By 1998, approximately 1 million of the total population of 60 million people in Thailand had been infected with HIV-1 since the explosive start of the epidemic among injection drug users (IDUs) in Bangkok in the late 1980s . Molecular epidemiologic studies have indicated that the initial epidemic among IDUs in Bangkok was due to envelope (env) subtype B strains and was largely unrelated to the much larger subsequent epidemic of HIV-1 subtype E among people at heterosexual risk of infection [2–4]. The Thai response to the epidemic [5–7] has resulted in a sharp slowing of transmission during the 1990s among young men [8,9] and women  at risk of sexually acquired infection. However, sentinel surveillance data indicate continued high HIV-1 prevalence (30–50%) among IDUs throughout the country.
In 1991, Thailand was selected by the World Health Organization as one of four developing countries with good potential for conducting efficacy trials of candidate HIV-1 preventive vaccines . The large population of IDUs in Bangkok at high risk for HIV-1 infection [12,13] coupled with an extensive network of outpatient drug treatment clinics operated by the Bangkok Metropolitan Administration (BMA) offered a potential setting in which to conduct a large-scale HIV-1 vaccine efficacy trial. Each year, approximately 8000 illicit drug users seek treatment in the BMA system; about 70% inject drugs and almost all use heroin. Methadone treatment has been offered since 1980 along with extensive HIV prevention efforts since the late 1980s. A 1994 cross-sectional study of Bangkok IDUs indicated that 80% of prevalent HIV-1 infections were subtype B and 20% were subtype E; however, information on the estimated time of seroconversion suggested that the proportion of infections due to subtype E had increased during the period 1988–1993 [14,15].
In this report we describe the cohort enrolled in 1995 and 1996 and present follow-up data collected through the end of 1998. This study was conducted to better define factors related to incident HIV-1 infection among IDUs in Bangkok to improve prevention efforts. We also set out to determine rates of follow-up and HIV-1 incidence to assess the feasibility of conducting an HIV vaccine efficacy trial in this setting and to characterize incident HIV-1 strains to guide vaccine design. A sub-study assessed the willingness of IDUs in Bangkok to participate in an HIV-1 vaccine efficacy trial . The results of these studies led to the initiation in Bangkok in March 1999 of a large-scale phase III efficacy trial of a bivalent recombinant gp120 HIV-1 vaccine (AIDSVAX B/E; VaxGen, Inc., Brisbane, California, USA), the first HIV vaccine efficacy trial in the developing world.
Voluntary HIV counselling and testing, with written informed consent, was offered to persons with a history of injection drug use aged 18 to 50 years who were attending any of 15 BMA drug treatment clinics, and who were not known to be HIV-seropositive. Blood for HIV antibody testing was collected from consenting volunteers who were then interviewed using a brief, standardized questionnaire. After the HIV test result was known, each volunteer was counseled confidentially about his or her HIV status. Eligible HIV-seronegative persons were offered enrollment, with written informed consent and repeat HIV antibody testing, into the prospective cohort study. On each study visit participants received 200 Thai baht (about US$ 8 in 1996). Screening and enrollment were conducted during two periods, May to November 1995  and May to December 1996.
Following a second HIV-seronegative test and enrollment into the prospective cohort study, participants were asked to return for follow-up visits every 4 months. On each visit participants received HIV risk-reduction counselling and health education, were interviewed using a standard questionnaire, and had blood collected for serologic testing. Male condoms and bleach for cleaning injection equipment were provided free of charge. Sterile syringes and needles were available at pharmacies and convenience stores throughout Bangkok without prescription for about 5–8 Thai baht (about US$ 0.20–0.32 in 1996). Participants who became HIV-seropositive received post-test counselling and were offered enrollment and follow-up under a separate study protocol on evaluation of HIV seroconverters; they received social and medical care according to BMA guidelines.
