Abstract P5-18-24: Population pharmacokinetics of trastuzumab emtansine, a HER2-targeted antibody-drug conjugate, in patients with HER2-positive metastatic breast cancer: clinical implications of the effect of various covariates

Abstract

Abstract Background: Trastuzumab emtansine (T-DM1) is a HER2-targeted antibody-drug conjugate composed of the humanized monoclonal antibody trastuzumab, the potent cytotoxic agent DM1 (a microtubule inhibitor), and a stable thioether linker. To estimate typical pharmacokinetic (PK) parameter values and interpatient variability, a population PK model for T-DM1 was previously developed from 1 phase 1 (0.3 to 4.8 mg/kg in qw or q3w regimens) and 2 phase 2 (3.6 mg/kg q3w) trials (Gupta, J Clin Pharmacol 2012). The model reported here has been updated with additional data from 2 randomized trials (phase 2 TDM4450g and phase 3 EMILIA, 3.6 mg/kg q3w). Another phase 2 trial (TDM4688g) was used for external validation of the model. The effect of demographic and pathophysiological covariates on the PK of T-DM1 was explored to better understand the clinical factors that might affect exposure and clinical outcome for individual patients. Methods: For the current analysis, 9934 T-DM1 serum concentration-time data points from 671 patients were simultaneously fitted using NONMEM® software. T-DM1 concentration-time data to date are best described using a 2-compartment linear model. All relevant and plausible covariates likely to have an effect on T-DM1 systemic exposure, or likely to have clinical relevance, were explored for possible correlation with the key T-DM1 PK parameters of clearance (CL) and central volume of distribution (Vc). These covariates include those related to demographics, renal and hepatic function, disease status, and treatment history. Results: The estimated CL for T-DM1 is 0.68 L/day, Vc is 3.13 L, and the terminal half-life is 3.94 days. Interindividual variability (IIV) of the base model is 25.6% and 17.5% for CL and Vc, respectively. Patients with greater body weight, sum of longest dimension of target lesions, serum concentration of shed HER2 extracellular domain, and aspartate aminotransferase concentrations, as well as patients with lower serum albumin and baseline trastuzumab concentrations, have statistically faster CL. Patients with greater body weight also have statistically larger Vc. Incorporation of these covariates (P<0.001 by likelihood ratio test) decreased IIV of CL and Vc to 19.1% and 11.7%, respectively. All covariates together explain 44.4% and 55.8% of IIV in CL and Vc, respectively. The model sensitivity analysis suggests that a patient with a statistically significant PK covariate value at the 5th or 95th percentile of the population will have a <20% difference in cumulative exposure (as represented by area under the T-DM1 concentration-time curve) compared with a typical patient with a median covariate value. Conclusions: A relatively small IIV for the estimated T-DM1 PK parameters of CL and Vc was observed. None of the evaluated covariates had a clinically meaningful magnitude of effect on T-DM1 exposure (<20% difference for patients with 5th and 95th percentiles vs patients with median value of covariates) that would justify a further dose adjustment. The body weight–based dose of 3.6 mg/kg q3w without further correction for other factors is considered appropriate in ongoing clinical trials. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P5-18-24.