerss for Disease Control and Prevention (USA) CI I Confidencee interval cpz z Chimpanzee e CSW W Commerciall sex worker CTL L Cytotoxicc T lymphocyte CVS S Cervico-vaginall secretions DNA A Deoxyribonucleicc acid DoSH H Departmentt of State for Health (The Gambia) ELISA A Enzyme-linkedd immunosorbent assay GNP P Grosss national product GU U Genito-urinary y GUM M Genito-urinaryy medicine HAART T Highlyy active antiretroviral therapy HIV-1 1 Humann immunodeficiency virus type 1 HIV-2 2 Humann immunodeficiency virus type 2 HIV-D D Infectionn with HIV-1 and HIV-2 HLA A Humann leukocyte antigen HSV-2 2 Herpess simplex virus type 2 HTLV-I I Humann T-cell lymphotropic virus type I IgA A Immunoglobulinn type A IR R Incidencee rate IRR R Incidencee rate ratio IVDU U Intravenouss drug user LNSP P Laboratórioo Nacional de Saude Püblica (Guinea-Bissau) LTNP P Long-termm non-progressor LSHTM M Londonn School of Hygiene and Tropical Medicine mac c Macaque e MHC C Majorr histocompatibility complex MIMOSA A Millenniumm Mother-and-child-health Study in Africa ml l millilitre e MRC C Medicall Research Council (UK) MTCT T Mother-to-childd transmission OR R Oddss ratio PAIGC C Partidoo Africano para a Independência da Guiné e de Cabo Verde PBMC C Peripherall blood mononuclear cells PCR R Polymerasee chain reaction PPP P Purchasingg power parity PSB B Projectoo de Saüde de Bandim (Guinea-Bissau) PTB B Pulmonaryy tuberculosis
183 3 HIV-22 IN WEST AFRICA
PYO O Person-yearss of observation pyo o Person-yearss of observation PVL L Plasmaa viral load RNA A Ribo-nucleicc acid RR R Ratee ratio SIV V Simiann immunodeficiency virus sm m Sootyy mangabey SSHS S Sichilii Sherlock Holmes Society SSI I Statenss Serum Institut (Denmark) STI I Sexuallyy transmitted infection STD D Sexuallyy transmitted disease TB B Tuberculosis s UNN AIDS Unitedd Nations AIDS Organization UNDP P Unitedd Nations Development Programme WB B Westernn Blot WHO O Worldd Health Organization ^ ^ Microlitre e
184 4 HIV-22 IN WEST AFRICA
Summary y
Thiss thesis is an investigation of the epidemiology of human immunodeficiency virus typee 2 (HIV-2) infection in two West African countries, Guinea-Bissau and The Gambia.
Chapterr 1 discusses the setting of the studies and puts them into the perspective of previouss research conducted at the Medical Research Council (MRC) Laboratories in The Gambia. .
Chapterr 2 is a review of the epidemiology of HIV-2, and describes the key differences of HIV-22 infection compared to HIV-1, including a lower transmission rate, a more limited geographicall spread, stable or decreasing prevalences and a better prognosis.
Chapterr 3 describes the results of the first round of the Gambian HIV sentinel surveillancee programme in 2000-1. The HIV-1 prevalence, 15 years after detection of the firstt case in the country, was still relatively low at 1.0%, and HIV-2 prevalence was 0.8%. Inn comparison to an earlier cross-sectional study among pregnant women carried out in 1993-5,, this shows a reversal, as then HIV-2 was the more prevalent virus (1.1%) and HIV-11 was less common (0.6%).
Chapterr 4 describes the incidence of HIV-2 in a rural area in north-west Guinea-Bissau. Thee incidence was 4.6 per 1000 persons per year, which is lower than expected in view of thee high prevalence (8%). Among the risk factors independently associated with incident infectionn in women were active syphilis, history of vaginal discharge, having been divorced,, living in the central part of the village, and having lived elsewhere in Guinea- Bissau.. Among men age below 30 years and active syphilis were significantly associated withh new infections.
185 5 HIV-22 IN WEST AFRICA
Chapterr 5 examines the hypothesis, put forward by researchers from Dakar and Boston, thatt HIV-2 infection might offer protection against subsequent infection with HIV-1. In thee same rural population in Guinea-Bissau as described in Chapter 3, we found that the incidencee of HIV-1 infections was higher among those already infected with HIV-2, so HIV-22 could possibly be a risk factor for HIV-1 acquisition, rather than a protective factor. .
