CEA as a monitor of gastrointestinal malignancy.

Ninety-four patients with carcinoma of the colon have been followed with serial determinations of plasma CEA (carcinoembryonic antigen) levels over a 3-year period using the Hansen assay. Nine hundred twelve CEA determinations have been made in these patients. Plasma CEA levels rose in 90% of the instances of clinical progression documented in these patients. In 30% of patients, this rise indicated progression 6 months or more before it was detected by standard clinical methods. Unfortunately, a few patients never developed elevated CEA levels even though disease clearly progressed. False positive results have also been encountered, with significant elevations occurring in patients who have since remained without evidence of disease for several months. Our data indicate that at least two sequential elevated CEA values, the second being higher, must be a minimal criterion for consideration of possible progression of disease. Even with this standard, we have encountered false positive results in 10% of our patients, indicating recurrence or progression where none has occurred clinically. CEA measurement is of limited usefulness for 30 days after curative surgery, because the elevation of CEA levels due to the original amount of tumor present as well as due to surgery per se may persist for this length of time in a significant number of patients. On the other hand, CEA levels have responded to chemotherapy in close correlation with observed clinical course in those patients with metastatic disease treated in this series. Initial pretherapy CEA values have so far proved to be good prognostic indicators of disease course following complete resection. With an initial CEA value of less than 2.5 ng/ml of plasma, recurrent has been rare (1/20). If the pretreatment CEA was greater than 7.0 ng/ml, it has been the rule (7/9).