Enabling Robust Concentration-QTc Assessment for Early Oncology Clinical Trials

Since the release of the ICH E14 Questions and Answers guidance document, concentration-QTc (C-QTc) modeling has become the preferred approach in early Phase 1 trials to assess the pro-arrhythmia and delayed cardiac repolarization risks of investigational drugs. While a white paper provided guidance for definitive QTc assessments, its focus on non-oncology drugs leaves gaps for early oncology trial settings. In Phase 1 oncology trials, model-assisted designs like the modified Toxicity Probability Interval (mTPI) or the Bayesian Optimal Interval (BOIN) design are widely used for dose-escalation. This study explores strategies for robust C-QTc analyses within these designs, focusing on sample size, electrocardiogram (ECG) collection, and model specification. Simulation studies evaluated the operating characteristics of frequently used C-QTc models under various scenarios in dose escalation setting. Results indicate that with sufficient ECG data collection, dose escalation trials can provide robust QTc assessment. A C-QTc model incorporating a time effect and double compound symmetry (CS) correlation structure demonstrated the most robust performance, improving power and reliability in slope variability estimation. These findings offer recommendations to enhance cardiac safety assessment in Phase 1 oncology studies.