Chapterr 6 examines the mortality in a cohort of patients attending a sexually transmitted diseasee clinic at the MRC facilities in Fajara. This study confirmed that the survival of HIV-22 infected patients was better than that of HIV-1 infected patients. In the group of patientss with normal CD4 count (> 500 cells / (0,1) the mortality of HIV-2 infected subjectss was half of that in HIV-1 infected subjects, but among those with a low CD4 countt (< 200 cells / (0,1), there was no difference in mortality between those infected with HIV-11 or HIV-2. This study also described the survival of patients who are infected with bothh viruses. These patients had a poor prognosis; their mortality rate was similar to thosee of patients with HIV-1. This indicates that HIV-2 is not able to mitigate the coursee of HIV-1 disease.
Chapterr 7 compares the survival of HIV-2 infected women with that of HIV-1 infected andd HIV-uninfected women. Unlike the study population in Chapter 6, these women weree not recruited because they were ill, but were screened for HIV at their first visit to ann antenatal clinic as part of a study examining mother-to-child transmission rates of HIV-11 and HIV-2 in 1993-5. This study population is more representative of the general populationn than the highly selected patients of Chapter 6. In 2001 we re-visited these womenn and established their vital status. The results confirmed some earlier findings: prognosiss with HIV-2 infection was better than with HIV-1 infection, and the mortality ratee is predicted by the plasma viral load in both infections. The surprising result was thatt in this seroprevalent cohort more than 50% of women with HIV-1 were still alive eightt years after recruitment. This indicates that the survival with HIV-1 in West African womenn is similar to that of women in Europe or North America prior to the use of highlyy active antiretro viral treatment (HAART). Another important finding was that it is
186 6 HIV-22 IN WEST AFRICA nott the virus type per se that is predictive of mortality, but the plasma viral load. In other words,, if a person has HIV-2 infection with a high plasma viral load, then the prognosis iss as bad as that of a person with HIV-1 and high plasma viral load. This suggests that thee start of treatment in HIV-2 disease should be guided by the same indicators as in HIV-11 disease.
Chapterr 8 describes survival of eight children who were infected with HIV-2 by their motherss around the time of delivery. Perhaps surprisingly, this is the world's largest cohortt of perinatally HIV-2 infected children. An earlier study following up these childrenn had found that all eight had survived to 18 months of age. After 5-7 more years threee of the eight children had died. This was a significandy higher mortality rate than in ann HIV uninfected control gtoup of children, but was lower than in an HIV-1 infected groupp of children. The numbers were very small though and few firm conclusions can bee drawn. It does suggest however, that children with perinatally acquired HIV-2 infectionn need the same care and prophylaxis as HIV-1 infected children.
Finally,, Chapter 9 draws conclusions from these studies, and identifies outstanding researchh questions. It remains unclear what initiated the HIV-2 epidemic, and why it is inn decline now. The plasma viral load in HIV-2 infected people is generally lower than in HIV-11 infected people, but the reasons for this are unknown. Another unresolved questionn is why only a limited proportion of HIV-2 infected people have a high plasma virall load and develop AIDS, and how large this proportion is. If a preventive vaccine weree to be developed, very large field trials would be needed to test the protective efficacyy of such a vaccine, as the incidence of HIV-2 is low. HIV-2 prevalence is stable orr declining in most countries, and HIV-2 is not a global threat. Nevertheless many peoplee in West Africa infected with this virus will develop immunodeficiency and die fromm AIDS. Trials and cohort studies are needed to identify appropriate treatment regimens. .
187 7 HIV-22 IN WEST AFRICA
188 8 HIV-22 IN Wi'ST AFRICA
Samenvatting g
Ditt proefschrift beschrijft onderzoek naar de epidemiologie van infectie met het humaan immunodeficiëntiee virus type 2 (HIV-2) in twee West-Afrikaanse landen, Guinee-Bissau enn Gambia.
Hoofdstukk 1 bespreekt de achtergrond van de verschillende onderzoeken en plaatst ze in hett perspectief van eerder onderzoek gedaan door de Medical Research Council (MRC) Laboratoriess in Gambia.
Hoofdstukk 2 geeft een overzicht van de epidemiologie van HIV-2 en beschrijft de belangrijkstee verschillen van HIV-2 to.v. HIV-1 infectie, onder andere een lagere transmissiekans,, een beperktere geografische verspreiding, een stabiele of zelfs dalende prevalentiee en een betere prognose.
Dee resultaten van de eerste ronde van het Gambiaanse HIV sentinel surveillance programmaa in 2000-1 worden beschreven in hoofdstuk 3. Vijftien jaar na de ontdekking vann de eerste HIV geïnfecteerde in het land, was de HIV-1 prevalentie nog steeds relatief laagg (1,0%); de HIV-2 prevalentie was 0,8%. In vergelijking met een eerder onderzoek onderr zwangere vrouwen in Gambia tussen 1993 en 1995 is dit een omkering, omdat toenn HIV-2 (1,1%) meer voorkwam dan HIV-1 (0,6%).
