Detection of hepatic metastases in colorectal cancer: A prospective study of laboratory and imaging methods

Objective: To assess and compare the accuracy of imaging methods (ultrasonography (US), computed tomography (CT), angiography, arterially enhanced CT with computed tomographic arterial portography (CTA)), biochemical analyses, and surgical assessment during the operation, in detecting the presence or absence of hepatic metastases in patients with colorectal cancer. Design: Open study. Setting: University of Lund, Sweden. Subjects: 71 consecutive patients without clinical evidence of liver metastases at the time of the operation for removal of the primary tumour. Interventions: Regular biochemical tests for five to seven years, and CT and US one year after operation in all patients without confirmed metastases. Main outcome measures: Presence or absence of hepatic metastases. Results: Accuracy of surgical assessment, angiography, US, CT and CTA was 90, 77, 80, 82 and 83%, respectively, and corresponding predictive values of a negative test were 87, 75, 77, 80 and 84%. Measurement of bilirubin concentration and hepatic enzyme activities were not helpful, and that of carcinoembryonic antigen had an accuracy of only 70%. Accuracy and predictive values were not improved by combining tests. Conclusion: The accuracy of US, CT or CTA was not good enough to merrit routine use before operations for colorectal cancer. They are potentially valuable for monitoring progress at follow up, but this remains to be confirmed.