Is oral sex safer than other forms of unprotected penetrative sex?

The evidence suggests that oral sex is less risky than unprotected anal intercourse. Oral sex is a common practice in both homosexual and heterosexual relationships. Although unprotected oral sex may be a less efficient means of transmitting HIV than unprotected anal or vaginal sex, the frequency of its occurrence may serve to increase its relative contribution to overall HIV transmission.

A case-control study nested within a large international cohort study of HIV-positive gay men has shown that having multiple partners for receptive oral sex increases the risk of seroconversion by one per cent (Page-Schafer).

Is HIV always present in semen?

Viral load testing suggests that traces of HIV can almost always be found in semen. Even if HIV is undetectable in blood it may still be present in high levels in semen. Viral load in blood may not be a good guide to the infectious potential of semen. See Infectivity, for further discussion of this issue. However, several studies have shown a trend towards higher levels of HIV in semen when plasma viral load is also high, and HIV can be isolated much more easily from the semen of individuals who have CD4 counts below 200.

Antiretroviral treatment has been shown to reduce HIV levels in semen in several small studies, but one study found detectable HIV in semen even after virus levels in blood became undetectable. This suggests that virus production in the genital tract may occur independently of virus production in the bloodstream and most other body tissues. This is because scientists describe the genital tract, and the testes, as an immunologically privileged site, where special immune processes take place.

Nevertheless, HIV levels are likely to be highest in the semen of individuals with CD4 counts below 200 not currently receiving antiretroviral therapy.

Is HIV always present in vaginal fluids?

Viral load testing suggests that HIV cannot be found in vaginal fluid with the same frequency as it has been detected in semen. HIV levels tend to be highest around the time of menstruation, or if a genitourinary infection or candida are present.

HIV can be isolated more easily from the vaginal fluids of women with CD4 counts below 200.

Antiretroviral therapy has been shown to decrease levels of HIV in vaginal fluid 10–50 fold when two drugs are used in combination.

Is HIV absorbed through the undamaged tissue of the mouth? What about cuts, sores or damaged gums?

The mucous membrane of the mouth contains few cells which are vulnerable to HIV infection. In contrast, cells which are vulnerable to HIV infection (possessing receptors which enable HIV to bind onto them) are common in the walls of the vagina and the rectum (Lehner).

Furthermore, little is absorbed through the tissue of the mouth even if it is damaged. The mouth is an entry point into the digestive system where enzymes in saliva are secreted to begin the task of digesting food. It is very thick compared with the mucous membrane in the rectum, and fluids stay in contact with it for a very short time because swallowing clears the mouth regularly. Thus the likelihood of semen, blood or vaginal fluid coming into contact with cells such as dendritic or Langerhans cells (also identified as receptors for HIV) in damaged tissue is small. Nevertheless, two cases have been reported where bleeding gums or receding gums are speculated to have been the route of infection (Lifson).

Is HIV absorbed through the tissue of the throat?

The tissue of the throat (the oesophagus) is similar to that of the mouth – thick and fairly unabsorbent. As in the mouth, continuous swallowing or peristalsis ensures that anything ingested will be in contact with damaged tissue for only a short time. However, several cases have been reported where researchers speculate that HIV infection may have occurred as a result of inflammation of the throat, caused either by infections or allergies (Chen, Murray).

It is possible that HIV could come into contact with the immune system cells of the tonsils if semen or vaginal fluid come into contact with them, particularly if the tonsils are swollen or inflamed.

Even where there is an apparently normal mouth, it would be unwise to assume that HIV could not be transmitted through oral sex. Research has shown that the tonsils of rhesus monkeys act as an entry point for SIV (Stahl-Hennig). However infecting the monkeys via this route involved swabbing tranquillised monkeys’ tonsils directly with monkey-adapted HIV after they had been given a mouth-drying agent. This is likely to have greatly increased the chances of infection compared with the normal oral environment.

Is HIV destroyed in the stomach?

