Since 1987, research on preventive HIV vaccines has evaluated a large number of vaccine candidates administered to many thousands of volunteers throughout the world in phase-I, phase-II, and phase-III clinical trials. However, the trials are of limited duration and are unable to evaluate the long-term safety of the vaccine candidates and the impact of participation in these trials. Only 2 studies in the United States in the 2000s,1 and more recently in Africa,2 looked at long-term tolerance to the vaccine candidates in the participants.
Given its experience since the 1990s, the research conducted by the ANRS in France may contribute to enriching the knowledge in this domain. Furthermore, the ANRS has developed a vaccine research strategy based on a specific network of volunteers, which is intended to develop HIV vaccines that are able to induce a cellular immune response. However, some vaccine candidates administered to volunteers from the network of the ANRS were able to induce a sustained humoral immune response, detected by standard screening tests, called “vaccine-induced HIV seropositivity.”3 This seropositivity linked to the production of HIV antibodies may involve certain problems regarding the volunteer's friends and family and his or her social life.4 Long-term tolerance to the vaccines, the effects of induced seropositivity and the psychological, social, and behavioral consequences of participation in these trials merit the study for the ANRS volunteers who received vaccine candidates from the early 1990s to the late 2000s.
The Creation of the COHVAC Postvaccine Trial Cohort
After a meta-analysis of the clinical tolerance of a class of vaccine candidates, HIV lipopeptides, which brought together the data from 10 trials conducted in France since 1996,5 the ANRS COHVAC cohort was created in 2008 (the study of a cohort of volunteers who participated in HIV preventive vaccine trials) with the main purpose of evaluating long-term tolerance. It provides access to the most exhaustive collection possible of adverse events that occurred in the long term among the volunteers who participated in vaccine trials of the ANRS. However, the objective is primarily descriptive because of the difficulty of relating adverse events to a vaccine candidate at a distance and conducting incidence comparisons with the general population.
One of the objectives of COHVAC is to evaluate the occurrence of long-term adverse social and behavioral events for the participants in the vaccine trials. This involves measuring whether or not the subjects, who were initially selected with a very low-risk profile for HIV infection, retained this profile, as well as collecting and measuring any negative experiences related with the trial which the participant had himself or in relation to his family members, colleagues, and friends.
Monitoring of the COHVAC cohort is annual and has a duration of 7 years. It also includes the collection of retrospective data from the inclusion of the volunteer in the vaccine trial and during the postvaccine trial period before monitoring in the cohort.
Volunteer Network of the ANRS
To conduct the first vaccine trials, the ANRS set up a network of volunteers across all of France in 1992. The volunteers became members of the “ANRS Volunteer Network” if they were between the ages of 21 and 54 years and had undergone a certain number of medical and psychological “selection” examinations that established whether the person was seronegative for HIV, was in good health, was motivated to participate in a trial that often requires 1 year of monitoring, and was at low risk of contamination by HIV. The selection was appreciated by a multidisciplinary committee (clinical practitioners and psychologists)6 (see also Linard and Giami in this Journal issue).
After being included in the network, the ANRS asked the volunteers to take part in a preventive vaccine trial for no financial compensation apart from their travel expenses. The inclusion criteria were the same as the selection criteria, particularly seronegativity and low exposure to HIV, and were validated for a second time at each investigation center. From 1992 to 2009, 17 trials were conducted with different types of vaccine candidates [recombinant HIV envelope proteins (rgp160) from 1992 to 1993, ALVAC-HIV canarypox vectors (vCP) expressing HIV-1 Env, Gag, Pro, and Pol genes from 1994 to 2001, associated or not with HIV-1 lipopeptides representing CTL epitopes of Gag, Pol, and Nef from 1997 to 2007 and finally, DNA vaccine carrying the genes of an HIV-1 clade C virus expressing the gp120 envelope protein and a Gag-Pol-Nef polyprotein combined with NYVAC-C, recombinant virus expressing Env, Gag, Pol, and Nef genes of an HIV-1 clade C virus in 2008 and 2009] and various routes of administration at 2 investigation centers of the ANRS (Hôpital Cochin and Hôpital Tenon in Paris), as well as at 4 additional centers as of 2004, in the Paris region (Hôpital Henri Mondor in Créteil) and outside the Paris region (Hôtel Dieu in Nantes, Hôpital Purpan in Toulouse, and Hôpital Sainte-Marguerite in Marseilles). Once the trials were completed, the ANRS organized a posttrial follow-up for the primary purpose of monitoring the humoral vaccine response and for the secondary purpose of recording serious adverse events, but without a sufficiently precise framework for the data to be analyzed. In addition, in 2007, this follow-up did not concern all the trials, and especially not the most recent ones, because of the planned creation of the cohort.
