Primary Outcomes

Description: Defined as the number of patients who are alive at 18 months after randomization divided by the total number of evaluable patients in each arm. An evaluable patient is defined as any patient who signed informed consent, deemed eligible by central review and received any protocol-defined treatment. 95% confidence interval will be estimated based on standard method. Chi-squared test (or Fisher's exact test if the data in contingency table is sparse) will be used to compare 18 month OS rates among treatment arms. OS within each arm will be summarized by Kaplan-Meier method. Median, 1-year and 2-year rates will be estimated based on Kaplan-Meier curves.

Secondary Outcomes

Description: Defined as the proportion of patients in whom an achieved R0 resection was achieved during surgery. Chi-square test (or Fisher's exact test if the data in contingency table is sparse) will be used to compare R0 resection rate between treatment arms. Sensitivity analysis will be conducted among patients in cohort 1) and cohort 2). The association between R0 resection rate and OS/progression-free survival will be assessed by log-rank test and Cox model.

Description: Defined as time from randomization to the first documentation of event where events considered are 1) disease progression, per RECIST, prior to surgery, 2) surgery with R2 resection, 3) recurrent disease following surgery, or 4) death due to any cause. Will be estimated using the method of Kaplan-Meier in each arm and compared between treatment groups using the log-rank test. The correlation between pathologic complete response (pCR) status and event-free survival time will be assessed by Cox model with landmark approach.

Description: Defined as the proportion of patients in whom a pCR was confirmed by histopathologic review of the surgical specimen. Chi-square test (or Fisher's exact test if the data in contingency table is sparse) will be used to compare pCR resection rate between two treatment arms. Sensitivity analysis will be conducted among patients in cohort 1) and cohort 2). The association between pCR rate and OS/progression free survival (PFS) will be assessed by log-rank test and Cox model.

Description: Overall adverse event rates will be compared between treatment groups using Chi-square test (or Fisher's exact test if the data in contingency table is sparse).