ABSTRACT
The specificity of the tumor markers used to date in patients with gastric cancer has not been satisfactory. For this reason we decided to evaluate the utility of TAG-72 in this disease. Between 1993 and 1998 we determined the levels of TAG-72 in 638 subjects (148 healthy volunteers, 33 patients with chronic renal failure (CRF), 149 patients with benign diseases of the liver, 95 patients with benign gastrointestinal diseases and 213 patients with gastric cancer). TAG-72 was measured using an IRMA method. Statistical analysis of the data was performed with the BMDP package. We established a cutoff for TAG-72 of 3 U/mL, corresponding to the 92.6th percentile of the healthy controls. We observed that neither CRF nor benign liver diseases affected TAG-72 levels, while certain benign gastrointestinal diseases did cause alterations of the marker. Using Cox multivariate analysis we discovered that the preoperative TAG-72 level was an independent prognostic variable associated with both disease-free and overall survival. We conclude that, although TAG-72 is not useful for the diagnosis of gastric cancer, it is a suitable tool for disease monitoring and prognostic assessment.