A pharmacogenetic predictive model for paclitaxel clearance based on the DMET platform

Area of Science:

• Pharmacogenomics
• Oncology
• Drug Metabolism

Background:

• Paclitaxel is a key chemotherapy for solid tumors.
• Significant inter-individual variability in paclitaxel exposure necessitates personalized treatment approaches.
• Low paclitaxel clearance is linked to increased toxicity, highlighting the need for predictive tools.

Purpose of the Study:

• To develop and validate a genetic prediction model for identifying patients with low paclitaxel clearance.
• To utilize the drug-metabolizing enzyme and transporter (DMET) platform for comprehensive genetic variant analysis.
• To assess the predictive capability of genetic variations in paclitaxel pharmacokinetics.

Main Methods:

• A population pharmacokinetic model (NONMEM) was employed to estimate paclitaxel clearance in 270 patients.
• The DMET platform identified 1,936 genetic variants across 225 genes.
• A 14 single-nucleotide polymorphism (SNP) model was built using a training set and validated on a separate patient cohort.

Main Results:

• The 14-SNP genetic model demonstrated high sensitivity (95%) in identifying patients with low paclitaxel clearance in the validation set.
• The positive predictive value of the model was 22%, indicating a high rate of false positives.
• The model's association with low paclitaxel clearance remained significant after adjusting for clinical factors (age, gender, hemoglobin).

Conclusions:

• A 14-SNP model derived from the DMET platform shows high sensitivity for detecting low paclitaxel clearance.
• Genetic variability alone, as captured by the DMET chip, is insufficient to fully predict paclitaxel clearance.
• Further research is needed to incorporate additional factors for improved prediction of paclitaxel pharmacokinetics.