Cohort of Volunteers Who Participated in HIV Preventive Vaccine Trials
Particularly, because of the absence of any record of data from the postvaccine trial follow-up of the ANRS, and to homogenize the evaluation of the long-term consequences of participation in vaccine trials, all the volunteers who received at least one dose of a vaccine candidate (excluding placebo) were asked to join the COHVAC cohort for a prospective annual follow-up over a period of 7 years and to collect their retrospective data since they began to take part in the vaccine trials. For volunteers who were deceased, could not be located, or who did not respond to invitation letters, data were collected without their consent from the files available at the investigation centers or databases of the trials by derogation of the Commission Nationale de l'Informatique et des Libertés (CNIL). Volunteers were also given the option to consent solely to the retrospective collection of their data without inclusion in the prospective follow-up to make the participation in the cohort as exhaustive as possible.
Social and Behavioral Questionnaires
The social and behavioral data of the participants in HIV preventive vaccine trials were analyzed using several questionnaires. Based on the file of each volunteer, a history of the selection data included a description of the sexuality of the volunteer, his motivation, and the selection committee's assessment before entry into the network. The volunteers were divided into 3 categories of approval: “without reservations,” “with reservations,” and excluded. Approval “with reservations” did not result in exclusion but was intended to draw attention to certain points for revision at the time of preinclusion in the trial. In particular, this reservation could concern the volunteer's behavior involving exposure to a sexually transmittable infection (epidemiological reservation), psychology, or biological tests.
On inclusion in the cohort, a face-to-face questionnaire was conducted to identify the risk factors for exposure to HIV as well as the psychological and social impact and the problems connected with donating blood and bone marrow. These data were collected beginning from the last visit made or until the date of the last contact for volunteers who were deceased, could not be reached, and did not respond or who agreed to the retrospective data collection only. A self-administered questionnaire was also offered to all the volunteers present in the inclusion in the prospective cohort regarding their emotional and social situation, their sexual behaviors, and the consequences of participating in a preventive HIV vaccine trial. By means of closed-ended questions, the participants were asked to specify the problems they had encountered without putting these into a hierarchical order or describing the particular details. A part of the self-administered questionnaire is based on a methodology used in the ANRS-VESPA investigation and measures risky sexual behavior regarding HIV exposure.7 The indicators evaluated are the frequencies of the subjects who had occasional partners or unprotected sex in the past 12 months outside of their stable partnership. This defines a risk of exposure to HIV through sexual practices, a definition that was chosen during the design of COHVAC. During the prospective follow-up, the face-to-face questionnaires and the self-administered questionnaires were conducted annually.
During the monitoring of the cohort, a self-administered “refusal” questionnaire concerning just the respondent's social and emotional situation and the consequences of previous participation was sent to the volunteers who had not responded or had refused to take part prospectively in the cohort.
Principles of Analysis
The analysis of the questionnaires was performed from inclusion to the end of monitoring in the cohort. The collection of historical data regarding selection involved all the volunteers who had participated in vaccine trials (excluding placebo), whether or not they had been included prospectively, and whose selection files were available at the investigation center. The data concerning the self-administered “refusal” questionnaires, including the reason for nonparticipation in the prospective cohort, which was subsequently sent to volunteers could not be reached, did not respond who or who had consented to a retrospective collection of data only, were analyzed separately from the volunteers who had been included prospectively to highlight any differences. Data regarding sexual behavior were presented according to sex. The characteristics were compared as much as possible with reference data of the general population or the results of studies on sexuality.
