Circumcision and HIV - Randomized Controlled Trials

Randomized Controlled Trials

Africa has a higher rate of adult HIV infection than anywhere else in the world. Three randomized controlled trials were commissioned in Africa to investigate whether circumcision could lower the rate of HIV contraction.

The first study to be published was named ANRS-1265. It was funded by the French government’s research agency, Agence Nationale de Recherches sur la SIDA (ANRS) and carried out in Orange Farm, Gauteng in South Africa. The purpose was to test the effect of adult male circumcision on HIV acquisition. The principal investigator was Dr. Bertran Auvert of Versailles University. The study enrolled 3,274 men aged 18–24. The participants were split into two equal groups. One group was circumcised straight away; the other group, serving as a control, was to be circumcised 21 months later. 146 of the original participants were found to have HIV at the start of the trial - to avoid stigmatization, they were not excluded. It was planned that all the men would visit the research clinic four times during this 21-month period, and that they would be tested for HIV each time. They were instructed not to have sex for six weeks after the operation, and asked at each clinic visit to provide detailed information about their sexual activity. The circumcision procedure used was the forceps-guided method, carried out by three local general practitioners in their surgical offices. After 17 months, 20 men had contracted HIV in the circumcised group and 49 in the control group. The trial was halted on ethical grounds. The results of the trial were published in November 2005.

The authors said, “Male circumcision provides a degree of protection against acquiring HIV infection, equivalent to what a vaccine of high efficacy would have achieved. Male circumcision may provide an important way of reducing the spread of HIV infection in sub-Saharan Africa.”

Williams et al.(2006) looked at the potential impact of circumcision on HIV in Africa, based upon the South African RCT, saying that that male circumcision (MC) could substantially reduce the burden of HIV in Africa, particularly in southern Africa where the existing prevalence of male circumcision is low and the existing prevalence of HIV is high. More specifically it predicted that if full coverage with MC was achieved in sub-Saharan Africa over the next ten years, MC could prevent approximately 2.0 (1.1 to 3.8) million new HIV infections over that ten-year period and a further 3.7 million in the ten years after that.

The above conclusions drawn from the Orange Farm study have been criticised by Michel Garenne (2006) of the Institut Pasteur. In his critique, published on the PLoS Journal of Medicine, he concludes that: "'male circumcision should be regarded as an important public health intervention for preventing the spread of HIV' appears overstated. Even though large-scale male circumcision could avert a number of HIV infections, theoretical calculations and empirical evidence show that it is unlikely to have a major public health impact, apart from the fact that achieving universal male circumcision is likely to be more difficult than universal vaccination coverage or universal contraceptive use."

Mills and Siegfried (2006) point out that trials that are stopped early tend to overestimate treatment effects. They argued that a meta-analysis should be done before further feasibility studies are done.

The NIAID, part of the NIH, supported two further trials, conducted in Kenya and in Uganda. The primary objectives of these studies were to determine whether adult male circumcision can be administered safely, and whether it would reduce the risk of acquiring HIV infection through heterosexual contact. After an initial HIV screening and a medical exam, eligible men were randomly assigned either to receive circumcision immediately or to wait two years before circumcision. All participants were closely followed for two years to collect information about their health, sexual activity, and their and their partners’ attitudes about circumcision; to counsel participants in HIV prevention and safe sex practices; and to check the HIV status of the volunteer. Participants in the Kenyan study were scheduled for six visits over the two-year follow-up, compared with four visits for the Ugandan trial participants. In addition to the study visits, men enrolled in the Kenyan trial were encouraged to receive all of their outpatient health care at the study clinics, which enabled researchers to collect information on the safety of the procedure and the number of other sexually transmitted diseases the men had during follow-up.

The Kenyan trial, also known as the UNIM trial (Universities of Nairobi, Illinois and Manitoba trial), began in February 2002, in Kisumu, Kenya. It was a collaborative effort between US, Canadian and Kenyan researchers, led by Dr. Robert Bailey, of the University of Illinois. Also involved were Stephen Moses, University of Manitoba, Jeckoniah Ndinya-Achola, University of Nairobi, and Kwango Agot, UNIM. The trial was funded by the NIAID and the Canadian Institutes of Health Research. This trial enrolled 2,784 men between 18 and 24 years old. The participants were assessed by HIV testing, medical examinations, and behavioural interviews during follow-ups at 1, 3, 6, 12, 18, and 24 months. The circumcision procedure used in the Ugandan trial is known as the sleeve method and takes about 30 minutes. The Ugandan trial used cauterization of the blood vessels to control bleeding and stitches to close the wound. 22 men in the intervention group and 47 in the control group had tested positive for HIV when the study was stopped on ethical grounds.

The Ugandan trial began August, 2003 in Rakai, Uganda, with 4,996 men aged between 15 and 49 years old. It was led by Ronald Gray and Maria Wawer of Johns Hopkins Bloomberg School of Public Health and David Serwadda and Nelson Sewankambo of Makerere University in Kampala, Uganda. The circumcision procedure used in the Kenyan trial was the foreskin clamp method. The Kenyan trial procedure took about 25 minutes and used stitches to control bleeding and improve wound closure. Trained and certified physicians performed the circumcisions in well-equipped operating rooms. Post-operative follow-up visits were scheduled at 24–48 hours, 5–9 days, and 4–6 weeks. HIV testing, physical examination, and interviews were repeated at 4–6 weeks, 6-, 12-, and 24-month follow-up visits. After 24 months, 964 of the original 2387 men of the circumcised men had been retained of whom 22 had contracted HIV. 980 of the 2430 uncircumcised men had been retained of whom 45 had contracted HIV.

