HiV Articles  
Antiretroviral Prophylaxis for HIV Prevention in Heterosexual Men and Women
  Download the PDF here

Download the PDF here

J.M. Baeten, D. Donnell, P. Ndase, N.R. Mugo, J.D. Campbell, J. Wangisi, J.W. Tappero, E.A. Bukusi, C.R. Cohen, E. Katabira, A. Ronald, E. Tumwesigye, E. Were, K.H. Fife, J. Kiarie, C. Farquhar, G. John-Stewart, A. Kakia, J. Odoyo, A. Mucunguzi, E. Nakku-Joloba, R. Twesigye, K. Ngure, C. Apaka, H. Tamooh, F. Gabona, A. Mujugira, D. Panteleeff, K.K. Thomas, L. Kidoguchi, M. Krows, J. Revall, S. Morrison, H. Haugen, M. Emmanuel-Ogier, L. Ondrejcek, R.W. Coombs, L. Frenkel, C. Hendrix, N.N. Bumpus, D. Bangsberg, J.E. Haberer, W.S. Stevens, J.R. Lingappa, and C. Celum for the Partners PrEP Study Team

NEJM July 11, 2012


Antiretroviral preexposure prophylaxis is a promising approach for preventing human immunodeficiency virus type 1 (HIV-1) infection in heterosexual populations.


We conducted a randomized trial of oral antiretroviral therapy for use as preexposure prophylaxis among HIV-1-serodiscordant heterosexual couples from Kenya and Uganda. The HIV-1-seronegative partner in each couple was randomly assigned to one of three study regimens - once-daily tenofovir (TDF), combination tenofovir-emtricitabine (TDF-FTC), or matching placebo - and followed monthly for up to 36 months. At enrollment, the HIV-1-seropositive partners were not eligible for antiretroviral therapy, according to national guidelines. All couples received standard HIV-1 treatment and prevention services.


We enrolled 4758 couples, of whom 4747 were followed: 1584 randomly assigned to TDF, 1579 to TDF-FTC, and 1584 to placebo. For 62% of the couples followed, the HIV-1-seronegative partner was male. Among HIV-1-seropositive participants, the median CD4 count was 495 cells per cubic millimeter (interquartile range, 375 to 662). A total of 82 HIV-1 infections occurred in seronegative participants during the study, 17 in the TDF group (incidence, 0.65 per 100 person-years), 13 in the TDF-FTC group (incidence, 0.50 per 100 person-years), and 52 in the placebo group (incidence, 1.99 per 100 person-years), indicating a relative reduction of 67% in the incidence of HIV-1 with TDF (95% confidence interval [CI], 44 to 81; P<0.001) and of 75% with TDF-FTC (95% CI, 55 to 87; P<0.001). Protective effects of TDF-FTC and TDF alone against HIV-1 were not significantly different (P=0.23), and both study medications significantly reduced the HIV-1 incidence among both men and women. The rate of serious adverse events was similar across the study groups. Eight participants receiving active treatment were found to have been infected with HIV-1 at baseline, and among these eight, antiretroviral resistance developed in two during the study.


Oral TDF and TDF-FTC both protect against HIV-1 infection in heterosexual men and women. (Funded by the Bill and Melinda Gates Foundation

The use of antiretroviral medications for the prevention of HIV type 1 (HIV-1) transmission is a promising strategy for reducing the spread of HIV-1.1-4 Antiretroviral treatment for persons infected with HIV-1 provides important clinical benefits and substantially reduces infectiousness.5-7 Antiretroviral prophylaxis is a potential HIV-1-prevention strategy for those not yet infected with HIV-1, administered either as postexposure prophylaxis after high-risk occupational or nonoccupational exposure or as preexposure prophylaxis in those with ongoing HIV-1 exposure.8,9 The rationale for antiretroviral prophylaxis in persons with ongoing exposure is based on its efficacy in infants exposed to HIV-1 during birth and breast-feeding10 and the partial or full protection it confers against mucosal simian HIV challenge in primates.11 In perinatal-transmission studies and animal models, the protective benefits of antiretroviral prophylaxis were maximized when the antiretroviral medication was administered both before and after HIV exposure.12

The efficacy of preexposure prophylaxis for HIV-1 protection in humans has been evaluated for tenofovir, in the form of a vaginal gel or as oral tenofovir disoproxil fumarate (TDF) or oral TDF coformulated with emtricitabine (TDF-FTC). Studies in animal models suggest that TDF-FTC provides greater protection against HIV-1 than TDF alone.11 The possibility of differential efficacy, safety, and cost suggests that TDF and TDF-FTC could be compared as potential preexposure prophylaxis agents. Persons at ongoing risk for HIV-1 acquisition in whom preexposure prophylaxis could be studied include persons who are HIV-1-seronegative but are in a partnership with a person already infected with HIV-1 (an HIV-1-serodiscordant partnership).13,14 We conducted the Partners Preexposure Prophylaxis (PrEP) Study, a multisite, phase 3, randomized, double-blind, three-group, placebo-controlled trial of daily oral TDF or TDF-FTC given as preexposure prophylaxis against HIV-1 acquisition among East African heterosexual men and women in HIV-1-serodiscordant partnerships.


