Clinical trials of circumcision against “HIV” “infection”

An earlier post [“Circumcision pseudo-science”, 2 September 2009] pointed out that the well-known immune-suppressing effect of surgery is a highly plausible explanation for the quantitatively concordant results of the 3 clinical trials of circumcision to prevent acquisition of “HIV-positive” status. But dubious interpretation is not by far the only flaw in these studies.

Auvert et al., “Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: The ANRS 1265 trial”, PLoS Medicine 2(11) (2005) e298.
Other mainstream researchers have criticized this study on a number of grounds:
— questions of randomization [Siegfried, “Does male circumcision prevent HIV infection?” PLoS Med 2(11): e393; Winkel, “Rush to judgment”, PLoS Med 3(1): e71];
— that the intervention and control groups were treated unequally in terms of instructions regarding intercourse [Young, “Two groups not on all fours”, PLoS Med 3(1): e75];
— that “the authors did not control for other sources of HIV transmission, such as exposure through blood transfusions or infected needles” [Vines, “Major potential confounder not addressed”, PLoS Med 3(1): e63].
— Others might question whether a study stopped after 12 months should be given much credence. Among 1582 controls, 49 new “HIV-positive” cases were observed whereas there were only 20 among the circumcised group of 1546; but 234 of the control group and 154 of the intervention group had been lost before the 12-month visits.
— The claimed incidence of 49 in the control group within a year bespeaks an incredibly high rate of intercourse, given that all estimates of “HIV” transmission report no more than a few per 1000 acts of unprotected intercourse with an infected partner.
— That claimed incidence (2.1% per year) also seems far too high when the overall prevalence of “HIV-positive” at baseline was only 4-5%; the prevalence would be reached after only 2 years!
— “In light of the anomalies and lacunae in Auvert and colleagues’ study, the protective effect of male circumcision they observed amounts to a faith lift for the empirically beleaguered paradigm of heterosexual HIV transmission in sub-Saharan Africa” [Potterat et al., “The protective effect of male circumcision as a faith lift for the troubled paradigm of HIV epidemiology in Sub-Saharan Africa”, PLoS Med 3(1): e64].
— Glass [“Rubbery figures?”, PLoS Med 3(1): e70] asked why 4 separate reports by Auvert et al. had given different numbers: “If we just look at the official figures — 15 to 45 at the International AIDS Conference and 20 to 49 in PLoS Medicine — between 1 August 2005 and 23 October 2005, it appears that there have been four seroconversions among the uncircumcised and five seroconversions among the circumcised. In less than three months, a 3:1 difference has shrunk to a 2.45:1 difference. Why are the numbers of seroconversions so much at variance in reports published by reputable journals?”
— This studied group of 18-24-year-old males was surely uncharacteristic in some fashion, since 596 of the 2236 participants observed during 21 months “received blood transfusions, were hospitalized, or received injections” [Auvert et al., “Authors’ reply”, PLoS Med 3(1): e67].

So there are ample reasons for not taking the Auvert study as definitive, yet its claim of 60% risk reduction through circumcision has become a shibboleth in the HIV/AIDS literature.
It is intriguing that other studies have found an increased risk of male-to-female “HIV” “transmission” when the male is circumcised [Sykes, “Male circumcision increases risk for females”, PLoS Med 3(1): e72; Chao et al., “Risk factors associated with prevalent HIV-1 infection among pregnant women in Rwanda”, Int J Epidemiology 23(#2, 1994) 371-80: “partner circumcision . . . remained strongly associated with HIV-1 infection even when simultaneously controlling for other covariates”].

***********************

The other two trials of circumcision are reported in Lancet, 369 (2007): Bailey et al., “Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial”, 643-56; Gray et al., “Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial”, 657-66.
These articles are honored by several commentaries in the same issue of Lancet, including a respectful bio sketch of Ronald Gray, who has been pursuing proof of circumcision as preventive for two decades: “His careful analyses of the data from that trial [an unsuccessful one to prevent “HIV” by treating women for sexually transmitted diseases] identified the importance of HIV viral load, lack of male circumcision, and genital ulcer disease on HIV transmission” (“Profile — Ronald Gray: collaborating with Ugandan researchers on HIV trials”, p. 635).
An editorial, “Newer approaches to HIV prevention” (p. 615) unblushingly states that they “show that male circumcision halves the risk of adult males contracting HIV through heterosexual intercourse. . . . a solid evidence-base to inform health policy. . . . Male circumcision might also directly protect against male-to-female transmission of HIV. A trial to test this hypothesis is under way in Uganda, with results expected in 2008” [emphases added].
Note the direct contradiction with the cited Chao study re male-to-female transmission.