Both trials were stopped early on December 13, 2006 on ethical grounds after it found that those belonging to the control group had a greater number of men with HIV than the circumcised group.

On Wednesday, March 28, 2007, the World Health Organisation (WHO) and UNAIDS issued joint recommendations concerning male circumcision and HIV/AIDS. These recommendations are:

• Male circumcision should now be recognized as an efficacious intervention for HIV prevention.
• Promoting male circumcision should be recognized as an additional, important strategy for the prevention of heterosexually acquired HIV infection in men.

Kim Dickson, coordinator of the working group that authored the report, commented:

• Male circumcision "would have greatest impact" in countries where the HIV infection rate among heterosexual males is greater than 15 percent and fewer than 20 percent of males are circumcised.
• WHO further recommends that the procedure must be done by a trained physician.
• Protection is incomplete and men must continue to use condoms and have fewer partners.
• Newly circumcised men should abstain from sex for at least six weeks.

The World Health Organization (WHO) said: “Although these results demonstrate that male circumcision reduces the risk of men becoming infected with HIV, the UN agencies emphasize that it does not provide complete protection against HIV infection. Circumcised men can still become infected with the virus and, if HIV-positive, can infect their sexual partners. Male circumcision should never replace other known effective prevention methods and should always be considered as part of a comprehensive prevention package, which includes correct and consistent use of male or female condoms, reduction in the number of sexual partners, delaying the onset of sexual relations, and HIV testing and counselling.”

Others have also expressed concern that some may mistakenly believe they will be fully protected against HIV through circumcision and see circumcision as a safe alternative to other forms of protection, such as condoms.

Dowsett et al. urged caution over using circumcision as a HIV prevention strategy saying that there were still questions that needed to be answered: "We need to investigate the effects of those other social and contextual factors that will be in play in real world settings – because the effectiveness of male circumcision will not be generated by the efficacy of the surgery alone." He contrasts the preventative effect of circumcision taken from the RCT's (55%) with the preventative effect of condoms (80-90%). He criticises the fact that the trials were not double-blinded - the participants knew their circumcision status and so this could have affected how the men responded behaviourally, psychologically and sexually. He criticised the randomisation measures used in the trial: sexual practices (number of partners, condom use) and sexual health measures (presence of STIs), saying that "Effective measures were not used, and differences related to sexual subjectivity, such as sexual network participation, pleasure preferences, body image, sexual history effects (e.g. abuse), partner preferences (younger, older, peers, groups) and so on were never assessed or analysed." He also asks how the extensive counselling and education might have influenced the participants' sexual activity. He adds that "all participants were subject to regular monitoring (e.g. behaviour surveys, clinical check-ups), which clearly might have enhanced compliance with suggested safety regimes and lowered risk-taking during the follow-up period. Such compliance cannot be guaranteed in real world settings." He also said the trials were subject to the Hawthorne effect.

An interim analysis from the Rakai Health Sciences Program in Uganda suggested that newly circumcised HIV positive men may be more likely to spread HIV to their female partners if they have sexual intercourse before the wound is fully healed. “Because the total number of men who resumed sex before certified wound healing is so small, the finding of increased transmission after surgery may have occurred by chance alone. However, we need to err on the side of caution to protect women in the context of any future male circumcision programme,” said Maria Wawer, the study's principal investigator.

Kalichman et al. (2007) argue that any protective effects circumcision could offer would be partially offset by increased HIV risk behavior, or “risk compensation" including reduction in condom use or increased numbers of sex partners. They note that circumcised men in the South African trial had 18% more sexual contacts than uncircumcised men at follow-up. They also said that because participants were given ongoing risk-reduction counseling and free condoms, it "reduced the utility of these trials for estimating the potential behavioral impact of male circumcision when implemented in a natural setting." They also criticised current models for failing to account for increased HIV risk behaviour. Increased HIV risk behaviour would mean more women would be infected which would consequently increase the risk of men. It would also mean that non-HIV STI's, which have been associated with increased HIV risk, would increase. Green et al. (2008) also disagree with using circumcision to prevent HIV, citing similar reasons.

Published meta-analyses, using data from the RCTs, have estimated the summary relative risk at 0.42 (95% CI 0.31-0.57), 0.44 (0.33-0.60) and 0.43 (0.32-0.59). (rate of HIV infection in circumcised divided by rate in uncircumcised men). Weiss et al. report that meta-analysis of "as-treated" figures from RCTs reveals a stronger protective effect (0.35; 95% CI 0.24-0.54) than if "intention-to-treat" figures are used. Byakika-Tusiime also estimated a summary relative risk of 0.39 (0.27-0.56) for observational studies, and 0.42 (0.33-0.53) overall (including both observational and RCT data). Weiss et al. report that the estimated relative risk using RCT data was "identical" to that found in observational studies (0.42). Byakika-Tusiime states that available evidence satisfies six of Hill's criteria, and concludes that the results of her analysis "provide unequivocal evidence that circumcision plays a causal role in reducing the risk of HIV infection among men." Mills et al. conclude that circumcision is an "effective strategy for reducing new male HIV infections", but caution that consistently safe sexual practices will be required to maintain the protective effect at the population level. Weiss et al. conclude that the evidence from the trials is conclusive, but that challenges to implementation remain, and will need to be faced.