Pharmacogenetic pathway analysis of irinotecan

Abstract

3073 Background: Irinotecan (IRN), a chemotherapeutic agent for various solid tumors, is the parent compound for SN-38, the active metabolite. The pharmacokinetics (PK) of IRN are complex and associated with variability in clinical outcomes. We sought to learn about the pharmacogenetics (PGx) of IRN by relating the drug’s PK to polymorphisms (SNPs) in drug-metabolizing enzyme and transporter genes potentially relevant to IRN. Methods: We measured concentrations of IRN and 3 metabolites in 86 patients. We fit a novel 7-compartment (15 parameter) model (including enterohepatic circulation) to the concentration data for IRN and metabolites. We applied principal components (PCs) analysis to the patient-specific PK model parameters to reduce the number of outcome variables and to identify appropriate linear combinations of the parameters that relate to IRN’s metabolic pathways. We then analyzed associations between SNPs and PCs to learn about functional polymorphisms. Associations are significant if p < 0.05. Results: We found that 8 PCs accounted for 90% of the total variation in the 15 fitted parameters. PC #1 corresponded to SN-38 in plasma and was associated with SNPs in the CYP3A4, MRP2, and ABCG2 genes. PC #2, the IRN compartments’ PC, was associated with a SNP in OATP-C*1b. The APC PC was associated with a CYP3A5 SNP. The SN-38 glucuronidation PC was associated with UGT1A1 and UGT1A9 polymorphisms. Conclusions: PC analysis is a useful way to reduce the dimension of multiple PK parameters and to produce pathway-specific and interpretable measures relating to PK. The use of PCs in PGx analysis allows identification of potential functional polymorphisms, which can be further evaluated in other data sets. [Table: see text]