On February 6, 1994, the Columbia Gay Health Advocacy Project (CGHAP) sponsored a conference entitled Oral Sex and Possible HIV Transmission. This conference was co-sponsored by the Gay Men's Health Crisis (GMHC). The event was held at MillerTheater on the Columbia University campus and ran from 2 PM to 6PM. More than 500 people attended the conference. The program consisted of a series of speakers presenting available scientific data, and a panel of community members discussing psychological, social, and educational aspects of the issue. There was extended time for questions and lively discussion from the audience.
The first group of speakers talked about the mechanics of transmission by oral sex. The speakers were Jeffrey Laurence, M.D., (New York Hospital-Cornell Medical Center), Alison Quayle, Ph.D. (Harvard University) and Peter Schlegel, M.D., and Gerard Ilaria, A.C.S.W. (New York Hospital-Cornell Medical Center).
Jeffrey Laurence, M.D.
Biology and Virology.
Dr. Jeffrey Laurence explained that studies have shown that a reliable laboratory can isolate HIV in the saliva samples of about 25% of HIV-positive people. The same labs can isolate HIV in 100% of blood samples from the same people. In those saliva samples in which virus can be found, the concentration of virus is much lower than in blood, semen, or the already low concentrations found in vaginal or cervical secretions. The low concentration of virus in saliva may mean that saliva is less likely to cause infection than blood or other body fluids. This is certainly supported by test tube and animal studies of infectivity.
Why is saliva less infectious than other fluids? It is believed that there is a substance in saliva that inhibits HIV. The inhibition observed may be due to large sugar-protein molecules in the saliva called glycoproteins. These glycoproteins apparently cause HIV to form giant clumps which are not capable of causing infection.
Animal studies also suggest that saliva is an unlikely source of HIV transmission. In studies, concentrated SIV (simian immunodeficiency virus; a virus similar to HIV) was rubbed on the vagina, rectums, and gums of monkeys. Infection occurred in monkeys that had been exposed via the rectum or vagina, but not those exposed via the gums.
In a test tube study, chimp saliva blocked the ability of HIV to infect T4 cells. Experiments with human saliva showed that it was less effective than chimp saliva at inhibiting the virus but still quite effective. In one study, 10-60% of the saliva samples could inhibit the virus, not completely, but by a fair amount. (Patricia Fultz, et al., CDC). At least ten studies have shown that saliva can inhibit HIV. HIV is present in ejaculate, pre-ejaculatory fluid, vaginal secretions, and cells in cervical fluid. None of these fluids contain the glycoproteins that inhibit HIV in saliva and all are more infectious than saliva.
Alison Quayle, Ph.D.
The next speaker, Dr. Alison Quayle, gave an overview of the mucous membranes and the way in which they protect the body from bacteria, viruses, and fungi. Because the mucous membranes are the surfaces of the body which communicate with the outside world, they are subject to a barrage of these organisms. Mucous membranes defend the body against infection by both nonspecific defenses and defenses specific to particular organisms. The nonspecific defenses include a sticky mucous that forms a physical barrier between the outside and the mucous membrane. The body also contains multiple "clearing mechanisms" for getting rid of infected material (cilia in the respiratory tract, peristalsis in the gut). Gastric acid in stomach kills most bacteria and viruses. Mucous surfaces also contain substances referred to as the body's disinfectants that either kill bacteria or viruses or prevent them from multiplying.
The specific defense tactic of the mucosal surfaces is the mucosal immune system. Mucosal secretions contain a class of antibodies called immunoglobulin A (IgA, or secreted antibodies). IgA secreted in mucous binds and neutralizes pathogens such as bacteria and viruses on mucosal surfaces. The mucous linings of the vagina and oral cavity have more layers than the lining of the rectum. The linings of the vagina and oral cavity are therefore more likely to provide protection against the entry of pathogens.
Dr. Quayle reviewed various body secretions, whether they contain virus, and their ability to transmit HIV. She distinguished between the presence of the virus in a body fluid and its infectivity, that is, its ability to multiply and transmit infection. Infectious virus can be identified by culturing susceptible cells in the laboratory and exposing these cells to the body fluid being tested.
