H.S. Kuan (1), N. Gupta (1), K. Venkatakrishnan (1), A. Milton (1), S. Girgis (2), I. Poggesi (2), D. Mould (3)
(1) Department of Clinical Pharmacology, Millennium Pharmaceuticals, Inc.; (2) Department of Clinical Pharmacology, Research & Development a Division of Janssen Pharmaceutica NV.; (3) Projections Research, Inc.
Objectives: Bortezomib is an ubiquitin-proteasome inhibitor indicated for the treatment of multiple myeloma and for patients with mantle cell lymphoma who have received at least one prior therapy. The population modelling approach was used to further understand the pharmacokinetic (PK) differences observed after first and repeated dosing.
Methods: Population models were fitted to bortezomib plasma concentration-time data using NONMEM 7.2. Various models were examined including linear disposition, empirical time dependent systemic clearance (CL), and combined linear and nonlinear distribution for a peripheral compartment. A log transform both sides approach was used with concentrations below the lower limit of quantification handled as censored observations[1]. Model goodness-of-fit and performance were evaluated by common diagnostic tools. For the full Markov Chain Monte Carlo Bayesian analysis (BAYES) method, model qualification was additionally performed using burn-in termination and post-run convergence testing.
Results: 142 patients receiving first (1-2mg/m2) and repeat (1-1.3mg/m2) dosing on days 1, 4, 8 and 11 of 21-day cycles in 5 Phase 1 studies provided 3004 concentration. A 3-compartment model in which the distribution from compartment 2 (peripheral) to compartment 1 (central) was parameterized by a Michaelis-Menten function was considered most adequate based on the BAYES method. Parameter estimates (inter-individual variability as CV) by this model were 10.8L/h (43%) for CL, 14.1L (81%) for central volume of distribution (V1), 349L (304%) and 895L (114%) for peripheral volumes of distribution (V2 and V3, respectively), 46.3L/h (93%) for clearance to compartment 2 (Q2), 0.06ng/mL (93%) for Michaelis constant (Km) with the maximum velocity (Vm) fixed at 0.0926ng/mL/h, 51.7L/h (70%) for clearance to and from compartment 3 (Q3). The residual variability was estimated to be 42% or 62% depending on the assay used to determine plasma concentrations.
Conclusions: The developed model attributes previously reported time dependence in bortezomib PK [2] to saturable distribution to rather than nonlinear elimination from the central compartment. All parameters except V2 are estimated with good precision. Moreover, V2 is found to exhibit the largest variability among patients. Visual predictive checks demonstrate that this model describes bortezomib disposition well and may be used to predict bortezomib plasma exposure following alternative dosing schedules.
References:
[1] Beal SL. Ways to fit a PK model with some data below the quantification limit. J Pharmacokinet Pharmacodyn. 2001;28(5):481-504.
[2] Reece DE, et al. Pharmacokinetic and pharmacodynamic study of two doses of bortezomib in patients with relapsed multiple myeloma. Cancer Chemother Pharmacol 2011;67(1):57-67.
Reference: PAGE 21 () Abstr 2308 [www.page-meeting.org/?abstract=2308]
Poster: Oncology