Preoperative staging and assessment of resectability of pancreatic cancer.

To study the accuracy of preoperative staging techniques for assessing resectability of pancreatic and ampullary adenocarcinoma, we entered 88 consecutive candidates into a prospective study of contrast-enhanced computed tomography, magnetic resonance imaging, angiography, and laparoscopy. Resectability was proved in 16 (29%) of 55 patients for the head of the pancreas, 1 (6%) of 17 for the body and tail of the pancreas, and 14 (88%) of 16 for the ampulla. The combined findings of computed tomography and angiography showed that more than 87% of pancreatic head tumors were unresectable because of vascular encasement, but neither modality sufficed alone. Small liver and peritoneal metastases were found in 15 (27%) of 55 cancers of the head of the pancreas, 11 (65%) of 17 cancers of the body and tail of the pancreas, and 1 (6%) of 16 cancers of the ampulla; computed tomography missed all but 2 of these instances of metastasis, but laparoscopy with biopsy identified 22 (96%) of 23 instances. Magnetic resonance imaging findings did not differ significantly from computed tomography and conferred no added benefit. Ninety percent of unresectable tumors were identified. Seventy-eight percent of pancreatic head cancers were resectable when all test results were negative vs 5% (2/37) when any test result was positive. This study demonstrates that accurate and efficient triage is possible for patients with cancer of the pancreas and ampulla.