In the tables presented here, the total numbers (N) for each item vary because of missing responses. The numbers of respondents (n) for the various responses to the items are presented as total numbers (N) to obtain a percentage per response. For certain items, the sum of the percentages is higher than 100 because multiple responses are possible. For the results of the assessment with reservations of the selection committee and for the study of the experience of problems connected with inclusion, univariate and multivariate logistic regressions were used to search for explanatory factors.
Inclusions in the Cohort
From 1992 to 2007, several thousand volunteers participated in the selection process in the “ANRS Volunteer Network” with a selection rate of 10%–15%. A total of 534 volunteers were then included in 1 of the 17 vaccine trials conducted under the aegis of the ANRS from 1992 to 2009, of whom 38 received a placebo in the ANRS VAC18 trial (Fig. 1). From December 2008 to January 2013, 488 (98%) of the 496 volunteers who received a vaccine candidate were included in the COHVAC cohort. One hundred thirty-three (27%) volunteers contributed to the retrospective follow-up only, of whom 5 died before the creation of the cohort (3 from cancer, 1 from amyotrophic lateral sclerosis, and 1 by drowning) and 355 volunteers (73%) also participated in the prospective follow-up beginning in year 0 (Y0). In 2016, in the absence of any safety signal since the start of inclusions in 2008, the monitoring in COHVAC was stopped prematurely and the last follow-up visit took place at the end of September 2016. The follow-up visits at 1, 2, 3, and 4 years were not affected by this termination; the follow-ups at 5, 6, and 7 years were not performed for this reason by 13%, 21%, and 26% of the volunteers, respectively.
Data on Selection in the Network
Of the 488 inclusions in COHVAC, 462 questionnaires (95%) were completed (Table 1). Fifty-four percent of the selected volunteers are men and 46% are women. Their median age is 44 years (22–54 years old). The median period between selection and participation in the vaccine trial is 1 year but can be as much as 7 years, and for 25% of volunteers, this period was more than 2 years. At the time of their inclusion in the vaccine trial, the volunteers had a median age of 45 years, and 50% of them were between 39 and 50 years old.
About 80% of the selected volunteers were part of a couple. Those who said that they had at least one partner of the same sex (homo/bi) made up 15.8% of the men and 10.1% of the women selected, and the volunteers with multiple partners (at least 2 partners) in the past year made up 4% of the male and 1.6% of the female heterosexuals. Of homosexual men, 11% said that they had had at least 2 partners in the past year, whereas 50% of bisexual men said that they had had at least 2 partners (no distinction of sex). Fifty-four percent of the selected volunteers are blood donors; 17% are registered in bone marrow registries. The motivations are altruism in 85% of cases, a current proximity with a person living with HIV in 47% of cases, and “for the next generation” in 48% of cases. These reasons do not differ between men and women. The other motivations reported concern and interest in research or medicine (n = 8), aid to Africa (n = 6), a connection with the HIV/AIDS community (n = 4), and a desire to be of service (n = 5).
Overall, the agreement of the committee was obtained “without reservations” for 78% of the volunteers, “with reservations” for 15%, and was not obtained for 7% of the volunteers; 17% of the men and 12% of the women were selected “with reservations” (not significant). The reasons given for the 69 volunteers selected “with reservations” are mainly biological (39%) and epidemiological (27%) in nature.
In the group of men, a study of the predictive factors of agreement “with reservations” vs “without reservations” highlighted a certain number of factors that tend to be associated (younger age, number of partners in life, partners in the past 12 months, not living as a couple, and homo/bisexual vs heterosexual sex) in the univariate analyses. The sole fact of having at least 2 partners in the past 12 months is close to the significant threshold of 5% in the multivariate analysis (data not presented). Because of missing data, this analysis is based on 193 volunteers out of 251 (77%) completed questionnaires. In the group of women, none of the behavioral factors are associated with agreement “with reservations.”
