O. Melnichenko (1), M. Savelieva Praz (2), A. M. Stein (3)
(1) Novartis Pharma, Moscow, Russian Federation; (2) Novartis AG, Basel, Switzerland; (3) Novartis Institute for Biomedical Research, Cambridge, USA
Objectives: Dose escalation trials in oncology are aimed to identify the Maximum Tolerated Dose and by design, adverse events often lead to dose interruptions and modifications. These dose modifications may impact significantly the dynamics of tumor shrinkage/growth. The standard approach for assessing the relationship between dose and efficacy characterizes the relationship between the dose at randomization and the best overall tumor response. This approach does not take into consideration the dynamics of dose changes and tumor shrinkage and may be sub-optimal for assessing the dose-efficacy relationship. The aim of this work is to apply a tumor growth modeling approach, which takes into account the full dosing history and longitudinal dose-response analysis, in order to establish a dose-efficacy relationship for a new drug.
Methods: The population dose-response analysis was performed with Monolix using tumor size measurements and dosing history from 84 Phase I clinical trial patients. Only patients with baseline and at least one post-baseline assessment of tumor size were included in the analysis. Several structural and dose-effect models were tested, the influence of demographic and clinical characteristics on tumor growth rate and efficacy parameters were examined.
Results: The model with a dose-proportional effect describes the data well. Of the covariates evaluated, the ECOG performance status had a statistically significant effect on the rate of tumor shrinkage. Model-based prediction showed that at the maximum dose level, 91% of patients will have tumor shrinkage, with 40% of them having shrinkage of more than 50%. If dose is halved, 77% of patients will have tumor shrinkage, with only 15 % of them having shrinkage of more than 50%.
Conclusions: The modelling results confirmed that the maximum dose tested was also the most efficacious one for shrinking tumors and was recommended to be taken forward in future trials.
Reference: PAGE 22 (2013) Abstr 2671 [www.page-meeting.org/?abstract=2671]
Poster: Oncology