Individual- and trial-level surrogacy in colorectal cancer:

摘要: Two conditions must be fulfilled for an intermediate endpoint to acceptable surrogate a true clinical endpoint: (1) there strong association between the and endpoint, (2) effects of treatment on endpoint. We test whether these are disease-free survival (DFS) progression-free (PFS) data from 20 trials comparing experimental treatments with standard early advanced colorectal cancer. The DFS (or PFS in disease) OS were quantified through log hazard ratios (log HR), estimated Weibull model stratified trial. rank correlation coefficients OS, trial-specific effects, using bivariate copula distribution endpoints. A linear regression was used compute “surrogate thresho...