Assuming an objective response rate (ORR) of 15% in the control group and 35% in the experimental group, with a two-sided α of 0.05 and power of 80%, the calculated sample size is 162 participants. Considering a dropout rate of 5%, approximately 86 participants are needed in each group, totaling about 172 participants. Assuming a median progression-free survival (mPFS) of 3 months in the control group and a hazard ratio (HR) of 0.55, 90 PFS events can provide at least 80% power to detect a significant improvement in PFS at a two-sided α of 0.05. Assuming a median overall survival (mOS) of 8 months in the control group and an HR of 0.6, 122 death events can provide at least 80% power to detect a significant improvement in OS at a two-sided α of 0.05.
Two interim analyses for OS are planned. The first interim analysis for OS will be conducted when the final required assessable sample size or number of events for ORR and PFS has been observed, with an estimated 30% of death events. The second interim analysis for OS will be conducted when 66% of death events have been observed. The O'Brien-Fleming type α-spending function (using the Lan-DeMets method for approximation) will be employed to control the overall Type I error rate, and the p-values for the interim and final analyses will be calculated based on the actual number of observed OS events. The interim analyses will be conducted by the Independent Data Monitoring Committee (IDMC). In this trial, the control of Type I error for multiple testing will use a sequential testing method to sequentially test ORR, PFS, and OS. If any endpoint does not pass the significance test, subsequent endpoints will not be tested for statistical significance.
Intention-to-Treat (ITT) Analysis Set: All randomized participants will be included in the ITT analysis set.
Safety Analysis Set (SS): This set includes all participants who received at least one dose of the investigational drug and have post-treatment safety evaluation data.
Per Protocol Set (PPS): This set includes all participants who did not experience major protocol deviations that significantly affected efficacy, and the analysis is based on the actual treatment received, supplementing the efficacy analysis of the ITT set.
Crossover Treatment Analysis Set: This set includes participants randomized to the control group who received at least one dose of MRG003 after disease progression (PD).
Anti-Drug Antibody (ADA) Analysis Set: This set includes all participants who received at least one dose of the investigational drug MRG003 and have evaluable ADA results.
The ORR for each treatment group will be calculated, and the two-sided 95% confidence interval (CI) will be estimated using the Clopper-Pearson method. The difference between the two groups will be tested using the Cochran-Mantel-Haenszel method, with stratification factors being liver metastasis and ECOG performance status. The median PFS and median OS for each treatment group, as well as the PFS and OS rates at different time points, will be estimated using the Kaplan-Meier method. The 95% CI for the median PFS and median OS will be estimated using the Brookmeyer-Crowley method, and the 95% CI for the PFS and OS rates at different time points will be calculated using the Greenwood formula. The differences in PFS and OS between the two groups will be compared using a stratified log-rank test, and the HR and 95% CI for PFS and OS between groups will be estimated using a stratified Cox proportional hazards model. Sensitivity and supplementary analyses for the primary endpoint are detailed in the supplementary material. Statistical analyses will be performed using SAS Version 9.4 or higher, with a two-sided alpha level of 0.05.