Hoofdstukk 4 is gewijd aan de incidentie van HIV-2 in een ruraal gebied in noordwest Guinee-Bissau.. De incidentie was 4,6 per 1000 personen per jaar, hetgeen laag is gezien dee hoge prevalentie (8%) Risicofactoren die significant geassocieerd waren met nieuwe infectiess in vrouwen waren onder meer actieve syfilis, abnormale vaginale afscheiding, ooitt zijn gescheiden, in het centrale deel van het gebied wonen, en ergens anders in Guinea-Bissauu gewoond hebben. Bij mannen waren leeftijd onder de 30 jaar en actieve syfiliss significant geassocieerd met nieuwe infecties.
189 9 HIV-22 IN WEST AFRICA
Hoofdstukk 5 onderzoekt de hypothese, geopperd door onderzoekers uit Dakar en Boston,, dat HIV-2 infectie bescherming zou kunnen bieden tegen infectie met HIV-1. Echter,, in dezelfde rurale bevolking in Guinee-Bissau als beschreven in hoofdstuk 3, vondenn wij dat de incidentie van HIV-1 infecties hoger was in de groep mensen die reeds mett HIV-2 besmet was. HIV-2 leek wellicht een risicofactor te zijn voor infectie met HIV-11 in plaats van een beschermende factor
Hoofdstukk 6 onderzoekt de sterfte in een cohort van HlV-geïnfecteerde patiënten die de polikliniekk van de MRC in Fajara bezochten. Deze studie bevestigde dat de overleving vann HIV-2 geïnfecteerden beter was dan die van HIV-1 geïnfecteerde patiënten. Sterfte inn de groep patiënten met normale CD4 waarden (> 500 cellen / (il) was twee keer zo laagg in HIV-2 geïnfecteerden als in HIV-1 geïnfecteerden . Echter, in de groep patiënten mett lage CD4 waarden (< 200 cellen / JA1) was er geen verschil in sterfte tussen personen mett HIV-1 of HIV-2 infectie. De overleving van patiënten die geïnfecteerd zijn met beidee virussen was vergelijkbaar met die van patiënten met alleen HIV-1. Dit betekent datt co-infectie met HIV-2 niet leidt tot een milder beloop van HIV-1 infectie.
Hoofdstukk 7 vergelijkt de overleving van HIV-2 geïnfecteerde vrouwen met die van HIV-11 geïnfecteerde en niet geïnfecteerde vrouwen. In tegenstelling tot de onderzoekspopulatiee in hoofdstuk 6 werden deze vrouwen niet gerecruteerd omdat ze ziekk waren. Zij werden getest voor HIV bij hun eerste bezoek aan een polikliniek voor zwangerenn in het kader van een onderzoek in 1993-5 dat de kans op moeder-kind transmissiee van HIV-1 en HIV-2 vergeleek. Deze onderzoekspopulatie is representatieverr voor de algemene bevolking dan de sterk geselecteerde patiëntenpopulatiee in Hoofdstuk 6. In 2001 werden deze vrouwen opnieuw bezocht en werdd nagegaan of zij (en hun kinderen - zie hoofdstuk 8) nog in leven waren. De resultatenn bevestigen enkele eerdere bevindingen: de prognose van HIV-2 infectie was beterr dan die van HIV-1 infectie, en de sterfte werd voorspeld door het plasma virus niveauu in beide infecties. Een verrassend resultaat was dat in dit seroprevalente cohort meerr dan 50% van de vrouwen met HIV-1 acht jaar na recruitering nog steeds in leven
190 0 HIV-22 IN WEST AFRICA was.. Dit wijst erop dat de overleving met HIV-1 in West-Afrikaanse vrouwen vergelijkbaarr is met die van vrouwen in Europa of Noord Amerika voor het tijdperk van dee zeer effectieve antiretrovirale therapie. Een andere belangrijke bevinding was dat het viruss type per se niet voorspellend is voor sterfte, maar het niveau van virus in het plasma.. Met andere woorden, als een persoon HIV-2 infectie heeft met een hoog plasma viruss niveau, dan is de prognose even slecht als die van een persoon met HIV-1 en een hoogg plasma virus niveau. Dit suggereert dat dezelfde criteria gebruikt zouden moeten wordenn voor het starten van behandeling van HIV-2 geïnfecteerde patiënten als gelden voorr HIV-1.
