Tumor Marker Trajectories and Survival Analysis in Patients with Normal Carcinoembryonic Antigen Ranges After Colorectal Cancer Resection

Abstract

BACKGROUND: Evidence regarding postoperative CEA for predicting long-term outcomes of colorectal cancer remains controversial, especially in patients with normal postoperative CEA. OBJECTIVE: To investigate the risk difference among different postoperative CEA trajectories in patients with normal postoperative CEA after curative colorectal cancer resection. DESIGN: This cohort study was conducted at a comprehensive cancer center and included data retrieved from a prospectively collected database between January 2006 and December 2018. SETTINGS: Retrospective cohort study. PATIENTS: Patients with colorectal cancer who underwent surgery for primary stage I–III colorectal adenocarcinoma were included. Patients with postoperative CEA >5 ng/mL were excluded. INTERVENTIONS: Standard curative radical resection was performed. MAIN OUTCOME MEASURES: Ten-year overall survival and disease-free survival were analyzed. RESULTS: The study population (n = 8156) was categorized into 6 trajectories, including persistently-ultralow (n = 2351), persistent-low (n = 2474), gradually-decrease (n = 401), persistent-medium (n = 1727), slightly-increase (n = 909), and around-upper-limit (n = 394). The median follow-up time was 7.8 years, and the median timeframe in which CEA was measured to determine trajectory was 2.6 years. The persistent-ultralow group had the highest 10-year overall survival (85.1%) and disease-free survival (82.7%). The around-upper-limit group had the lowest 10-year overall survival (55.5%) and disease-free survival (53.4%). The adjusted hazard ratio trend was comparable to crude hazard ratio of the persistent-ultralow group. Consequently, the higher initial serum CEA groups had higher hazard ratios of overall survival and disease-free survival. The adjusted hazard ratios of overall survival/disease-free survival were 2.96/2.66 (95% confidence interval: 2.39–3.66/2.18–3.69) for the around-upper-limit groups. LIMITATIONS: The study was limited by its retrospective design. CONCLUSIONS: The postoperative serum CEA trajectory is an independent factor associated with long-term outcomes. Although CEA levels were all within normal range, higher levels of postoperative serum CEA trajectory correlated with worse long-term oncological outcomes..