In this study of heterosexual men and women with a partner known to have HIV-1 infection, once-daily oral TDF and TDF-FTC were associated with risk reductions of 67% and 75%, respectively, against HIV-1 infection when provided in conjunction with other HIV-1 prevention services. Both TDF and TDF-FTC showed significant, and a similar magnitude of, HIV-1 protection for both women and men.

Clinical trials of tenofovir-based preexposure prophylaxis have had conflicting results. Once-daily oral TDF-FTC reduced the risk of HIV-1 acquisition by 44% in a multicountry study among men who have sex with men and by 62% among young heterosexuals from Botswana,20,21 and the use of 1% tenofovir vaginal gel decreased the incidence of HIV-1 among South African women by 39%.22 Biologic and behavioral hypotheses have been proposed to explain the failure of two trials of preexposure prophylaxis among African women to show protection against HIV-1 infection,23,24 including a lack of adherence to daily doses of preexposure prophylaxis, vaginal concentrations of tenofovir achieved with oral dosing that may be particularly sensitive to nonadherence,25 sexually transmitted infections or other cofactors affecting infection with HIV-1 in young women, high HIV-1 concentrations in the seropositive partner during primary HIV-1 infection, and innate or acquired immunologic factors that may provide adjunctive protection in long-term couples with HIV-1 serodiscordance. Further study is needed to understand which, if any, of these factors influence the efficacy of preexposure prophylaxis.

Although we studied established couples known to be HIV-1-serodiscordant, all HIV-1 transmissions ultimately occur between serodiscordant partners. Our findings provide proof of concept that preexposure prophylaxis can reduce HIV-1 acquisition in heterosexual populations.

High adherence is essential to achieve clinical benefits from antiretroviral agents for HIV-1 treatment,26 and emerging evidence suggests that adherence to preexposure prophylaxis is also important for HIV-1 prevention. In the Preexposure Prophylaxis Initiative trial involving men who have sex with men, the relative reduction in the risk of HIV-1 infection from TDF-FTC preexposure prophylaxis was 44% overall but was 73% among the participants with an adherence of 90% or more (as measured by means of pill counts) and 92% among the participants with detectable tenofovir levels in the blood - although only half the participants had detectable levels.20

In our study, retention and pill-count adherence were high, tenofovir was detected in 82% of samples from randomly selected participants, and detectable tenofovir levels were associated with a reduction in the relative risk of HIV-1 infection of more than 85%. The high proportion of samples with detectable tenofovir levels is consistent with the 92% study-drug coverage we calculated on the basis of missed visits, withholding of the study drug, and nonadherence, with the absolute difference of 10 percentage points most likely reflecting the fact that pill counts can overestimate adherence if pills are not returned.

Analyses of objective adherence measures across preexposure prophylaxis trials will be informative for understanding the relationship between adherence and protection against HIV-1 infection. In a subgroup of our study cohort, intensive monitoring of adherence by means of pill bottles with caps that electronically monitor bottle openings and monthly unannounced visits to the home for purposes of pill counting supported high adherence,27 and in-depth interviews have emphasized that trust and the support of a partner reinforce high adherence.28 Strategies to promote and achieve high adherence outside clinical-trial settings will be necessary to achieve maximum public health benefits of preexposure prophylaxis.

We found similar degrees of protection against HIV-1 with TDF and TDF-FTC, in contrast to findings in studies of animal models.11 Dual-agent preexposure prophylaxis would most likely be more expensive than single-agent preexposure prophylaxis, and the potential for differential tolerability and antiretroviral resistance in persons with HIV-1 seroconversion despite the use of preexposure prophylaxis should be considered in decision making with regard to public health policies regarding preexposure prophylaxis. We are continuing the TDF and TDF-FTC groups of our study, including offering randomization to TDF or TDF-FTC to participants originally assigned to the placebo group, to gather additional information on the relative safety, efficacy, and HIV-1 resistance of TDF as compared with TDF-FTC.

