Diagnosis of esophageal varices on liver CT: is thin-section reconstruction necessary?

The aim of this study was to evaluate the effect of 1 mm, 3 mm, and 5 mm section thicknesses of liver CT on the detection and grading of esophageal varices in cirrhotic patients.

A total of 219 consecutive cirrhotic patients who had undergone both upper endoscopy and 64-channel liver CT were included. Portal phase images of each CT were reconstructed with a section thickness of 1 mm, 3 mm, and 5 mm. Four radiologists independently reviewed the 3 image sets. The observers evaluated the presence of esophageal varices on a 5-point confidence scale and measured the maximal short diameter of the largest varix identified. Sensitivity, specificity and predictive value calculation and receiver operating characteristic analysis were performed. Correlation between CT measurements and endoscopic grading as the reference standard was assessed.

The averaged Az values at 1 mm, 3 mm, and 5 mm image sets were 0.936, 0.946, and 0.935, respectively, and they were not significantly different among the 4 observers. When a 3 mm cut-off criterion for large varices was applied, sensitivity, specificity, and predictive values were comparable among the 3 image datasets.

Routine liver CT protocol is sufficient for evaluation of esophageal varices in cirrhotic patients without adding thin section reconstruction images.