There is no evidence to prove or disprove this. Researchers assume that as with other viral infections, HIV will be inactivated by the digestive enzymes and by changes in pH (acidity/alkalinity balance). It is not possible to say whether semen which enters the stomach may enter the bloodstream through stomach ulcers or other inflamed tissue, but it seems highly unlikely that HIV will have survived the process of neutralisation in saliva by the time it reaches the stomach.

Does saliva kill HIV?

There is strong evidence to suggest that an enzyme in saliva called secretory leukocyte protease inhibitor (SLPI) prevents HIV infection of lymphocytes, and therefore inhibits HIV infection in the mouth and throat. A 1989 study showed that even when greatly diluted, saliva was still capable of preventing HIV infection of white blood cells. It is still unclear how this happens and what agent inhibits infection (Fox).

The experiments showed that saliva taken from all 34 participants, nine of whom were seropositive, exhibited inhibitory effects (that is, it prevented infection); in 29 cases (85%) a complete inhibitory effect was observed. Inhibition is measured by incubating saliva with an equal quantity of HIV concentrate and then attempting to culture HIV over 18 days. If HIV cannot reproduce itself a 100% inhibitory effect can be reported. There was little correlation between health status, age or gender in the inhibitory effect of saliva (Fox). Previous experiments had shown that complete inhibition does not take place immediately, but this was more to do with the quantity of the virus than the infectivity of the saliva. Concentrations of HIV much higher than those occurring in blood were used.

Some researchers have also cited the presence of HIV antibodies in saliva as evidence of further protection, but since there is considerable ambiguity over the neutralising effect of HIV antibodies it is not possible to say that this is significant (Fultz).

Is HIV present in saliva?

One experiment took more than two months to culture HIV from the saliva of HIV–positive people, and in the four samples in which the virus was detected, the level was extremely low – one infectious particle per ml. This measure should be treated with caution but nevertheless suggests that saliva contains very low levels of infectious HIV (Piazza).

However, another study by the same research team found high levels of HIV in the saliva of several patients with AIDS when they used a more sensitive HIV PCR viral load test. Indeed, in several cases they found higher levels of HIV in saliva than in semen.

These findings should be treated with caution because the researchers tested only one sample from each patient, and the study was conducted in 1994, when PCR viral load testing was still at an experimental stage. The researchers reported that when they re–tested five patients two months later they found large differences in viral load from the previous study. This may be a consequence of changes in levels of virus production, or it may be due to variations which are a feature of the testing technology (Piazza 1995).

A more recent study found that infectious HIV can be detected at high levels in saliva during the early weeks of HIV infection (the ‘window’ period before antibodies appear), but levels fall rapidly after this point.

Free floating infectious viruses and virus-infected lymphocytes could be detected in saliva taken from individuals with primary HIV infection attending clinics in North Carolina. In 7 out of 8 cases, free floating infectious virus could be detected at an average level of 2,000 copies per ml, and in 5 out of 8 cases cell associated virus could be detected at an average level of 20,000 copies per ml.

The researchers said that virus levels in blood and semen were much higher, and it is not clear how much of a transmission risk the virus levels in saliva might pose, given that studies of transmission through vaginal intercourse have shown that transmission did not occur in serodiscordant couples when the infected partner had a viral load below 1500 copies . However, some individuals in this study had virus levels as high as 500,000 copies in saliva, suggesting that during the early weeks of infection some individuals may be ‘super-excretors’ of HIV, and may play a significant role in the ongoing amplification of the HIV epidemic (Pilcher).

Another point to bear in mind is that viral load of 500,000 copies is associated with a high risk of severe seroconversion illness. A person experiencing such a severe flu-like illness is unlikely to be interested in engaging in any form of sexual activity whilst so unwell.

What about blood in saliva?

An Italian study showed that while very small quantities of blood may be released into the saliva by kissing or eating, large quantities of blood can be released into saliva after brushing the teeth; at least ten times as much as after kissing or eating. Whether the HIV contained in blood would be diluted and inactivated by saliva is unclear (Piazza).

