Accuracy of multidetector computed tomography in differentiating primary epiploic appendagitis from left acute colonic diverticulitis associated with secondary epiploic appendagitis

The objectives of this study are to evaluate the prevalence of left acute colonic diverticulitis (LACD) associated with secondary epiploic appendagitis (SEA) detected by computed tomography (CT); to describe CT features that distinguish LACD associated with SEA from primary epiploic appendagitis (PEA); and to assess the accuracy of CT in diagnosing LACD associated with SEA versus PEA. Institutional review board approval was obtained. We retrospectively identified 46 consecutive patients with LACD between July 2004 and July 2005 and 26 patients with PEA between 2000 and 2005 investigated using multidetector CT. Two radiologists blinded to the final diagnosis reviewed the CT images for findings of LACD-associated SEA or PEA. Each reader classified each CT scan into one of four categories: PEA, LACD-associated SEA, LACD without SEA, and indeterminate. Fisher's exact test and Wilcoxon test were performed to compare the groups. The prevalence of LACD-associated SEA was 71% (33/46) in the LACD group. The accuracy of CT was 100% for diagnosing LACD-associated SEA (33/33), 100% for diagnosing LACD without SEA (13/13), and 96% for diagnosing PEA (25/26). Colon wall thickening, "inflamed diverticulum", extraluminal gas, abscess or phlegmon, multiple paracolic fatty lesions, and a thin hyperattenuated rim were significantly associated with LACD-associated SEA. Neither the dot sign nor parietal peritoneal thickening showed good accuracy for differentiating PEA from LACD-associated SEA. CT is accurate for distinguishing LACD-associated SEA from PEA. The findings that perform best for diagnosing SEA are evidence of diverticulitis, multiple fatty lesions, and a thin hyperattenuated rim.