Is frequent measurement of tumor markers beneficial for postoperative surveillance of colorectal cancer?

Abstract

To determine whether frequent measurement of tumor markers triggers early detection of colorectal cancer recurrence. Of 1,651 consecutive patients undergoing colorectal cancer surgery between 2010 and 2016, 1,050 were included. CEA and CA 19-9 were considered to be postoperative tumor markers and were measured every 3months for 3years, and then every 6months for 2years. Sensitivity analysis of elevated CEA and CA19-9 levels and multivariate analysis of factors associated with elevated CEA and CA19-9 levels were performed. The proportion of triggers for detecting recurrence was determined. The median follow-up period was 5.3years. After applying the exclusion criteria, 1,050 patients were analyzed, 176 (16.8%) of whom were found to have recurrence. After excluding patients with persistently elevated CEA and CA19-9 levels before and after surgery from the 176 patients, 71 (43.6%) of 163 patients had elevated CEA levels and 35 (20.2%) of 173 patients had elevated CA19-9 levels. Sensitivity/positive predictive values for elevated CEA and CA19-9 levels at recurrence were 43.6%/32.3% and 20.2%/32.4%, respectively. Lymph node metastasis was a factor associated with both elevated CEA and CA19-9 levels at recurrence. Of the 176 patients, computed tomography triggered the detection of recurrence in 137 (78%) and elevated tumor marker levels in 13 (7%); the diagnostic lead interval in the latter 13 patients was 1.7months. Tumor marker measurements in surveillance after radical colorectal cancer resection contribute little to early detection, and frequent measurements are unnecessary for stage I patients with low risk of recurrence.