Characteristics of the Participants Included in COHVAC
The inclusions of 355 volunteers in the prospective cohort occurred between December 2008 and July 2012. The median period between the start of the trial and inclusion in the cohort is 5.1 years (2.2–18.4 years). A self-administered questionnaire for inclusion in the prospective follow-up was collected for 353 subjects, that is, 72% of the 488 volunteers who received a vaccine candidate during the trials (Table 2). These respondents are 52 years old on average (25–71 years old), 55% are men, 87% have a level of education higher than or equal to the Baccalaureate, 80% are in employment (of whom 93% are 24–54 years old and 69% are 55–64 years old), 72% live with a partner, 43% have children living in the household, 74% own their home, 76% are involved in an association, political party, or labor organization, 40% are registered blood donors and 18% as bone marrow donors, and 62% say that religion is not important at all in their lives. About half of the subjects say that they know personally one or more people living with HIV. This person is most frequently a friend. Of the 14 volunteers who refused to take part in the prospective follow-up and the 119 participants who gave no response regarding inclusion in the cohort, 8 and 10, respectively, refused to do the self-administered questionnaire. Of this limited sample, the characteristics of the volunteers who took part in the trials but not in the follow-up of the cohort are similar, except for a longer period since participation in the trial (Table 2).
Psychological and Social Impact Connected With Participation in a Trial
The experience of the participants and the social consequences are collected by means of a self-administered questionnaire, and a face-to-face questionnaire conducted at the time of inclusion in the cohort and annually for a period of up to 24 years after the start of the trial (Table 3).
Although they were encouraged to do so once they were selected, 99% of the respondents to the self-administered questionnaire talked about their participation in the trial at the time of inclusion to at least one person, their spouse (94%), their family (95%), friends (94%), and colleagues (81%), with no change in these percentages over the course of the years. Twenty-six percent of the participants encountered problems from their spouse (6%), at work (7%), or from family members (11%) at least once during the follow-up. Fifteen volunteers (4%) reported problems from banks or insurance companies, of whom 9 are participants in the first trials inducing HIV seropositivity. In the face-to-face questionnaire, 75 volunteers (21%) reported a problem connected with the donation of blood and 18 volunteers (5%) reported a problem connected with the donation of bone marrow. When these 2 problems are excluded, the same small percentage of volunteers (19%–26%) who express at least one problem with people in their lives or society was recorded in the self-administered questionnaire and the face-to-face questionnaire. Four volunteers reported other problems related to medical practitioners (with the treating or occupational health doctor in cases of pregnancy or surgery).
Approximately 82% of the volunteers believed that the trial was a “little constraining” or “not constraining at all” at every visit in the self-administered questionnaire. However, 18% believed that participation was “very constraining” or “constraining” at least once. The free responses explaining these problems were about the travel required to attend the follow-up visits (distance and frequency during the trial) and the required availability, but rarely did they mention painful examinations. In the study of the factors associated with an experience of constraints at inclusion, and by grouping the volunteers according to 3 levels of constraints (a little constraining to very constraining), this is significantly related to recent participation in trials, a high level of education, living as a couple, and having a job (data not presented).
Only 21 volunteers (6%) regretted participating in a trial because of their induced seropositivity and blood donation concerns (n = 5) or subjectively associated health problems (n = 4), lack of information after the trial (n = 1), an interruption in one of the trials (n = 1) or concerns about the people in their lives (n = 1), a fear of difficulty obtaining a loan (n = 1), “I would not do it again” (n = 1), a previous episode of depression (n = 1), and problems with the hospital (n = 1). Overall, concerns of people in their lives were reported at least once by 77% of the participants from family (60%), spouse (43%), friends (44%), and work colleagues (24%). On a positive note, support or admiration from people in their lives was experienced by 86% of the volunteers from friends (77%), spouse (70%), family (72%), and at work (53%). Finally, 20 volunteers (6%) only believed at least once that the experience was not positive. For volunteers who refused the prospective follow-up and those who did not respond regarding inclusion but responded to the self-administered “refusal” questionnaire, the psychological and social impact is comparable with that of the volunteers in follow-up, apart from constraints (very constraining or constraining) expressed by 5/17 volunteers (29%) compared with 9% of volunteers who responded at the start of their follow-up in the prospective cohort. The most frequently mentioned reasons for nonparticipation in the prospective cohort were the required availability, travel, and nonreceipt of the invitation letter. However, some volunteers expressed a desire to be tested nevertheless for seropositivity.
