Abstract B193: Analysis of the concentration-QTc relationship for eribulin mesylate.

Abstract

Abstract Introduction: Characterization of cardiac safety liability has become a critical part of clinical development plans for investigational drugs. For oncologic compounds, the standard “thorough QTc” study, which is typically conducted in healthy volunteers and includes placebo and a supratherapeutic dose, may not be possible due to drug toxicity. In the development of eribulin, a Phase 1 study was conducted to assess the potential effect of eribulin mesylate for QTc interval prolongation in female and male patients with recurrent and/or metastatic cancer. Results for two objectives are reported herein: to identify the more appropriate method of adjusting the QT interval for differences in heart rate (HR) and to develop an exposure-response model to characterize the relationship, if any, of time-matched change from baseline in QTc with exposure to eribulin. Methods: Subjects were administered 1.4 mg/m2 of eribulin mesylate on Days 1 and 8 of a 21-day Study Phase. The dataset for the PK/PD analysis consisted of measured drug concentrations and QTcF values from triplicate 12-lead ECGs extracted from the continuous Holter. The timepoints in the study used in the PK/PD analysis were pre-dose, after eribulin infusion, and 15 minutes, 30 minutes, 1, 1.5, 2, 3, 4, 5, 6, 10, 24, and 48 hours after the start of eribulin administration on Cycle 1 Days 1 and 8, with time-matched points for baseline from Day 0. In the correction of QT interval for HR, the Fridericia method (QTcF=QT/(RR)0.33) was compared to an individually-derived method, QTcNi (QTcNi = QT + bi*(1000-RR), where RR is 60/(HR) and bi subject-specific slope). The performance of each correction algorithm was assessed via linear regression of individual subjects' QTc versus RR intervals at on-treatment visits. The HR correction method with the mean of squared slopes closer to zero was preferred. The concentration-QTc relationship was assessed using a linear mixed effects analysis of the relationship between baseline-adjusted QTcF interval (QTcF) and observed eribulin mesylate concentration, according to the equation: QTcF = α + β* [eribulin], where interindividual variability was estimated for both intercept () and slope (). The effect of treatment period (i.e., Day 1 or Day 8) on slope and intercept was also evaluated. Consideration was also given to nonlinear relationships. Results: Both QTcF and QTcNi corrected for HR reasonably well, with QTcNi providing a correction that was less dependent on HR. The means of squared slopes for QTcF were 0.0097 and 0.0083 on Day 1 and Day 8, respectively, while corresponding means for QTcNi were 0.0053 and 0.0069. Development of mixed-effects models for both QTcF and QTcNi indicated no relationship with eribulin concentrations. The observed difference in typical QTc between Day 1 and Day 8, estimated as 5.7 msec for QTcNi and 7.6 msec for QTcF, was modeled with a fixed effect on the intercept. Within-subject residual deviation was modeled additively, with zero mean and estimated standard deviation of 13 msec. Conclusion: QTcNi was more effective than QTcF in removing the HR dependence of corrected QT values. No relationship was found between eribulin concentration and QTc. No QTc prolongation was observed on Day 1, whereas QTc values on Day 8 were small (9 msec) and independent of eribulin plasma concentration. A QTc interval increase of this magnitude is not expected to be of clinical concern in this subject population. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr B193.