Abstract 3767: Use of pharmacokinetic-pharmacodynamic modeling to characterize platelet response following inotuzumab ozogamicin treatment in patients with follicular or diffuse large B-cell non-Hodgkin's lymphoma

Abstract

Abstract Background: Inotuzumab ozogamicin (CMC-544) is an immunoconjugate chemotherapy agent composed of a CD22-directed IgG4 antibody linked to calicheamicin, a potent cytotoxic anti-tumor antibiotic. Thrombocytopenia was the most frequent adverse event observed in phase 1/2 studies. This integrated population pharmacokinetic-pharmacodynamic (PK/PD) analysis was undertaken to a) elucidate the relationship between exposure of inotuzumab or total calicheamicin (tCali) and the decrease in platelet count; b) identify the covariates that influence the PK of inotuzumab and platelet response; and c) optimize the dosing schedule. Methods: Serum concentrations of inotuzumab, tCali (sum of conjugated and unconjugated forms), and platelet count were available from 5 phase 1/2 studies with follicular or diffuse large B-cell non-Hodgkin's lymphoma (NHL) who received inotuzumab alone or in combination with rituximab. Most patients received 1.8 mg/m2 of inotuzumab infused over 1 hour every 3 to 4 weeks (dose range, 0.4-2.4 mg/m2). The sequential PK/PD analysis was performed by nonlinear regression using NONMEM, version 7, level 1.2 (Icon Development Solutions, Hanover, MD). Covariates (factors for demography, hematology measures, renal and hepatic function, and baseline tumor size) were evaluated using graphical analysis, a general additive model (GAM), and a stepwise covariate model by forward addition and backward elimination process (SCM). Results and Conclusion: A 2-compartment model with linear elimination adequately described the PK of inotuzumab and tCali. The time course of platelet response was modeled using a semi-mechanistic transit compartment model model with drug effect as a sigmoidal Emax function. Significant covariate effects identified for PK estimates of inotuzumab included baseline body surface area (BSA) on clearance (CL) and central volume (V1), and baseline creatinine clearance on inter-compartment clearance (Q). For tCali, significant covariate effects included baseline BSA on CL and V1, treatment occasion (cycle) on CL and V1, and gender on peripheral volume (V2). The CL of tCali decreased from 0.73 L/h in Cycle 1 to 0.3 L/h in Cycles 2 and 3, suggesting a nonlinear disposition process. For the platelet model, the EC50 estimates for effect of inotuzumab and tCali on production of platelets were 219 and 50.2 ng/mL, respectively. Baseline platelet count was identified as a significant covariate. Integrated modeling of drug concentration and platelet count provides a means to quantify the magnitude of platelet suppression and to identify dosing schedules that mitigate the incidence of thrombocytopenia. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3767. doi:1538-7445.AM2012-3767