Background:  The in vivo activity of an ART drug may depend on its plasma concentration relative to the susceptibility of the virus.

Methods:  This non-randomized, open-label study evaluated the relationship of IQ (ratio of calculated protein binding-corrected plasma protease inhibitor [PI] concentration at 12 hours [C₁₂] to baseline IC₅₀) to week-2 and -24 virologic outcomes for 3 ritonavir (RTV)-enhanced PI regimens chosen by site investigators. At entry, subjects had HIV RNA >2500 copies/mL on a PI-based regimen for ≥12 weeks and phenotypic resistance to ≥2 PI. Following a phenotype (Monogram), their pre-entry PI was substituted with indinavir (IDV)/RTV 800/100 mg, fosamprenavir (FPV)/RTV 700/100 mg, or lopinavir (LPV)/RTV 400/200 mg twice daily for 2 weeks. On day 14, a 12-hour pharmacologic evaluation of PI was performed. Background therapy was then optimized using pre-entry phenotype findings and follow-up continued for 24 weeks. Associations between IQ and 2-week or 24-week change in viral load from baseline were examined by Spearman’s correlation test.

Results:  We enrolled 53 subjects:  12 on IDV, 33 on FPV, and 8 on LPV, of whom 49 completed 2 weeks and 46 completed 24 weeks of study. Average baseline CD4 counts were 104 (IDV), 174 (FPV), and 231 (LPV) cells/mm³ and average baseline HIV RNA were 5.0, 4.8, and 4.3 log₁₀ copies/mL. The median fold changes in IC₅₀ for IDV, FPV, and LPV were 33, 72.5, and 157 in the IDV arm; they were 42, 12.0, and 79 in the FPV arm; and 16, 37, and 20.5 in the LPV arm. Median 2-week changes in viral load were –0.7, –0.1 and –1.0 log₁₀ for the IDV, FPV, and LPV arms; median C₁₂ were 806, 2060, and 6880 ng/mL and protein binding corrected IQ were 3.6, 2.0, and 2.4, respectively. Spearman r values for 2-week correlation were 0.02, –0.45, and –0.38 for IDV, FPV, and LPV arms; only the FPV arm value was significant (p = 0.01). Correlations between baseline IC₅₀ fold change and 2-week viral load change were of similar magnitude; correlation for FPV was 0.57 (p = 0.001). Median 24-week changes in viral load in the IDV, FPV, and LPV arms were similar (–0.4, –0.3, and –0.3 log₁₀, respectively). Correlations between IQ and 24-week changes in viral load (–0.03, –0.31, and –0.52) were not significant.

Conclusions:  For RTV-enhanced FPV, 2-week viral load responses correlated with calculated protein binding-corrected IQ, though correlation with baseline IC₅₀ fold changes was similar, suggesting that day 14 C₁₂ added little incremental information. Evaluation of IDV and LPV were limited by sample size. Pending analyses of measured free drug concentrations may further describe the IQ/viral load relationship.