465   Plasma Trough Levels Correlate with Distinct Genetic Mechanisms During the Development of Amprenavir Resistance.

R. Elston*¹, S. Randall¹, R. Myers¹, M. Maguire¹, A. Rakik¹, M. Ait-Khaled¹, D. Stein², and W. Snowden¹.
Glaxo Wellcome,¹Stevenage, UK and²Res. Triangle Park, NC.

Background:The development of amprenavir (APV) resistance in the clinic is associated with four distinct genetic mechanisms, involving either the protease substitution I50V or I54L/M or, less commonly, V32I + I47V or I84V. Limited information is available about factors that influence which of these genetic mechanisms is adopted during resistance development. The purpose of this analysis was to investigate whether specific genetic mechanisms correlate with variations in observed plasma levels of APV.

Methods:Genotypic analysis was performed on virus in clinical trial samples from previously PI-naïve subjects who experienced viral rebound or inadequate virologic suppression whilst receiving APV/NRTI combination therapy.

Results:The I50V pathway was predominant in virus from subjects who received APV 1050 mg BID or 1200 mg BID (I50V n = 15, I54L/M n = 11, V32I + I47V n = 9, I84V n = 6). However, in virus from subjects who received a lower APV dosage (900 mg BID), the I54L/M pathway predominated (I50V n = 0, I54L/M n = 6, V32I + I47V n = 2, I84V n = 0). This observation that different APV dosages were associated with the predominance of specific resistance mechanisms was confirmed by analysis of measured APV steady-state plasma C_minconcentrations, where available (n = 9). This analysis revealed a significant difference (p < 0.05) in exposure according to viral resistance genotype: The development of mutation I50V, which is known to result in reduced replicative fitness, occurred in virus from subjects with higher C_minconcentrations (n = 4, range 451-877 ng/ml), whereas mutation I54L/M developed where C_minlevels were lower (n = 5, range 43-183 ng/ml). Consistent observations were made in a study in PI-experienced adults, where APV-efavirenz interactions resulted in sub-optimal APV exposure. APV resistance developed predominantly through the I54L/M pathway (n = 7), whereas I50V developed in virus from only two subjects, one of whom had APV levels boosted by ritonavir.

Conclusions:This is the first evidence that APV plasma level may influence the preference for a specific genetic mechanism by which resistance develops and has potential implications in terms of mechanisms of resistance which may be observed with boosted PI regimens.