Limitations of CEA monitoring as a guide to second-look surgery in colorectal cancer follow-up

Abstract

Forty-two patients with localized colorectal cancer (Dukes' A, B, C stages) were treated with potentially curative surgery and controlled with a follow-up program, which included CEA monitoring, for a period ranging from 12 to 48 months (median 33 months). During this period, we observed recurrent neoplastic disease in 14 patients. A retrospective analysis of the results showed that: 1. patients with a preoperative CEA value greater than 20 ng/ml have a significantly higher risk of recurrence than the patients with CEA less than 20 ng/ml; 2. sensitivity of the CEA test was good for metastatic recurrent disease, fairly good for residual neoplastic disease, but insufficient for local recurrence; 3. test-specificity was poor, as demonstrated by the negative results of four exploratory laparotomies performed exclusively on the basis of increased CEA levels. Since the principal aim of a second-look operation is the cure of local recurrence, this type of surgery cannot be proposed only on the basis of increased CEA levels.