Sexual Behavior and Risk-Taking Regarding HIV
During the different follow-ups (Table 4), the percentage of respondents who reported one (or more) occasional partner(s) during the past 12 months changed from 15% to 8% among men and from 4.3% to 1.6% among women. Of these, 33%–70% of men and 33%–100% of women always used a condom with their occasional partners. Of the 192 male and 162 female respondents, the cumulative number of volunteers who had had at least one occasional partner was 48 men (25%) and 17 women (11%). As such, the same volunteers were not concerned every year. In addition, 36 men (19%) and 11 women (7%) reported unprotected sex. For 20 men (10%) and 2 women (1%), this was with occasional partners of unknown serological status, and for 2 of these men, with partners seropositive for HIV at the time of inclusion in the cohort. In the face-to-face questionnaire, the percentages of volunteers who said that they had new partners (since the last visit) were stable at around 10% of men and 2%–4% of women, with systematic use of a condom by about half of the volunteers concerned. It should be noted that one participant was infected with HIV by homosexual transmission after his third follow-up visit in the cohort. At the time of his selection, he had been classified as “with epidemiological reservation” and during the follow-up in the cohort, he reported protected relations with occasional partners. Finally, a small and stable percentage of volunteers indicated risky sexual practices several years after their selection with a low exposure to HIV.
DISCUSSION AND CONCLUSION
The French cohort of ANRS volunteers provides original data regarding the characteristics of the volunteers and the social and behavioral consequences of their participation in preventive HIV vaccine trials. The data collection began before entry into the trial and ended several years or even 2 decades afterward for a population of close to 3/4 of the trial participants.
Volunteers were first recruited for inclusion in the “Volunteer network of the ANRS” before the start of a trial, and the process selected volunteers whose age at the start of the trial was relatively advanced (median age of 45 years, with 50% of volunteers between the ages of 39 and 50 years), more advanced than that of volunteers recruited in other countries.1,2,8 The selection period lasted for about 15 years between 1992 and 2007, and we compare the data with the results of 2 studies performed among the general population: the ANRS-KABP study (knowledge, attitude, beliefs, and practice) on the AIDS virus, which was conducted from 1992 to 2004,9 and the CSF study (Context of Sexuality in France)10 in 2006. Similar percentages of the volunteers selected were in a couple.8 The percentages of people who reported at least one partner of the same sex were 16% of men and 10% of women, that is, higher than the survey figures (4% of men and 2% of women in 1992 and 4% of men or women in 200610). The percentages of heterosexual volunteers with multiple partners in the previous year (7% of men and 3% of women) were lower in the survey results, where these percentages varied between 21.9% and 14.0% in men and between 9.2% and 8.5% in women from 1992 to 2004.8 Regarding homosexual men, during selection, 11% said that they had had more than one partner in the past 12 months, which is significantly lower than the figure of 63% from the study in 200610 (these comparisons are all statistically significant). The process of the ANRS thus selected volunteers who were mostly in a couple, with a higher proportion of homosexuals than in the people surveyed during similar periods, but the latter reported a lower number of sexual partners and so are probably less exposed to the risk of infection by HIV. In addition, the motivations of the volunteers recorded at the time of selection were mainly altruistic, with a concern for future generations or a relationship with people living with HIV, which is supported by the results of a systematic review of volunteers' motivations to take part in vaccine trials.11 Protection from HIV, which was mentioned inter alia in the early-phase trials,11 was not an acceptable motivation at the time of entry into this network. Financial motivations are obviously absent from our study.
The sociodemographic characteristics of the volunteers at the time of inclusion in the cohort (level of education, employment status, home ownership, involvement in associations, and blood or bone marrow donation) indicate that social status and engagement in society are higher than in the French population (reference data in general population can be found in the legend of Table 2).
After describing the behaviors and motivations of volunteers at the time of selection and their sociodemographic characteristics several years after the trials, we then showed that for the large majority of the volunteers this participation was seen as a positive experience and negative consequences were reported by a minority of participants at the same level as among the volunteers of one of the first efficacy trials in the United States12 and phase I–II trials in Africa2 and Thailand.13 Even if only experienced by a minority, negative consequences have prompted sponsors of vaccine trials to define a prevention, evaluation, and intervention model to take these into account.14,15 Problems with banks or insurance companies due to vaccine-induced seropositivity were reported by 4% of the volunteers in the cohort. Regrets are very rarely mentioned and are mainly connected with health problems, induced seropositivity or the impossibility of donating blood during the trial, which is the first problem mentioned in the face-to-face questionnaire. This is consistent with the predominantly altruistic profile observed in the network's volunteers. In terms of their personal experience, the participants who were the least available because of their work or partner and who had taken part more recently logically expressed more problems related to their participation. Some volunteers who did not take part in long-term follow-up in the cohort reported the same problems in the self-administered “refusal” questionnaire. It should be noted that because the recruitment campaigns were national, the participants come from all over France and spent hours traveling long distances to take part in the trials and the cohort.
