LB-018 - MODEL BASED META ANALYSIS OF IMATINIB PHARMACOKINETICS: IN ABCB1, ABCG2, AND CYP3A5 GENOTYPES.

H. Kim¹, J. Hong², S. Jung¹, H. Lee¹, K. Ryu², J. Lim³, H. Kim³, M. Kim², H. Sung², J. Park², M. Chang⁴; ¹Department of Pharmaceutical Medicine and Regulatory Science, Yonsei University, , , ²Department of Pharmacy and Yonsei Institute of Pharmaceutical Sciences, Yonsei University, , , ³Graduate Program of Industrial Pharmaceutical Science, Yonsei University, , , ⁴Yonsei University, , .

Background: Imatinib mesylate (Gleevec) is a Tyrosine kinase inhibitor (TKI) used in the treatment of chronic myeloid leukemia (CML), where it targets the abnormal BCR-ABL tyrosine kinase associated with the Philadelphia chromosome. Imatinib is a substrate for efflux transporters such as ABCB1 and ABCG2, and it is metabolized by CYP3A4 and CYP3A5. Previous studies have shown a failure to achieve adequate remission rates in patients with gastrointestinal stromal tumors (GIST) and in receiving CYP3A4 inducers or inhibitors. Therefore, predicting plasma concentrations using personalized data such as genomic information is crucial. We aim to provide a rationale for more effective and safer use of imatinib through population pharmacokinetic modeling with appropriately populated samples by Model-Based Meta-Analysis (MBMA).

Methods: This study employed a MBMA approach using data obtained from the literature. We searched the population PK studies using PubMed and ("imatinib" [Title]) AND ("population pharmacokinetics" [Title]) were used as search terms, and the search date was January 16, 2023. Mrgsolve R package was used for the data extraction and pharmacokinetic data of 1500 patients per study were simulated. After the data extraction, non-linear mixed-effect modeling was done using 'Monolix2023R1' software. The evaluation of the final model was done with goodness-of fit plot (GOF), visual predictive check (VPC) and bootstrap (n=5,000).

Results: Two studies were selected for MBMA. Imatinib concentration-time profiles were adequately explained by a one-compartment model with first-order absorption. We also investigated the influences of four covariates: ABCB2, CYP3A5*3, ABCB1 genotypes, and weight, and the central volume of distribution was positively correlated with body weight and negatively correlated with adult GIST population. Clearance was positively correlated with body weight and negatively correlated with both ABCG2 421C>A and group 2. Objective function value (OFV) of the final model decreased from the initial model with 723,669.36 to 719,635.20 (∆OFV = -4,034.16).

Conclusion: This study suggests initiating imatinib treatment with a lower dose in high-risk patients or implementing regular Therapeutic Drug Monitoring (TDM) as a beneficial approach to prevent excessive plasma concentrations and enhance drug safety.