Hoofdstukk 8 beschrijft de overleving van acht kinderen die rond de geboorte via hun moederss met HIV-2 werden besmet. Het is wellicht verbazingwekkend dat dit 's werelds grootstee cohort van perinataal HIV-2 geïnfecteerde kinderen is. Alle acht kinderen leefdenn nog toen ze 18 maanden oud waren. Vijf tot zeven jaren nadien waren drie van dee acht kinderen gestorven. Dit was een significant hogere sterfte dan in een groep ongeïnfecteerdee kinderen, maar lager dan in een groep kinderen die met HIV-1 waren besmet.. De getallen zijn echter klein en er kunnen weinig definitieve conclusies uit wordenn getrokken. Desalniettemin suggereren deze gegevens dat kinderen met perinataal verworvenn HIV-2 infectie dezelfde zorg en profylaxe nodig hebben als kinderen die met HIV-11 geïnfecteerd zijn.
Inn hoofdstuk 9 worden ten slotte conclusies getrokken en openstaande onderzoeksvragenn gesignaleerd. Het blijft onduidelijk wat de HIV-2 epidemie heeft ingeluidd en waarom de epidemie nu een neerwaartse trend te zien geeft. Het plasma virus niveauu in HIV-2 geïnfecteerde personen is over het algemeen lager dan in HIV-1 geïnfecteerdee personen, maar de redenen hiervoor zijn niet bekend. Een andere onopgehelderdee vraag is waarom slechts een beperkt deel van de met HIV-2 geïnfecteerdee personen een hoog plasma virus niveau hebben en AIDS krijgen, en hoe groott dit percentage is. Indien een preventief vaccin tegen HIV-2 ontwikkeld zou worden,, zijn zeer grote trials noodzakelijk om het beschermend effect van zo'n vaccin te meten,, omdat de incidentie van HIV-2 laag is. De HIV-2 prevalentie is stabiel of dalend
191 1 HIV-22 IN WEST AFRICA inn de meeste landen en HIV-2 vormt geen mondiale bedreiging. Desalniettemin zullen velee West-Afrikanen die met dit virus besmet zijn immuno-deficiëntie ontwikkelen en stervenn aan AIDS. Trials en cohort studies zijn nodig om geschikte behandelingregimes tee identificeren voor HIV-2 infecüe.
192 2 HIV-22 IN WKST AFRICA
Listt of Publications
Ariyoshii K, Schim van der Loeff MF, Sabally S, Cham F, Corrah T, Whittle H. Does HIV-22 infection provide cross-protection against HIV-1 infection? [letter] AIDS 1997; 11:: 1053-4.
Ariyoshii K, Schim van der Loeff M, Cook P, Whitby D, Corrah T, Jaffar S, Cham F, Saballyy S, O'Donovan D, Weiss RA, Schulz TF, Whittle H. Kaposi's Sarcoma in the Gambia,, West Africa is less frequent in Human Immunodeficiency Virus type 2 than in Humann Immunodeficiency Virus type 1 infection despite a high prevalence of Human Herpesviruss 8. J Human Virology 1998; 1: 193-9.
Gambiaa Government / Medical Research Council Joint Ethical Committee. Ethical issuess facing medical research in developing countries. Lancet 1998; 351: 286-7.
Grasslyy NC, Xiang Z, Ariyoshi K, Aaby P, Jensen H, Schim van der Loeff M, Dias F, Whitdee H, Breuer J. Mortality among Human Immunodeficiency Virus type 2-positive villagerss in rural Guinea-Bissau is correlated with viral genotype. / Virol 1998; 72: 7895- 9. .
Schimm van der Loeff MF, Aaby P. Towards a better understanding of the epidemiology off HIV-2 [review]. AIDS 1999; 13 (suppl A) :S69-S84.
Ariyoshii K, Schim van der Loeff MF, Berry N, Jaffar S, Whitde H. Plasma HIV viral loadd in relation to season and to Plasmodium falciparum parasitaemia. [letter] AIDS 1999; 13:1145-6. .
Ariyoshii K, Jaffar S, Alabi A, Berry N, Schim van der Loeff MF, Sabally S, N'Gom PT, Corrahh T, Tedder R, Whitde H. Plasma RNA viral load predicts the rate of CD4 T cell declinee and death in HIV-2 infected patients in West Africa. AIDS 2000; 14: 339-44.
Schimm van der Loeff MF, Corrah T, Whitde HC. Low and stable infection rates in Senegal?? [letter] ^4/Ltf2000; 14: 1276-7.
Schimm van der Loeff MF. HIV in Africa [review]. Medicine Magazine 2001; 29: 35-7.
Schimm van der Loeff MF. Reducing the risk of HIV infection through blood transfusion, [letter]] Tropical Medicine & International Health 2001; 6: 325.