Is it safer to spit out or swallow semen or vaginal fluid?

It probably makes little difference, although it is often suggested that while swallowing semen or vaginal fluid keeps it away from any damaged tissue in the mouth, such as bleeding gums, swallowing also makes it possible for contact with any damaged or inflamed tissue in the throat to occur.

Is oral sex with a male partner safer if ejaculation does not occur in the mouth?

The evidence of all the case reports to date suggests that this is likely. Only two cases have been reported where ejaculation into the mouth did not occur.

What is the risk of infection from precum? Is there a lower concentration of HIV in this fluid?

Research suggests that high concentrations of HIV can sometimes be detected in pre–cum. Several studies show that HIV can be isolated from seminal fluid, and that in the event of male genital tract inflammation, much higher levels of the cells commonly infected with HIV (CD4 lymphocytes and macrophages) are likely to be present in semen (Borzy; Ilaria). The stage of disease does not seem to make a great difference to the level of HIV in pre–cum. Ilaria noted HIV in semen of some men with AIDS he studied, but not others, and found the same unpredictable distribution of HIV in pre–cum amongst asymptomatic men.

Despite this evidence, it is very difficult to judge whether pre–cum is likely to contain a sufficient quantity of HIV for infection to occur. Remember that contact with HIV alone is not sufficient to guarantee infection. The conditions for infection are discussed in The mechanisms of HIV transmission.

Is oral sex more risky with a partner who has AIDS?

Whatever risk is attached to oral sex is probably increased if the person who ejaculates into the mouth of his partner has a CD4 count below 100 and/or is symptomatic. The likelihood that someone with a low CD4 count has HIV in his semen increases as the CD4 count falls below 200; in one study 87% of those men with CD4 counts below 100 had HIV in their semen, compared with none who had CD4 counts above 500 (Vernazza).

What is the risk of infection from oralanal contact (rimming)?

It is possible that the presence of blood in the rectum (for instance as a result of fingering, fisting, having intercourse with or use of sex toys) may present a risk of infection for the person doing the rimming. Faeces may contain traces of blood containing HIV, and HIV infection of parasites found in faeces has been demonstrated in the laboratory. It is also possible that blood may be introduced into the rectum by the tongue, from bleeding gums. One case has been reported where a man rimming another man may have infected his partner by this route. The man who was infected was rimmed by an HIV – infected partner who is speculated to have had gum disease (Gill).

What are the risks of kissing?

There are no reported cases of HIV transmission through kissing, either social kissing or deep passionate kissing. Admittedly it might be difficult to separate out possible cases of transmission as a result of kissing, because it is rarely separate from other sexual activities. See HIV in saliva above for more detailed discussion of this issue, including reasons why transmission through kissing is unlikely.

How soon is it safe to have oral sex after brushing my teeth?

Brushing the teeth and gums often causes mild abrasions, and stimulates bleeding where gingivitis is present. Although a recent report by the Department of Health's Expert Advisory Group On AIDS suggested that such bleeding will probably stop within ten minutes, vigorous fellatio can easily stimulate bleeding in the unhealthy mouth. The best long-term precaution is to pay close attention to gum hygiene and change your toothbrush regularly. In the short term, if your gums normally bleed a lot when you brush your teeth, this means that you are at greater risk of infection if semen containing HIV gets into your mouth. If you are HIV-positive, it also means that you could pass on HIV to partners if you are sucking.

What risk is posed by mouth ulcers?

Mouth ulcers allow food particles and bacteria into the bloodstream, so they will let HIV in too. Mouth ulcers are most common in those under 30, and oral sex should be avoided if you have an ulcer or other cuts or lesions in your mouth.

Can HIV be passed from the mouth to the genitals?

There are no reliable reports in the scientific literature. An American report of a case in which a woman was alleged to have transmitted HIV through performing oral sex on a man with no other apparent risk factors was later shown to have been false. The man did have other risk factors which he didn't reveal to researchers in an initial interview. Similarly, doubts have been raised about the French report of two men infected through receiving oral sex (Rozenbaum).

