Comparing predictive decision rules in postoperative CEA monitoring

Area of Science:

• Oncology
• Clinical Chemistry

Background:

• Colorectal carcinoma is a significant cause of cancer-related mortality.
• Serial carcinoembryonic antigen (CEA) assays are commonly used postoperatively to monitor for tumor recurrence.
• Predictive decision rules for tumor recurrence based on CEA levels exist but require validation.

Purpose of the Study:

• To retrospectively evaluate the clinical utility of established decision rules for predicting tumor recurrence in colorectal cancer patients using serial CEA assays.
• To assess the adequacy of CEA assays in detecting late recurrences and informing surgical decisions.

Main Methods:

• A retrospective analysis of CEA values from 214 patients who underwent surgery for colorectal carcinoma.
• Comparison of seven previously published decision rules for predicting tumor recurrence.
• Patient follow-up ranged from 36 to 120 months post-surgery.

Main Results:

• Decision rules using CEA cutoff values were found inadequate for predicting recurrence in asymptomatic patients.
• The prognostic usefulness of CEA assays was attenuated, particularly for late recurrences.
• Late-recurring tumors frequently did not present with elevated CEA levels.

Conclusions:

• Elevated CEA levels alone, without other objective evidence, may be insufficient justification for second-look surgery in colorectal cancer patients.
• Serial CEA assays have limitations in predicting late tumor recurrences, necessitating careful interpretation in clinical practice.