Vaccine and Infectious Disease Division

Biomarkers of study drug adherence

At high (80 to <95%) and very high (95 to <102%) adherence, a statistically significant greater reduction in log10 plasma HIV-1 RNA was observed in acyclovir compared to placebo arm participants. Compared to the very high adherence group, both the ‘over-adherence’ (≥102 %) group and the incalculable group had significantly smaller changes in viral load, consistent with poorer compliance to taking the medication. Low and moderate adherence were also associated with smaller plasma HIV-1 RNA change.

The Partners in Prevention study, a large study conducted in Africa among 3,408 HIV discordant couples, in which one partner had HIV and herpes simplex virus 2 (HSV-2) and the other was HIV negative, concluded in 2009 but its analysis continues to yield new insights into HIV, HSV-2 and other aspects of viral infection and study design. Recently, VIDD principal staff scientist Dr. Deborah Donnell and colleagues from the University of Washington analyzed drug adherence data from the Partners in Prevention study to assess whether pill counts is an effective measure of drug adherence.

The study’s primary goal was to determine whether the antiviral drug acyclovir, which suppresses genital ulcers due to HSV-2, could reduce HIV transmission. The study researchers found that acyclovir lowered blood levels of HIV virus but did not reduce HIV transmission. The co-infected partner was asked to take two acyclovir pills daily, was given a pill bottle with 80 pills every month and instructed to return the bottle with any remaining pills at their next visit. The researchers then counted the remaining pills to determine how many doses the participants may have missed. Such pill counts are commonly used as a measure of drug adherence but they are not a direct measure, as participants could throw away extra pills to appear to be compliant. Donnell and her colleagues asked whether HIV virus levels in the blood and genital ulcer symptoms corresponded with the pill-count measure of adherence, and found that participants with less than 80 percent adherence did not have as marked reduction in HIV levels or genital ulcers as participants with higher adherence. They also noted that “over adherence” (meaning too few pills returned) also corresponded to less reduction in virus levels and genital ulcers, likely a result of poor adherence. The correlation of these biomarkers with pill counts indicates that rigorous pill counting is a good measure of study drug adherence, although periods with missing pills also likely correspond to periods of poor adherence.

Donnell DJ, Baeten JM, Hong T, Lingappa JR, Mujugira A, Nakku-Joloba E, Bangsberg D, Celum C. Correlation Between Pill Counts and Biologic Effects in an HIV-1 Prevention Clinical Trial: Implications for Measuring Adherence. AIDS Behav. 2012 Aug 7.

Fred Hutchinson Cancer Research Center is a world leader in research to prevent, detect and treat cancer and other life-threatening diseases.