Population Pharmacokinetic Modeling of Investigational Agent MLN4924 in Cancer Patients
H. Steve Kuan (1), R. Donald Harvey (2)
(1) Department of Clinical Pharmacology, Millennium Pharmaceuticals, Inc.; (2) Winship Cancer Institute of Emory University
Objectives: MLN4924 is a novel investigational inhibitor of NEDD8-activating enzyme [1] that is being evaluated in patients with hematologic and non-hematologic cancer. A population model was developed to characterize the pharmacokinetics of MLN4924 administered as a 1-hour intravenous infusion following multiple dosing schedules and to assess the effects of patient covariates and dose.
Methods: The population pharmacokinetic model was developed on plasma concentration-time data using NONMEM (version 7.1.2). Two- and three-compartment disposition and zero-order input models were tested with or without logarithmic transformation of concentrations. Following identification of the base model, the effects of demographics (age, sex, body size), clinical laboratory parameters (hepatic transaminases and creatinine clearance) and assigned dose (mg/m2) were assessed in stepwise procedures using likelihood ratio testing of nested models. Model goodness-of-fit and performance were evaluated by common diagnostic tools.
Results: One hundred and four patients enrolled in four ongoing Phase 1 dose-escalation clinical studies provided 1262 concentrations with a median of 11.5 observations per patient (range 2 to 19). A two-compartment model with log-transformed concentrations most reasonably described the data. The final model estimated systemic clearance (CL), central volume of distribution (V1), inter-compartmental clearance (Q) and peripheral volume of distribution (V2) to be 59.1 L/hr, 136 L, 35.2 L/hr and 187 L, respectively, after centering on the median. The inter-individual variability (IIV as CV) was estimated to be 26.5 % for CL, 43.3 % for V1, 53.7 % for Q and 28 % for V2. Age (29.3-90.8 years) and body surface area (BSA; 1.48-2.72 m2) were significant covariates on CL reducing the parameter's IIV by 7.3 %. Assigned dose (25-278 mg/m2) and BSA were significant covariates on V1 reducing the parameter's IIV by 17.7 %. Sex (37 women, 67 men) was a significant covariate on V2 reducing the parameter's IIV by 7.9 %.
Conclusions: The effect of BSA was significant on both CL and V1, lending support to MLN4924 administration on the basis of body size. In addition, CL was decreased with increasing age; V1 was decreased with increasing assigned dose; and females had a smaller V2 compared with males. Clinical relevance of these statistically significant covariates will be further evaluated in the context of safety and efficacy.
References:
[1] Milhollen MA, et al. MLN4924, a NEDD8-activating enzyme inhibitor, is active in diffuse large B-cell lymphoma models: rationale for treatment of NF-kappaB-dependent lymphoma. Blood, 2010. 16(9):1515-1523.
