Using antiretroviral treatment as a preventative measure is one of the best hopes for slowing HIV's spread. But there's a problem: HIV is thought to be vastly more infectious in the first few weeks after someone has contracted the virus, so that most of the transmission hazard will have passed by the time doctors diagnose the disease. A new analysis, however, may be cause for hope—it seems HIV might not be as infectious early on as researchers previously thought.
Treatment as prevention is a simple enough idea. Modern antiretroviral drugs can cut HIV counts to undetectable levels, and if there's hardly any copies of HIV floating around in your blood, it's a lot harder to pass them on to anyone else. A 2011 study conducted in nine countries showed that treatment as prevention cut new infections by 89 percent. Similarly, a six-week regimen for expectant mothers reduced the number of mother-to-child infections by about two thirds.
Those are promising results, but they come with a caveat: The risk that a person with HIV infects someone else is highest early on in the course of the disease. According to some, that risk is so much higher in the first few weeks that it would be impossible to diagnose and treat HIV before most of the infection hazard has passed. If so, treatment as prevention might not be the best idea, especially given the high cost of antiretroviral drugs in Africa and other regions that would most benefit.
The resulting estimate of the infection hazard during HIV's earliest stages was nine times smaller than the currently accepted value.
Enter Steve Bellan, a postdoctoral fellow at the University of Texas-Austin, and his colleagues from UT and Yale University. As others have pointed out, almost all of the evidence concerning HIV infectiousness in its early stages comes from a single study of 23 couples with one infected partner in Rakai, Uganda, but that and subsequent re-analyses had several drawbacks. First, previous analyses of the Rakai data didn't adequately take into account the study's methodology, such as the fact it was a study of couples or how the original researchers had selected or excluded couples for their research. Second, none of the analyses of the data took into account how infectivity might vary across the couples—because of genetics, say, or simply how often they had sex.
To address those drawbacks, Bellan and his team developed a computer simulation tailored to the circumstances of the Rakai study. By running the model many times with different parameters—different levels of infectivity, for example—they honed in on a version that best matched the patterns of infection in the study's 23 couples. The resulting estimate of the infection hazard during HIV's earliest stages was nine times smaller than the currently accepted value. Much of the disparity, the team argues, stems from not taking variation across couples into account.
The researchers argue that their findings support using treatment as prevention and ought to be considered when planning prevention efforts. "Our findings cautiously suggest that the population-level benefits might be larger than predicted by earlier estimates," they write in PLoS Medicine.
Quick Studies is an award-winning series that sheds light on new research and discoveries that change the way we look at the world.