Role of carcinoembryonic antigen and liver function tests in the detection of recurrent colorectal carcinoma

Abstract

The optimal laboratory evaluation for the early detection of liver metastases from colorectal cancer is controversial. This investigation was undertaken to compare the efficacy of liver function tests (LFTs) with that of carcinoembryonic antigen (CEA) levels for the early detection of liver metastases. Patients who developed liver metastases after potentially curative resections of adenocarcinoma of the colorectum between 1974 and 1988 were reviewed. The following laboratory tests were serially evaluated during the follow-up period: CEA, alkaline phosphatase (AP), serum glutamic oxaloacetic transaminase (SGOT), serum glutamic pyruvic transaminase (SGPT), and lactic dehydrogenase (LDH). These values were retrospectively assessed from the time of documented liver metastases to identify which lab value(s) were elevated initially. Ninety-two patients were available for study. Average time for the occurrence of liver metastases was 20 months (range, 3-72 months). The incidence of elevation of individual tests at the time of suspicion of liver metastasis was: CEA, 94.6 percent (P less than 0.25, chi-squared); AP, 18.5 percent; SGOT, 12.0 percent; SGPT, 5.4 percent; and LDH, 29.3 percent. When comparing CEA with a battery of LFTs at the time of suspicion of liver metastasis, CEA was elevated with normal LFTs in 64.1 percent (P less than 0.05, chi-squared), the most frequent occurrence. At least one LFT was elevated with a normal CEA in only 2.2 percent; CEA and at least one LFT were increased in 30.4 percent; and both tests were normal in only 3.3 percent. These results indicate that, of the individual laboratory tests performed, CEA elevation heralds liver metastases significantly more frequently. LDH is the liver function test most frequently elevated when liver metastases are first suspected. When CEA is directly compared with a battery of LFTs, CEA is statistically significantly more frequently elevated. In fact, suspicion of liver metastases would have been delayed by the omission of LFTs in only 2.2 percent of patients. Therefore, we conclude that LFTs should be deleted from the follow-up of colorectal cancer patients, decreasing costs without significantly decreasing accuracy.