CPS + EG and Neo-Bioscore predict individual outcomes but not trial results after neoadjuvant therapy for early breast cancer.

CPS + EG and Neo-Bioscore are prognostic scores estimating the risk of recurrence after neoadjuvant chemotherapy for early breast cancer, but can they predict the long-term outcomes in a clinical trial? In this study, we assessed their surrogacy values at patient level and trial level for long-term outcomes, using data from 8,359 patients across nine neoadjuvant randomized trials conducted by the German Breast Group (GBG) and AGO-B study group. We found that at the patient level, both scores were significantly associated with outcomes, but at the trial level, the associations with distant disease-free survival (DDFS), invasive disease-free survival (iDFS), and overall survival (OS) were weak for both scores (R2 range = 0.03 to 0.24, with wide confidence intervals). Thus, our findings suggest that CPS + EG and Neo-Bioscore are not reliable surrogate endpoints for predicting long-term treatment benefit in neoadjuvant trials but remain valuable for estimating individual patients' residual risk after neoadjuvant treatments and guiding subsequent treatment decisions for early breast cancer patients.