HIV treatment as HIV prevention: where are we now?

During the last two years, there has been widely reported scientific, public health and community debate about the extent to which HIV treatment (through decreasing blood and genital fluid HIV viral load) may decrease the sexual transmission of HIV. This current debate originated in studies published a decade ago. These studies, in African heterosexual couples where the HIV-positive partner was not treated, demonstrate that there is a very strong relationship between blood HIV viral load and risk of onwards HIV transmission to the sexual partner. Based on a small number of couples, there were no cases of HIV transmission when the HIV-positive partner had an undetectable viral load. These data led some researchers to speculate that the use of HIV therapy to decrease blood viral load to undetectable may also markedly reduce onwards HIV transmission.

Other evidence in favour of using HIV treatment as prevention comes from mother-to-child HIV transmission studies. Randomised controlled trials have demonstrated that the use of HIV therapies by pregnant women greatly reduces the risk of HIV transmission to her newborn baby. As a consequence, HIV transmission from mother to child is now very uncommon in settings where HIV treatment is widely available.

For sexual transmission of HIV, the data relating to reduction of HIV transmission are less conclusive, but two recent highly controversial reports have energised the field. First, a World Health Organization (WHO) research group reported their results of a mathematical model showing that universal voluntary HIV testing with immediate treatment of all those diagnosed, a so-called ‘test and treat’ strategy, could substantially reduce severe generalised heterosexual HIV epidemics.(1) Second, a consensus statement released by the Swiss Federal Commission for HIV/AIDS (the Swiss Statement, see Talkabout #164 [8]) states that HIV-positive people on effective HIV treatment with undetectable blood viral load for six months or more who are free of other sexually transmitted infections cannot transmit HIV through sexual contact.

However, there is substantial concern that these reports may overstate the case for HIV treatment as HIV prevention. In response to the Swiss Statement, the joint United Nations Programme on HIV/AIDS (UNAIDS) and national public health authorities around the world and in Australia emphasised that the effect of undetectable viral load is uncertain, and that condoms are the only proven method of prevention of sexual transmission of HIV. Whatever one thinks about the quality of these two reports, they have outlined a research agenda to definitively determine the role of HIV treatment in the prevention of sexual transmission of HIV.

Data from heterosexual couples
The use of HIV treatment to prevent sexual transmission of HIV is based on the assumption that there is a very strong association between reducing blood viral load and reducing sexual transmission. For example, the WHO model assumes a 100-fold (ie 99%) reduction in infectiousness associated with HIV treatment, whereas other researchers have assumed a 2- to 10-fold reduction. Unfortunately, we just don’t have the studies to tell us which of these assumptions are true. This is fundamentally important: if treatment reduces transmission by 99%, then it is at least as effective as condom use. If it is only 50% effective, then continued condom use would be vital to HIV prevention.

How can we find the answer to this puzzle? The highest quality studies of viral load and the risk of sexual transmission are in HIV serodiscordant (ie different HIV status) couples because such studies can directly measure exposure to and transmission of HIV. However, a recent review article reported only two studies, with fewer than 300 person years of follow up, which followed heterosexuals on HIV therapy with undetectable blood viral load and their HIV-negative sexual partners. Although no transmissions were recorded, statistically this study highlights that there just isn’t enough data to be anywhere near certain about this.

The HIV Prevention Trials Network is conducting a definitive large study to examine HIV treatment and HIV prevention in heterosexuals. HPTN 052 is a randomised controlled trial of early versus standard HIV therapy of more than 1,500 people in serodiscordant heterosexual couples. The primary outcome in this study will be the rate of HIV infection in the initially HIV-negative partners. It is anticipated that the results of HPTN052 will be available around 2015. These results will allow the formulation of evidence-based policy on the use of HIV treatment as an additional means of preventing sexual transmission of HIV in heterosexuals.(2)

Data from male homosexual couples
While the data from heterosexual couples are sparse, they are virtually non-existent in homosexual male couples. Unfortunately, HPTN 052 has enrolled fewer than 40 male homosexual couples, so it will not be highly informative for gay men. Given that transmission risk in anal intercourse is around 20-fold higher than vaginal intercourse, it is quite plausible that the relationship between undetectable viral load and HIV transmission is substantially less strong in homosexual men. Data on recent HIV trends in homosexual men are not encouraging. In Australia, as in most of the developed world, HIV transmission has recently increased markedly in homosexual men, despite the increasingly large majority of HIV-positive men on HIV therapy with undetectable viral load. Recently published Australian research demonstrates that the per-contact probability of HIV transmission through anal intercourse in gay men in Sydney has not decreased since the early 1990s,(3) when almost no-one with HIV was receiving effective HIV therapy. This is suggestive – although by no means conclusive – evidence that HIV therapy is substantially less than 100% effective in preventing HIV transmission between homosexual men.

