Pharmacokinetic/pharmacodynamic analysis of romidepsin used as an HIV latency reversing agent.

OBJECTIVES To develop a population pharmacokinetic model for romidepsin given as an HIV latency reversing agent (LRA) and to explore the relationship between romidepsin exposure and its in vivo effects on viral gene expression and antiviral immunity. METHODS A population pharmacokinetic analysis was performed in 15 HIV-1-infected patients who received three weekly infusions of romidepsin (5 mg/m2) within the BCN02 clinical trial. A full pharmacokinetic profile was obtained for each participant at the first dose, and additional samples thereafter. A population pharmacokinetic model was developed. Bayesian estimates of the individual pharmacokinetic parameters of romidepsin were used to simulate individual time-concentration curves on each occasion. The relationship between romidepsin AUC0-∞ and its in vivo effects was assessed. RESULTS Romidepsin pharmacokinetics were best described by a three-compartment model with linear kinetics. Body weight influenced romidepsin disposition. A significant relationship was observed between romidepsin AUC0-∞ and increases in expression of exhaustion markers by CD4+ and CD8+ T cells and apoptosis markers in CD4+, but not with histone acetylation levels or HIV-1 cell-associated RNA in CD4+ T cells. For each increase of 100 ng·h/mL in romidepsin AUC0-∞, CD4+ counts decreased by a mean (95% CI) of 74 (42-94) cells/mm3 after dosing. CONCLUSIONS A population model describing the pharmacokinetics of romidepsin as an HIV LRA was developed. Higher exposure to romidepsin resulted in higher expression of apoptosis markers and declines in CD4+ count but did not increase viral reactivation levels. These observations have important implications for the optimization of effective kick-and-kill strategies for an HIV-1 cure.