Utility of postoperative CEA for surveillance of recurrence after resection of primary colorectal cancer.

Introduction: To evaluate the usefulness of postoperative CEA levels in the surveillance of colorectal cancer patients.

Methods: Over a 56 month period a total of 569 patients with measured CEA levels underwent curative resection for colorectal cancer. The median follow up was 40 months, during which period recurrence occurred in 149. Serum CEA levels were measured at 6 monthly intervals starting from 3 months post resection. ROC was used to calculate the optimum cut-off of CEA (5 ng/ml).

Results: Postoperative elevation of CEA levels were more frequent in patients with an aggressive primary colorectal cancer (grade, T stage and nodal disease; p < 0.05). In patients found to have colorectal recurrence, a significantly higher proportion of patients were resectable in the group with a non-elevated CEA (diagnosed by CT with PET imaging p < 0.05). The median interval between CEA elevation and diagnosis of recurrence (diagnostic interval) was 4 weeks. CEA elevation led to a change in the routine surveillance program by bringing imaging forward by 2 months. CEA levels were a significant predictor of survival following resection of colorectal primary (CEA ≤5-38 months, CEA >5-27 months; p < 0.05). CEA (p < 0.05) retained its significance on multivariate analysis along with the T stage (p < 0.05).

Conclusion: CEA is a predictor of recurrence, resectability and survival following resection of colorectal cancer. Furthermore, an elevated CEA has a short diagnostic interval (4 weeks) for detecting recurrent disease and therefore should mandate adjustment of the routine surveillance program with the next planned imaging being brought forward (2 months).