Feng Jin, Brian Kirby, Yuying Gao, Brian Kearney, and Anita Mathias
Gilead Sciences, Foster City, CA 94404
Objectives: Sofosbuvir (SOF), a potent NS5B polymerase inhibitor with broad HCV genotype coverage, is approved for the treatment of genotype 1, 2, 3, and 4 chronic HCV infection in treatment-naïve and -experienced patients and HIV/HCV co-infected patients. As a prodrug, SOF undergoes extensive sequential metabolism to form its major metabolites GS-566500 and GS-331007. The objective of this analysis was to develop a mechanistic integrated population pharmacokinetic model for SOF, GS-566500 and GS-331007, and to evaluate the impact of covariates on their PK.
Methods: Data from 8 clinical studies conducted in HCV-infected patients (n=1288) were included in this analysis. Data were analyzed using NONMEM 7.3 with first order (FO) method. Stepwise forward addition followed by backward elimination was implemented in the covariate (age, gender, race, body weight, ethnicity, creatinine clearance (CLCR) or estimated glomerular filtration rate (EGFR), cirrhosis status, IL28B status, Ribavirin usage, food, and concomitant medications) model building process to develop final PopPK model. Various model assessment methods were used to evaluate the model performance.
Results: Plasma PK of SOF and its metabolites was well described by the integrated model. The PK model was parameterized in clearances (CL), volumes (V), absorption rate constants (ka), lag times, and relative fraction absorbed of each analyte. Between-subject random effects were tested on clearances, volumes, and absorption rate constants. The following statistically significant parameter-covariate relationships were identified: gender on SOF and GS-331007 CL; food on SOF ka, GS-331007 V and ka; and CLCR on GS-566500 and GS-331007 CL. The population median estimated clearances were 78.26 L/hr, 81.45 L/hr, and 27.66 L/hr for SOF, GS-566500, and GS-331007, respectively, with the relative molar % of dose absorbed at 13.2%, 23.6%, and 63.2% respectively. The covariates tested did not have a clinically meaningful impact on SOF and its metabolites exposures. PK/PD results utilizing exposures from this integrated model were consistent with those from individual models previously established for SOF and GS-331007.
Conclusions: SOF, GS-566500, and GS-331007 PK were adequately described by an mechanistic integrated model. Relative to the individual model, this integrated model provided further insight on the conversation of SOF to its metabolites.
Reference: PAGE 24 (2015) Abstr 3323 [www.page-meeting.org/?abstract=3323]
Poster: Drug/Disease modeling - Infection