The study protocols were approved by the Ethical Review of Research Committee, Ministry of Public Health, Nonthaburi, Thailand; an Institutional Review Board, Centers for Disease Control and Prevention, Atlanta, GA, USA; and the Global Programme on AIDS, World Health Organization, Geneva, Switzerland.
Serum specimens were tested for HIV antibodies by enzyme immunoassay (EIA) at the BMA laboratory . At enrollment, serum specimens were tested for hepatitis B surface antigen [HBsAg; Monolisa HBsAg (second generation), Sanofi Diagnostics Pasteur, Marnes-La-Coquette, France], hepatitis B surface antibodies (HBsAb; Monolisa Anti-HBV; Sanofi Diagnostics Pasteur, Chaska, Minnesota, USA), and hepatitis C antibodies [Monolisa Anti-HCV; Sanofi Diagnostics Pasteur (USA)]. Serum specimens from enrollment were also evaluated for syphilis with a rapid plasma reagin (RPR) test (Sypal; Diagast Laboratories, Lille Cedex, France). RPR-reactive specimens were tested with the Treponema pallidum hemagglutination assay (TPHA; Immutrep TPHA; Omega Diagnostics, Alloa, Scotland, UK).
During cohort follow-up, specimens that were HIV-seropositive on EIA were tested with another EIA (Genetic Systems, Redmond, Washington, USA) and Western blot (Novapath HIV-1 Immunoblot; BioRad, Hercules, California, USA) at the HIV/AIDS Collaboration laboratory; and specimens that were positive on both EIA and Western blot were considered HIV-seropositive. Following an HIV-seropositive test and enrollment into the study of HIV seroconverters, additional blood specimens were collected for genetic characterization of the infecting HIV strains by nucleic acid sequencing and phylogenetic analysis .
Data management and statistical analyses
We used Epi-Info (version 6.04c; Centers for Disease Control, Atlanta, Georgia, USA) and SAS (version 6.12; SAS Institute, Cary, North Carolina, USA) for data management and analyses. Incidence rates for HIV-1 infection were calculated per 100 person–years (PY) of HIV-seronegative observation; seroconversions were assumed to have occurred at the mid-point of the interval between the last HIV-seronegative and the first HIV-seropositive tests. Exact 95% Poisson confidence intervals (CI) were calculated for the estimated HIV-1 incidence rates.
Categorical variables were analyzed by χ2 or Fisher's exact test or χ2 for trend. Differences in distributions of continuous variables were evaluated by the t test or by the Wilcoxon rank sum test. To estimate relative hazard rates (RR), a proportional hazard regression model was used. The discrete or grouped time analogue of the continuous time proportional hazards model was employed ; the grouped time points of infection being each 4-month follow-up visit. The assignment of status for each risk factor was based on the answer regarding exposures preceding the current visit. Multivariate modeling was carried out using a backward step-down elimination procedure; risk factors that were significant (P < 0.1) on univariate analyses were used in the initial model. Interactions were assessed among covariables in the final model.
We used a competing risk model to analyze risk factors for HIV-1 incidence for each specific HIV-1 subtype (B and E), assuming that infection with the two subtypes were independent . When analyzing incidence of one subtype, seroconversion to the other subtype was treated as a censored observation. To overcome computational problems in the subtype B analysis due to the small number of events, it was necessary to approximate the discrete time proportional hazards model by the logistic model for discrete failure time analysis. As the probability of infection during each 4-month follow-up period was low, the odds ratio was virtually identical to the hazard ratio, this was verified in the subtype E analyses by fitting both models.