Vann der Sande MAB, Coleman RL, Schim van der Loeff MF, McAdam KPWJ, Nyan OA,, Thien Th, Dolmans WMV, Walraven GEL. A template for improved prevention andd control of cardiovascular disease in sub-Saharan Africa. Health Policy and Planning 2001;16:345-50. .
193 3 HIV-22 IN WEST AFRICA
Schimm van der Loeff MF, Aaby P, Ariyoshi K, Vincent T, Aveika AA, Da Costa C, Pembreyy L, Dias F, Harding E, Weiss HA, Whitde HC. HIV-2 does not protect against HIVV 1 infection in a rural community in Guinea-Bissau. AIDS 2001; 15: 2303-10.
Holmgrenn B, Andersson S, Harding E, Schim van der Loeff M, Vastrup P, Aaby P, Ariyoshii K, Whitde H. Increased prevalence of HTLV-I among HIV-2 infected women butt not HIV-2 infected men in rural Guinea-Bissau. / Acq Imm Defic Syndr 2002; 30:342- 350. .
Schimm van der Loeff MF, Jaffar S, Awasana AA, Harding E, Sarge-Njie R, Alabi A, Saballyy S, Corrah T, Ariyoshi K, Whittle HC. Mortality of HIV-1, HIV-2 and HIV-1 / HIV-22 dually infected patients in a clinic-based cohort in The Gambia, West Africa. AIDSAIDS 2002; 16: 1775-83.
Berryy N, Jaffar S, Schim van der Loeff M, Ariyoshi K, Harding E, N'Gom PT, Dias F, Wilkinss A, Ricard D, Aaby P, Tedder R, Whitde H. Low level viremia and high CD4% predictt normal survival in a cohort of HIV type-2-infected villagers. AIDS Res Hum RetrovirusesRetroviruses 2002; 18:1167-73.
Westt B, Morison L, Schim van der Loeff M, Gooding E, Aveika AA, Demba E, Mayaud P.. Evaluation of a new rapid diagnostic kit (Femexam) for bacterial vaginosis in patients withh vaginal discharge syndrome in The Gambia. Sexually Transmitted Diseases 2003; 30: 483-489. .
Alabii AS, Jaffar S, Ariyoshi K, Blanchard T, Schim van der Loeff M, Akum AA, Corrah T,, Sabally S, Sarge-Njie R, Cham-Jallow F, Jaye A, Berry N, Whitde H et al. Plasma viral load,, CD4 % and survival of HIV-1, HIV-2, and dually infected Gambian patients. AIDS.AIDS. 2003; 11: 1513-1520.
Schimm van der Loeff MF, Sarge-Njie R, Ceesay S, Aveika AA, Jaye P, Sam O, Jaiteh KO, Cubittt D, Milligan P, Whitde HC. Regional differences in HIV trends in The Gambia: resultss from sentinel surveillance among pregnant women. AIDS 2003; 17: 1841-6.
Lagardee E, Schim van der Loeff M, Enel C, Holmgren B, Spira R, Pison G, Piau JP, Delaunayy V, M'Boup S, Ndoye I, Coeuret-Pellicer M, Whitde H, Aaby P for the MECORAA group. Mobility and the spread of human immunodeficiency virus into rural areass of West Africa. International'Journal of'Epidemiology. In press.
Schimm van der Loeff MF, Hansmann A, Aveika AA, Ota MO, O'Donovan D, Sarge-Njie R,, Ariyoshi K, Milligan P, Whitde H. Survival of HIV-1 and HIV-2 perinatatly infected childrenn in The Gambia. AIDS. Accepted.
Ariyoshii K, Berry N, Cham F, Jaffar S, Schim van der Loeff M, Jobe O, N'Gom PT, Larsenn O, Andersson S, Aaby P, Whitde H. Human T-Lymphotropic Virus Type I (HTLV-I)) proviral load does not enhance Human Immunodeficiency Virus Type 2 (HIV-2)) disease but may associate with mortality.///?. Accepted.
194 4 HIV-22 IN WEST AFRICA
Jayee A, Sarge-Njie R, Schim van der Loeff M, Todd J, Alabi A, Sabally S, Corrah T, Whitdee H. No differences in cellular immune responses between asymptomatic human immunodeficiencyy type 1 and type 2- infected Gambian patients. JID Accepted.
Vann der Sande M, Schim van der Loeff M, Aveika AA, Sabally S, Togun T, Corrah T, Njiee R, Kaye S, Alabi A, Whittle H. Baseline body mass index (BMI) is an independent predictorr of survival time in HIV infected patients. Submitted.
