Background:  The phenotypic clinical cutoffs, above which virologic response is observed to decrease, for atazanavir (ATV) and atazanavir with ritonavir (ATV/r) have been defined using week 12 and 24 data from BMS AI424-043 and AI424-045 (045) studies as 2.2 and 5.2, respectively. Study 045 is a randomized, placebo controlled, double-blind multinational study of protease inhibitor (PI)-experienced patients treated with one of the following regimens: Tenofovoir (TDF) + 1 nucleoside reverse transcriptase inhibitor (NRTI) + ATV/r, TDF + 1 NRTI + lopinavir/r, or 1 NRTI + saquinavir + ATV/r. At baseline in 045, ATV/r was substituted for PI/NNRTI therapy. Background NRTI therapy remained unchanged to week 2. Thus, analyses of week 2 HIV RNA outcomes to define ATV/r clinical cutoffs eliminate the effect of changes in background treatment and diminish the potential effect of archived mutations on the ATV/r clinical cutoffs . We sought to validate the prior phenotypic clinical cutoffs determination using the week 2 virologic outcome data from the 045 study.

Methods:  Data from all 111 subjects with prior PI failure in 045 who had available baseline phenotype, baseline HIV RNA ³400 copies/mL, and available post-baseline HIV RNA data were included. ATV phenotyping was performed at baseline and correlated with virologic outcomes. The week 2 virologic outcomes evaluated were a change in HIV RNA from baseline of ³0.5 and ³1.0 log₁₀ copies/mL. Fisher’s exact test was used to define the ATV/r fold change (FC) cutoff.

Results:  The median baseline HIV RNA was 4.5 log₁₀ copies/mL and the median (inter- quartile range, IQR) baseline ATV FC was 1.2 (2.4). The median week 2 HIV RNA change was –1.3 log₁₀ copies/mL. At week 2, 88% and 65% of subjects experienced ³0.5 and ³1.0 log₁₀ copies/mL reduction, respectively. The greatest discrimination between responders and non-responders occurred at ATV FC of 5.2 by Fisher’s exact test for both ³0.5 (p <0.0001) and ³1.0 (p <0.0001) log₁₀ copies/mL reduction endpoints. This is in agreement with the prior week 24 analyses which also defined the ATV/r clinical cutoffs at 5.2 FC.

Conclusions:  Both the week 2 and 24 analyses defining the ATV/r clinical cutoffs were concordant. The best response to ATV/r was observed at baseline FC <5.2.