Purpose: To retrospectively assess the sensitivity and specificity of multidetector computed tomographic (CT) angiography in the preoperative evaluation of pancreatic adenocarcinoma by using surgical findings as the reference standard.
Materials and methods: The institutional review board approved this HIPAA-compliant study; informed consent was waived. We reviewed CT reports, surgical notes, and pathology reports from 114 patients with pancreatic or distal cholangiocarcinoma who underwent multidetector CT angiography and surgery at our institution between March 2003 and March 2006. When CT findings and surgical reports were discordant, radiologists experienced in pancreatic imaging retrospectively reviewed images for lesion resectability; four-, eight-, 16-, and 64-row CT scanners were used in 54, 19, 25, and 16 patients, respectively. Collimation of 1.25 mm was used for four- and eight-row CT and 0.5 or 0.625 mm for 16- and 64-row CT. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for resectability were calculated for initial clinical interpretation and blinded retrospective review.
Results: Eighty-eight patients had resectable lesions according to CT angiographic criteria (group A: 46 women, 42 men; mean age, 67 years; age range, 39-85 years): resection was aborted in 10 patients (11%). Twenty-six patients underwent surgery despite lesion unresectability assessed according to CT angiographic criteria (group B: 16 women, 10 men; mean age, 62 years; age range, 33-83 years); all lesions were confirmed as unresectable. The initial clinical interpretation of CT angiographic scans in all 114 patients had 100% sensitivity in the detection of resectability, 72% specificity, 89% PPV, and 100% NPV. These parameters did not appear to vary among different types of scanner. With the blinded retrospective evaluation by experienced readers, specificity increased to 94% and PPV to 98%, with no difference in sensitivity and NPV.
Conclusion: Multidetector CT angiography is an effective preoperative tool that reduces the number of aborted pancreatic resections; there is no evidence from this retrospective study suggesting varying results from the various generations of multidetector CT scanners used.
(c) RSNA, 2007.