Evaluation of clinical validity of an S-1 dosage formula based on renal function using data of the SPIRITS and the G-SOX trials

Background: The aim of this study was to evaluate clinical validity of the S-1 dosage formula based on body surface area (BSA) and creatinine clearance (CLcr) to achieve the target area under the concentration–time curve of 5-FU, which we had developed and refined in each prospective pharmacokinetic study. Methods: The recommended dose determined by the refined formula was assessed using data of the SPIRITS (S-1 vs. S-1 plus cisplatin [SP]) and the G-SOX (SP vs. S-1 plus oxaliplatin [SOX]) trials. Nine hundred and thirty-eight patients in these trials were classified into three groups according to their actual S-1 starting doses compared with the recommended doses (under-dosed, <recommended dose; equal-dosed, =recommended dose; over-dosed, >recommended dose). Results: The patients in the under-dosed group in both trials showed similar tendencies: male, younger, higher BSA, and higher CLcr. The incidence of any grade neutropenia was significantly greater in the over-dosed group compared with the equal-dosed group in the S-1 and the SOX arms. The hazard ratios (HR) of overall survival (OS) (under-dosed vs. equal-dosed) were 1.361 (S-1 arm), 1.259 (SP arm) in the SPIRITS trial, and 1.381 (SOX arm), 0.999 (SP arm) in the G-SOX trial. Multivariate analysis in all the patients demonstrated that OS of the over-dosed group was equivalent (HR 1.002, 95% confidence interval [CI] 0.850–1.182, p = 0.980) and that of the under-dosed group was inferior (HR 1.267, 95% CI 1.005–1.597, p = 0.045) to the equal-dosed group. Conclusions: It is suggested that the refined S-1 dosage formula can recommend optimal dose in terms of safety and efficacy.