Regarding sexual behavior and risk-taking in relation to HIV, the data from the selection process and the self-administered questionnaire in the cohort were not directly comparable because of the differences between the questionnaires. Nevertheless, the percentage of men who reported having at least 2 partners in the 12 months before selection is 7%, and the percentage of men who reported occasional partners in the past 12 months in the self-administered questionnaire is 15% at the time of inclusion in the cohort. This difference may be due to the way in which the data were collected given that the self-administered questionnaire minimizes the desirability bias, especially because the volunteers had already participated in the vaccine trial. Nevertheless, even when doubled, this behavior is not associated with a high risk of contamination by HIV unless the sex was unprotected. As such, the percentage of volunteers who reported risky practices with partners of unknown serological status or who were seropositive for HIV is low (10% of men and 1% of women) even several years after their participation in the trial. However, more recent data show that the risk of HIV transmission from seropositive partners with an undetectable viral load is negligible.16,17 In the data in the literature, only one study from 1997 reports an increase in risky behavior.18 All the later studies show no increase in this behavior among participants in vaccine trials in Canada,19 South Africa,20 and Tanzania.21
The long-term follow-up of participants in preventive HIV vaccine trials of the ANRS shows that most of the participants and the people in their lives readily accepted the trials, and that they understood the information regarding the objectives, risks, and problems. In addition, participation in vaccine trials did not lead to an increase in exposure to HIV infection. Through taking part in the Volunteer Network of the ANRS, they voluntarily demonstrated a remarkable engagement and altruism in their partnership with vaccine research.
The authors thank all the volunteers and the clinical centers: AP-HP Hôpital Cochin, CIC Cochin-Pasteur, Paris (O.L., P. Duchet Niedziolka, L. Belarbi, B. Phung, P. Loulergue, and H. Bodilis), AP-HP Hôpital Tenon, Paris (G. Pialoux, L. Slama, J. Chas, and S. Le Nagat), AP-HP Hôpital Henri Mondor, Créteil (J-D Lelièvre and S. Dominguez), CHU Hôtel-Dieu, Nantes (B. Bonnet, P. Morineau Le Houssine, and N. Feuillebois), Hôpital Sainte-Marguerite CISIH, Marseilles (I. Poizot-Martin, O. Zaegel, and N. Cloarec), and Hôpital Purpan, Toulouse (L. Cuzin and M. Chauveau). Coordinating Investigator: O.L., Hôpital Cochin, Paris. Coordinating Coinvestigators: Benjamin Silbermann, Hôpital Cochin, Paris, and Jean-Daniel Lelièvre, Hôpital Henri Mondor, Créteil. Scientific Committee: O.L., B. Silbermann, J-D Lelièvre, J-P. Aboulker, L. Meyer, C. Durier, C. Desaint, V. Meiffredy, S. Grabar, C. Lewden, L. Slama, B. Bonnet, L. Cuzin, I. Poizot-Martin, A. Brézin, A. Moulignier, O. Lidove, J-P Viard, F. Linard, H. Bertone, V. Doré, B. Spire, A. Krivine, H. Fleury, E. Ziegler, L. Clot, M. Molina, H. Pollard, and A. Bouakane. ANRS: A. Bouakane, V. Doré, and A. Collin. Methodology and Management Unit: INSERM, SC10-US19, Villejuif (J-P. Aboulker, L. Meyer, C. Desaint, C. Durier, Z. Sumer-Yalcin, C. Lascoux, Y. Saïdi, B. Abdelkader, and E. Moreau).
. Gilbert PB, Chiu YL, Allen M, et al. Long-term safety analysis of preventive HIV-1 vaccines evaluated in AIDS vaccine evaluation group NIAID-sponsored phase I and II clinical trials. Vaccine. 2003;21:2933–2947.
