Purpose: The aim of the present study was to determine the accuracy of yearly postoperative monitoring of serum tumor markers to either detect or rule out recurrence in colorectal cancer patients.
Methods: A total of 127 colorectal cancer patients who underwent curative surgery were enrolled. The serum carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) levels were assayed, and radiological examinations were performed routinely for 5 years after surgery or until recurrence was detected. Yearly recurrence rates (number of recurrences/number of patients assessed in a given year), sensitivities, specificities, and likelihood ratios were calculated. Post-test probabilities were calculated from these values.
Results: Recurrences tended to show almost the same frequencies in the first and second year after surgery (20 of 127 patients and 18 of 107 patients, respectively). However, the post-test probability of recurrence in patients with positive and negative serum CEA levels was significantly lower in the second year than in the first year (test positive: 40.0% and 76.0%; test negative: 9.3% and 0.5%, respectively).
Conclusions: Measuring CEA can help to identify patients likely to demonstrate recurrence with high accuracy only within the first year after surgery. Another examination, such as imaging, is therefore necessary for monitoring patients at 2 or more years after surgery.