Unprotected sex between long-term partners with HIV: no evidence for superinfection

A study of long-term HIV-positive partners who do not use condoms when they have sex with each other has found no evidence of HIV superinfection, and instead a clear relationship between long-term frequent exposure to their partner’s virus and a strong immune response to that virus, suggesting that repeated exposures eventually build immunity against superinfection.

The findings, published in the October 2008 edition of PLoS Pathogens, a free access online journal, are important, say the University of California researchers, because of the growing practice of `serosorting` - unprotected sex between people of the same HIV status.

Current guidance for people with HIV from many sources is that unprotected sex poses a risk of superinfection – infection with a new strain of HIV that over-runs the existing virus population due to lack of immunity to that virus.

However there is limited evidence about the frequency of superinfection and little evidence that it has harmful effects even when it takes place.

Superinfection seems to happen not only in people who have been recently infected with HIV, but also in those with longstanding HIV infection. A recently reported study in Kenyan women estimated an annual incidence of superinfection of at least 4%, but no evidence of disease progression as a consequence of superinfection. A study in gay men reported an incidence of 5% per year.

Although superinfection has been associated with CD4 cell declines, and a handful of cases of transmitted drug resistance, it does not appear to have a widespread compromising effect on either the health or treatment outcomes of people with HIV.

There is much stronger evidence that unprotected sex with other HIV-infected people is harmful for people with HIV where it involves the risk of exposure to sexually transmitted infections – especially syphilis – or to hepatitis C, both of which have been on the rise among men who have sex with men in Europe and North America.

The study

Researchers from the University of California and San Francisco General Hospital examined HIV-specific T-cell responses in 49 individuals with undetectable viral load on antiretroviral therapy from a prospective cohort of HIV-positive couples receiving treatment through the Positive Health Program at San Francisco General Hospital.

Individuals were divided into two groups: those with partners who had undetectable viral load (n=29), and those with partners who had detectable viral load (n=20). There were no significant differences between the two groups in terms of relationship length, time on treatment, age, duration of infection or CD4 cell count.

Analysing T-cell HIV-specific interferon-gamma responses to various HIV epitopes (portions of viral protein that elicit an antibody response from the host), they found that significantly stronger responses to protease, reverse transcriptase and integrase peptides were detectable in those with viremic partners, with the strength of responses in this group strongly correlated with a greater frequency of unprotected sex and in particular with the frequency of receptive exposure to HIV (being the passive partner in intercourse). There was no correlation with the number of events of insertive intercourse.

There was evidence from two patients in this group whose partners started antiretroviral therapy of a subsequent decline in HIV-specific responses over the course of a year, suggesting that the control of viral load led to a loss of immune stimulus.

There was no evidence of superinfection, as measured by phylogenetic analysis of HIV DNA from patients and HIV RNA from partners.

However the authors speculate that, in order for HIV-specific immunity to be induced, a degree of superinfection must have taken place, but that it is probably confined to the mucosal surface of the body – most likely the rectum and the gut.

This phenomenon is similar to the pattern seen in exposed but uninfected sex workers, who appear to lose HIV-specific immunity if they have unprotected sex less frequently. A study in Nairobi sex workers found that women who took a break from sex work and lost HIV-specific CD8 cell responses were more likely to seroconvert subsequently, suggesting that localised mucosal infection ceased to be contained (although the study could not rule out renewed sexual exposure).

These findings suggest that even where an individual has drug-resistant virus and a detectable viral load, the risks of superinfecting an HIV-positive partner with that drug-resistant virus are low even if that partner is receptive.

The findings provide no information about what happens if an individual with HIV has receptive intercourse with many different partners with HIV, but the study’s lead author Chris Willberg, now of the Biomedical Research Centre at the University of Oxford, told aidsmap: “We would speculate that it is regular exposure to the same epitopes that is required to stimulate the responses. What we did not explore is the ability for new [epitope] responses to be developed through exposure.”

Would women exhibit the same responses if exposed to HIV from a viremic partner?

“The most logical explanation for the maintained responses that we observed is that they were driven by receptive exposure to HIV antigen derived from the viremic partner,” Chris Willberg commented. “Therefore, we would expect to see the same results in women also receptively exposed to viremic partners. If the mechanisms responsible for driving the responses in this study are the same as those that drive responses in exposed uninfected individuals, then there is plenty of evidence to suggest women would respond in a similar manner.”