Is HIV Less Virulent Today, or More So?

Conflicting Research Spurs Debate, Calls for Further Research

IAVI_research.jpg used under a Creative Commons license at
Photograph © International AIDS Vaccine Initiative

In December 2014, two studies investigating the comparative virulence of HIV in Southern Africa and Europe, respectively, drew two very different conclusions.

The first, conducted in Botswana and South Africa, suggested that the virus' adaptation to certain varieties of HIV-resistant genes—called human leukocyte antigen B (HLA-B)—effectively weakens the virus' ability to replicate, thereby slower disease progression.

The second, which followed a cohort of European patients over several years, looked specifically at the average viral load and CD4 count just after the acute stage of infection and concluded that, in terms of disease progression alone, HIV has become far more virulent with faster disease progression.

How is it possible that the two studies ended with such strikingly different interpretations? Is it simply the case of questionable study design, or is it possible that the variability of the virus from continent to continent—or even country to country—has led the teams of scientists in completely opposite directions?

Measuring HIV Virulence in Botswana and South Africa

In the first study, scientists at Oxford University, headed by lead investigator Rebecca Payne, questioned whether the presence of certain HLA-B genes—strongly associated with slower disease progression and better viral control—might have induced mutations in HIV which could effectively weaken its "viral fitness."

Earlier research had shown that certain populations had a higher percentage of individuals with this rare, HIV-resistance mutation, ranging from 75% in Japan to 20% in South Africa. In looking at the disparity, the investigators began to wonder whether this might contribute, in some part, to the vast differences in the epidemic between low-prevalence countries like Japan and hyper-prevalent regions like sub-Saharan Africa.

Since HIV rates are still relatively low in Japan, the researchers focused their research on a cohort of patients in Botswana, a country in which the HIV epidemic reached its peak in 2000, and compared it to a matched cohort in South Africa, which only reached its peak in 2010.

The initial survey revealed that the average viral load among untreated patients in Botswana, where the disease is "older," was far lower than that of South Africa, where the disease is ten years "younger" (15,350 copies/mL versus 29,350 copies/mL, respectively). Furthermore, despite having a CD4 count 50 cells/mL lower than that of South Africa, Botswanans with HIV tended to live longer, suggesting a less virulent subtype.

With this evidence in hand, the investigators then looked at the genetic structure of the patients' HIV and found that a higher number of Botswanans had an HLA-B "escape" mutation (meaning that the virus had adapted to the presence of the HLA molecule to escape detection). In doing so, the scientists believed that the "fitness" of the virus may have been weakened, slowing its replicative capacity as well as its ability damage a patient's immune system.

All told, 46% of the Botswanan cohort had key HLA-B mutations compared to only 38% of South Africans. Test tube assays seemed to support the hypothesis, with HIV from the Botswanan sample replicating 11% slower than that from South Africa.

Based on statistical data from antenatal clinics, Payne and her team have further suggested that HIV virulence may have begun to wane in South Africa, as well, with the average viral load among untreated women decreasing from 13,550 in 2002-2005 to 5,750 in 2012-2013.

Measuring HIV Virulence in the European CASCADE Cohort

The European study took a far simpler, real-world approach, wherein patient data from the long-standing, pan-European CASCADE cohort was analyzed from 1979 to 2002. In their research, the CASCADE investigators focused on two key factors:

  • the average CD4 count after seroconversion (which determines how profoundly HIV infection has weakened a person' s immune system), and;
  • the average viral "set point" (where the viral load settles after the acute stage of infection, with higher viral loads generally correlating to faster disease progression).

In their retrospective analysis, the researchers found that the average CD4 count fell from 770 cells/mL in 1979 to 570 cells/mL in 2002, while the average viral set point nearly tripled from 11,200 in 1979 to 31,000 in 2002.

Even more concerning was the speed by which the disease seemed to advance, year on year, in people with HIV. According to the research, the average time it took for a patient's CD4 count to drop below 350—the stage by which antiretroviral therapy is recommended—decreased from seven years in 1979 to a mere 3.4 years by 2002.

Key Differences in the Research

Both pieces of research ultimately have their limitations, with study designs that will likely spur debate among scientists and policy makers alike. Among the key differences:

  • While the African study reviewed statistical data from over 2,000 patients in Botswana and South Africa, the actual number of patients included in the replicative capacity sample was not only small (16 from South Africa and 63 from Botswana) but taken at a single point in time.  By contrast, nearly 16,000 patients were included in the CASCADE cohort, all of whom were investigated over far longer periods of time.
  • While Payne and her team focused on the impact of HLA-induced mutations on a patient's viral load, they could not show that presence of these mutations had any impact on CD4 depletion. By contrast, the CASCADE researchers considered the CD4/viral load dynamic central to establishing HIV virulence. They also limited inclusion to only those patients who had been diagnosed within three months of infection, providing a clearer start-point by which to measure disease progression/CD4 depletion.
  • It is important to note, however, that the CASCADE team only conducted sensitivity analyses on white, gay men (to better ensure commonality in treatment history and viral subset). While the analysis suggested that virulence may be leveling off in Europe as a whole—with the community viral load dropping from 31,000 in 2002 to 25,500 in 2008—the same cannot be said for gay men. Since it is known that the rapid spread of HIV through the gay male population (alongside higher levels of treatment exposure) has resulted in greater genetic diversity and transmitted resistance, it is possible that the viral subset affecting this group may, in fact, be more virulent.
  • By contrast, the African study was conducted in countries where heterosexual sex was not only the primary mode of transmission but where, until recently, far fewer individual s were exposed to HIV therapy. As a result, the genetic diversity of HIV within Southern Africa is considered to be far less, with some research suggesting that the regional variability of the virus may allow for profound differences in HIV virulence.

In short, despite shortcomings in the African study and limitations of the CASCADE research, both conclusions could very well be correct. Further investigations are expected from both teams.


Payne, R.; Muenchhoff, M.; Mann, J.; et al. "Impact of HLA-driven HIV adaptation on virulence in populations of high HIV seroprevalence." PNAS. December 16, 2014; 111(50):E5393-5400.

Pantazis, N.; Porter, K.; Costagliola, D.; et al. "Temporal trends in prognostic markers of HIV-1 virulence and transmissibility: an observational cohort study." The Lancet HIV. December 2014; 1(3):e119-126.

Continue Reading