Unanswered questions in circumcision research

[Guest guest]

Colleagues,

Lancet recently published reports from two randomized controlled trials of

circumcison for HIV prevention in Rakai, Uganda (Gray et al, Lancet 2007; 359:

657-666), and in Kisumu, Kenya ( et al, Lancet 2007; 359: 643-656). In

2006, PLoS Medicine published a report

from a third trial in South Africa (Auvert et al, PLoS Med 2006; 2: e298).

Reports from these 3 claim that circumcision cuts sexual acqusition of HIV

infection by around 50% or more. However, the reports to date leave some

questions. For example:

1. Did the project test and protect the wives? During the Rakai trial, 29 of 67

incident infections occurred in men who were currently married. How many of the

wives were HIV-negative and at risk? While it may be illegal in Uganda to tell

wives that their husbands are HIV-positive without their husbands' permission,

it is not illegal to urge the husbands to do so. What happened?

2. What were the men's non-sexual exposures to HIV? A lot of infections occurred

in men who reported no sexual exposures without a condom.

(a) In Rakai 16 (of 67) incident infections occurred in men who reported no sex

partner or 100% condom use during the observation interval. The rate of

incidence in men with no unprotected sex was 0.72 per 100 person-years vs 1.00

per 100 person-years in the full

cohort.

( The trial in Kisumu collected information on sexual exposures, but reported

HIV incidence against sexual exposures for the first 3 months only. During that

period, 5 of 7 seroconverions (PCR did not find HIV in blood at baseline)

occurred in men who denied sexual activity, including 3 of 4 men who had been

circumcised (all 4 of whom had HIV in their blood within one month after

circumcision).

© In South Africa 23 (of 69) incident infections occurred in men who reported

no unprotected sex during the observation interval. The rate of incidence in men

with no unprotected sex was 1.11 per 100 person years vs 1.48 per 100 person

years in the full cohort.

3. In these 3 trials, did so many men lie? Are there other possible connections

between lack of circumcision and HIV infection? Can we tell men considering to

get circumcised what really happened in

Rakai, Kisumu, and South Africa without more complete reports from these trials,

and without resolving some of these questions?

4. Did any of the trials try to trace infections to sex partners, or to blood

exposures? Could this still be done? With a few exceptions, men in all 3 trials

reported few sexual partners. Some men might not have wanted to bring in sex

partners for tests, but some might have been and might still be willing. HIV

infections might also be traced to blood exposures at clinics, by offering tests

to others attending the same clinics, sequencing viruses, etc.

5. Did the research procedures infect any of the men? Is anyone looking at that

possibility? All 3 trials reported incident HIV in men who denied sexual

exposures but had blood exposures connected to the research (circumcision, blood

tests, injections).

Considering conflict of interest, should the word of the research teams that

they were careful be the final word on this matter? Investigations could include

sequencing of viruses to see if there was any epidemiological linkages between

the HIV found in men

visiting clinics on the same day, etc.

5. Overall, have these trials been adequately reported? If not, could something

be done about it at this stage?

e-mail: <david_gisselquist@...>