A population model for the leukopenic effect of etoposide

Area of Science:

• Pharmacometrics
• Mathematical Modeling
• Oncology

Background:

• Hematologic toxicity is a significant concern in anticancer therapy.
• Accurate quantification of toxicity is crucial for optimizing treatment strategies.
• Existing models may not fully capture inter-individual and intra-individual variability in drug response.

Purpose of the Study:

• To develop and validate a novel model-dependent approach for quantifying hematologic toxicity in cancer patients.
• To identify key covariates influencing the population's response to anticancer drugs.
• To characterize the variability associated with drug-induced white blood cell (WBC) count decline.

Main Methods:

• A population pharmacokinetic/pharmacodynamic (PK/PD) model was developed, integrating structural, covariate, and variance submodels.
• The model was simultaneously fit to data from etoposide therapy in 71 cancer patients (118 courses).
• Key covariates investigated included baseline WBC count, drug concentration, serum albumin, and serum bilirubin.

Main Results:

• A typical response profile showed a 4.5-day lag time before WBC decline, a 22-day duration below baseline, and a C50 of 3 mg/L for etoposide.
• Lower serum albumin and higher serum bilirubin were associated with increased toxicity.
• Significant inter-individual and intra-individual variability in lag time, duration, and C50 was observed.

Conclusions:

• The developed population model provides a robust framework for quantifying hematologic toxicity.
• The model can predict treatment outcomes and inform optimal dosing and sampling strategies.
• Understanding variability is key to personalizing chemotherapy and managing toxicity risks.