Nevertheless, if a large amount of blood is present in saliva, there may be some risk.

Is HIV present in urine?

No. Although antibodies to HIV can be detected in urine, infectious HIV has never been isolated from urine (Skolnik). No cases of HIV transmission have ever been reported through taking urine into the mouth, or onto broken skin.

What role for the pharynx?

Whilst debate continues about exactly how risky oral sex really is, researchers from the University of Washington in Seattle and the Imapcta Salud y Educacion, Lima, Peru have successfully cultured infectious HIV from pharyngeal swabs from four men with high pharyngeal HIV RNA.

HIV-1 RNA can be detected in saliva, yet culture of HIV-1 from saliva is only successful in less than 1% of samples due to the inactivation of the virus by saliva. Tonsillar biopsy specimens have previously been shown to harbour both HIV-1 RNA and DNA. The researchers set about describing the predictors and variability of HIV-1 RNA in the pharynx and attempted to cultivate HIV-1 from oropharyngeal surfaces.

In total 64 HIV-positive men without bacterial STIs from Seattle, USA and Lima, Peru were evaluated prospectively at week 0, 2 and 4 to assess viral load in plasma and in swab specimens obtained from the pharynx. A subset of 17 men with high pharyngeal viral load were evaluated one year later for the recovery of infectious virus from blood, tonsil and buccal surfaces.

The median CD4 count of the 64 participants was 290 cells/ml and 45% were currently receiving antiretrovirals. The median baseline viral load was similar in plasma (4.24 log) and the pharynx (4.22 log). Each one log increase in plasma viral load was associated with a 0.323 log increase in pharyngeal viral load whilst both antiretroviral therapy and tonsillectomy were associated with reductions in pharyngeal viral load. There was a statistically significant study site effect between Seattle and Lima with respect to age, treatment, viral load and tonsillectomy; men were older in Seattle (median 39 vs 27 years), 24% in Seattle were using antiretroviral therapy compared with 5% in Lima, median viral in Seattle was 3.12 log compared to 4.75 log in Lima, and 42% of the men in Seattle had a tonsillectomy compared with 13% in Lima.

Infectious HIV was cultured from the pharynx of 4 of the subset of 17 men with high pharyngeal viral load but was not isolated from the buccal mucosa (the inside of the mouth) or the saliva of any of the men. The observation that the oropharynx appears to harbour higher levels of HIV than the buccal mucosa is likely due to the proximity of lymphoid tissue in the posterior oropharynx. Three of the four men were not using antiretroviral therapy and the other was taking dual nucleoside therapy which was not fully suppressive.

Median mucosal RNA was 5.85 log in the culture positive men versus 4.69 log in culture negative men.

Both tonsillectomy and use of antiretroviral therapy were associated with a reduction in pharyngeal viral load. Pharyngeal HIV shedding was higher among persons with tonsils and detectable viral load.

Whilst the detection of culturable HIV RNA in the posterior pharynx may indicate the potential for the oral transmission of HIV further behavioural data are needed if we are to understand the role sex acts such as "deep throating" may play in the oral transmission of HIV.

What role do the cells lining the mouth play?

When exposed to high levels of human immunodeficiency virus (HIV), the cells from the lining of the mouth can develop a low-level infection, according to research from Charles R. Drew University and the University of California, Los Angeles published in the March 2003 issue of the Journal of Virology.

In the study, the researchers tested the ability of HIV to infect oral mucosal cells known as normal human oral keratinocytes (NHOK) in the test tube. They found that when exposed to high concentrations of the virus, the cells established a low-level, productive infection that could subsequently transfer to activated lymphocytes.

“Human saliva contains several types of anti-HIV activity that may help protect an individual against a small virus inoculum. However, if individuals are exposed to inocula containing a heavy viral load, it is conceivable that the oral epithelium could be infected and thus serve as a beachhead for HIV-1 infection,” say the researchers.