Bloodborne HIV: Don't Get Stuck!

Mass circumcision based on assumption, not evidence

Programs to circumcise millions of men in Africa are based on selected evidence from three trials in South Africa, Kenya, and Uganda.[1-4] Many people think these trials show that getting circumcised reduces men’s risk to get HIV from sex by more than 50%.

The trials don’t say that; specifically, the trials don’t say men’s infections came from sex. None of the studies collected all information needed to measure the impact of circumcision on men’s risk to get HIV from sex — asking about all blood and sexual risks and tracing and testing sexual partners.[5]

Trials in South Africa and Uganda report new infections in men according to whether the men reported any possible sexual exposure to HIV (any sex without a condom). The evidence from these two trials suggests less than 1/3rd of men’s infections came from sex (see details below). It is, of course, possible that men misreported their sexual behavior. In other words, programs to circumcise men in Africa are not based on evidence, but rather on assuming away evidence.

Were men’s sexual partners HIV-positive? We don’t know. The Ugandan trial (including a linked study) tested most of the men’s spouses and live-in partners for HIV (including women who did not want to know their HIV status [see “eligibility criteria” in reference 6]). The trial observed most new infections in men who reported no non-marital partner. But the trial has not reported the HIV status of the long-term partner of any circumcised or intact man who got HIV during the trial.

The South African trial found most new infections in men who reported only 0-1 sexual partner after the last HIV-negative test. The Kenyan study collected but has not reported information on men’s sexual behavior during the trial; at baseline a majority reported 0-1 partners in the previous 6 months. There is no indication either trial made any effort to trace and test the sexual partners of men who got HIV during the trial. This would have been not only good research, but was also ethical — protecting women who were not the source of men’s new infections.

Evidence from these trials supports other recommendations

Beware skin-piercing risks: As noted above (and detailed below), evidence from two trials says less than 1/3rd of men’s infections came from sex. Moreover, two studies report new infections in men with specific healthcare encounters. In the South African trial: men who “attended a clinic for a health problem related to the genitals” vs. men who did not were 5.7 times more likely to get HIV; and men who reported an injection, transfusion, and/or hospitalization vs. men with none of these risks were 1.7 times more likely to get HIV.  The Kenyan trial found four men new infections one month after they were circumcised. (See details below.)

Trust your body’s natural defenses: In Uganda, not washing one’s penis after sex cut men’s risk (whether circumcised or intact) for HIV. Thus, one message from the Ugandan trial is that natural defenses against HIV in sexual fluids may be as effective as circumcision (see details below and on two other pages).

Beware unethical research: The circumcision trial in Uganda (together with research in a linked trial) tested many of the men’s wives and stable partners for HIV, but did not insist that women learn their HIV status or warn their husbands if women were HIV-positive. The trial followed men to see them get HIV, without warning them. The trial in South Africa followed men, at least some of whom did not know their HIV status. None of the trials required testing and warning wives and other partners when men got new HIV infections (see details below).

More details from the three trials

South Africa: The first study of the three studies to report was carried out in South Africa during 2002-05.[2] The study team solicited men willing to be circumcised, then on a random basis assigned half the men to an intervention group to be circumcised first and the other half to a control group to remain intact until the end of the study. The study team then followed and retested the men – circumcised and intact – at scheduled visits over as long as two years.

During follow-up, 20 men in the intervention (circumcision) group got HIV at the rate of 0.85% per year, while 49 men in the control (intact) group got HIV at the rate of 2.11% per year. If all the men’s infections came from sex, this says that circumcision cuts men’s risk to get HIV from sexual partners from 2.11% to 0.85% per year.

But did all or even most infections come from sex? Not according to evidence the study collected and reported. Twenty-three of the 69 men with new infections said they had no sexual partner or always used condoms from their last HIV-negative test to their first HIV-positive test. Men reporting no sexual risks got HIV at the rate of 1.11% per year. If these men are telling the truth, they apparently got HIV from blood during injections, dental care, tattooing, and other skin-piercing events. If men with no sexual risks got HIV from blood, we can estimate that men with sexual risks also got HIV from blood contact at the same rate. The rate of getting new HIV infections in men who reported at least one unprotected (without a condom) sex event was 1.86% per year. The modestly faster rate to get HIV in men reporting vs not reporting any sexual risk explains less than 1/3 of the men’s infections (using standard epidemiological terms, the crude population attributable fraction of incident HIV associated with reporting any vs. no unprotected sex is 27%).