. Schmidt C, Jaoko W, Omosa-Manyonyi G, et al. Long-term follow-up of study participants from prophylactic HIV vaccine clinical trials in Africa. Hum Vaccin Immunother. 2014;10:714–723.
. Silbermann B, Tod M, Desaint C, et al. Long-term persistence of vaccine-induced HIV seropositivity among healthy volunteers. AIDS Res Hum Retroviruses. 2008;24:1445–1448.
. Voronin Y, Zinszner H, Karg C, et al. HIV vaccine-induced sero-reactivity: a challenge for trial participants, researchers, and physicians. Vaccine. 2015;33:1243–1249.
. Durier C, Launay O, Meiffrédy V, et al. Clinical safety of HIV lipopeptides used as vaccines in healthy volunteers and HIV-infected adults. AIDS. 2006;20:1039–1049.
. Giami A, Lavigne VAX, g. Motivations et représentations chez les volontaires engagés dans des essais vaccinaux contre le VIH (phase 1). Psychol Française. 1996;41:173–188.
. Bounhik AD, Préau M, Lert F, et al. Unsafe sex in regular partnerships among heterosexual persons living with HIV: evidence from a large representative sample of individuals attending outpatient services in France (ANRS-EN12-VESPA Study). AIDS. 2007;21(suppl 1):S57–S62.
. Schmidt C, Smith C, Barin B, et al. Background morbidity in HIV vaccine trial participants from various geographic regions as assessed by unsolicited adverse events. Hum Vaccin Immunother. 2012;8:630–638.
9. Beltzer N, Lagarde M, Wu-Zhou X, et al. Les connaissances, attitudes, croyances et comportements face au VIH/sida en France. Evolutions 1992–1994–1998–2001–2004. Paris, France: Observatoire Régional de Santé d'Ile de France; 2005.
10. Bajos N, Bozon M. (dir.), Enquête sur la sexualité en France. Pratiques, genre et santé. Paris, France: La Découverte; 2008.
. Dhalla S, Poole G. Motivators to participation in actual HIV vaccine trials. AIDS Behav. 2014;18:263–277.
. Fuchs J, Durham M, McLellan-Lemal E, et al. Negative social impacts among volunteers in an HIV vaccine efficacy trial. J Acquir Immune Defic Syndr. 2007;46:362–368.
. Jenkins RA, Thapinta D, Morgan PA, et al. Behavioral and social issues among volunteers in a preventive HIV vaccine trial in Thailand. J Acquir Immune Defic Syndr. 2005;40:592–599.
14. Allen M, Lau CY. Social impact of preventive HIV vaccine clinical trial participation: a model of prevention, assessment and intervention. Soc Sci Med. 2008;66:945–951.
15. Milford C, Barsdorf N, Kafaar Z. What should South African HIV vaccine trials do about social harms? AIDS Care. 2007;19:1110–1117.
16. Cohen MS, Chen YQ, McCauley M, et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365:493–505.
17. Rodger AJ, Cambiano V, Bruun T, et al. Sexual activity without condoms and risk of HIV transmission in serodifferent couples when the HIV-positive partner is using suppressive antiretroviral therapy. JAMA. 2016;316:171–181.
18. Chesney MA, Chambers DB, Kahn JO. Risk behavior for HIV infection in participants in preventive HIV vaccine trials: a cautionary note. J Acquir Immune Defic Syndr Hum Retrovirol. 1997;16:266–271.
19. Lampinen TM, Chan K, Remis RS, et al. Sexual risk behaviour of Canadian participants in the first efficacy trial of a preventive HIV-1 vaccine. CMAJ. 2005;172:479–483.
20. Gray G, Metch B, Churchyard G, et al. Does participation in an HIV vaccine efficacy trial affect risk behaviour in South Africa? Vaccine. 2013;31:2089–2096.
21. Tarimo EAM, Munseri P, Aboud S, et al. Experiences of social harm and changes in sexual practices among volunteers who had completed a phase I/II HIV vaccine trial employing HIV-1 DNA priming and HIV-1 MVA boosting in dar es salaam, Tanzania. PLoS One. 2014;9:e90938.