Abstract / Summary
Background: Certain antiretroviral drugs, including efavirenz, have been associated with mitochondrial and oxidative effects in hepatic cells, but quantitative translation from nominal in vitro concentrations to an in vivo effect-site concentration remains uncertain. Methods: A hypothesis-generating physiologically based pharmacokinetic-pharmacodynamic (PBPK-PD) framework integrated simulated EFV concentration-time profiles for six pregnancy/CYP2B6 scenarios with published Hep3B MitoSOX concentration-response data fitted with the Hill–Emax model. Translational exposure was evaluated as total hepatic EFV concentration relative to the nominal in vitro concentration range and as unbound hepatic concentration approximated from total plasma concentration and fraction unbound. Results: Simulations predicted lower total plasma exposure during pregnancy in the reference normal-metaboliser scenario (Cp,total Cmax: decreasing from 9.94 to 7.15 µM). Under the free-drug scenario (fu,p = 0.22%), the estimated Cmax,u decreased from 0.0219 to 0.0157 µM, remaining far below the lowest non-zero nominal in vitro concentration (10 µM). Benchmarked against the fu,vitro-corrected in vitro EC50 (0.324 μM via Kalvass–Maurer protein-binding correction), this margin narrowed to approximately 15- to 21-fold at the reference fu,p and to approximately 2-fold at the highest literature-reported value of fu,p (1.5%). Conclusions: Pregnancy and CYP2B6 phenotype altered simulated EFV exposure, whereas quantitative translation to the Hep3B oxidative stress response depended strongly on the concentration metric used. Because unbound EFV was not reported in the in vitro system and intracellular hepatic EFV was unavailable, neither comparison establishes a definitive effect-site match. This framework supports an exploratory QIVIVE interpretation and identifies the exposure measurements required for quantitative validation.