Hansmannn A, Schim van der Loeff M, Kaye S, Aveika AA, Sarge-Njie R, O'Donovan D, Ariyoshii K, Alabi A, Milligan P, Whitde H. Contrasts in plasma viral load, CD4% and survivall in a community-based cohort of HIV-1 and HIV-2 infected women in The Gambia.. Submitted
vonn Seidlein L, Walraven G, Milligan PJM, Alexander N, Manneh N, Deen JL, Coleman R,, Jawara M, Lindsay SW, Drakeley C, De Martin S, Olliaro P, Bennett S, Schim van der Loefff MF, Okunoye K, Targett GAT, McAdam KPWJ, Doherty JF, Greenwood BM, Pinderr M. The effect of mass administration of pyrimethamine / sulfadoxine combined withh artesunate on malaria transmission: a double blind, community randomised, placebo controlledd trial in The Gambia. Submitted.
Enell C, Delaunay V, Wade A, Holmgren B, Schim van der Loeff M, Pison G, Lagarde L. Sexuall behavior, social and demographic characteristics and the HIV epidemic: comparisonn of three rural sites in West Africa. Submitted.
Schimm van der Loeff MF, Aaby P, Todd J, Aveika AA, da Costa C, Ariyoshi K, Vincent T,, Dias F, Whittle H. Incidence of HIV-2 in a rural community in Guinea-Bissau, West Africa.. Submitted.
Schimm van der Loeff MF, Holmgren B, Todd J, Vincent T, Dias F, Ariyoshi K, Anderson S,, Aaby S, Whitde H. Survival of HIV-2, HTLV-I or dually infected adults in a rural area off West Africa. Results from a community-based study with 11 years of follow-up. Ms in preparation. preparation.
Bolithoo P, Price P, Schim van der Loeff MF, Jaye A, Awasana AA, McDermid JM, Yindomm L-M, Chase D, Sirugo G, Burke V, French MA, Whitde HC. The TNFA-308 andd IL12B 3'UTR loci and the outcome of HIV Disease in Gambians. Ms in preparation.
Vann der Sande MAB, Schim van der Loeff MF, Cashdollar R, Dowley M, Aveika AA, Togunn T, et al. Incidence of TB and survival after TB diagnosis is similar in HIV-2 comparedd to HIV-1 patients after adjustment for CD4+ cell count. Ms in preparation.
195 5 HIV-22 rN WKST AFRICA
196 6 HIV-22 IN WEST AFRICA
Curriculumm vitae
Maartenn Schim van der Loeff werd geboren als derde van vijf zoons in Venray, waar hij gymnasium-PP deed op het Boschveld College. In Amsterdam volgde hij vanaf 1977 de studiee Geneeskunde aan de Vrije Universiteit, en werkte er als student-assistant bij de vakgroepenn medische ethiek en sociale geneeskunde. Hij deed artsexamen in juni 1987 enn vervulde vervolgens zijn vervangende dienstplicht in de Spaarne kliniek te Haarlem, eenn ziekenhuis voor behandeling van verslavingsziekten. Daarnaast was hij docent anatomiee aan de School voor Gezondheidszorg Amsterdam, in Amsterdam.
Inn 1990 vertrok hij naar Zambia, waar hij vanuit Sichili Mission Hospital werkzaam was inn primary health care en patiëntenzorg in een uitgestrekt en leeg district. In 1994 begon hijj aan de master's opleiding in epidemiologie aan de London School of Hygiene and Tropicall Medicine. Na voltooing in 1995 keerde hij terug naar Afrika, dit keer naar The Gambia,, waar hij onderzoek deed naar de epidemiologie van HIV-2 voor de Medical Researchh Council. Hij was er lid van de nationale ethisch commissie en was wetenschappelijkk adviseur van deze commissie van 2000 tot 2002.
Inn 2002 keerde hij terug naar de London School, waar hij werkt als lecturer in epidemiologie,, en vanwaar hij forenst naar Gambia en Guinea-Bissau. Hij is getrouwd en heeftt drie kinderen.
197 7 HIV-22 IN WEST AFRICA
198 8 HIV-22 IN WEST AFRICA
Acknowledgements s
Writingg a thesis is as much about cutting words as about writing them. First drafts were usuallyy short, second and third drafts very long, and the hard work on draft four onwards waswas to reduce the word count. It is painful to delete wonderful phrases that were formulatedd after many days of pondering, but it helps to focus the paper, and, perhaps as aa welcome by-product, the mind of the writer.
Thiss thesis has been slighdy long in coming. When I started doing research in 1995, I thought,, naively, that science was done in quietness, without haste, and with time for endlesss tinkering until perfection. I have learned since that the research world is full of buzz,, hurry, and deadlines. This thesis is a compromise of these two approaches; I wouldd have preferred to tinker for a few more years, but I was advised it was time to handd in. I hope it shows more signs of perfection than of haste.