Cohort characteristics at enrollment
A total of 3643 persons with a history of injection drug use were screened for enrollment, 1674 in 1995 and 1969 in 1996. The overall HIV-seroprevalence was 29.9% and was similar in the two years, being 29.3 and 30.5%, respectively. In total, 1209 persons were enrolled in the study, 499 in 1995 and 710 in 1996. All participants were Thai, 93.5% were male, the mean age was 31.3 years (range 18–52 years) and 80.2% were between 20 and 39 years old (Table 1). Some form of employment was reported by 71.2% and 54.4% had a monthly income of at least 5000 Thai baht (US$ 200 in 1996). Most IDUs had begun injecting drugs before they were 25 years old, and about half had been injecting for more than 6 years. About two-thirds of IDUs had been arrested during the time they had been injecting drugs. In the previous 6 months, 96.4% had reportedly used heroin, 16.6% had used stimulants (primarily methamphetamine), and 32.1% had used tranquilizers or hypnotic agents. Almost all heroin users had injected, about one-third of stimulant users had injected, whereas few tranquilizer users injected (Table 1). Heroin injection was frequent; 78.5% reported at least daily injection. Almost all (95.9%) IDUs reported that they had obtained new and unused injection equipment at pharmacies or drug stores. However, about one-third reported that in the previous 6 months they had used a syringe or needle that had previously been used by someone else. The level of sexual activity was modest; 44.3% reported no sex with someone of the opposite sex in the previous 6 months, and only 11.2% reported sex with a casual partner. History of urethritis was reported by 15.0% and genital lesion by 23.1%.
At enrollment, prevalence was 95.6% (95% CI, 94.3–96.6) for HCV antibodies, 48.2% (95% CI, 45.0–51.3) for HBsAb, 9.7% (95% CI, 7.9–11.6) for HBsAg, and 6.4% (95% CI, 5.0–7.8) for syphilis RPR and TPHA positivity.
Participants enrolled into the study were similar (P > 0.07) to the HIV-seronegative persons who were screened but not enrolled with regard to sex, education, incarceration history, and history of using syringes/needles used by others. However, in comparison with those not enrolled, the participants who were enrolled were older (mean age 31.7 versus 29.0 years;P < 0.01), more likely to be married (42.4 versus 37.3%;P < 0.01), more likely to be in a methadone maintenance program (18.4 versus 8.2%;P < 0.01), less likely to inject more than once per day [52.6 versus 62.1% (of those reporting injection frequency);P = 0.02], and less likely to report casual sex in the previous 6 months (13.5 versus 17.6%;P = 0.02).
Overall, 1124 (93.0%) participants made at least one visit after enrollment and contributed time to follow-up. There was a total of 2308 PY of HIV-negative follow-up. The median time on study was 29 months (range 4–40 months). Follow-up time was contributed by 88.2% of eligible participants at 12 months, 75.9% at 24 months, and 71.2% at 36 months. However, attendance at specific follow-up visits was made by 81.9% at 12 months, 62.2% at 24 months, and 47.6% at 36 months. Thirty-two HIV-negative cohort members died, yielding a mortality rate of 1.4 per 100 PY (95% CI, 0.95–2.0). Causes of death were drug overdose (34%), traumatic accident (28%), homicide (6%), heart failure (6%), and various other medical conditions (25%).
HIV-1 incidence and subtypes
During follow-up, 133 participants seroconverted to HIV-1, yielding an incidence rate of 5.8 per 100 PY (95% CI, 4.8–6.8). The largest number of seroconversions (35) and the highest incidence rate (9.5 per 100 PY; 95% CI, 6.6–13.2) occurred at the first (4-month) follow-up visit (Table 2). Following the first HIV-seropositive test, additional blood specimens were obtained from 126 (94.7%) of the 133 seroconverters for genetic characterization of the infecting HIV-1 strains. Of these, 99 (78.6%) were env subtype E and 27 (21.4%) were subtype B. Incidence rates were 4.3 per 100 PY (95% CI, 3.5–5.2) for subtype E and 1.2 per 100 PY (95% CI, 0.8–1.7) for subtype B.
