Pharmacogenetic prediction for tumor response, toxicity, and survival of NSCLC patients treated with irinotecan and cisplatin chemotherapy

Abstract

3581 Background: Inherited genetic polymorphisms of the genes involved in the activation and metabolism of irinotecan can affect toxicity and clinical outcome of patients after treatment with irinotecan. We examined 14 polymorphisms within six genes putative influence on irinotecan metabolism to define a pharmacogenetic model for predicting tumor response, toxicity, and survival in NSCLC patients treated with irinotecan/cisplatin chemotherapy. Methods: Blood samples from 81 NSCLC patients treated with irinotecan (80 mg/m2 D1 & D8) and cisplatin (60 mg/m2 D1) chemotherapy were used for genotyping UGT1A1 211G>A (*6), UGT1A1*28, UGT1A9 -118(T)9>10 (*22), ABCB1 1236C>T, 2677G>T/A, 3435C>T, ABCC2 -24C>T, 1249G>A, 3972C>T, ABCG2 34G>A, 421C>A, and SLCO1B1 -11187G>A, 388A>G, and 521T>C polymorphisms. Genotypes were correlated with tumor response, toxicity, and survival by multivariate logistic or Cox regression analysis with adjustment for clinical and histopathologic features. Results: The overall response rate was 47% (36/77), the median progression-free survival was 5.1 months, and the median overall survival (OS) was 14.3 months. The most common severe toxicity was NCI-CTC grade 4 neutropenia, which occurred in 22 (27%) patients. Grade 3 diarrhea was developed in 8 (10%) patients. The presence of ABCC2 3972T allele was predictive for tumor response (odds ratio=3.244, 95% CI 1.219–8.629, p=0.018). The UGT1A1 211AA (*6/*6) genotype was a significant risk factor for grade 4 neutropenia (odds ratio=6.111, 95% CI 1.032–36.197, p=0.046). The UGT1A9*22 allele showed a trend for less grade 3 diarrhea (odds ratio=0.205, 95% CI 1.032–36.197, p=0.054). In multivariate Cox regression analysis, the UGT1A1*6/*6, UGT1A9*22/*22, and squamous cell histology were predictive for worse PFS (p=0.015, 0.049, and 0.026, respectively). Whereas, ABCC2 1249A allele, ECOG PS 2, male gender, and stage IV were predictive for worse OS (p=0.017, 0.014, 0.045, and 0.015, respectively). Conclusions: Patients’ selection based on pretreatment genotyping can minimize irinotecan-related severe toxicity and maximize the clinical benefit of irinotecan-based chemotherapy for patients with advanced NSCLC. No significant financial relationships to disclose.