Study challenges current thinking regarding viral load as a predictor of CD4 cell loss

Edwin J. Bernard
Published: 26 September 2006

Plasma HIV RNA (viral load) measurements predict no more than 9% of the rate of CD4 cell loss in untreated HIV-positive individuals, according to the findings of a large, retrospective multiple cohort study published in the September 27th edition of the Journal of the American Medical Association. An accompanying editorial argues that these findings “have several important clinical implications.” Not only should viral load measurements play a diminishing role in informing decisions regarding when to start antiretroviral therapy, it argues, but now that we understand that unique genetic and immunological factors affect more than 90% of CD4 cell depletion on an individual level, “future improvements in the treatment of HIV infection and AIDS,” are bound to follow.

Since the results of the groundbreaking MACS study were published in 1996 – which found that plasma viral load broadly predicted the rate of CD4 decline – most clinicians have agreed that the CD4 count represents the current state of the immune system, and HIV viral load reflects the rate at which the immune system will further decline.

The hypothesis that followed on from this observation was that HIV was directly responsible for CD4 cell death, although many studies have since suggested other indirect methods of CD4 cell destruction.

Since the MACS study was published it has been observed in clinical practice that the rate of CD4 decline varies enormously between individuals with similar viral load measurements. In addition, we now know that a variety of unique genetic and immunological factors can affect an individual’s response to HIV infection. Consequently, HIV treatment guidelines have begun to play down the role of viral load as a diagnostic tool for deciding when to start treatment.

In order to assess just how much of an individual’s CD4 cell count decline can be predicted by plasma viral load levels – and, therefore, how much these other factors may be responsible - investigators from five United States cities overseeing three demographically diverse HIV cohorts examined viral load measurements and CD4 counts from 1289 chronically-infected individuals not receiving antiretroviral therapy.

The study population comprised 891 individuals (16% of whom were female, 34% of whom were non-White, 57% of whom did not acquire HIV via sex between men) from four HIV treatment centres comprising part of the Centers for AIDS Research Network of Integrated Clinical Systems (CNICS); 312 gay men from the San Francisco Men’s Health Study (10% of whom were non-White); and 89 participants from the Research in Access to Care for the Homeless Cohort (REACH), also in San Francisco (13.5% of whom were female, 56% of whom were non-White, and 20% of whom did not acquire HIV via sex between men).

Participants were eligible for inclusion in this analysis if they had at least six months of follow-up, at least one plasma HIV RNA (viral load) test, and at least two CD4 cell count measurements. Follow-up averaged 2.3 years.

Although 1289 individuals from these three cohorts were included in the analysis, another 712 were not included because they did not have at least six months of follow-up, primarily because they began antiretroviral therapy within six months of entering the cohort, due to more advanced HIV disease. In contrast, only 10% of those who were included in the analysis initiated antiretroviral therapy within a year of entering the cohort.

When the participants were stratified according to broad viral load measurements (below 500 copies/ml; 501-2000; 2001-10,000; 10,001-40,000; above 40,000) average CD4 cell count decline was found to be consistently greater as viral load increased, similar to that observed in the 1996 MACS study. This ranged from an average CD4 cell loss of 20 cells/mm3 per year in the below 500 copies/ml group, to a loss of 78 cells/mm3 per year in the above 40,000 copies/ml group.

However, within each viral load group the investigators observed a large interpatient variability of individual CD4 loss. This was true across all of the different viral load strata, and included individuals who gained CD4 cells as well as losing far fewer (or more) than the average of the strata as a whole.

Using a statistical method to determine what is known as the coefficient of determination – which calculates the proportion of an effect (in this case CD4 decline) "explained" by the predictor variable (in this case viral load) – they found that only 4% (95% CI, 2-6%) of CD4 cell count changes over six months could be accounted for by the viral load measurement taken at the start of the six months.

To confirm this surprising finding, the investigators used the same statistical methods on the 1512 individuals in the MACS cohort (whose data is available in the public domain) and found that just 6% (95% CI, 4-8%) of CD4 cell count changes over six months could be accounted for by the viral load measurement taken at the start of the six months.