Risk factors for HIV-1 seroconversion
On univariate analysis, females, although comprising only 6.5% of the cohort, were at substantially increased risk for HIV-1 seroconversion (Table 3). Compared with males, females tended to be younger (mean age 28.5 versus 31.5 years;P = 0.001), have an income of less than 5000 Thai baht per month (72.1 versus 44.8%;P = 0.001), be married (59.5 versus 42.7%;P = 0.003), and were less likely to have had sex with a casual partner in the previous 6 months (2.5 versus 11.9%;P = 0.01). In total, 43.1% of participants reported incarceration at least once during follow-up, and 11.6% of all follow-up time involved incarceration since the previous visit. In general, increased sexual activity was inversely associated with HIV-1 seroconversion. For women only, as well as for men only, sex with a man was not associated with HIV-1 seroconversion. When the subset of seroconverters who did not report injecting (at the time of seroconversion) were analyzed separately, sexual risk behaviors remained unassociated with HIV-1 seroconversion. A 'heroin injection/sharing scale’ was created to further evaluate the association of frequency of heroin use and sharing of injection equipment with seroconversion (Table 3). On this scale, the four strata were: (1) no heroin injection; (2) less than daily heroin injection and no sharing of drugs or injection equipment; (3) at least daily heroin injection and no sharing; and (4) at least daily heroin injection and reported sharing of drugs or injection equipment. Seroconversion risk increased greatly with the levels on this scale. In a multivariate model the factors that remained independently associated with seroconversion were low income, heroin injecting and sharing as reflected in the 'heroin injection/sharing scale', and incarceration, and especially injecting while incarcerated.
Risk factors for HIV-1 subtypes B and E
Risk factors for seroconversion with HIV-1 subtypes B and E were similar (Table 4). Low income was significantly associated only with subtype E infection. For both subtypes, heroin injection and sharing injection equipment and incarceration, especially with injection, were the strongest risk factors for infection. Sexual activity was not associated with infection for either subtype and there was a trend for sexual activity to be inversely associated with seroconversion for both subtypes.
These data indicate ongoing high HIV-1 infection risk for IDUs in Bangkok a decade after the initial explosive epidemic in this population. Ninety-six percent of participants reported obtaining new syringes and needles from pharmacies. Despite the relatively easy availability of inexpensive, sterile syringes and needles in pharmacies throughout Bangkok and education on injection risk reduction, HIV-1 risk was largely related to the frequency of heroin injection and sharing of injection equipment. A great deal of additional HIV-1 risk was associated with incarceration, especially if drug injection was reported while in jail. Incarceration was reported by 66% of participants before enrollment and by 43% during study follow-up. There is an urgent need for additional prevention efforts to reduce HIV-1 transmission among IDUs in Bangkok, especially in jails and correctional facilities.
The HIV-1 incidence rate of 5.8 per 100 PY among Bangkok IDUs is high in comparison with that of other long-term IDU epidemics. Recently reported HIV-1 incidence rates among IDUs were three to four per 100 PY in Amsterdam  and less than one per 100 PY in New York City . High HIV-1 incidence rates were observed in Bangkok for persons regardless of sex, age, marital status, employment status, duration of drug use, or type of drug treatment program. It should be noted that this study had limited ability to assess the protection provided by methadone treatment as assignment to detoxification or maintenance was not randomized. There was probably some under reporting of risk behaviors as substantial HIV-1 incidence among those reporting no injecting was not explained by sexual behaviors. These data suggest a diffuse, broad HIV-1 epidemic among the drug using population in Bangkok. Earlier studies have indicated that Bangkok IDUs are well informed about HIV/AIDS and are motivated to reduce risk  and the present study indicates that Bangkok IDUs have access to inexpensive, clean injection equipment. However, we suspect that there are substantial barriers to ready access to injection equipment at the time of injection. For example, IDUs may avoid carrying injection equipment when they procure heroin to avoid police scrutiny and arrest. The great risk associated with incarceration warrants special attention. Although the risk associated with incarceration is not fully characterized, it is likely that a large proportion of this risk results from the sharing of drug injection equipment in settings where access to clean syringes and needles is severely limited. However, there may also have been sexual transmission risk that was not fully assessed by the study questionnaire. It is also possible that being incarcerated was a marker for some other risk that occurred before or after incarceration. These data can be used to alert policy makers to the ongoing HIV-1 transmission in jails and prisons and to develop prevention strategies.