In our study, 25% (two of eight) of participants who had acute HIV-1 infection at the time of study-drug initiation had viral resistance develop (through the M184V mutation in one and the K65R mutation in the other). The initiation of preexposure or postexposure prophylaxis in persons with acute HIV-1 infection can select for resistance; strategies to improve the recognition of acute infection are needed.20,29,30 Resistance was rare in partners in whom seroconversion occurred after randomization, of whom a minority had detectable tenofovir levels.

Adherence to preexposure prophylaxis, protection against HIV-1 infection, and antiretroviral resistance appear to be tightly interwoven. Low adherence provides little HIV-1 protection but little risk of resistance if infection is acquired. High adherence potentially blocks most transmissions, and the few persons who acquire HIV-1 despite preexposure prophylaxis potentially have an increased risk of drug resistance. Four participants with HIV-1 seroconversion in our study became infected with HIV-1 that was resistant to nonnucleoside reverse-transcriptase inhibitors, which should not have been selected for by the study medication and instead probably reflects circulating resistance, which is increasingly being detected in Africa.31

When used for HIV-1 treatment, TDF is known to cause small decreases in glomerular filtration that are of uncertain clinical significance.32 In our population of HIV-1-seronegative participants without preexisting renal impairment, we found no evidence of clinically significant elevations in serum creatinine. Additional studies are needed of proximal renal tubular function, bone mineral density, and other aspects of long-term safety of TDF-based preexposure prophylaxis, as well as safety in pregnant, breast-feeding, or adolescent women, among whom HIV-1 rates are high.33,34

For couples known to have HIV-1 serodiscordance, antiretroviral treatment of the partner infected with HIV-1 provides substantial, though incomplete, protection against HIV-1 transmission; 25 to 30% of HIV-1 infection in serodiscordant couples are from infected partners outside the couple.5,7 Mathematical modeling may help guide policy decisions regarding optimal targeting and timing of treatment and preexposure prophylaxis for reducing HIV-1 incidence in couples.35 Antiretroviral-based HIV-1-prevention strategies may be particularly important for couples seeking to have children.36-38 In addition, preexposure prophylaxis offers an HIV-1 prevention strategy for uninfected persons with partners who do not know their HIV-1 status or who are infected with HIV-1 but have not begun antiretroviral therapy.

Successful prevention of HIV-1 infection on a population scale will need to incorporate multiple, evidence-based biomedical and behavioral strategies to achieve maximum benefits. The HIV-1 incidence in this study was lower than that seen in previous studies of HIV-1-serodiscordant African couples,14,39 emphasizing the importance of and synergy among HIV-1 testing in individuals and couples, risk-reduction counseling, and other prevention services, in combination with antiretroviral preexposure prophylaxis, for reducing the risk of HIV-1 infection in heterosexual populations. Potential implementation of preexposure prophylaxis as a public health measure will require clinical monitoring, methods for encouraging adherence, and ensured access to antiretroviral therapy for HIV-1-infected persons. Nonetheless, to stem the global HIV-1 epidemic, effective primary HIV-1 prevention strategies are critical.


Study Participants

We screened 7856 couples with discordant HIV-1 serostatus. We enrolled 4758 couples and followed 4747: 1584 randomly assigned to TDF, 1579 to TDF-FTC, and 1584 to placebo (Figure 1 and (Table 1Among participants seropositive for HIV-1, the median CD4 count was 495 cells per cubic millimeter (interquartile range, 375 to 662), 80% had a CD4 count of 350 cells or more per cubic millimeter, and the median plasma HIV-1 RNA level was 3.9 log10 copies per milliliter (interquartile range, 3.2 to 4.5). Overall, baseline characteristics were similar across the three study groups.

Follow-up and Adherence

Retention was 96% or greater during the study period (Figure 1), with 4722 of the 4747 followed participants (99.5%) completing at least one postrandomization HIV-1 test, for a total of 7830 person-years of follow-up for the assessment of HIV-1 incidence accrued (median, 23 months; interquartile range, 16 to 28; range, 1 to 36). Study medication was dispensed at 96% of the attended visits. The most common reason for not dispensing study medication was pregnancy (with an incidence of 11.9, 8.8, and 10.0 pregnancies per 100 woman-years in the TDF, TDF-FTC, and placebo groups, respectively; P>0.05). Time off the study medication due to pregnancy and breast-feeding accounted for 5.3% of the follow-up time among women (2.0% among all participants). Study-medication interruptions for safety-related reasons accounted for less than 1% of the overall follow-up time: 0.6% in the TDF group, 0.7% in the TDF-FTC group, and 0.6% in the placebo group.

The primary study measure of adherence was monthly counts of the returned study bottles and tablets: 98% of the dispensed study bottles were returned, and 97% of dispensed study tablets were taken (see Table S7 in the Supplementary Appendix). Factoring in missed visits, all reasons for nondispensation of study medication, and nonadherence to dispensed study pills, we calculated that study medication was in use during 92.1% of the total follow-up time.

