HIV treatment as prevention and HPTN 052

INTRODUCTION

The use of antiretroviral therapy (ART) for the treatment of HIV-1 is currently one of the most popular ideas to reduce transmission of HIV [1 – 8]. The strategy has generated a plethora of mathematical models that, for the most part, predict success [9–14,15*,16,17,18**,19 – 21], and several ecological studies that argue that the prevention benefit of ART can already be seen in a few communities where ART usage is high [22,23,24*,25*].

These public health aspirations are truly important. However, policies favoring the use of treatment as prevention have been rapidly moving forward in the absence of a critical piece of information: the degree to which ART actually prevents the transmission of HIV and the durability of the effect(s) of ART for this purpose. The purpose of this article is to review briefly the development and results of a single, multicenter, Phase III, clinical trial designed to ask whether the treatment of HIV infection also prevents the sexual transmission of HIV-1 in HIV-serodiscordant couples.

BIOLOGICAL PLAUSIBILITY

In 2000, researchers from the Rakai (Uganda) project [26] reported that the concentration of HIV in blood plasma could be correlated with HIV transmission probability in HIV-serodiscordant couples. The study concluded that the risk of transmission of HIV with less than 1500 copies of HIV RNA is negligible.

Prior to the Rakai study, however, several reports showed that HIV can also be recovered in genital secretions, as would be predicted based on the sexual transmission of the virus. With the development of the first antiretroviral agent – azidothymidine – interest surfaced in reducing replication of HIV in the genital tract with ART [27]. Many groups working on this problem were focused on developing “treatment as prevention'. Indeed, in 1997, in a study from our group on this matter, we stated, `… antiretroviral therapy, by decreasing levels of HIV-1 in semen, may lower the infectious inoculum of treated men and possibly reduce the likelihood of sexual transmission to uninfected partners (p. 60) [28].

This work led to a series of reports about the concentration of HIV in semen using quantitative culture [29 – 31] and development of a PCR technique for more accurate measurement [32]. With these tools, studies demonstrated that ART reduced the concentration of HIV in both semen [33, 34] and female genital secretions [35]. Evidence also emerged that antiretroviral drugs penetrate the male and female genital tract with different efficiency, [36, 37] helping to explain the sexual transmission of resistant viruses [38 – 41].

HPTN 052: TRIAL DESIGN

Consistent with the Rakai study [26], other observational studies supported the critical importance of viral load in HIV transmission [42]. However, blood viral load never correlates perfectly with genital tract HIV [43 – 45]. In addition, the exact mode of HIV transmission – whether from cell-free or cell- bound virus remains obscure. Although inductive logic suggested that ART that suppresses HIV replication should also reduce HIV transmission, the magnitude and durability of such an effect could not be determined without empirical evidence. In 2000, the National Institute of Allergy and Infectious Diseases/National Institutes of Health (NIAID/NIH)-sponsored HIV Prevention Trials Network (HPTN) approved the development of a protocol to study the use of ART to reduce HIV in blood and genital secretions and, thus, prevent the sexual transmission of HIV in HIV-serodiscordant couples.

After careful consideration, the only plausible and ethical design was to randomize HIV-serodiscordant couples into two study arms: early vs. delayed therapy. In the early arm, ART would be offered at CD4 cell counts higher than when ART was recommended by the WHO in the respective countries where the study would take place. In the delayed arm, ART would be offered when the CD4 cell count fell to the general standard of care, that is, below 200 cells/ml in the pilot phase of the study, modified to at or below 250 cells/mm3 in the full trial (Fig. 1). Through a detailed statistical plan, including modeling of the timeframe for both the decline of CD4 cell counts and potential trans- mission events, it was determined that 1750 participants with HIV (and CD4 cell counts between 300 and 500 cells/ml in the pilot phase, and 350 – 550 cells/ml in the full study) and their HIV-negative partners would be needed to answer the primary objective, which was to compare the rates of HIV infection between the two arms (Fig. 1). As already noted, the CD4 cell count inclusion criterion changed over the course of the study. Thirteen clinical sites worldwide participated (nine initially, and four more in 2009 in order to increase the rate of enrollment), starting with a pilot phase in 2005, followed by the initiation of the full study in 2007. A list of the sites, the number of couples enrolled at each site, and the enrollment inclusion/exclusion criteria are shown in Tables 1 and 2 [46**]. The current version of the protocol is available at http://www.hptn.org.

HPTN 052: PRE-IMPLEMENTATION REQUIREMENTS

Having designed a plausible trial, the study faced two difficult challenges. First, NIAID policy in the early 2000s required that US pharmaceutical companies donate drugs for use in the study. We calculated that 1750 discordant couples would be recruited and followed on-study until the last couple completed 5 years of study follow-up, and the statistical plan assumed that every HIV-infected person in the delayed arm would eventually be placed on ART (statistical plan available in the protocol at http://www.hptn.org). These assumptions led to the recognition that large amounts of drug were required to conduct the study successfully, no company would receive actual benefit from the trial because no treatment comparisons were planned, and multiple anti-retroviral combinations would be required for first- and second-line therapy. Through an iterative process, we were able to establish a collaborative agreement with each of the companies listed in Table 3 to provide their drug products. Second, our collaborating partners in the NIH-sponsored AIDS Clinical Trials Group (ACTG) raised concerns that subjects with low CD4 cell count screening out of HPTN 052 required an option for treatment. This finding led to the development and implementation of an inter- national ACTG protocol (ACTG 5175), launched at all participating HPTN 052 sites at the time the HPTN 052 study began [47,48].