The study reported two health care exposure in men with and without new HIV infections: (a) men who “attended a clinic for a health problem related to the genitals” were 5.7 times more likely to get HIV than men without this risk;[2] and (b) men who reported injections and/or transfusion and/or hospitalization were 1.7 times more likely to get HIV than men who reported none of those events (see Authors’ reply in this link).[7]

The study team has not said what procedures men got at clinics treating genital health problems; it has also not reported HIV incidence separately for injections, transfusions, or hospitalizations. The study has not reported other healthcare procedures, such as infusions and dental care. The study has also not reported any data on skin-piercing cosmetic procedures; did they ask?

The study team could have done a better job identifying the sources of men’s infections if they had asked more questions and reported more data. But where is the failure? Did they not ask, or are they not telling? Like most studies on HIV in Africa, this study does  not give other researchers’ free access to collected data (with safeguards to protect participants’ confidentiality). Also, like most studies on HIV risk in Africa, this study has not disclosed its questionnaire and data collection forms – so there is no public record of what information they collected and chose not to report (chose to withhold).

Table: What information on sex and blood risks did the three studies collect and report for men with and without new HIV infections?

Sources: see references in the text.

Kenya: The Kenya study,[3] 2002-06, was similar in design to the South Africa study. The study circumcised some men, then followed and retested circumcised and intact men for as long as two years to see who got HIV. Nineteen men in the intervention (circumcised) group got HIV at the rate of 1.9% over two years, while 46 men in the control (intact) group got HIV at the rate of 4.1% over two years.

How many of the 65 men got HIV from sex? The study asked men about sexual partners and condom use, but reports this information for only seven men infected during the first three months of follow-up. Five of the seven reported no sexual partners from the time they entered the trial (using sensitive tests, the study could not find HIV in blood collected then) until their first HIV-positive test after 1-3 months. The study team has said nothing about tracing and testing men’s sexual partners – did they do it and not report it, or just not do it?

As for blood-borne risks, the study reports four men circumcised in the trial had new HIV infections one month later. Contaminated local anaesthetic or instruments could have infected the men during circumcision, but the study team says nothing about what might have gone wrong. Aside from circumcisions, the study says nothing about blood-borne risks. This is a glaring oversight, because the study team had in hand evidence linking HIV in the community to blood-borne risks: While recruiting men for the trial, men who reported one or more injections in the previous six months were 2.5 time more likely to be HIV-positive vs other men; men with tattoos were 2.2 times more likely to be infected; men who had ever received saro (traditional blood-letting) were 2.1 times more likely to be infected,[8] and men who reported “blood exchange” were 18.6 times more likely to be HIV-positive.[9]

Uganda: The study in Uganda, 2003-06, followed the same design as studies in South Africa and Kenya: assigning men randomly to be circumcised or to remain intact, then following them for as long as long as two years to see who got HIV. Twenty-two men in the intervention (circumcised) group got HIV at the rate of 0.66% per year, and 45 men in the control (intact) group got HIV at the rate of 1.33% per year.[4]

During follow-up, researchers asked men whether and how they cleaned their penis after sex, expecting that washing or wiping might be protective. Intact men who cleaned their genitals after sex, but waited at least 10 minutes to do so, got HIV at the rate of 0.39% per year. Also, men who wiped only without using water got less HIV than men who used water. The authors proposed that acid in vaginal fluids “may impair HIV survival,” so that washing these away with water may “facilitate viral survival and possible infectivity.”[10]

How did the men get HIV? Six men with new HIV infections reported having no sex partners during the period between their last HIV-negative and first HIV-positive test; 10 others with new infections reported always using condoms. Taken together, the 16 men who reported no possible sexual exposure to HIV got HIV at the rate of 0.72% per year, presumably from skin-piercing events that exposed them to HIV in blood. Men who reported any unprotected sex got HIV at the rate of 1.17% per year. As in South Africa, the marginally faster rate at which men who reported sexual risks got HIV explains less than a third of the new infections observed during the trial (using standard epidemiological analyses and terms, the crude population attributable fraction of incident HIV associated with having any vs no unprotected sex is 29%).

