Viral load: small change by sixth day of treatment can often predict poor response.

A U.S. National Institutes of Health study of 124 pediatric and adult patients taking protease inhibitors for the first time found that the change in viral load in the first six days of the treatment was able to predict many cases of poor response of the regimen by week 12. Therefore treatment could be changed quickly in these cases (instead of at 4 or 8 weeks, as recommended by current U.S. guidelines), reducing the development of drug resistance by minimizing the time on an ineffective therapy.

In this study, "reduction in plasma HIV-1 of less than 0.72 log by day 6 after initiation of therapy predicted poor long-term responses in more than 99% of the patients." For those with less than a 0.96 log reduction, the chance of poor response at 12 weeks was 95%. But while very good at predicting some cases of drug failure, 6-day viral load was not as good at assuring long-term success. This is because unpredictable events can occur after day six, such as new mutations that cause viral resistance.

In clinical practice, some patients will not get their blood drawn on exactly the sixth day. Presumably the cut-off value for counting drug failure should be adjusted if the second blood draw is at, for example, 7 days, although the paper did not discuss this. However, this study found that samples taken at day 13 or day 28 were not as predictive as those taken at day 6 --probably because of the more complex factors affecting viral load after the initial period of rapid decline.

Prediction may be better if there is also a viral load test on day 3, 4, or 5 -- due to natural fluctuations of viral levels, and also due to the variability of test results. The researchers did not seem to think that this extra test would be necessary in clinical practice.

This research team previously reported that the trough (lowest) drug concentration in blood plasma at the end of the first week, but not the dose, correlated with viral decline, and predicted long-term response. But it would be hard to measure blood levels of three or more drugs, or to know how sensitive the virus was to them. The authors suggest that the much simpler measurement of early viral load change is good enough.

Comment

Physicians might want to look at viral load decline by day six of certain new treatments -- in order to quickly change a clearly ineffective regimen. But the patients and treatments in this study were often not representative of what physicians see today. We hope the guidelines committees will study all the available information, and decide if a day-six viral load test should now be recommended for patients starting or changing antiretroviral therapy -- or if additional studies should be done first.