Effect of TDF and TDF-FTC on HIV-1 Acquisition

HIV-1 seroconversion was observed in 96 participants, of whom 14 had plasma HIV-1 RNA retrospectively detected in specimens obtained at enrollment (5 receiving TDF, 3 receiving TDF-FTC, and 6 receiving placebo) (Figure 1, and Figure S1 in the Supplementary Appendix). Of 82 HIV-1 infections developing after randomization, 17 were in the TDF group, 13 were in the TDF-FTC group, and 52 were in the placebo group, indicating relative reductions in the rates of HIV-1 acquisition of 67% due to TDF (95% confidence interval [CI], 44 to 81; P<0.001) and 75% due to TDF-FTC (95% CI, 55 to 87; P<0.001), each relative to placebo (Figure 2). The HIV-1-protective effects of TDF-FTC and TDF were not significantly different (P=0.23). With both TDF (P=0.003) and TDF-FTC (P<0.001), efficacy of less than 30% was ruled out in the primary modified intention-to-treat analysis. The intention-to-treat analysis including data for participants who were infected with HIV-1 at randomization yielded similar results (Figure 3).

As compared with placebo, among women, the efficacy of TDF was 71% (P=0.002) and of TDF-FTC 66% (P=0.005); among men, the efficacies were 63% (P=0.01) and 84% (P<0.001), respectively. The HIV-1-protective effects of TDF and TDF-FTC were not statistically different according to sex. Protection against HIV-1 was generally similar between subgroup categories for other prespecified subgroups analyses (Figure 3). During the follow-up period, 21% of the partners seropositive for HIV-1 (22% in the TDF group, 20% in the TDF-FTC group, and 21% in the placebo group) started combination antiretroviral therapy; the HIV-1-protective effects of TDF and TDF-FTC were similar to those observed in the primary modified intention-to-treat analysis if follow-up time after the HIV-1-seropositive partner started antiretroviral therapy was excluded (see Table S8 in the Supplementary Appendix).

Antiretroviral Resistance

Of the 96 persons who had seroconversion to HIV-1 positivity, 92 (96%) had plasma samples available for amplification of the HIV-1 RNA to assess for resistance (Table S9 in the Supplementary Appendix). Among the 8 participants in the TDF and TDF-FTC groups who were found to have been infected at randomization, HIV-1 with resistance to the study medications developed in 2 participants: 1 in the TDF group had a TDF-resistant virus (K65R mutation), and 1 in the TDF-FTC group had an FTC-resistant virus (M184V mutation). No participants who acquired HIV-1 after randomization were infected with an HIV-1 strain with the K65R or M184V mutation.

Detection of Tenofovir and Prophylactic Effect

Among 29 participants in the TDF and TDF-FTC groups who became infected with HIV-1, 31% had a detectable tenofovir level in a plasma sample obtained at the seroconversion visit, as compared with 82% of 902 samples from a random subgroup of 198 participants who did not acquire HIV-1 (Table S10 in the Supplementary Appendix). A detectable level of tenofovir, as compared with an undetectable level of the drug, was associated with estimated reductions in the relative risk of acquiring HIV-1 of 86% (with TDF) and 90% (with TDF-FTC).

Sexual Behavior

At enrollment, 27% of partners seronegative for HIV-1 reported having sex without a condom with their HIV-1-seropositive partner during the prior month. This percentage decreased during the follow-up period (to 13% and 9% at 12 and 24 months, respectively) and was similar across the study groups (see Figure S2 in the Supplementary Appendix). The proportions of participants reporting outside partnerships and acquiring sexually transmitted infections during the follow-up period did not differ significantly across the study groups (see Table S11 in the Supplementary Appendix).

Safety and Adverse Event Profiles

There were no significant differences in the frequency of deaths, serious adverse events, or serum creatinine or phosphorus abnormalities across the study groups (Table 3, and Table S12 in the Supplementary Appendix). Neutropenia was seen more commonly in the TDF-FTC group (17% of participants with a grade 1 or 2 event and 1% with a grade 3 or 4 event) (see Table S13 in the Supplementary Appendix) than in the TDF group (15% of participants with a grade 1 or 2 event and 1% with a grade 3 or 4 event) or the placebo group (12% of participants with a grade 1 or 2 event and 1% with a grade 3 or 4 event). The active study medications were associated with modestly increased reports of gastrointestinal side effects and fatigue as compared with placebo, primarily during the first month of administration (see Table S14 in the Supplementary Appendix).

  iconpaperstack view older Articles   Back to Top