A prospective study of serodiscordant male homosexual couples would result in the most conclusive evidence, but conducting transmission studies is challenging for several reasons.

First, a high proportion of these relationships are not monogamous, so assigning the regular partner as the source of infection is not straightforward. However, this can usually be resolved by taking a detailed history of sexual behaviour combined with phylogenetic testing of HIV.

Second, the current legal framework around HIV transmission in some Australian states is problematic. There is a theoretical possibility that an HIV-positive participant could be prosecuted for HIV transmission or exposure under public health and/or criminal legislation. However, this risk could be ameliorated if the consent process for entry into the study includes acknowledgment by the HIV-negative partner that he is aware of the HIV-positive status of his partner, and that he is aware of how HIV is transmitted. If HIV infection occurred, it would be clear that the exposure to HIV was with the informed consent of the HIV-negative partner. In addition, if data on sexual risk behaviours are collected only from the HIV-negative partner, there would be no data that could possibly be construed as an ‘admission’ by the HIV-positive partner of knowingly exposing a person to HIV.

Of course, it would be vastly preferable if laws relating to consenting sexual practices leading to HIV transmission were repealed. However, the existence of such laws should not be allowed to stop research of importance to HIV-positive people and their partners, as long as the welfare of study participants can be maintained.

Plans for a new study of HIV treatment and prevention in Australia
A new study of HIV treatment as prevention in male homosexual couples is being planned in Australia. About 10% of gay community attached men in steady relationships in Australia report having a regular serodiscordant partner. Unprotected anal intercourse is reported in about half of these partnerships, and has increased significantly over time. Data from cohort studies in Sydney suggest that the average rate of break-ups in serodiscordant relationships is similar to that in seroconcordant relationships at about 30% per year. HIV incidence is about 2.2% per year in such couples, despite most men having undetectable viral load. HIV incidence is much higher – around 5% per year – in men whose serodiscordant relationship is less than one year old, or are under 35 years of age or who report unprotected anal intercourse with their HIV-positive partner.

A study of HIV transmission in serodiscordant male homosexual couples would be a world first, and researchers at Sydney’s National Centre in HIV Epidemiology and Clinical Research (NCHECR) have been busy preparing. They held a series of community consultations with AIDS councils and people living with HIV/AIDS organisations in NSW and Victoria. They also conducted a survey of clinical practices in Sydney and Melbourne to determine their willingness and capacity to take part in the study. They held extensive legal consultations on how to secure confidentiality and protection for the participants. These preparations have revealed a very high level of support for the study among gay and HIV community organisations, and from clinicians working in HIV medicine.

The researchers submitted a grant application to conduct the study in 2011–2014. The proposed study will collect evidence about the use of viral load by homosexual men in negotiating sexual practices. It will estimate the association of receiving HIV therapy, with and without undetectable blood plasma viral load, in preventing transmission of HIV and explore factors that may influence the effect of HIV therapy and undetectable viral load on HIV transmission (including the type of therapy, adherence and the presence of concurrent sexually transmitted infections). The study is also designed to investigate the relationship of semen viral load with HIV transmission risk and the reasons why it may be detectable despite undetectable blood plasma viral load. Evidence from this study will be vital in the development of guidelines on how HIV therapy may help in controlling HIV epidemics in homosexual men.

Depending on the result of funding applications, the study could start as early as 2011, and the first evidence about the role of HIV therapy in HIV prevention among homosexual men could be available by 2015, at roughly the same time as the results of HPTN 052 are reported. The study’s results will allow the development of practical recommendations on the applicability of the Swiss Statement and ‘test and treat’ strategies for HIV prevention in gay men.

For more information on NCHECR’s proposed study, please contact Dr Iryna Zablotska, izabtoska@nchecr.unsw.edu.au [9], or Professor Andrew Grulich, agrulich@nchecr.unsw.edu.au [10]

References:

Granich RM, Gilks CF, Dye C, De Cock KM, Williams BG. Universal voluntary HIV testing with immediate antiretroviral therapy as a strategy for elimination of HIV transmission: a mathematical model. Lancet 2009;373(9657):48–57.
Cohen MS, Kaleebu P, Coates T. Prevention of the sexual transmission of HIV-1: preparing for success. J Int AIDS Soc 2008;11(1):4.
Jin F, Jansson J, Law M, Prestage GP, Zablotska I, Imrie JC, et al. Per-contact probability of HIV transmission in homosexual men in Sydney in the era of HAART. AIDS 2010;24:907–13.

2010

Aug-Sep