HIV-infected blood is highly infectious. HIV-infected semen is also infectious, but less so than blood, according to Dr. Quayle. Studies have reported the presence of HIV in 30% of semen samples. These studies also report that at any particular point in time only about 10% of HIV-infected semen samples contain HIV capable of infecting cells in culture. Semen appears to be more likely to be infectious in men whose HIV infection is recent (6-12 weeks) and in those with advanced HIV disease. However, studies show that HIV-infected men who are asymptomatic are also capable of producing infectious ejaculate. The presence of an inflammation in an HIV-infected man increases the number of infected white blood cells in the semen and thus increases the infectivity of the ejaculate. Treatment with zidovudine (AZT) can decrease the amount of infectious virus in semen.
Studies have shown that about 20% of the HIV-infected women studied have infectious virus in their vaginal secretions at any particular point in time. HIV-infected women are more likely to transmit the virus during their menstrual period. This is because women secrete about a half-cup of blood during menses and blood is very infectious.
Saliva, as Dr. Laurence described above, contains very low concentrations of virus, possibly due to an unidentified inhibitory factor in saliva. Nevertheless, infectious virus has been found in the saliva of some HIV-infected people. The presence of blood in saliva (or any body fluid) probably increases infectivity. Lesions, ulcers, and inflammations caused by other sexually transmitted diseases make the mucous membranes more susceptible to infection. Urine rarely contains infectious virus. HIV can be detected in feces but cannot be cultured in the lab and may not be infectious.
Peter Schlegel, M.D., and Gerard Ilaria, A.C.S.W.
Peter Schlegel and Gerard Ilaria presented information from their study of pre-ejaculatory fluid and HIV. This study (and a second similar one done by Dr. Jeffrey Pudney at Harvard) showed the presence of HIV-infected cells in pre-ejaculatory fluid. However, the cells were not cultured and may or may not have been infectious. Schlegel and Ilaria described the difficulties in performing studies of pre-ejaculatory fluid and the pressing need for further research about this neglected but important topic.
The next section of the conference consisted of presentations by three epidemiologists, Dr. Alan Lifson (University of Minnesota), Dr. Michael Samuel (New Mexico Department of Health), and Rebecca Young (Columbia University).
Michael Samuel, Dr. P.H., and Alan R. Lifson, M.D., M.P.H.
What Does Risk Mean?
Dr. Alan Lifson began by talking about the differing uses of the term "risk." In talking about infection, epidemiologists use the term "increased risk" to mean an increased likelihood that a particular exposure will lead to infection. This increased likelihood is meaningful only if it is statistically significant -- that is, unlikely to have occurred by chance alone. Statistical associations can be significant to varying degrees ranging from small to large.
There are three different questions about HIV transmission via oral sex.
- Is there any chance of infection?
- Can this chance be meaningfully quantified?
- If so, is this chance small or large as compared to other sexual activities?
Dr. Lifson and Dr. Samuel interpret the data now available to indicate that it is possible to transmit HIV during fellatio but the degree of risk is uncertain but probably low as compared to unprotected vaginal or anal intercourse. Multiple detailed case reports indicate that HIV infection can be transmitted from the insertive partner's penis to the receptive partner's mouth via fellatio with ejaculation in the mouth. There are no reports of transmission occurring without ejaculation, but some of the reports do not indicate whether ejaculation in the mouth occurred.
There are no case reports that conclusively show reports of HIV infection occurring from the receptive partner (the mouth) to the insertive partner (the penis) via fellatio. A couple of cases have been reported, but these accounts are not been detailed and it is not known whether these are reliable reports.
There is little data available about the possibility of becoming infected by putting the mouth on the vagina of an HIV-infected woman. One case of woman-to-man transmission via oral sex and several women-to-women cases have been reported, but in little detail. Dr. Lifson believes that these reports indicate that there is a possibility of infrequent transmission via cunnilingus, from vagina to mouth, but that it has not been conclusively proven.
