Model Based Neutrophil Guided Dose Adaptation in Chemotherapy
Johan E Wallin , Lena E Friberg , Mats O Karlsson
Deptartment of Pharmaceutical Biosciences, Uppsala University, Sweden;
Objectives: Severe neutropenia is one of the most important dose limiting events in many anticancer regimens. One of the most employed approaches to this problem has been to reduce the consecutive dose in fixed steps, commonly by 25%. Another approach has been to use pharmacokinetic (PK) sampling to tailor dosing, but only rarely have model-guided computer-based approaches utilizing PK and/or pharmacodynamic (PD) data been used. A previously described semi-mechanistic model for myelosuppression has been used to characterize a wide range of anticancer drugs(1), and both interindividual and interoccasion variability (IIV/IOV) has been described for a number of agents(2). This knowledge could be used in a clinical setting to make model-based dose individualization, which compared to current stepwise procedures, may tailor doses in a more precise manner, and allow increased overall dose intensity in the population without increasing the risk for severe toxicity. In this study we investigated by simulations the outcome of model-based dose adaptation, and the influence of type and amount of data provided to the model. We also investigated the influence of IIV and IOV magnitudes for adaptation outcome.
Methods: PK and PD data were simulated for one thousand patients in five treatment courses. Different portions of data were used to obtain empirical Bayes estimates that were subsequently used to adjust the dose to a level predicted to result in a target neutrophil nadir. Performance was evaluated with different levels of IIV and IOV.
Results: In the presence of PD measurements, PK data provided little additional information. By a limited PD sampling the number of patients on target could be increased with the model-based approach compared to standard dose-adjustment methods. Thereby the model-based dose-adjustment method could facilitate increased overall dose intensity in the population, without a corresponding increase in patients experiencing severe neutropenia. The number of patients achieving target range neutropenia was increased by 27% compared to the standard method. Successful dose adaptation seemed to be more sensitive to IOV magnitude in the drug efficacy parameter than in other PK or PD parameters, whereas IIV magnitude was of little importance.
Conclusions: A model-based dose adaptation procedure with a limited neutrophil measurement schedule may increase the chance of success in treatment as it allows for increased dose intensity. When neutrophil counts are available PK data provide little additional information on the expected myelosuppression time-course.
References:
[1] Friberg et al. "Model of chemotherapy-induced myelosuppression with parameter consistency across drugs" J Clin Oncol 2002 20(24): 4713-21
[2] Hansson et al. Comparison of Inter-Occasion and Inter-Individual Variability in Chemotherapy- Induced Myelosuppression PAGE 2008:1328
