Abstract
This study assessed the potential risk of QT prolongation associated with the dual enhancer of zeste homolog 1/2 inhibitor valemetostat. An evaluation of the relationship between plasma valemetostat concentration and heart-rate-corrected QT (QTc) interval was performed. Time-matched plasma concentration and 12-lead electrocardiogram data were collected from the phase I studies DS3201-A-J101, in patients with relapsed/refractory B-/T-cell non-Hodgkin lymphomas (NCT02732275), and DS3201-A-U102, in patients with relapsed/refractory acute myeloid leukemia and acute lymphoblastic leukemia (NCT03110354). A prespecified linear mixed-effects model was used to assess the effect of valemetostat on change in QTc corrected by the Fridericia method (ΔQTcF). A population-specific method (ΔQTcP) was also used to remove the heart rate interval (RR) dependence. The final dataset contained 769 electrocardiogram measurements from 100 patients. Linear mixed-effects modeling found no significant demographic or clinical covariate effects. The slope versus concentration was significant (95% confidence interval [CI] of the coefficient excluded 0) in the final models for ΔQTcF, but not ΔQTcP, while the relative standard error of the slope was > 50% for both models. Baseline QTc had a negative effect on ΔQTc in all models. At the steady-state geometric mean maximum concentrations in the dose range of 100-700 mg tested in the DS3201-A-J101 and DS3201-A-U102 studies, the 90% CI upper bounds for model-predicted ΔQTcF and ΔQTcP were 1.52-8.38 ms, all of which were below the clinically significant threshold of 10 ms. The analysis supports a lack of a clinically meaningful effect on the QTc interval for valemetostat.