PI Perspective #18

April 1996     View PDF     En español

Viral Load Testing Confirmed

Compelling new data from three major studies show that viral load (HIV RNA levels) is an excellent predictor for risk of disease progression, easily superior to CD4+ counts in some circumstances. The new data also suggest that viral load is a better predictor of clinical outcome than CD4+ cell counts when used to monitor the effects of antiviral therapy and that combining the two markers may be even more predictive. Based on these and other data, makers of the new viral load tests (Chiron’s “b-DNA” and Roche’s “Q-PCR”) have applied to the FDA seeking wide licensing approval for the tests. The applications sought for licensing include disease staging and monitoring response to antiviral therapy.

The Upjohn/Pharmacia Studies
Viral load data from two studies by Pharmacia and Upjohn indicate that a decrease in viral load after 8 weeks of therapy predicts the risk of disease progression for the next 52 weeks. One study compared AZT to AZT + delavirdine in people with 200–500 CD4+ cells who had not received previous therapy. The other study compared ddI alone to ddI + delavirdine for people with fewer than 300 CD4+ cells who had received prior therapy. A total of 1740 people participated in these studies. Although the study is still blinded as to which therapy people are receiving, the viral load data is compelling. It shows that people who experienced a half log increase in HIV RNA levels after 8 weeks of therapy were 70% more likely to suffer disease progression than people who did not have such a rise. Conversely, people who experienced a half log decrease in HIV RNA levels after 8 weeks of therapy had a 56% lower risk of disease progression than those who did not have such a decrease.

HIV RNA levels at study entry (baseline) also predicted the risk of disease progression. People with lower HIV RNA levels at baseline were less likely to progress compared to those with higher HIV RNA levels. Small short-term increases in CD4+ cell counts did not decrease the risk of progression, but sustained increases in CD4+ cell counts did correlate with a lower risk of disease progression.

Analysis of the Upjohn/Pharmacia data suggests that there were different viral load threshold levels in the two different clinical studies which indicated increased risk of disease progression over the subsequent year. For the group with CD4+ cell counts under 300 at study entry, a viral load threshold above 100,000 copies of HIV RNA was associated with short-term disease progression. For the group entering the study with CD4+ cell counts between 200 and 500, the threshold level was one million copies of HIV RNA. This suggests that for people with advanced disease (CD4+ counts under 300), even modest elevations in HIV RNA increased the short-term risk of progression. For people in earlier disease (CD4+ cell counts between 200 and 500), a considerably higher level of HIV RNA was needed to trigger short-term progression.

The Mellors/Pittsburgh Study
A very striking study conducted at the University of Pittsburgh looked at the role played by the “set point” of virus established early in the course of infection. Most people suffer an initial burst of HIV RNA shortly after they become infected, but the level quickly decreases as the body stabilizes the infection. This level is called the “set point.” Scientists have wondered whether or not the level of RNA at which the body “stabilizes” might play a role in predicting disease progression. It is known that some people stabilize at extremely low or even undetectable levels, while others stabilize with HIV RNA levels in the thousands, tens of thousands, or even hundreds of thousands. Could this difference in the level of HIV RNA help explain the difference between “long-term non-progressors,” “normal” progressors, and “rapid” progressors?

To answer this question, HIV RNA levels were taken from 181 people who were enrolled in the Multicenter AIDS Cohort study (MACS) between April 1984 to March 1985 in Pittsburgh. Blood samples were collected and stored every six months. Chiron’s second generation branched chain DNA (bDNA) test was used to measure viral load. This test is able to accurately detect counts greater than 500 HIV RNA copies. Perhaps the most striking finding from this study is that even small differences in the early baseline or “setpoint” HIV RNA levels were highly predictive of survival ten years later. This study found that as little as a two fold difference in baseline HIV RNA levels was enough to show a difference in survival. Study data, not yet published, showed declining survival curves over time any time the HIV RNA level was above 5000 copies. This does not mean that such levels of HIV RNA were predictive of short term disease progression, but only that they affected the overall long-term survival of the patients. Similar to other studies, high CD4+ cell counts at baseline were not, by themselves, predictive of survival. However, as would be expected, people with low CD4+ cell counts were indeed shown to have a much higher risk for death in the short-term. Among people with CD4+ counts over 500, there was a strong correlation between higher viral load and risk of death. In other words, a high CD4+ count and a low viral load together predicted long-term non-progression, but people with a high CD4+ and a high viral load had a much greater risk of disease progression and death over time. Changes in viral load over time were also predictive of survival outcomes. People whose viral load doubled from the baseline level were 28% more likely to die and 29% more likely to progress to AIDS than those whose viral load continued at the original levels.

The O’Brien/VA Study
The third study was from the Veterans Affairs Cooperative Studies Program trial 298 conducted in the late 1980’s. This study compared AZT to placebo for people with 200–500 CD4+ cells. Earlier analysis of this study showed that the AZT-treated patients were less likely to suffer disease progression over 18 months. Blood samples were collected and stored every four months during the study. The Roche PCR (polymerase chain reaction) test was used to measure the viral load. Originally 338 people participated in this study, however blood samples from only 270 of those people could be used. This study found that both baseline CD4+ cell counts and viral load were predictive of progression to AIDS. Changes in viral load after treatment, however, were more predictive of progression and death than CD4+ counts. CD4+ counts alone accounted for only a small portion of the effect of treatment, while viral load accounted for nearly 60% of the treatment effect. Using the two markers together, however, accounted for most of the effect seen from treatment. One possible way to interpret these data is to suggest that physicians consider both markers in making treatment decisions. Since it is unlikely that viral load measures would be useful in guiding decisions regarding the prevention and treatment of opportunistic infections, CD4+ monitoring remains essential. A new standard-of-care should include both tests.

The Roche PCR is likely to receive FDA approval within the next few months, with the Chiron bDNA test following on its heels later this year. Another test, called NASBA, is largely used in Europe and appears to be gaining wider utilization in the United States as well.

Other Antiviral Update

LAST      NEXT

Viral Load: Time Has Come