P385. Population pharmacokinetic modelling of vedolizumab in patients with ulcerative colitis or Crohn's disease
M. Rosario1, N. Dirks2, M. Gastonguay2, I. Fox1, A. Milton1, 1Takeda Pharmaceuticals International Company, Clinical Pharmacology, Cambridge, United States, 2Metrum Research Group LLC, Biotechnology, Tariffville, United States
Vedolizumab (VDZ), currently in development for treating ulcerative colitis (UC) and Crohn's disease (CD), exhibits target-mediated drug disposition. Described here are a VDZ population pharmacokinetic (PK) model, an estimation of typical VDZ parameter values, and effects of covariates that may be predictors of VDZ PK variability.
The population PK data set included data from 5 VDZ clinical studies (1 phase 1 study [healthy subjects] and 1 phase 2 and 3 phase 3 studies [GEMINI 1, 2, and 3 - NCT00783718, NCT00783692, and NCT01224171] [UC or CD patients]). Analysis of repeated measures was conducted via nonlinear mixed-effects modelling using NONMEM (Version 7.2). The population PK base model was developed using first-order conditional estimation with an eta-epsilon interaction method and extensively sampled phase 1 and 2 data. Base model results were used to develop the full covariate model, which was fit to sparse phase 3 data using the full Bayesian Markov Chain Monte Carlo method. Model selection was based on standard goodness-of-fit criteria; a full covariate modelling approach, emphasizing parameter estimation rather than stepwise hypothesis testing, was implemented for the population PK analyses.
The population PK data set included 2554 individuals (18,427 evaluable VDZ serum concentrations). VDZ PK was best described by a 2-compartment model with parallel linear and nonlinear elimination. Reference covariates for the final PK model included baseline age, body weight, albumin level, disease activity scores (CD Activity Index score [CD]; partial Mayo Clinic score [UC]), prior anti-tumour necrosis factor alpha treatment, and diagnosis (CD or UC). Presence of anti-VDZ antibodies (AVAs) and adjuvant therapy use were evaluated. Overall, PK parameter estimates were similar for UC and CD patients using reference values for individual covariates. Linear clearance of VDZ (CLL) was 0.159 L/day for UC and 0.155 L/day for CD. Volume of distribution for the central compartment (Vc) was 3.19 L and 1.66 L for the peripheral compartment. Interindividual variabilities (coefficients of variation) in CLL and Vc were 35% and 19%, respectively; the proportional residual variance estimate was 24%. While baseline albumin had the potential to have a clinically relevant effect on VDZ CLL, defined as an effect size of more than ±25% from the reference value, effects of other covariates tested were not considered clinically meaningful.
Population PK parameters were similar in patients with UC and CD. No change in dosing recommendations is needed based on baseline covariates (eg, age, gender, weight, presence of AVA, and albumin).