Langerhans' cells in the preputial mucosa are nothing new: all mucosal tissues have Langerhans' cells. Szabo and Short did not report Langerhans' cell concentrations in comparison with other mucosal tissues, their concentration in the glans, foreskin remnant, and circumcision scar in circumcised men, the presence of associated T cell infiltration (which may be necessary for viral transmission), or how findings in elderly cadavers correlate to sexually active young men. Szabo and Short state that the inner surface of the foreskin and the frenulum must be regarded as the most probable sites for viral entry of primary HIV infections in men; but without quantitative comparative data their statements are pure speculation.

The only reports of preputial Langerhans' cells have been in specimens from neonates³ and elderly cadavers. If normal genital mucosa is at risk, we need to know the concentration of Langerhans' cells in healthy men, men with multiple sexual partners, men with genital infections, men with HIV, and men of differing races and ages before any recommendations can be made.

Szabo and Short dismiss the complications of circumcision as having a low incidence; but the rate of immediate complications in the United States is between 3.1% and 9%,⁴ and another 5% will later develop meatal stenosis.⁵ A higher rate of complications is believed to follow circumcisions performed in the developing world, where circumcision has been linked to tuberculosis, tetanus, penile amputation, and death.

HIV transmission is heavily dependent on certain sexual behaviours, not anatomy. The authors have not provided any new information to alter this fact but have taken the discussion off on a needless tangent. Indiscriminate mass circumcision, which is currently popularised by some Western researchers, is unproved and does not address the core behavioural issues that have fuelled this pandemic. Therefore, it will not alter the course of AIDS in Africa.

Cruz, Rio (International Coalition for Genital Integrity, Box 8462, Santa Cruz, CA 95061, USA riocruz@cruzers.com).

Editor—Szabo and Short's article on male circumcision and HIV places them in splendid solidarity with Victorian notions of sex and hygiene, together with tribal initiation rituals.^1-1,1-2 They are completely isolated from their major peer medical organisations, not one of which endorses routine infant circumcision as a prophylactic measure despite over 100 years of pressure brought to bear by circumcisers.

To accept that circumcision is a really good idea, we first have to believe that nature made some huge design error in human anatomy that requires removal by force. This is a great leap of faith given the fact that not just humans but all mammals, both male and female, have evolved over millions of years to end up with a prepuce. But for some reason known only to religious types and medicalised capitalism the only mammal to be benefited by summarily removing this omnipresent organ through surgery is the human male.

The history of medicalised circumcision is a fascinating study in Victorian medicine and anti-sexuality.^1-4 Amputating the normal prepuce of human beings started in the English speaking countries as a measure to prevent masturbation. It did not work, but circumcisers have learnt that the pretexts for penile pruning are inexhaustible. Simply by playing on the fears of the culture they can keep the practice going—and the income flowing. At the turn of the 20th century better hygiene was the big issue, followed by penile cancer in the 1930s,^1-3 cervical cancer in the '50s, sexually transmitted diseases in the '60s, urinary tract infections in the '80s,^1-4 and, perhaps the most dreaded of all, AIDS in the '90s. If it looks as if routine infant circumcision is an operation in search of a disease, that's because it is. Every single claim for legitimate medical benefit justifying this routine has been discredited.^1-5

But still the amputations go on. Every 26 seconds another penis is reduced in the United States. This is in sharp contrast with the rest of the world, where over 80% of the male population are left whole and intact—including all of Europe, most of non-Muslim Asia and Latin America—their genitals as nature designed them before the collective wisdom of Szabo and Short and other pro-circumcision proponents had a “better” idea.

Dalton, John D (Howgate Farm, Linglabank, Frizington, Cumbria CA26 3SU).

Editor—Szabo and Short have concluded that circumcision of male children should be seriously considered as an additional means of preventing HIV.^2-1 Whether they have a valid argument for the circumcision of consenting adults, they have certainly not made a case for circumcising unconsenting children who are not sexually active. Furthermore, there are certain failings of the article that should not have escaped the attention of the peer reviewer.

Szabo and Short conducted a Medline search for relevant literature, but they present no full listing of the search results. An objective review of the literature would have shown that there was no consensus that male circumcision protects against HIV.^2-2 One meta-analysis showed circumcised men to be more at risk of HIV than those with the normal, intact penis.^2-3

No evidence is presented by Szabo and Short to confirm their claim that HIV enters the body through CD4 and CCR5 receptors on Langerhans' cells located in the penis. As such their proposed mechanism for prevention of HIV by male circumcision is little more than supposition.

It is unacceptable for Szabo and Short to claim that circumcision has a low incidence of complications on the basis of a booklet favouring circumcision that has had no peer review.^2-4 Although a complication rate as low as 0.06% has been claimed for circumcision, rates as high as 55% have also been reported.^2-5 A detailed literature review of the complication rate for circumcision concluded that a realistic rate of significant complications is 2-10%.^2-6 It seems possible that any programme of child circumcision would cause more serious complications than it would prevent cases of HIV.

We believe that we live in an enlightened age. What is most surprising is that we still believe that we should ward off disease by cutting children's genitals. Publishing the opinion of Szabo and Short will do more to perpetuate non-therapeutic circumcisions of unconsenting children in North America and Australia than it will for the prevention of HIV in Africa.

Oliver, R T D (St Bartholomew's Hospital, London EC1A 7BE e.m.davies@mds.qmw.ac.uk). Oliver, Josephine (University of Birmingham medical School, Birmingham B15 2TT). Ballard, Ron C (National Reference Centre for Sexually Transmitted Diseases, South African Institute for Medical Research, Johannesburg, South Africa).

