Establishment and application of a predictive model for gefitinib-induced severe rash based on pharmacometabolomic profiling and polymorphisms of transporters in non-small cell lung cancer

摘要

BackgroundRash is a well-known predictor of survival for patients with gefitinib therapy with non-small cell lung cancer (NSCLC). However, whether patients with more severe rash obtain the more survival benefits from gefitinib is still unknown, and predicted model for severe rash is needed.MethodsThe relationship between gefitinib-induced rash and progression free survival (PFS) was primarily explored in the retrospective cohort. The association between rash and gefitinib/metabolites concentration and genetic polymorphisms were determined by pharmacometabolomic and pharmacogenomics methods in the exploratory cohort and validated in an external cohort.ResultsThe survival for patients with rash was significantly higher than that of patients without rash (p?=?0.0002,p?=?0.0089), but no difference was found between grade 1/2 or grade 3/4. Only the concentration of gefitinib, but not its metabolites, was found to be associated with severe rash, and the cutoff value of gefitinib was 204.6?ng/mL conducted by ROC curve analysis (AUC=0.685). A predictive model for severe rash was established: gefitinib concentration (OR?=?11.523, 95% CI?=?2.898-64.016,p?=?0.0016),SLC22A8rs4149179(CT vs CC, OR?=?3.156, 95% CI?=?0.958–11.164,p?=?0.0629),SLC22A1rs4709400(CG vs CC, OR?=?10.267, 95% CI?=?2.067–72.465,p?=?0.0087; GG vs CC, OR?=?5.103, 95% CI?=?1.032–33.938,p?=?0.061). This model was confirmed in the validation cohort with an excellent predictive ability (AUC?=?0.749, 95% CI?=?0.710–0.951).ConclusionsOur finding demonstrated that the incidence, not the severity, of gefitinib-induced rash predicted improved survival, the gefitinib concentration and polymorphisms ofSLC22A8andSLC22A1were recommended to manage severe rash.