Abstract
We evaluated the usefulness of a newly devised tumor marker index (TMI), namely, the geometric mean of normalized carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9), in determining colorectal cancer (CRC) prognosis. This retrospective cohort study included 306 patients with stages I-III CRC who underwent elective laparoscopic resection between April 2010 and March 2020. Survival rates and risk factors of relapse-free survival (RFS) and cancer-specific survival (CSS) were analyzed using Kaplan-Meier curves and Cox proportional hazards model. High-TMI group (122 patients) had significantly lower rates (95% confidence interval [95% CI]) for 5-year RFS (89.7%, 83.9-93.5 vs. 65.8%, 56.3-73.8, p < 0.001) and CSS (94.9%, 89.4-97.6 vs. 77.3%, 67.7-84.4, p < 0.001) than low-TMI group. Multivariate analysis (hazard ratio [95% CI]) indicated ≥ T3 disease (RFS: 2.69, 1.12-6.45, p = 0.026; CSS: 7.64, 1.02-57.3, p = 0.048), stage III CRC (RFS: 3.30, 1.74-6.28, p < 0.001; CSS: 6.23, 2.04-19.0, p = 0.001), and high TMI (RFS: 2.50, 1.43-4.38, p = 0.001; CSS: 3.80, 1.63-8.87, p = 0.002) as significant RFS and CSS predictors. Area under the curve (AUC) of 5-year cancer deaths (0.739, p < 0.001) was significantly higher for TMI than for CEA or CA19-9 alone. Preoperative TMI is a useful prognostic indicator for patients with resectable CRC.