Investigating the contribution of residual unexplained variability components on bias and imprecision of parameter estimates in population pharmacokinetic mixed-effects modeling

Abstract

In a nonlinear mixed-effects modeling (NLMEM) approach of pharmacokinetic (PK) and pharmacodynamic (PD) data, two levels of random effects are generally modeled: between-subject variability (BSV) and residual unexplained variability (RUV). The goal of this simulation-estimation study was to investigate the extent to which PK and RUV model misspecification, errors in recording dosing and sampling times, and variability in drug content uniformity contribute to the estimated magnitude of RUV and PK parameter bias. A two-compartment model with first-order absorption and linear elimination was simulated as a true model. PK parameters were clearance 5.0 L/h; central volume of distribution 35 L; inter-compartmental clearance 50 L/h; peripheral volume of distribution 50 L. All parameters were assumed to have a 30% coefficient of variation (CV). One hundred in-silico subjects were administered a labeled dose of 120 mg under 4 sample collection designs. PK and RUV model misspecifications were associated with relatively larger increases in the magnitude of RUV compared to other sources for all levels of sampling design. The contribution of dose and dosing time misspecifications have negligible effects on RUV but result in higher bias in PK parameter estimates. Inaccurate sampling time data results in biased RUV and increases with the magnitude of perturbations. Combined perturbation scenarios in the studied sources will propagate the variability and accumulate in RUV magnitude and bias in PK parameter estimates. This work provides insight into the potential contributions of many factors that comprise RUV and bias in PK parameters.