Manyy people have contributed to the work described in this thesis, and I would like to thankk them.
II thank Keith McAdam for taking a step in the dark by employing me in 1995. I was an unknownn quantity in science, and my only assets were a good reference from the London Schooll of Hygiene & Tropical Medicine, having worked in Zambia, and being married to Marianne. .
Hiltonn Whittle once introduced himself to a group of visiting scientists as someone being "interestedd in viruses and humans", and I have learned many tilings from Hilton about both.. Hilton has supervised this project from its inception in 1997, and has been central too its success. I thank him for his stable and enduring support. This thesis is as much a productt of his guidance as of my work.
199 9 Hiv-22 IN WEST AFRICA
Att the London School I learned the theory of epidemiology; from Peter Aaby I learned thee practice. Peter has provided me with opportunities, unconventional ideas, and directionn since 1996. I thank him for his support and his inspiration, and both Peter and Idaa for the warm welcome that I always received in Bissau.
Whenn I approached Roel Coutinho first about the idea of a thesis on HIV-2 in West Africa,, he was immediately enthusiastic, having both an interest in HIV and a history in Guinea-Bissau.. He has coached me long-distance, and the emails and meetings to the Nieuwee Achtergracht gave me direction, new ideas, and invaluable suggestions.
Thankss to Albert Meijer, who advised me strongly to do a PhD and who convinced me thatt it was worthwhile. Thanks to my paranimfen Jan van der Mannen and Casper Schim vann der Loeff for support and friendship.
II thank Koya Ariyoshi, who introduced me to concepts of virology and immunology, and whoo has been a stimulating and creative force through the years.
Sincee 1997, when he was appointed as data manager, I benefited from Akum Aveika Awasana'ss clear mind, which translated my tentative questions into focused queries for FoxPro.. I thank him for his loyalty and support through the years. Also many thanks to hiss data team: Awa Kenda, Majulla Saidy, Sulayman Sowe, Aji-Fatou Jallow, Fatou Njie, andd Ami Sarjo.
Thee MRC Unit in The Gambia is a special place with highly talented and inspired scientistss as well as unconventional free-thinkers, and I have enjoyed the discussions and exchanges,, especially with Robin Bailey, Tom Blanchard, Roz Coleman, Tom Dohertv, Kevinn Hughes, Assan J aye, Steve Kaye, Michel Klein, Sam McConkev, Diarmuid O'Donovan,, Lorenz von Seidlein, Matthew Shaw, Giorgio Sirugo, Gijs and Ralf Walraven,, David Warndorff, and Martin Weber.
200 0 HIV-22 IN WEST AFRICA
Thankss to the team in the GUM Clinic in Fajara: Saihou Sabally, Toyun Togun, Marie- Clairee Mendy, Helen Manneh, Ken Joof, Ramou Jagne, Alieu Jatta, Saikou Badgie, Baba Danso,, Babucarr Jawneh. Also thanks to Pamela Njai and her team on die clinics and ward. .
Speciall thanks to Tumani Corrah, who coached me subtly, and whom I admire as a caringg clinician and a trusted colleague, to Ramu Sarge-Njie for her dedication, support, andd friendship, and to Andreas Hansmann, whose doggedness in field work turned a projectt about which I was sceptic into a resounding success.
Thee work in Guinea-Bissau has always been particularly challenging and rewarding. Glynnn Taylor introduced me to Caio (Guinea-Bissau), Bryan Savage led the team for a fulll year, and Tim Vincent has been the mainstay in Caio ever since mid-1997. Tim's resiliencee and culinary skills have made staying and working in Caio an even greater pleasure. .
Manyy thanks to Saihou Ceesay, director of the National AIDS Secretariat in The Gambia, forr his encouragement and support, and for the fruitful collaborations. Thanks to Kebba 00 Jaiteh, Omar Sam, and Alhagie Kolley from the Department of State for Health for thee pleasant collaboration on the sentinel surveillance study.
11 have had the pleasure to work with five statisticians, and these have been stimulating andd instructive collaborations. Thank you: Shabbar Jaffar, Helen Weiss, Jim Todd, Paul Milligan,, and Richard Hayes. I would like to thank my colleagues at the London School off Hygiene & Tropical Medicine for accepting me in their midst and providing me with thee academic environment that I needed to complete the thesis. Special thanks to Simin Bahrainipur,, Sara Thomas, and Claire Cook for making me feel at home in Keppel Street.
Thankss to the collaborators at home and overseas: Judy Breuer, Catherine Peckham, Giselaa Schneider, Neil Berry, Jane Rowley, Catherine Enel, Gilles Pison, Emmanuel
201 1 HII V-2 IN WEST AFRICA
Lagarde,, Carlos da Costa, Francisco Dias, Andrew Wilkins, Olav Larsen and Birgitta
Holmgren. .
