Superinfection occurs at 5% a year in recently infected gay men not on therapy.

The Eleventh Conference on Retroviruses and Opportunistic Infections in San Francisco opened with a bang, as a group of investigators from Los Angeles and San Diego provided long-awaited information on the incidence of sexually-acquired superinfection, concluding that, amongst their cohort of 78 recently infected individuals not on therapy, the annual rate of superinfection was 5%.

Superinfection is the re-infection of an HIV-positive person with a slightly different version or strain of HIV. The first documented report of sexually transmitted HIV superinfection emerged in 2002, when Jost and colleagues from the University of Geneva, Switzerland, presented a case study of a 38-year-old gay man who was infected with two different subtypes of HIV on two different occasions more than two-and-a-half years apart.

Bruce Walker and colleagues at Harvard Medical School reported the first case of sexual superinfection within the same subtype later that year. Since then, a series of case reports have documented that infection with a second strain of HIV is definitely possible, under certain circumstances, but until now there had been no data regarding how often this occurs.

Smith and colleagues retrospectively analysed blood plasma samples from 78 individuals - the majority of whom (90%) were gay men - who were enrolled in the San Diego and Los Angeles Acute HIV Infection and Early Disease Research Programmes between December 1997 and June 2003, and looked for independent clusters of the pol gene. When superinfection was suspected, they rigourously tested the samples using four different lines of molecular investigation: clonal (V3) and dye-primer (pol) sequencing, and length polymorphism analysis (V1-2 and V4-5).

Three cases of superinfection were identified in the cohort, two of which have previously been reported as individual cases - by Daar at the 9th Annual Retrovirus Conference in Seattle two years ago, and Koelsch in the journal AIDS last year. In each case, superinfection occurred between five to 13 months after the estimated date of initial infection. Each superinfecting HIV strain was associated with a change in susceptibility to antiretrovirals, even though none of the men were on therapy. Two were initially infected with drug-resistant HIV and then became superinfected with a wild-type strain, while the other was initially infected with a wild-type strain and then was superinfected with a drug-resistant strain. Within six months of acquiring the superinfecting strain, their viral loads increased by an average of 1.6 logs and their CD4 counts decreased by an average of 132 cells/mm3 (p< .05).

All three of the cases were gay men who had multiple sexual partners, making identification of the person who superinfected them (the index case) impossible. This is frustrating, because without the index cases, the possibility of co-infection - that is, infection with two different viruses at the same time - cannot be completely ruled out. However, the researchers' rigourous testing appeared to satisfy the eminent members of the San Francisco audience, who concurred that Smith provided compelling evidence in his oral presentation that these were, indeed, cases of HIV-1 clade-B superinfection.
In fact, since the method they used to identify superinfection is considered to be somewhat conservative, Smith noted that their 5% annual rate may actually be an underestimate, although previous studies in chimpanzees and injection drug users found a similar 5% incidence rate.
Last month, Chakraborty and colleagues asked the question, 'Can HIV-1 superinfection compromise antiretroviral therapy?' in a letter in the journal AIDS. A group of highly respected researchers, including Miguel Quinones-Mateu - who was interviewed on this subject in the December 2002 issue of AIDS Treatment Update - closely examined virus samples from seven individuals whose previously successful HAART regimen failed unexpectedly, in order to see if their hypothesis was correct: that superinfection with a drug-resistant virus caused therapy to fail. They identified only one person in whom this was likely, but upon further investigation discovered that this person had taken an unauthorised treatment interruption during the same month that their viral load had skyrocketed from 12,000 to over 600,000 copies. "Was this the result of transient superinfection with a wild-type (more fit) HIV-1 isolate during treatment interruption or the selection of a pre-existent virus established during an original HIV-1 co-infection?" they ask.

The evidence is mounting, however, that sexually transmitted superinfection appears only to occur in people who are not on antiretroviral therapy. Last year, Gonzales and colleagues looked for superinfection in 718 people, the majority of whom were on therapy - and representing over 1000 person-years of follow-up - finding no evidence of it. They suggested that superinfection may be prevented as a result of partial immunity, viral interference from the original virus strain or the effect of antiretroviral therapy.

This has several implications and begs several more questions. If superinfection is unlikely to occur when both positive partners are on therapy, does that mean that two positive men who 'bareback' with each other and are on successful HAART are now only risking passing on hepatitis C and other sexually transmitted infections, and not superinfection with HIV? Does it also depend on their seminal viral load, and whether they shed HIV in the semen? Does it make a difference if one or both have a concurrent STI? It is impossible to say for certain, but it does seem likely that the 5% annual rate reported here does not apply to people on HAART.

Even if the risk is low for those on HAART, the rapid disease progression that can occur with superinfection might not be considered a risk worth taking. A poster to be presented in San Francisco on Wednesday by Gottlieb and colleagues documents that superinfection led to the rapid onset of AIDS (within three-and-a-half years) and death (within six) despite an initial good immune response to the first infection in a member of the MACS cohort.

Still, if antiretroviral therapy does prevent superinfection, there are positive implications for the as-yet-untested concept of pre-exposure prophylaxis (PREP) for the sexual transmission of HIV. Studies are about to begin on high-risk gay men in San Francisco and Atlanta (as well as on high-risk women in Africa and Asia), where tenofovir monotherapy is being used in an attempt to prevent them becoming infected when exposed to the virus.

Next month's AIDS Treatment Update will feature an article on PREP.

References
Smith D et al. Incidence of HIV Superinfection Following Primary Infection. 11th CROI, San Francisco, abstract 21, 2004.
Gottlieb G et al. HIV-1 Superinfection in a Rapid Disease Progressor: Rapid Replacement of the Initial Strain with the Superinfecting Virus by Natural Selection 11th CROI, San Francisco, abstract 454, 2004.
Chakraborty B et al. Can HIV-1 superinfection compromise antiviral therapy? AIDS 18 (1): 132-134, 2004.