Prognostic value of carcinoembryonic antigen kinetics in predicting distant metastatic recurrence of breast cancer: a multicenter cohort study

Purpose: Distant metastasis of breast cancer significantly affects the prognosis. Serum markers such as CEA and cancer antigen (CA) 15-3 are used for surveillance. However, the nonspecificity and nonmalignant elevation of CEA limit its absolute value in predicting metastatic recurrence. We reevaluated the clinical value of CEA kinetics in predicting metastatic recurrence in breast cancer. Methods: In this multicenter retrospective cohort study, we utilized electronic medical record data from eight institutions (2008-2023), including 4,872 surgically treated patients with invasive breast cancer. Novel CEA kinetic indices were defined. Predictive capabilities were assessed using logistic regression, Kaplan-Meier survival curves, and Cox proportional hazards models. Results: CEA velocity (vCEA) and last-interval vCEA (vCEA-L) showed higher accuracy and sensitivity in predicting distant metastasis than absolute CEA indices. All tumor marker indices were independent predictors; vCEA, vCEA-L, and maximum vCEA (vCEAmax) showed the highest association. In the 5-year distant disease-free survival analysis of patients, vCEA showed the most significant difference based on whether its cutoff value was exceeded (84.1% vs. 47.2%) (P < 0.001). Cox analysis showed that vCEA was the strongest predictor (hazard ratio, 4.509; P < 0.001). vCEAmax remained prognostic even when CEAmax was below its cutoff value. Conclusion: Analysis of dynamic CEA changes, particularly velocity-based indices, offers superior predictive power than the static-based CEA indices for metastatic recurrence in breast cancer. These kinetic markers enhance prognostic accuracy when combined with CA 15-3. Future research should integrate these markers with imaging, liquid biopsy, and artificial intelligence for personalized follow-up.