Furtherr thanks to the following, who helped me in numerous ways: Joop, René Keet,
Peterr Langfield, Malcolm Clarke, Mark Derveeuw, Oliver Razum, Elizabeth Harding,
Abrahamm Alabi, M'Bemba Ceesay, Bernard Ndeye, Lucy Pembrey, Michael Correa,
Mamadyy Njai, and all those that I have not mentioned by name here.
Myy practice of science has been shaped by the work in The Gambia, Guinea-Bissau and
Londonn over die last 8 years. Two people have shaped my thinking about science well beforee that, and I would like to acknowledge them here: Drs. W. Rambags, my chemistry teacherr at the Boschveld College in Venray, and Prof. J.M. Broekman, professor of philosophyy and medical ethics at the Vrije Universiteit in Amsterdam.
II thank the people of Caio, the patients at the GUM clinic in Fajara, and the women of thee sentinel and MIMOSA studies. Without their participation there would have been no studiess at all.
Myy children Laura, Rutger and Agnes have been all too aware that I was writing a thesis.
Theyy accepted that as a valid reason for me to move to London and become a commuter too The Gambia. I hope that they are satisfied with the result. Finallv I'd like to thank
Marianne,, for encouragement, practical help, endurance, advice, moral support, proof reading,, and for much more.
202 2
Propositionss belonging to the thesis Stellingenn behorende bij het proefschrift
HIV-22 IN WEST AFRICA - EPIDEMIOLOGICAL STUDIES HIV-22 iisr WEST AFRICA - EPIDEMIOLOGICAL STUDIES
HIV-22 biedt geen bescherming tegen HIV-1 infectie HIV-22 does not protect against HIV-1 infection. Hett plasma virus niveau en het CD4 cel percentage zijn belangrijke en Plasmaa viral load and CD4 cell percentage are important and independent onafhankelijkee predictors voor sterfte aan HIV-2 infectie. predictorss of mortality in HIV-2 infection. Inn HIV-2 geïnfecteerde patiënten met een CD4 waarde lager dan 200 per Amongg HIV-2 infected patients with CD4 counts below 200 cells per [d, the microliter,, is de prognose vergelijkbaar met die van patiënten met HIV-1 prognosiss is similar to that of patients infected with HIV-1. infectie. .
"Simiann Immunodeficiency Virus sooty mangabey" is a misnomer. "Simiann Immunodeficiency Virus SOOty mangabey" is een foutieve benaming
HIV-22 prevalence is falling in West Africa. Dee prevalentie van HIV-2 in West Afrika is dalend.
Inn monitoring HIV epidemics, HIV-1 and HIV-2 prevalence should be given Bijj het bestuderen en beschrijven van HIV epidemieën, moeten HIV-1 en separately,, and not combined as "HIV prevalence". HIV-22 prevalenties apart worden weergegeven, en niet worden opgeteld als "HIVV prevalentie". Evenn in areas where HIV-2 is highly prevalent, the major public health danger iss posed by HIV-1, not by HIV-2. Zelfss in gebieden waar HIV-2 veel voorkomt, is HIV-1 een groter gevaar voor dee volksgezondheid dan HIV-2. Treatmentt of AIDS patients with HAART will not change the prevalence of HIV-11 in Africa. Behandelingg van AIDS patiënten met HAART zal de prevalentie van HIV-1 (Gray(Gray 2002) inn Afrika niet veranderen. (Gray(Gray 2002) Thee discovery that sodium transport and glucose transport are coupled in the smalll intestine, so that glucose accelerates absorption of solute and water, was Dee ontdekking dat natrium en glucose transport in de dunne darm zijn potentiallyy the most important medical advance of this century gekoppeld,, zodat glucose de absorptie van zout en water vernelt, was (Editorial,(Editorial, Lancet 1978; it: 300-1) mogelijkerwijss de belangrijkste medische vondst van deze eeuw. (Editorial,(Editorial, Lancet 1978; ü: 300-1) Twinss are a geneticist's dream but a data manager's nightmare. Tweelingenn zijn voor de geneticus een droom, maar voor de data manager een Sciencee is not an objective, truth-directed machine, but a quintessential!y nachtmerrie. . humann activity, affected by passions, hopes, and cultural biases. (Stephen(Stephen Jay Gould, The Panda's Thumb) Wetenschapp is niet een objectieve, waarheid-gerichte machine, maar een puur menselijkee activiteit, beinvloed door passie, hoop en culturele waardes. (Stephen(Stephen Jay Gould, The Panda's Thumb)
*»» '•»•<