Clinical Pharmacology Lead Pfizer Ltd., China (People's Republic)
Disclosure(s):
Jiaye Lin: No relevant disclosure to display
Objectives: The use of concentration-QTc (C-QTc) modeling is gradually being accepted as the primary approach to assess the risk of QTc prolongation. In this analysis, the effects of ponsegromab on heart rate (HR) and QT interval corrected using Fridericia’s formula (QTcF) were evaluated using combined data from 3 phase 1 studies in healthy participants.
Methods: A total of 89 participants who received subcutaneous injection of a single dose of 3, 10, 30, 100, 300, and 400 mg ponsegromab or placebo with intensive concentrations and matched triplicate electrocardiograms (ECGs) collected were included in the C-QTc analysis. A pre-specified linear mixed effects model [1] was applied using either unbound or total ponsegromab concentrations with the change from baseline in QTcF (ΔQTcF) as the dependent variable; assumptions for the pre-specified linear model were checked using model-independent exploratory graphics. Data from multiple studies were combined to cover a wide dose range, and the heterogeneity between studies was evaluated by examining the goodness-of-fit (GOF) plots by study and assessing the study effect as a covariate on intercept or slope in alternative linear mixed concentration-QT models.
Results: The results suggested no clinically relevant effects of ponsegromab on HR or QTcF over the 3 to 400 mg single dose range evaluated. Graphical exploration suggested QTcF adequately corrected heart effect on QT with no time delay between drug concentration and ΔQTcF concordance. In the concentration-ΔHR/ΔQTcF plots, apparently no consistent drug effect was observed and the linear model adequately characterized the relationship between ponsegromab concentrations and ΔHR/ΔQTcF. Based on the pre-specified model (also the final model), the slopes of ΔΔQTcF and ΔΔHR versus ponsegromab concentrations were not statistically significant from zero. In addition, no statistically significant changes in ΔΔQTcF and ΔΔHR were predicted at the unbound and total Cmax observed up to a 400 mg single dose of ponsegromab, the highest dose studied. The predicted ΔΔQTcF and ΔΔHR at the maximum geometric mean Cmax of unbound ponsegromab observed in the study data (~20900 ng/mL) were 0.88 msec (90% CI: -3.22, 4.98) and -1.7 bpm (90% CI: -4.62, 1.22), respectively. The upper bound of the two-sided 90% CI for predicted ΔΔQTcF at the maximum geometric mean Cmax observed in the study data, was below 10 ms.
The standardized residual errors from the pre-specified C-QT model with pooled data were similar across the 3 studies. The alternative models including study effect on slope or intercept were tested and compared to the pre-specified model. However, the alternative models either had higher Akaike Information Criteria (AIC)/Bayesian Information Criterion (BIC) values or failed to converge due to error. In addition, the estimated study effect on intercept or slope was not statistically significant. Based on these assessments, no heterogeneity between studies that could have impacted the <-QT analysis was observed.
Conclusions: The analysis with data pooled from 3 phase 1 studies indicated that ponsegromab (up to 400 mg) is not associated with clinically meaningful QTc prolongation.
Citations: [1] Garnett, C., Bonate, P.L., Dang, Q. et al. Scientific white paper on concentration-QTc modeling. J Pharmacokinet Pharmacodyn 45, 383–397 (2018). https://doi.org/10.1007/s10928-017-9558-5
