NCCN胃癌指南解读__培训课件.ppt

* A superior response rate was observed for Xeloda/cisplatin vs. 5-FU/cisplatin (p=0.030), supporting the general trend to superiority seen in other efficacy endpoints. Tumor responses were assessed according to Response Evaluation Criteria in Solid Tumors (RECIST) specifications. Overall response is the sum of confirmed complete and partial responses. Three complete responses were observed in Xeloda/cisplatin groups vs. 4 complete responses in the 5-FU/cisplatin group. Patients in the Xeloda/cisplatin group had 1.4 times the chance of achieving an overall response compared to patients in the 5-FU/cisplatin group. The results of an independent review did not contradict these findings. * In the per protocol analyses without covariate adjustments, Xeloda/cisplatin was found to be non-inferior to 5?FU/cisplatin; therefore, the primary endpoint of non-inferiority in progression-free survival (PFS) in the ML17032 study was met with high significance. The Kaplan-Meier plot of PFS shows a clear separation of curves for Xeloda/cisplatin after the first month and this trend continues until month 13 when only a few patients were left in the risk sets. The comparison of the hazard ratio (HR) limit of 1.25 shown here is a more rigorous test than that vs. the limit of 1.4, and shows a clear difference that is statistically significant (p=0.0008) based on the per protocol analysis of non-inferiority. Although not shown, the p value for the less rigorous HR cut-off of 1.4 was 0.0001 based on the per protocol unadjusted analysis of non-inferiority. In addition, the protocol-specified primary analysis for a non-inferiority margin of 1.4 based on the adjusted analysis was also statistically significant (p=0.003). Every test yielded a statistically significant result of non-inferiority in PFS for Xeloda/cisplatin in comparison to 5-FU/cisplatin. * Xeloda/cisplatin and 5?FU/cisplatin were similar in terms of the percentage of patients experiencing hematologic adverse ev