CT Findings for Detecting the Presence of Gangrenous Ischemia in Cholecystitis

OBJECTIVE. The objective of our study was to evaluate the diagnostic accuracy of CT in differentiating gangrenous cholecystitis from uncomplicated cholecystitis and healthy gallbladders. MATERIALS AND METHODS. We performed a retrospective review of 308 patients with histologically proven gangrenous cholecystitis who underwent preoperative CT 1 month before gallbladder removal over a 4-year period. Two readers who were blinded to the histologic diagnosis independently recorded CT features of and overall likelihood of gangrenous cholecystitis on a 5-point scale (1, definitely absent; 5, definitely present). Ratings were dichotomized such that a diagnosis was considered present at a rating of 4 or 5 and considered absent at lower ratings. Interobserver variability for individual CT findings was also assessed. RESULTS. Patients had gangrenous cholecystitis (n = 28), acute cholecystitis (n = 98), chronic cholecystitis (n = 118), or healthy gallbladders (n = 64). Multivariate analysis showed that CT findings of gallbladder distention greater than 4.0 cm (odds ratio [OR], 9.63; p < 0.01), mural striation (OR, 11.39; p < 0.01), and decreased mural enhancement (OR, 3.55; p < 0.05) independently predicted gangrenous cholecystitis. Using these CT features, the diagnosis of gangrenous cholecystitis was made with a specificity of 93.9% and 89.6% for readers 1 and 2, respectively, and accuracy of 90.9% and 87.0%, respectively. Good agreement was seen between the two readers with respect to gallbladder distention greater than 4.0 cm (kappa = 0.77) and decreased mural enhancement (kappa = 0.64). CONCLUSION. A markedly distended gallbladder associated with decreased wall enhancement is highly specific for gangrenous cholecystitis.