Response to commentaries.

The comments of Dr. Fisher provide us with a welcome opportunity to clarify concerns that some may have regarding the use of testosterone therapy in HIV+ men. We have studied the effects of testosterone in nearly 200 HIV+ men and have found the treatment to be effective not only in improving sexual desire and function, but also in alleviating depression, fatigue, and wasting symptoms--all of which greatly enhance the quality and possibly duration of life. Through systematic research, such as ours and others', more physicians are viewing testosterone as a safe, effective treatment for people with HIV. The concerns of Dr. Fisher and others need to be addressed, through both discussion and further research, so that if testosterone is determined to be an effective, safe treatment, patients desiring the treatment will not be met with unnecessary resistance.

Contrary to Dr. Fisher's perception, the primary goal of our research with testosterone is not to "stimulate sexual desire and function," but rather to treat a medical deficiency state and the accompanying clinical manifestations of hypogonadism. We consider it the obligation of physicians and other health professionals to treat conditions that they have diagnosed (or refer to experts if they themselves are unqualified to treat) when treatments are available. We consider it the proper role of research scientists to assess the safety and efficacy of treatments that have not yet been established for a particular diagnosed condition or a new patient population. The medical use of testosterone was becoming increasingly common in HIV practice here in New York even prior to our research, despite the absence of systematic research into the effects of the treatment. As research scientists, we hope that our work may make the approach to testosterone treatment more rational, identify risks and benefits, and document and measure the extent of changes in mood, sexual functioning, energy level, and weight, if they occur.

Dr. Fisher considers testosterone therapy as "ill-advised" for people with HIV because he believes that the value of improving the quality of life and daily functioning of people with HIV is outweighed by the risk to society, as HIV+ men who receive testosterone may infect others. We believe that this view is erroneous and misguided. The logical extension of Fisher's argument is that any treatment for people with HIV increases societal risk, because the longer they are alive, the longer they pose a potential threat to the community as disease vectors. Would Dr. Fisher recommend withholding the new antiviral combination treatments, which appear to have dramatic results for many patients, given that such treatment may lead such patients to resume sexual activity? We consider withholding treatment for a diagnosed condition for the purpose of societal control an ethically indefensible position.

The rates of unprotected anal or vaginal sex within our sample (4%, 3/80 at baseline; 8%, 6/73 at Week 12) are far below those found in other studies describing the prevalence of high-risk behavior among HIV+ samples; therefore, our data add no credence to the suggestion that testosterone increases the likelihood that people with HIV will engage in unprotected sex. Nearly half the men in our sample remained sexually abstinent even after three months of successful treatment. As described in our article, most of the eight men in our study who did report unprotected anal sex at some point during the study engaged in such behavior with other HIV+ men; each occasion of unprotected anal sex occurred in the context of mutual agreement and consent.

Dr. Fisher cites references (some of which have not been published) to suggest that people with late stage HIV, such as those in our sample, are more likely to have AIDS dementia and be depressed, which may increase their likelihood to have unsafe sex. Several studies have shown that depression does not increase as illness stage progresses (Joseph et al., 1990; Perry et al., 1993; Rabkin, 1996), and there is no evidence that depression or dementia is associated with high-risk sexual activity. Although a portion of our sample was depressed, none had significant symptoms of dementia or cognitive impairment, as we screened out men who had such symptoms.

Not only is there no evidence that testosterone therapy increases the likelihood that people with HIV will engage in unprotected sex, the treatment may actually enhance the probability of condom use. Our qualitative interviews with patients and colleagues indicate that erectile dysfunction leads some men to become frustrated with condoms and occasionally not use condoms at all because they lose their erection when they try to use a condom. By improving erectile function, testosterone may increase the likelihood that such men will be able to maintain an erection when using condoms, and hence reduce the frequency of unprotected sex.

As expressed by Dr. Tiefer, our ethical responsibilities as research scientists are focused primarily on the welfare of our patients. However, we do take the public health risk seriously and have implemented safeguards into our study protocol, as described in our article. Dr. Fisher questions the ethics of our decision to continue treatment with patients who reported unprotected anal or vaginal sex with people who are not HIV+ once treatment has already begun. We believe that it is ethically questionable to terminate treatment for a patient when continuing the treatment provides us with the opportunity to counsel such patients concerning safer sex; terminating treatment certainly would do nothing to reduce their risk behavior, and they could still continue treatment through their primary care physician. Dr. Fisher questions the efficacy of primary prevention and safer sex counseling, yet one of the most dramatic examples of behavioral change in the annals of public health is the decline in the incidence rates (new cases) of HIV seropositivity among gay men once the mode of transmission was identified and safer sex methods disseminated. In 1982, the annual incidence rate of HIV among gay men in San Francisco was 21%, but by 1986 it had declined to less than 1% (AIDS Monitor, 1988)--a rate that has been Maintained least through the late 1980s (Kingsley et al., 1991).

Finally, the comments of Dr. Tiefer concerning the unfortunate tendency of confusing sexual moralism with sex theory, research, and education are well targeted in response to Dr. Fisher's view. Dr. Fisher's sentiments remind us of a comment made by a member of our institute's IRB when our protocol was first reviewed: "the whole thing [testosterone therapy for HIV+ men] is a little like helping Typhoid Mary find more jobs as a cook." It is not sexuality that constitutes risk, but rather, ignorance. Although further research is needed on the effects of testosterone, results to date do not support Dr. Fisher's strong reservations. Rather, the data indicate that testosterone has the potential to improve the quality of life of HIV+ individuals in several domains and therefore should be supported by the scientific and professional communities.

References

AIDS Monitor. (1988). Safe sex stops the spread of the virus. New Scientist, 117, 36.

Joseph, J., Caumartin, S., Tal, M., Kirscht, J., Kessler, R., Ostrow, D., & Wortman, C. (1990). Psychological functioning in a cohort of gay men at risk for AIDS: A three-year descriptive study. Journal of Nervous and Mental Disease, 178, 607-615.

Kingsley, L. A., Zhou, S. Y., Bacellar, H., Rinaldo, C. R., Chimel, J., Detels, R., Saah, A., Van Raden, M., Ho, M., & Munoz, A. (1991). Temporal trends in human immunodeficiency virus type 1 seroconversion 1984-1989. American Journal of Epidemiology, 134, 331-339.

Perry, S. W., Jacobsberg, L., Card, C., Ashman, T., Frances, A., & Fishman, B. (1993). Severity of psychiatric symptoms after HIV testing. American Journal of Psychiatry, 150, 775-779.

Rabkin, J. G. (1996). Prevalence of psychiatric disorders in HIV illness. International Review of Psychiatry, 8, 157-166.