Background: There is evidence to suggest a pharmacokinetic-pharmacodynamic relationship in HIV-infected patients receiving protease inhibitor (PI)-containing highly active antiretroviral therapy (HAART); however, the effective trough PI plasma concentrations achieved have not been precisely determined.
Methods: The relationship between HIV viral load and concomitant PI trough plasma concentration (C(trough)) was evaluated in 101 patients receiving at least 4 months of thrice daily indinavir (IDV)-containing (n=68) or nelfinavir (NFV)-containing (n=33) HAART. The more discriminating C(trough) efficacy thresholds were determined statistically for each PI by using the raw C(trough) and the time-corrected C(trough), using the precise delay since the last PI intake and the half-life of each PI.
Results: For IDV (P=0.002) and NFV (P=0.019) median C(trough) levels were higher in patients with undetectable viral load [0.23 mg/L (n=30) and 2.3 mg/L (n=16) respectively] than in patients with detectable viral load [0.11 mg/L (n=38) and 0.6 mg/L (n=17) respectively]. C(trough) levels of IDV (r=-0.45; P<0.0001) and NFV (r=-0.43; P=0.011) were correlated with the concomitant viral load. The more discriminating C(trough) efficacy thresholds were estimated statistically as 0.12 mg/L for IDV and 0.5 mg/L for NFV. When C(trough) values were time-corrected, the C(trough) efficacy thresholds, 8 h after the last intake, were 0.15 mg/L for IDV and 0.65 mg/L for NFV.
Conclusions: These results support the importance of achieving minimal effective C(trough) to improve the virological efficacy of PI-containing HAART, and specify the target concentrations for IDV and NFV.