Dynamic contrast-enhanced MR imaging and dual-phase helical CT in the preoperative assessment of suspected pancreatic cancer: a comparative study with receiver operating characteristic analysis.

Abstract

OBJECTIVE: The purpose of this study was to compare dynamic contrast-enhanced MR imaging and dual-phase helical CT in the preoperative assessment of patients with suspected pancreatic carcinoma. SUBJECTS AND METHODS: Thirty-three consecutive surgical candidates (20 men, 13 women; 39-81 years old) were included. MR imaging comprised fast spin-echo (TR/TE 4000/91), fat-suppressed T1-weighted spin-echo (500/15), and T1-weighted breath-hold gradient-echo fast low-angle shot (100/4; flip angle, 80 degrees) images before and after the administration of gadopentetate dimeglumine. Helical CT used 5-mm collimation with a pitch of 1:1.5-1.7; images were obtained 20 and 65-70 sec after injection of 150 ml of contrast material. Two pairs of interpreters who were unaware of the results of the other imaging method independently scored each examination for the presence of a lesion and for surgical resectability using a five-point scale. Results were correlated with surgery (n = 25) or consensus review (n = 8). Receiver operating characteristic methodology was used to analyze the results for resectability, and positive predictive values were calculated. RESULTS: Both MR imaging and helical CT revealed 29 of 31 lesions. In determining lesion resectability, the mean areas under the receiver operating characteristic curves were 0.96 and 0.81 (p = .01) and the positive predictive values were 86.5% and 76% (p = .02) for MR imaging and helical CT, respectively. CONCLUSION: MR imaging and helical CT performed equally well in lesion detection. MR imaging was significantly better in the assessment of resectability of pancreatic tumors.