Newell and Bärnighausen (pp. 617-9) are also enthusiastic: “We now have proof” that circumcision, “a permanent intervention . . . can reduce the risk of HIV infection in men, which is positive news about prevention after past and current disappointments.”

But there are similar problems with the Bailey and Gray articles as with the Auvert study, for instance that the incidence of new “HIV-positive” cases was extraordinarily high, respectively 2.1% and 1.33% “infections” per year, difficult to reconcile with the low transmissibility of “HIV-positive” — a few per thousand with unprotected sex with an “HIV-positive” partner — as well as with the overall prevalence of “HIV-positive”. In the Bailey study the prevalence was 8%, which would be reached within 4 short years at an incidence of 2.1%, so unlikely a situation as to call the study into question on that ground alone. (The Gray study did not cite a baseline prevalence.)

The Bailey study was halted prematurely after a year, on the basis of 1232 and 1234 results for the initial 1391 and 1393 enrollees. Again as with the Auvert study, a high proportion (751 of 2778) had received injections in the 6 months before the study. During the study, “10,154 unrelated adverse events were recorded among 1979 (71%) participants. The most frequent unrelated adverse events were upper respiratory tract infections (3189 events, 1184 participants, 43%), malaria (2271 events, 1076 participants, 39%), skin or mucous membrane infections (1011 events, 682 participants, 24%), and gastroenteritis (456 events, 327 participants, 12%). Study groups did not differ with respect to these common illnesses”.
With all due respect: It seems unbelievable that the incidence of each one of these was similar in the two study groups. Skeptics remain free to suggest that those adverse events most likely to stimulate a positive “HIV” test might have been more frequent in the control group, since the treated (circumcised) group had rather intensive post-operative medical attention that the control group did not, including “free medical care, were counselled about safe sexual practices, had unrestricted access to condoms, were tested for sexually transmitted infections, and were treated for bacterial infections.”
A definite difference in the two groups was that “sexual abstinence in the circumcision group . . . returned to baseline level at month 24”. Presumably sexual abstinence — lack of it — had remained at baseline in the control group, which was therefore exposed more frequently to all sorts of contagious infection, not only sexually transmitted ones.
It was also reported that herpes infection correlated with “HIV-positive”; the skeptical explanation is, of course, that herpes is one of the many conditions that can yield a positive “HIV” test.

The Gray study in Rakai began with 2474 and 2522 in the intervention and control groups respectively, of whom only 2253 and 2250 were available to the 12-month follow-up when the trial was suspended. About 400 in each group reported symptoms of a sexually transmitted disease within the previous year. During the follow-up period, the control group reported more than twice as many different sexual partners than the intervention group and 3 times as many non-marital partners, with the actual numbers comparable to the numbers of seroconversions. Again, the controls were exposed much more often to all sorts of contagious conditions.
The variation of “HIV” incidence with age was the familiar one: highest at an intermediate age, lower at both lower and higher ages, in this case a maximum in the range 25-29, which is earlier than in American cohorts but not so different from 25-34 reported from Kenya and Lesotho [HIV demographics further confirmed: HIV is not sexually transmitted, 26 February 2008]  or South Africa (25-29 among females, ~35 among males — HIV demographics are predictable; HIV is not a contagious infection, 27 August 2008].
Despite the flaws in the study, the authors claim that “circumcision must now be deemed to be a proven intervention for reducing the risk of heterosexually acquired HIV infection in adult men” [emphasis added] even as it is admitted that “trials that are stopped early could overestimate efficacy”. It is also admitted that circumcision has significant risks, especially in rural areas: “the rate of moderate and severe adverse events related to surgery was almost 4%, which is comparable with rates in the South African and Kenyan trials.6,9 One should note that there were cases in which appropriate follow-up management was required to prevent more serious sequelae. Furthermore, substantially higher complication rates have been reported when surgery is done in rural clinics or by traditional circumcisers.24” [emphasis added].

Risks from circumcision are far from negligible, in other words.

***************************

The numerous flaws in these trials demonstrate that they cannot be regarded as definitive, to put it as mildly as possible. Yet the HIV/AIDS Establishment has treated as gospel the “HIV”-preventive effect of male circumcision, and the Centers for Disease Control and Prevention is even contemplating recommending universal circumcision of male babies in the United States even though these flawed trials were done in Africa and data from the United States show no association between circumcision and “HIV-positive” status.