Background: P ACTG 1020A, a phase I/II safety and pharmacokinetics study, is enrolling patients from South Africa and the United States, aged 3 months to 21 years, naive to atazanavir (ATV). Patients are treated with ATV ± ritonavir (RTV) combined with 2 nucleoside reverse transcriptase inhibitors (NRTI).

Methods:  Cholesterol (CH) and triglycerides (TG) (nonfasting) measured at baseline and at weeks 24 and 48. CH and TG were log₁₀ transformed for statistical analyses of whether changes occurred during treatment and whether they varied as a function of country, prior treatment experience with or withou RTV. We used t-tests (2 groups), paired t-tests, and McNemar’s test to evaluate significance ( a = 0.05).

Results:  There were 124 patients who had a baseline CH; geometric mean was 132 mg/dL. Mean levels varied significantly (p ≤0.05) between South Africa (31) and the United States (93), and between experienced (88) and Naïve (36). A multivariate regression analysis, with country, experience, and RTV-B as predictors of baseline CH revealed that only the effects of country remained significant. Data from weeks 24 and 48 were available for 84 and 50 patients, respectively. There was a 10% increase in CH from baseline to week 24 (p <0.001) and a 7% increase to week 48 (p = 0.06). These changes did not vary significantly with country or experience. However, RTV was significantly associated with changes to weeks 24 (p = 0.01) and 48 (p = 0.03):  unboosted patients showing 5% and 2% increases from baseline to weeks 24 and 48, respectively, while the RTV patients’ CH increased by 20% and 32%. Between baseline and week 24, 8 of 81 patients went from a CH <180 to ≥180; at week 48, 8 of 48 went from CH <180 to ≥180. At weeks 24 and 48, 3 of 82 and 4 of 49 went from CH <200 to ≥200. Baseline TG values (n = 126) had a geometric mean of 112 mg/dL. Mean levels varied significantly (p ≤0.05) between subjects from South Africa (31) vs the United States (95), and between experienced (90) vs naïve (36) subjects, reflecting the effects of prior treatment and that experienced patients were overrepresented in the United States. Data from weeks 24 and 48 were available for 85 and 52 patients, respectively. Among these patients there were small, non-significant decreases in TG from baseline to weeks 24 and 48 (p = 0.26 and 0.08). These changes did not vary significantly with country, experience status, or RTV.

Conclusions:  Patients treated with ATV showed no significant change over time in TG, but showed a significant increase in CH, which was associated with RTV. Despite this, most patients CH remained <200 mg/dL. The fact that baseline CH levels were significantly higher among patients from the United States, controlling for treatment experience, is an area that needs further study.