Further evidence of super-infection found in African sex workers and Swiss drug injectors

Super-infection - that is, infection with a genetically diverse strain of HIV once chronic infection is established - was the topic of three presentations on the first full day of the Second International AIDS Society Conference on HIV Pathogenesis and Treatment in Paris.

Although none of the studies provided any major new insights in addition to the five published studies that provided the ‘proof of principle’ that chronic infection with one strain may not provide protection against challenge from another, they did add something further to the knowledge-base that might begin to answer the burning question of how often super-infection occurs and what the clinical relevance of super-infection might be in people on HAART.

The first study looked at the emergence of new recombinant viruses - one that combines genetic material from two HIV subtypes or recombinants to become a new type of HIV - in sex workers in Africa. Long after her death in 1998, a long-term survivor from the Nairobi Pumwani Sex Workers Cohort was found to have been super-infected with subtype C virus at least ten years after being infected with subtype A. This subsequently became a new recombinant virus that dominated the other strains of HIV in her body. This woman had been a commercial sex worker for four years prior to entering the cohort in 1985, aged 22, saw two or three clients a day, and used condoms only 5% of the time, despite ongoing counselling. In 1992, she suffered from what appeared to be seroconversion illness and her CD4 cell count plummeted from around 800 to around 200 cells/mm³.

A US lab painstakingly performed serial, complete RNA sequence analysis on three samples of her blood: from 1986, 1995 and 1997, and provided strong scientific proof that indeed super-infection had led to recombination.

Another study also in African sex workers appeared to find a super-infection rate of 1.3% over four years, although the methodology was not as rigourous as the previous report.

This study, from Burkina Faso, screened 152 high risk sex workers (i.e. those that continued to practice unprotected sex with their clients) out of a cohort of 447 that were enrolled in a prospective study between 1998 and 2002. Using a procedure based on the Heteroduplex Mobility Assay they found that four of the 147 DNA samples analysed had dual or co-infection. Retrospective analysis of stored blood samples found that two of the four had been co-infected with two strains of the virus during initial infection. Two, however, appeared to have acquired a second strain during the four year study.

One of the women presented with two separate recombinant strains - CRF02-AG and CRF06-cpx - and the other with CRF02-AG and a divergent clade A virus. The first woman had a rise in viral load when she was apparently super-infected, from 55,287 to 187,927 copies/ml. The second woman, however, only had a very slight rise in viral load when she was apparently super-infected - from 134,173 to 155,421 copies/ml. Although super-infection was the most likely cause of these dual infections, RNA sequence analysis was only done on the env gene, and could have missed co-infection at the start of the study. In the absence of virus from the infecting partner, it is impossible to know whether super-infection occurred here at all.

The most intriguing piece of data came from a Swiss study of intravenous drug users (IDUs) using both injected cocaine and heroin, that looked at both new and chronic infections. No super-infection was found during follow-up of 52 newly infected IDUs over a year or longer, but amongst the chronically infected IDUs who had an unexpected rise (> 1 log) in viral load, three were found to have been super-infected. Two were long-term non-progressors with subtype B not on HAART who subsequently experienced acute retroviral syndrome followed by subtype CRF11 becoming the predominant virus.

A third, however, was found to have been super-infected transiently - that is, only one of the eight blood samples examined contained both CRF11 and B subtypes, where previously there had only been subtype B. Unfortunately this intriguing data was found during a systematic search of all stored blood samples in the cohort, and no RNA tests could be done on stored cells. This means that sample contamination or some other factor may be at play. Since this person had no change in viral load or CD4 count at the time of the apparent transient infection, it is difficult to gauge what factors, if any, might have helped this person fight off super-infection. If transient super-infection is found again, this might provide the ‘proof of principle’ that super-infection may not always lead to chronic infection with the new clade of virus: an intriguing idea.

Both injecting drugs and being a sex worker put the people reported on in this article into the highest possible risk category of being exposed to new HIV subtypes, which may or may not be equivalent to gay men barebacking with multiple partners on a weekly basis. Therefore, no conclusions should be drawn from these studies about the relevance to the barebacking debate currently raging amongst gay men in the West.