The accuracy of EUS and helical CT in the assessment of vascular invasion by peripapillary malignancy

Background: The relative accuracy of helical CT and EUS for defining the local resectability of peripapillary malignancies is undefined. Methods: Fifty-one patients with a peripapillary malignancy and no metastatic disease were prospectively valuated with helical CT and EUS. Imaging results were compared with surgical staging, and a tumor was defined as resectable when there was no macroscopic or microscopic residual tumor. Results: Nine patients had surgically confirmed locally unresectable disease, which was accurately predicted by EUS in 6 patients (sensitivity 67%) and by helical CT in 3 patients (sensitivity 33%; p = 0.35). When only patients with complete EUS examinations were included, the sensitivities of EUS and helical CT for vascular invasion were 100% and 33% (p = 0.06), respectively. When all patients not undergoing surgery because of imaging evidence of locally unresectable disease were included, the sensitivities were 100% and 62.5% (p = 0.02), respectively. One of 15 patients with a tumor amenable to surgical resection was labeled as unresectable by EUS but subsequently had a local recurrence of the tumor. The specificities of EUS (93%) and helical CT (100%) were similar. Conclusion: EUS is more sensitive than helical CT for detecting vascular invasion by peripapillary malignancies and should be added to staging protocols, particularly when findings on helical CT are equivocal.