Other pre-study implementation requirements included the development of detailed clinical and counseling materials. Regional trainings were conducted over the course of a year on the clinical management of ART and couples risk-reduction counseling. In addition, on-site training on the protocol requirements was conducted at all 13 sites over the course of 2 years.

HPTN 052: ETHICS AND EQUIPOISE

The ethical considerations related to HPTN 052 are substantial and will be the subject of a separate detailed report (Cohen and Sugarman, unpublished). In brief, the initial idea of the study met with three criticisms: offering ART could be coercive due to a lack of availability of ART to the general population in the majority of the countries participating in the study, ART should not be offered at high CD4 cell counts (i.e., above current guidelines), and ART would undermine prevention counseling – the proven intervention – resulting in negative risk compensation behaviors. These concerns are entirely valid. The informed consent process was used to define alternative strategies for potential study subjects to obtain ART at a CD4 cell count consistent with local country guidelines and to describe the potential costs and benefits of earlier ART. Repeated couples counseling during the course of the study was used to reduce increased risk-taking behavior. Indeed, the trial was designed to study the potential benefit of risk-reduction counseling through the measurement of sexual behavior and condom use.

Other concerns arose as the trial was moving forward, including the issue of separation of the role of an investigator vs. their role as a clinical care provider. Because healthcare workers are a rare commodity in countries with a heavy HIV disease bur- den, the participants and caregivers might find it difficult to separate clinical care from research. This problem is sometimes referred to as the `therapeutic misconception' [49]. To mitigate this conflict of interest among the investigators, an `on-call' clinical management team was established, charged with advising all sites on general clinical management and complicated clinical cases within the parameters dictated by the protocol.

Evolving science created yet another challenge for the study team. With increased interest in HIV prevention, a large number of modeling studies [8] and observational studies [50,51,52*,53] purported to resolve both the value of ART for prevention and the `when to start ART' question [54*,55,56,57*]. These reports forced consideration of the equipoise balance required for the trial to proceed [58]. Ultimately, observational treatment studies led to changes in WHO [59] and other guidelines, [60] leading to both practical and ethical concerns about HPTN 052. At the same time, the Swiss AIDS Commission concluded that the benefits of ART had been resolved and offered the `Swiss Statement' to make this point [61]. However, the issues considered (by some) `resolved' by the observational trials were being directly addressed by HPTN 052. The weight of evidence for both early treatment and ART for prevention has been summarized in two Cochrane reviews [62*,63].

Publication of the Comprehensive International Program of Research on AIDS (CIPRA) Haiti study [64*] had additional and important effects on HPTN 052 because of concern that clinical equipoise had been breached, favoring earlier ART. The CIPRA Haiti study was designed to compare early vs. delayed ART. In this study, however, the subjects were enrolled with a mean CD4 cell count of 287 cells/ml, and the delayed arm received ART at a CD4 cell count of 160 cells/ml, which were far different degrees of immunosuppression than studied in HPTN 052.

The results of the CIPRA Haiti study led to an urgent ad-hoc meeting of the Data and Safety Monitoring Board (DSMB) overseeing HPTN 052 in July 2009. The DSMB recommended that HPTN 052 continue without modification to the study design. In November 2009, WHO updated their guidelines for the initiation of ART to at or below a CD4 cell count of 350 cells/ml [59] for the general population. We communicated changes in WHO treatment guidelines to all study participants, along with information about the provision of ART avail- able in their local setting. The study itself underwent no changes, as it was designed specifically to provide `Level 1 evidence' for policy changes predicated on less robust results.

THE EVENTUAL HPTN 052 TRIAL: A SUMMARY

HPTN 052 began in a pilot phase in April 2005 and enrolled 1763 HIV-serodiscordant couples, the vast majority of which (97%) were heterosexual. The study was conducted at 13 sites across Africa, Asia, and the Americas. During a regularly scheduled review in April 2011, the study's DSMB recommended that current results be released to the public as soon as possible. The primary results of HPTN 052 have been published [46**]. For each transmission event, genetic analysis was conducted to determine whether the event was `linked', meaning that the virus had been transmitted between the members of the enrolled couple, or `unlinked', meaning that the new infection had been acquired outside of the primary relationship. Among the 877 couples in the delayed ART group, 27 linked HIV transmissions occurred. This finding was in contrast to only one linked transmission that occurred in the immediate ART group. This difference was highly statistically significant. At least 10 unlinked transmission events also occurred, suggesting that the couples counseling provided through the study was not entirely effective. Clinical benefit was also observed in participants receiving early ART. HPTN 052 provides unambiguous proof that ART can stop the transmission of HIV.

HPTN 052: THE WAY FORWARD

The HPTN 052 study is still ongoing. Given the current results of the trial, participants on the delayed ART arm are being offered ART. With the continuation of the study, we will be able to reliably answer the other important question the study was designed to answer: is the reduction of sexual transmission of HIV by ART durable?

CONCLUSION

HPTN 052 is a complicated and challenging study design to prove a single concept that treating HIV- infected people with effective ART will render them less contagious. This idea is central to the `test and treat' strategy [6]. The results of the study were met with great enthusiasm upon their release in May 2011, and at the 6th International AIDS Society HIV

Pathogenesis Meeting in Rome, July 2011. HPTN 052 represents a nearly 20-year investment in marrying exceptionally strong biological plausibility to a clinical trial. The findings of HPTN 052 were heralded as a `game changer', [65] and became the foundation for The Economist's `The End of AIDS?' issue [66]. The goal now is to determine whether ART can be used so broadly and so effectively as to reduce the spread of HIV within a population. At least two community randomized trials that employ ART as their cornerstone are planned [6], and the results should determine the true benefit of this provocative and very effective intervention.