The peer-reviewed results of the first trial of the effect of male circumcision on HIV infection, which some experts call “a landmark trial,” will be published in PLoS Medicine on October 25. The trial found that circumcision reduced the rate of new infections among heterosexual men in South Africa by about 60%.

Because HIV infection rates are generally lower among African groups where circumcision is a traditional practice, compared with largely non-circumcised populations, researchers had suspected that circumcision might offer some protection against HIV transmission. However, because the lower infection rates in the circumcised groups might be due to some other difference between circumcised and uncircumcised populations, the only way to test whether circumcision has a protective effect is a randomized intervention study. Such trials are under way in Uganda and in Kenya but the trial reported in PLoS Medicine is the first to be completed, peer-reviewed, and published in a medical journal.

The large protective effect of male circumcision observed has surprised many HIV experts, and it will be important to see whether the related studies under way confirm the results of this trial.

The trial, conducted by a team of French and South African researchers and sponsored by ANRS (the French National Agency of Research on AIDS), took place in the Orange Farm area near Johannesburg, where male circumcision in adulthood is a common but not universal practice.

The researchers offered young sexually active uncircumcised heterosexual men the chance to be circumcised and then monitored for HIV infection. The participants (3,274 men) were randomly allocated to two different groups-one group was immediately circumcised and the other group was to remain uncircumcised until after the completion of the trial 21 months later. The circumcised men were instructed to abstain from sexual activity for 6 weeks after the operation. The plan was to follow the participants over 21 months, testing them for HIV at months 3, 12 and 21, to see whether there was a difference in the rate of new infections between the two groups.

After 18 months, the number of new HIV infections in the control group was 49, compared with 20 in the treatment group. The results suggested that circumcision can reduce female-to-male HIV infection by about 60% (95% CI: 32%-76%).

The committee monitoring this study considered that the protective effect of male circumcision was so large that it would be unethical to continue the study. The trial was therefore stopped and the uncircumcised men were offered circumcision.

The trial, results of which were first reported at an international AIDS conference in July, has attracted huge attention among the global health community.

Bertran Auvert and his co-authors have called for the promotion of male circumcision as part of AIDS prevention efforts in Orange Farm and in other parts of Africa. However, other HIV experts take a more cautious view, believing that the results of this one study must be confirmed by further studies before action can be recommended, especially given the relatively small number of new infections (20 among the circumcised men versus 49 among the controls) and the relatively large number of men who did not return for all the follow-up testing (100 and 125, respectively).

In a statement published shortly after Auvert and colleagues’ trial was presented at the July conference, UNAIDS emphasized that “although the trial shows promising protective effects of adult male circumcision in HIV acquisition, more research is needed to confirm the reproducibility of the findings and of this trial and whether or not the results have more general application.”

Furthermore, adult circumcision carries risks, especially if performed by medical personnel or traditional healers without proper training. Moreover, immediately after circumcision, men may be at a higher risk of acquiring infections (which is why the participants were asked to not to have sex for six weeks after the operation). A further concern is that circumcised men, considering themselves to be “protected,” might be more likely to engage in unsafe sex. Research is also needed to find out whether male circumcision only has a preventive effect on female to male transmission, or whether it may also reduce male to female transmission or male to male transmission. In addition, it will be important to determine the mechanism by which circumcision exerts its apparent protective effect.