Heroin is relatively inexpensive in Bangkok  and was the main drug used by this population; however, at enrollment, use of stimulants (mainly methamphetamine) and sedatives/hypnotics was reported by 17 and 32%, respectively. Although opium and heroin production has sharply declined in Thailand in recent years, the nation remains a major transit route for heroin from the Golden Triangle region of Burma (Myanmar) and Laos . Thailand is also experiencing a large epidemic of methamphetamine use, especially among adolescents and young adults. Although most of this drug use is not by injection, health officials need to be aware of changing patterns of drug use in Thailand.
This study clearly documents the emergence of HIV-1 subtype E as the predominant subtype in the recent epidemic among Bangkok IDUs. Subtype E now accounts for four of five new infections, a dramatic reversal from 10 years earlier when subtype B accounted for a large majority of infections among IDUs . Risk factors were similar for both HIV-1 subtypes, indicating similar parenteral modes of transmission. Increased sexual risk behavior was inversely associated with HIV-1 infection risk for both subtypes B and E. It is likely that the level of risky sexual activity is inversely related to the frequency of heroin injection, which is a far more risky behavior. Factors leading to the increase in subtype E are not completely defined by this study. It is possible that the increase in subtype E is due to the interaction of this IDU population with the larger Thai population that is at sexual risk for HIV-1 infection. As about 2% of the adult Thai population is HIV-infected  and as more than 95% of sexually acquired infections in Thailand are due to subtype E strains , there is ample possibility for mixing of HIV-1 subtype E-infected persons with the IDU population. This increase in subtype E could also be due to greater infectiousness of persons infected with subtype E compared with subtype B. However, it would require additional assumptions for us to estimate the relative transmissibility of subtypes B and E [25,26] as we were not able to directly measure the prevalence of infection with these two subtypes among the population not enrolled in the study. The ongoing study of the persons newly infected in this cohort will evaluate the course of infection with these two subtypes and will perhaps add to our understanding of transmissibility.
In addition to other HIV prevention strategies, a safe and effective HIV-1 vaccine would be extremely useful for this difficult-to-control epidemic in Thailand and elsewhere. This study has demonstrated that a large number of IDUs can be enrolled with reasonably high rates of follow-up. Despite an active HIV counselling and prevention program, this population remains at high-risk for HIV-1 infection . In 1996 we conducted a study in a subset of this cohort to assess comprehension and willingness to participate in an HIV-1 vaccine efficacy trial of a candidate recombinant gp120 bivalent B/E vaccine . Comprehension was high following initial explanation about a trial and improved at follow-up a week later. At baseline, 51% were definitely willing to participate, and at follow-up 54%; with only 3% being unwilling to participate at either time. This study also provided valuable insight into barriers to and positive motives for participation. These efforts led to the development of a proposal to conduct a phase III HIV-1 vaccine efficacy trial among IDUs in this setting. Based on the data from the genetic characterization of the early incident HIV-1 strains from this study [18,28], a bivalent recombinant gp120 HIV-1 vaccine (AIDSVAX B/E; VaxGen, Inc.) based on both a subtype B strain (MN) and a subtype E strain (A244) was formulated for use in this population in Thailand [29,30]. In March 1999, enrollment began into a phase III efficacy trial of this vaccine in Bangkok and full enrollment of 2500 HIV-negative IDUs was completed in August 2000. Results from this trial may be available in 2002.
