Hibma, Jennifer; Boni, Joseph; Ruiz, Ana
Pfizer
Objectives: Inotuzumab ozogamicin (InO; PF-05208773) is a humanized anti-Cluster of Differentiation 22 (CD22) monoclonal antibody conjugated to calicheamicin, a potent cytotoxic antibiotic. InO is being studied for the treatment of adult patients with relapsed or refractory B-cell acute lymphoblastic leukemia (ALL). The objective of this analysis was to characterize the exposure-response relationship between heart rate corrected QT interval (QTc) and InO concentrations.
Methods: Data from clinical studies with InO monotherapy included 2743 observations from 250 patients with relapsed or refractory B-cell ALL and relapsed or refractory NHL. Serial 12-lead triplicate-ECGs were collected at screening, just prior to first dose (baseline), at predefined time points post dose and collected within 24 hours of scheduled PK collections. First the effect of InO on heart rate interval (RR interval) was evaluated prior to correcting QT. Then, the most appropriate correction factor(s) were selected (ie, Bazett’s (QTcB), Fridericia’s (QTcF), or a Study Specific (QTcS)) to obtain QTc values independent of the underlying heart rate. Next, a linear mixed effects model was used to describe the exposure response relationship between InO concentration and QT interval using the appropriate correction factor(s). Analyses were conducted using a population analysis approach and linear mixed effects models (NONMEM, Version 7.3.0, ICON Development Solutions, Dublin, Ireland). Full model development involved testing for covariates (both continuous and categorical) with the goal of describing the inter-individual variability (IIV) and improving predictive performance. Parameterization of covariate models was guided by examination of the plots of the IIV on slope and intercept versus covariates from the final base model.
Results: InO did not affect RR intervals. Using both the QTcF and QTcS correction factors, the QTc intervals had a positive correlation with InO concentration. The typical population slope estimate was 0.00649 (95% CI: 0.003-0.013) for the relationship between QTcF and InO serum concentration, and was 0.00695 (95% CI: 0.002-0.015) for the relationship between QTcS and InO serum concentration.
Conclusions: Based on 1000 simulations, the expected changes from baseline (QTcF and QTcS, median and upper 95% CI), at both therapeutic and supratherapeutic concentrations, were below 10 msec (QTcF 2.53 msec (4.92 msec) and 2.70 msec (5.40 msec), respectively; QTcS 3.87 msec (7.54 msec) and 4.14 msec (8.28 msec), respectively.
Reference: PAGE 25 (2016) Abstr 6026 [www.page-meeting.org/?abstract=6026]
Poster: Drug/Disease modeling - Oncology