Surrogate endpoints for overall survival in neoadjuvant randomized clinical trials of early-stage breast cancer

Individual patient data analysis of 11 neoadjuvant randomized clinical trials (RCTs) of early breast cancer (EBC) found that pathological complete response rate was a weak surrogate endpoint for overall survival (R² 0.07; 95% CI 0.00-0.48). Overall, the R² for the association of HR for OS and the HR for invasive disease-free survival was 0.46 (95% CI 0.08-0.71), which is just below the cutoff for moderate surrogacy. The R² ranged from 0.5-<0.7 in the majority of subgroups explored, while it was ≥ 0.07 (strong surrogacy) in HR-negative/HER2-negative subtype, histological grade 1-2 tumors, and lobular tumors. The surrogacy value of iDFS for OS was affected by length of follow-up, with substantially increase up to 36 month of follow-up and little further improvement after 48 months of follow-up. The authors conclude that iDFS with sufficient follow-up is an acceptable surrogate endpoint for OS in neoadjuvant RCTs for EBC; this holds true across many subgroups with the notable exception of HR-positive disease.