Various sub-analyses within the investigators’ own cohorts, and the MACS cohort – to exclude possibilities of inclusion or exclusion bias or of viral load or CD4 measurement error – came to broadly similar conclusions: a 5.2% role of viral load in predicting CD4 count changes over six months in the investigators' cohorts, and a 7.8% role in the MACS cohort.

The highest coefficient of determination was calculated when the investigators focused on individuals with an initial CD4 cell count between 200 and 350 cells/mm3. This is the group of individuals who face the decision of exactly when to start antiretroviral therapy, based on current guidelines. Amongst the 175 participants in the investigators’ cohorts (with a median follow-up time of 1.9 years) viral loads at baseline still only explained 9% of CD4 cell count changes.

“These findings represent a major departure from the notion that plasma HIV RNA level is a reliable predictor of rate of CD4 cell loss in HIV infection and challenge the concept that the magnitude of viral replication (at least as reflected by plasma levels) is the main determinant of the speed of CD4 cell loss at the individual level,” they write.

Using an individual’s viral load as a prognostic tool to assess when to start therapy is “therefore of limited clinical value,” the investigators conclude.

The investigators then discuss what factors other than the direct effects of HIV replication (as measured by standard plasma viral load testing) may be responsible for the more than 90% of CD4 cell count changes. They note that “HIV infection is associated with heightened T-cell activation and cellular turnover, and expression of immune activation markers is associated with both clinical disease progression and rate of CD4 depletion...Our results provide further support for additional studies exploring the relative contribution of immune activation to the pathogenesis of immune deterioration in treatment-naïve HIV-infected persons.”

One such study published in 2004 found that the immunosuppressive glucocorticoid drug, prednisolone, can raise CD4 counts in HIV-positive individuals not on antiretroviral therapy for at least two years, thus delaying the depletion of CD4 cells by HIV. The suggestion is that it does this by reducing the chronic immune activation caused by HIV infection.

An accompanying editorial in the same issue of JAMA by Dr Keith Henry from the University of Minnesota and colleagues from the University of Pennsylvania and the National Institute of Allergy and Infectious Diseases, calls the study’s findings “provocative” and notes that “25 years into the HIV epidemic, a complete understanding of what drives the decay of CD4 cells – the essential event of HIV disease – is still lacking.”

The editorial argues that the study “may have several important implications”, alongside the main conclusion that “baseline measurements of viral load alone should have less of a role in driving decisions on when to start antiretroviral therapy for an individual patient.”

A “potentially more exciting implication…is that future improvements in the treatment of HIV infection and AIDS may result from improved understanding of the [more than] 90% of CD4 cell depletion that remains enigmatic.”

It concludes by discussing the possibility that immunomodulating therapies – as well as those which target unique genetic factors affecting HIV disease progression – have “the potential over the decades ahead to build on the success of antiretroviral therapy and expand access to sustainable effective therapy.”


Rodriguez B et al. Predictive value of plasma HIV RNA level on rate of CD4 T-cell decline in untreated HIV infection. JAMA 296 (12): 1498-1506, 2006.

Henry WK et al. Explaining, predicting, and treating HIV-associated CD4 cell loss. JAMA 296 (12): 1523-1525, 2006.

Community Consensus Statement on Access to HIV Treatment and its Use for Prevention

Together, we can make it happen

We can end HIV soon if people have equal access to HIV drugs as treatment and as PrEP, and have free choice over whether to take them.

Launched today, the Community Consensus Statement is a basic set of principles aimed at making sure that happens.

The Community Consensus Statement is a joint initiative of AVAC, EATG, MSMGF, GNP+, HIV i-Base, the International HIV/AIDS Alliance, ITPC and NAM/aidsmap

This content was checked for accuracy at the time it was written. It may have been superseded by more recent developments. NAM recommends checking whether this is the most current information when making decisions that may affect your health.

NAM’s information is intended to support, rather than replace, consultation with a healthcare professional. Talk to your doctor or another member of your healthcare team for advice tailored to your situation.