We thank the voluntary participants of this study and the following personnel of the BMA for their contributions: Dr Chaiyan Kampanartsanyakorn, Dr Thamnoon Vaniyapongs, Dr Orasa Apaiwongse, Dr Raphiphat Kasemsook, Dr Wonchat Subhachaturas, Dr Krit Hiranras, and the staff of the 15 BMA clinics, the BMA laboratory and the BMA IDU cohort study office. From The HIV/AIDS Collaboration we thank Dr Khanchit Limpakarnjanarat, Dr Jaranit Kaewkungwal, Ms Nartlada Chantharojwong, Ms Pongsri Virapat, Ms Natapakwa Skunodom, Ms Thongpoon Chaowanachan, Mr David Puckett, Mr Manas Yaimak, and the staff of the laboratory and administrative units. We also thank Dr Ruengpung Sutthent, Dr Abu Abdul-Quader, Dr Kevin O'Reilly, Dr Saladin Osmanov, and Ms Patricia Friedmann for valuable support; Dr Frits van Griensven, Dr Dale Hu, Dr Ira Longini, Dr R. J. Simonds, and Lynne Stockton for helpful comments and suggestions on the paper; and Ms Orawan Kongrawd for assistance in preparing the manuscript.
1. Schwartländer B, Stanecki KA, Brown T. et al. Country-specific estimates and models of HIV and AIDS: methods and limitations.
AIDS 2000, 13: 2445 –2458.
2. Ou C-Y, Takebe Y, Weniger BG. et al. Independent introduction of two major HIV-1 genotypes into distinct high-risk populations in Thailand.
Lancet 1993, 341: 1171 –1174.
3. Subbarao S, Limpakarnjanarat K, Mastro TD. et al. HIV-1 in Thailand, 1994-1995: persistence of two subtypes with low genetic diversity.
AIDS Res Hum Retroviruses 1998, 14: 319 –327.
4. Limpakarnjanarat K, Ungchusak K, Mastro TD. et al. The epidemiologic evolution of HIV-1 subtypes B and E among heterosexuals and injecting drug users in Thailand, 1992–1997.
AIDS 1998, 12: 1108 –1109.
5. Ungphakorn J, Sittitrai W. The Thai response to the HIV/AIDS epidemic.
AIDS 1994, 8 (Suppl 2): S155 –S163.
6. Phoolcharoen W, Ungchusak K, Sittitrai W, Brown T. Thailand: lessons from a strong national response to HIV/AIDS.
AIDS 1998, 12: S123 –S135.
7. Rojanapithayakorn W, Hanenberg R. The 100% condom program in Thailand.
AIDS 1996, 10: 1 –7.
8. Nelson KE, Celentano DD, Eiumtrakul S. et al. Changes in sexual behavior and a decline in HIV infection among young men in Thailand.
N Engl J Med 1996, 335: 297 –303.
9. Mason CJ, Markowitz LE, Kitsiripornchai S. et al. Declining prevalence of HIV-1 infection in young Thai men.
AIDS 1995, 9: 1061 –1065.
10. Bunnell RE, Yanpaisarn S, Kilmarx PH. et al. HIV-1 seroprevalence among childbearing women in northern Thailand: monitoring a rapidly evolving epidemic.
AIDS 1999, 13: 509 –515.
11. Esparza J, Osmanov S, Kallings LO, Wigzell H. Planning for HIV vaccine trials: the World Health Organization perspective.
AIDS 1991, 5: S159 –S163.
12. Mastro TD, Kitayaporn D, Weniger BG. et al. Estimating the number of HIV-infected injection drug users in Bangkok: a capture-recapture method.
Am J Public Health 1994, 84: 1094 –1099.
13. Kitayaporn D, Uneklabh C, Weniger BG. et al. HIV-1 incidence determined retrospectively among drug users in Bangkok, Thailand.
AIDS 1994, 8: 1443 –1450.
