Modeling the HIV Protease Inhibitor Adherence–Resistance Curve by Use of Empirically Derived Estimates

This work statistically modeled the relationship between adherence and development of drug resistance, using empirically defined relationships of the rate of viral suppression and drug-resistance-mutation accumulation derived from patients receiving protease-inhibitor-based therapy.

Abstract

The standard view postulates a bell-shaped relationship between adherence to therapy and development of drug-resistant human immunodeficiency virus (HIV), with a resistance peak at a moderate level of adherence. This relationship has not been confirmed empirically. We statistically modeled the relationship between adherence and development of drug resistance, using empirically defined relationships of the rate of viral suppression and drug-resistance-mutation accumulation derived from patients receiving protease-inhibitor-based therapy. We found that the maximal rate of drug resistance occurs at 87% adherence and declines modestly at 100% adherence. Higher levels of viral suppression at 100% adherence (a marker of greater regimen potency) progressively reduce the overall population rate of drug resistance and shift the peak resistance rate to lower levels of adherence.