Unlike the other two studies, the Uganda study team traced and tested most of the men’s wives (and other long-term partners) for HIV.[6] However, the study has not reported the wives’ HIV status for men who got HIV. Did they get HIV from their wives? The study team is sitting on that information.

The Uganda study team provides no data on injections and other skin-piercing events for men with and without new HIV infections, and does not say if they collected any such data.

Ethical short-comings

The three studies treated research participants in ways that would not be allowed in the US, Canada, and France, non-African countries that funded the studies:

• The South Africa study recruited men and the Uganda study recruited wives without insisting they hear their HIV test results. Neither study has said how many participants did not hear their results.
• The Uganda study followed men who did not know the study had found some of their wives to be HIV-positive (some of the wives also did not know) to watch the men get HIV.
• None of the studies insisted that men who acquired HIV bring their wives for couple counseling. This ethical lapse – leaving wives with unknown risks – undermined the objective of the research, which was to see how much circumcision reduced sexual transmission.

Study teams for two of the three trials did not register the trials (ie, document what they were planning to do in their human subjects research) before the trial began. Both were registered only after follow-up was completed and less than one month before publication.[11, para 35 in reference 12]

References

1. WHO, UNAIDS. New data on male circumcision and HIV prevention: policy and programme implications. Geneva: WHO, 2007. Available at: http://apps.who.int/iris/bitstream/handle/10665/43751/9789241595988_eng.pdf?sequence=1 (accessed 2 May 2018).

2. Auvert B, Taljaard D, Lagarde E, et al. Randomized controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS t1265 trial. PLoS Med 2005; 2: 1112-1122. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1262556/pdf/pmed.0020298.pdf (accessed 2 May 2018).

3. Bailey RC, Moses S, Parker CB, et al. Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial. Lancet 2007; 369: 643-656. Available at: https://www.ncbi.nlm.nih.gov/pubmed/17321310 (accessed 2 May 2018).

4. Gray RH, Kigozi G, Serwadda D, et al. Male circumcision for HIV prevention in men in Rakai, Uganda: a randomized controlled trial. Lancet 2007; 369: 657-666. Available at: https://www.ncbi.nlm.nih.gov/pubmed/17321311 (accessed 3 May 2018).

5. Brewer, D. D., Rothenberg, R. B., Potterat, J. J., Brody, S., & Gisselquist, D. (2004). HIV epidemiology in sub-Saharan Africa: rich in conjecture, poor in data (reply to letter by Boily et al.). International Journal of STD & AIDS, 15, 63-65. Available at: http://www.interscientific.net/ijsa2004.html (accessed 2 May 2018).

6. Wawer M. Trial of male circumcision: HIV, sexually transmitted disease (STD) and behavioral effects men, women and the community. ClinicalTrials.gov, 2007. Available at: https://clinicaltrials.gov/ct2/show/NCT00124878 (accessed 2 May 2018).

7. Auvert B, Sobngwi-Tambekou J, Taljaard, et al. Author’s reply. PLoS Med 2006; 3: 0141-0143. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1360639/pdf/pmed.0030065.pdf (accessed 2 May 2018).

8. Mattson CL, Bailey RC, Agot K, et al. A nested case-control study of sexual practices and risk factors for prevalent HIV-1 infection among young men in Kisumu, Kenya. Sex Transm Dis 2007; 34: 731-736. Available at: https://www.ncbi.nlm.nih.gov/pubmed/17321310 (accessed 2 May 2018).

9. Mattson et al. Sexual practices and risk factors for HIV among young men in Kisumu, Kenya. Available at: http://www.abstract-archive.org/ (accessed 2 May 2018).

10. Makumbi et al. Male post-coital penile cleansing and the risk of HIV-acquisition in uncircumcised men, Rakai District, Uganda. Abstract WEAC1LB, 4^th IAS Conference on HIV Pathogenesis, Treatment and Prevention, Australia, 25 July 2007. Available at: http://www.abstract-archive.org/ (accessed 3 May 2018).

11. Thornton J. Circumcision and HIV. Blog post 18 September 2012. Available at: https://ripe-tomato.org/2012/09/18/circumcision-and-hiv/ (accessed 2 May 2018).

12. World Medical Association. Declaration of Helsinki 2013. Available at: https://www.wma.net/policies-post/wma-declaration-of-helsinki-ethical-principles-for-medical-research-involving-human-subjects/ (accessed 3 May 2018).