Dr. Michael Samuel presented the information available from cross-sectional and prospective studies done concerning fellatio. All of these studies except one showed that there was no statistically significant increased risk of HIV transmission associated with fellatio. The exception was the study done by Dr. Samuel which looked at the combined data from the San Francisco Men's Health Study, The San Francisco City Clinic cohort, and the San Francisco General Hospital cohort. These data indicate that there is a statistically significant increased risk of infection via fellatio.
Dr. Samuel thinks that the difference between the results of his study and others can be explained in several ways. There may be a mistake in the Samuel's study. Alternatively, because the Samuel's study was the most recent, the increase in the relative number of people who have oral sex but not anal sex may allow the risk of oral sex to be "unmasked" and measured.
Comments on Lack of Data Regarding Women
Ms. Young addressed some of the complex issues regarding transmission from women-to-women or women-to-men via cunnilingus. Clearly, most lesbians have been infected through sex with an infected man or needle-sharing. The relatively small number of lesbians infected make it difficult to draw conclusions from the available data. There have been no large scale studies. Surveys that have been done tend to ask questions in language that is vague and confusing. The scientific community tends to have a lack of knowledge and bias about lesbians that tend to make it difficult to formulate useful surveys.
The final portion of the program was devoted to addressing the complex psycho-social aspects of oral sex and HIV transmission and particularly the difficulty created by ambiguity. We purposely chose a diverse panel of individuals (including both men and women who are HIV-positive and HIV-negative) to discuss these questions with the audience. The panelists were AIDS educators, activists, or policy-makers. The panel consisted of Sally Cooper, Carlos Cordero, Spencer Cox, Richard Elovich, Tonya Hall, Michael Isbell, Wendell O'Neill, and Sarah Schulman.
Some of the issues addressed by the panelists and audience were:
There is a conflict between the standard medical advice (to use condoms and dental dams) versus the reality of practice (few people seem to use barriers for oral sex).
AIDS educators face a dilemma. The cautious approach (use barriers for oral sex) may theoretically be the safest. However, it is possible that in practice, discouraging unprotected oral sex may lead to attitudes of frustration and despair that could increase the amount of (far more dangerous) unprotected vaginal or anal intercourse.
Our ability to conceptualize useful approaches to safer sex is limited by the societal difficulty in discussing sexual experience in all its complexity. We may need to move beyond the "cook book" approach and begin to develop more flexible models of harm reduction in specific contexts.
Lesbians face a difficult conflict. Women-to-women transmission seems fairly unlikely. There are many HIV infected lesbians who became infected either through sex with a man or needle-sharing. However, ignoring the possibility of women-to-women transmission leaves infected women (who are primarily women of color) without any guidelines. Additionally, given the long history of scientific ignorance and bias regarding lesbians, an attitude of skepticism about the data available is reasonable.
The largest agreement among audience, scientists and community panelists was the need for increased discussion about sexual practices and safer sex. We must seek creative ways of allowing these discussions to occur in an atmosphere that encourages honesty and allows for the subtlety, diversity, and complexity of human behavior.
The conference was probably the first large-scale, open, public discussion on the topic of oral sex and HIV transmission and aroused a great deal of both interest and anxiety among the organizers and speakers. GMHC joined CGHAP as a co-sponsors after some debate about the utility of such a conference, given the fact that no clear-cut answers are available. In the end, the representatives from GMHC thought that the conference was a highly useful first step and encouraged increased discussion of the topic among the safer sex educators at GMHC. Most speakers and panelists shared this view. The informal feedback we have received from the audience has been positive.
The AIDS Reader (a free medical update on AIDS funded by Roerig) will publish portions of the conference transcript in the July issue of this journal. The AIDS Reader has a circulation of 25,000 physicians and other health care professionals. An audio tape was also made and can be ordered from GMHC by calling 212-337-3505. The conference was also videotaped and portions of it have been shown on New York cable TV. GMHC reports that it has received an unprecedented amount of requests for tapes of the conference.
We wish to thank all of our generous sponsors who made this conference possible. If you have questions or comments please contact Laura Pinsky (the conference organizer) at:
Columbia University Health Service
John Jay Hall
519 W. 114 St.
New York, NY, 10025