Abstract Background: The relationship between pathological complete response (pCR) and improved long-term prognosis was first reported in the landmark NSABP B-18 and B-27 trials and then validated in large cohort studies and in meta-analyses. Whether pCR, as surrogate end point, should capture the full effect of neoadjuvant chemotherapy on survival is still unproven. To assess the role of pCR as surrogate of disease-free survival (DFS) and overall survival (OS) at a trial level, we performed a trial-based meta-regression of randomized studies comparing different neoadjuvant therapies with data on both pCR rates and outcome end points. Methods: The systematic search included electronic databases and proceeding of international meetings. No restriction regarding treatment types was set up except for endocrine therapy trials that were excluded. We did not include studies comparing neoadjuvant versus adjuvant chemotherapy. Relevant data were extracted from each selected trial: patients randomly assigned to treatment and analyzed per arm, tumor stage, type of chemotherapy and targeted therapy, cycles number, additional post-surgical treatments, number of patients achieving pCR and number of outcome events per arm. For each trial, the experimental and the control arm were defined. Treatment effects on DFS and OS were expressed as hazard ratios (HRs) and on pCR as odd ratios (ORs). A weighted regression analysis was performed on log-transformed treatment effect estimates to test the association between the treatment effects on the surrogate outcome and the treatment effects on the clinical outcome. Results: A final set of 28 trials, 57 arms and 29 contrasts for a total of 13,738 patients was included in the analysis. As one trial had 3 arms, it contributed two contrasts to the analysis. The regression of the logHR for DFS on the logOR for pCR, using the complete set of data, demonstrated a weak but significant association between the two effects (R2 = 0.16, p-value = 0.03). Sixteen% of the variability among the effects on DFS can be explained by the observed effects on pCR. The regression of the logHR for OS on the logOR for pCR demonstrated a weak association between the two effects (R2 = 0.15, p-value = 0.04). Conclusions: This trial-based meta-regression showed a possible role of pCR as surrogate end point for DFS and OS in breast cancer patients treated with neoadjuvant chemotherapy. The surrogacy of pCR warrants to be further evaluated in more homogeneous subsets of clinical trials in the neoadjuvant chemotherapy-based setting. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P3-14-03.