Editor—Szabo and Short suggest that the increased number of Langerhans' cells on the surface of the foreskin explains why circumcised men are less likely to become infected with HIV.^3-1 They did not mention an issue that has long dogged debate on the protective effect of circumcision on the incidence of cervical cancer and now increasingly prostate cancer^3-1,3-2—that is, the extent to which improved hygiene and affluence are confounding variables to the benefits of circumcision. This is exemplified by the lower incidence of cervical cancer in educated high caste women in India whose husbands were not circumcised than in less educated Muslim women with circumcised husbands.^3-3 Undoubtedly the increased numbers of Langerhans' cells with HIV receptors in the foreskin may well contribute to an increased susceptibility to HIV.

Evidence that nutritional state and other sexually transmitted diseases also play a part in acquiring HIV infection prompted us to examine the role of the foreskin in the occurrence of HIV infection in a series of 83 new patients (40 positive for HIV) attending a urethritis clinic at East and West Drakefontein Gold Mines Carltonville, Gauteng, South Africa, as part of a study of the impact of HIV and sexually transmitted diseases on serum concentrations of prostate specific antigen.^3-3 After giving their signed informed consent the miners received a questionnaire and were examined to ascertain whether they were circumcised (including whether the glans penis was visible) and the retractability of the foreskin. In addition, a limited history of sexual activity was recorded.

As expected, the frequency of HIV infection was significantly lower in those who were circumcised (table). The small subgroup who had been circumcised after puberty seemed to show some benefit in reducing the incidence of HIV infection. Even more interesting in the light of Szabo and Short's hypothesis about the increased numbers of Langerhans' cells in the foreskin, we found, contrary to what might be expected if their hypothesis was correct, that the frequency of HIV infection was less in men with long foreskins that were difficult to retract than in those with short easily retractable short foreskins.

Clearly this observation is based on too small a sample size for us to be totally confident in the results. However, these observations, added to those on the role of hygiene versus circumcision in reducing cervix cancer from India,^3-4 suggest that further studies would help to clarify Szabo and Short's hypothesis and need to be done before widespread use of circumcision is implemented to try to reduce the spread of HIV infection. Furthermore, work needs to be done on the influence of circumcision after puberty because performing such a procedure after the first infection of a sexually transmitted disease could be more effective than circumcision based on the total population.

Szabo, Robert (Mildura Base Hospital, Mildura, Victoria, Australia). Short, Roger V (r.short@unimelb.edu.au Department of Obstetrics and Gynaecology, University of Melbourne, Royal Women's Hospital, Carlton, Victoria 3053, Australia).

Editor—Our review article was primarily concerned with exploring the mechanisms by which male circumcision protects against heterosexually acquired HIV infection in men. We concluded that it is the inner aspect of the foreskin, which is poorly keratinised but well supplied with Langerhans' cells, that is likely to be one of the principal sites of HIV entry into the penis.

We have now developed an active research programme on this topic, and, together with our collaborators, we hope to publish a number of papers in the near future on the distribution of Langerhans cells in the foreskins of young men with and without balanitis, the degree of keratinisation of the various penile epithelia, and the uptake of live HIV virus applied to the inner and outer aspects of adult human foreskins in vitro. Some of our histological findings were shown on the BBC Horizon/Discovery television programme “The Valley of Life or Death” on 16 November.^4-1

The claim by Van Howe et al that a meta-analysis of the many papers that show a significant correlation between lack of male circumcision and HIV infection is unscientific and misleading makes little sense, since most of the 40 studies that show such an association have incorporated multivariate analysis to correct for confounding variables such as different sexual practices. Furthermore, Van Howe's own meta-analysis has been invalidated because of several major methodological errors.^4-2,4-3

Male circumcision, like all minor surgical procedures, carries a small risk of postoperative complications. But this should not detract from the twofold to eightfold protective effect that circumcision provides against HIV infection, which, unlike the surgical complications, is almost invariably fatal. Other than recommending that male circumcision should be seriously considered as an additional means of preventing HIV in all countries with a high prevalence of infection, we have avoided all discussion about the relative advantages and disadvantages of neonatal male circumcision as a routine procedure in developed countries, where the prevalence of HIV infection is low. We do not intend to enter that debate, where objectivity is hard to find.

It is pleasing to note that organisations are now beginning to give serious consideration to the policy implications arising from the protective effect of male circumcision against HIV infection. In June 2000 the Horizons Project of the Population Council published a report of an international discussion meeting entitled “Male Circumcision and HIV Prevention: Directions for Future Research,”^4-4 and in July the World Health Organization held a similar consultation in Durban at the time of the international AIDS conference, although its findings have yet to be published.

It would be unfortunate if the zealous opponents of neonatal male circumcision in developed countries, however well meaning, distracted attention from the glaring fact that in central and southern Africa, where 24.5 million people are infected with HIV,^4-5 circumcision could offer some immediate protection against spread of the disease until such time as effective vaccines become available.

In all, 41 correspondents or groups contributed 50 responses to this education and debate article. Of the 28 correspondents who gave their address, 10 were from Canada, eight from the United Kingdom, four from the United States, four from Australia, one from new Zealand, and one from India. Broadly speaking, most of the rapid responses posted were against male circumcision and the hypothesis that it protects against HIV infection, with the remainder calling for more research on the subject.^5-1