OBJECTIVE: Our goal was to determine the value of MRI in the assessment of vascular invasion in the preoperative staging of pancreatic carcinoma. MATERIALS AND METHODS: In 53 consecutive patients with an established diagnosis of pancreatic ductal adenocarcinoma, SE T1-weighted and breath-hold gradient echo images at 0.5 T were obtained before and after bolus injection of Gd-DTPA. Major peripancreatic vessels were evaluated for the presence of (a) no tumor invasion, (b) contiguity with tumor, and (c) tumor encasement. All patients subsequently underwent surgery. Results of unenhanced and contrast-enhanced MRI studies were compared with the histologic findings in the resected specimens in 34 cases and with the surgical findings in 19 nonresected cases. RESULTS: In six patients, pathologic examination showed the presence of tumor confined to the pancreas with no vascular invasion. With MRI, five of these cases were correctly evaluated; in the remaining case, tumor-vessel contiguity was erroneously diagnosed. In 21 patients, tumor contiguity with adjacent vessels was found at pathologic examination. At MRI, 16 of these cases were correctly assessed, and 5 were understaged as tumors confined to the pancreas. Of the remaining 26 patients, 7 had vascular encasement by tumor at pathologic examination of the resected specimen; the other 19 patients had unresectable tumors encasing the adjacent vessels at surgical evaluation. MRI detected vascular encasement in 21 of these 26 cases; in the other 5, tumor contiguity was incorrectly diagnosed. The overall accuracy of MRI for determining vascular invasion was 79%. CONCLUSION: MRI is a helpful method for preoperative assessment of vascular involvement due to pancreatic carcinoma.
OBJECTIVE: Our goal was to determine the value of MRI in the assessment of vascular invasion in the preoperative staging of pancreatic carcinoma. MATERIALS AND METHODS: In 53 consecutive patients with an established diagnosis of pancreatic ductal adenocarcinoma, SE T1-weighted and breath-hold gradient echo images at 0.5 T were obtained before and after bolus injection of Gd-DTPA. Major peripancreatic vessels were evaluated for the presence of (a) no tumor invasion, (b) contiguity with tumor, and (c) tumor encasement. All patients subsequently underwent surgery. Results of unenhanced and contrast-enhanced MRI studies were compared with the histologic findings in the resected specimens in 34 cases and with the surgical findings in 19 nonresected cases. RESULTS: In six patients, pathologic examination showed the presence of tumor confined to the pancreas with no vascular invasion. With MRI, five of these cases were correctly evaluated; in the remaining case, tumor-vessel contiguity was erroneously diagnosed. In 21 patients, tumor contiguity with adjacent vessels was found at pathologic examination. At MRI, 16 of these cases were correctly assessed, and 5 were understaged as tumors confined to the pancreas. Of the remaining 26 patients, 7 had vascular encasement by tumor at pathologic examination of the resected specimen; the other 19 patients had unresectable tumors encasing the adjacent vessels at surgical evaluation. MRI detected vascular encasement in 21 of these 26 cases; in the other 5, tumor contiguity was incorrectly diagnosed. The overall accuracy of MRI for determining vascular invasion was 79%. CONCLUSION: MRI is a helpful method for preoperative assessment of vascular involvement due to pancreatic carcinoma.
Authors: M Trede; B Rumstadt; K Wendl; J Gaa; K Tesdal; K J Lehmann; H J Meier-Willersen; P Pescatore; J Schmoll Journal: Ann Surg Date: 1997-10 Impact factor: 12.969