14. Wasi C, Herring B, Raktham S. et al. Determination of HIV-1 subtypes in injecting drug users in Bangkok, Thailand, using peptide-binding enzyme immunoassay and heteroduplex mobility assay: evidence of increasing infection with HIV-1 subtype E.
AIDS 1995, 9: 843 –849.
15. Kalish ML, Baldwin A, Raktham S. et al. The evolving molecular epidemiology of HIV-1 envelope subtypes in injecting drug users in Bangkok,Thailand: implications for HIV vaccine trials.
AIDS 1995, 9: 851 –857.
16. MacQueen KM, Vanichseni S, Kitayaporn D. et al. Willingness of injection drug users to participate in an HIV vaccine efficacy trial in Bangkok, Thailand.
J Acquir Immun Defic Syndr 1999, 21: 243 –251.
17. Kitayaporn D, Vanichseni S, Mastro TD. et al. Infection with HIV-1 subtypes B and E in injecting drug users screened for enrollment into a prospective cohort in Bangkok, Thailand.
J Acquir Immun Defic Syndr Hum Retrovirol 1998, 19: 289 –295.
18. Subbarao S, Vanichseni S, Hu DJ. et al. Genetic characterization of incident HIV type 1 subtype E and B strains from a prospective cohort of injecting drug users in Bangkok, Thailand.
AIDS Res Hum Retroviruses 2000, 16: 699 –707.
19. Prentice RL, Gloeckler LA. Regression analysis of grouped survival data with applications to breast cancer data.
Biometrics 1978, 34: 57 –67.
20. Lunn M, McNeil D. Applying Cox regression to competing risks.
Biometrics 1995, 51: 524 –532.
21. Coutinho R. Interventions among injecting drug users: the Amsterdam experience.
Presented at Fourth Science Forum: Research Symposium on the Prevention of HIV in Drug Users
. Flagstaff, Arizona, 1997.
22. Des Jarlais DC, Marmor M, Friedmann P. HIV incidence among injecting drug users in New York City, 1992–1997: evidence for a declining epidemic.
Am J Public Health 2000, 90: 352 –359.
23. Des Jarlais DC, Choopanya K, Vanichseni S. et al. AIDS risk reduction and reduced HIV seroconversion among injection drug users in Bangkok.
Am J Public Health 1994, 84: 452 –455.
24. Beyrer C, Razak MH, Lisam K, Chen J, Lui W, Yu X-F. Overland heroin trafficking routes and HIV-1 spread in south and south-east Asia.
AIDS 2000, 14: 75 –83.
25. Mastro TD, Kitayaporn D. HIV-1 transmission probabilities: estimates from epidemiological studies.
AIDS Res Hum Retroviruses 1998, 14 (Suppl 3): S223 –S227.
26. Mastro TD, Kunanusont C, Dondero TJ, Wasi C. Why do HIV-1 subtypes segregate among persons with different risk behaviors in South Africa and Thailand?
AIDS 1997, 11: 113 –116.
27. Des Jarlais DC, Vanichseni S, Marmor M. et al. "Why I am not infected with HIV": implications for long-term HIV risk reduction and HIV vaccine trials.
J Acquir Immune Defic Syndr Hum Retrovirol 1997, 16: 393 –399.
28. Phan K-O, Callahan ME, Vanichseni S. et al. A comparison of full-length glycoprotein 120 from incident HIV type 1 subtype E and B infections in Bangkok injecting drug users with prototype E and B strains that are components of a candidate vaccine.
AIDS Res Hum Retroviruses 2000, 16: 1445 –1450.
29. Berman PW. Development of bivalent rgp120 vaccines to prevent HIV type 1 infection.
AIDS Res Hum Retroviruses 1998, 14: S277 –S289.
30. Francis DP, Gregory T, McElrath MJ. et al. Advancing AIDSVAXTM to phase 3.
:Safety, immunogenicity, and plans for phase 3.
AIDS Res Hum Retroviruses 1998, 14: S325 –S331.