RESUMO
AIM OF THE STUDY: This prospective study was undertaken to evaluate the accuracy of ultrasonography combined with colour Doppler and endoscopic ultrasonography for predicting superior mesenteric and portal vein involvement in pancreatic diseases. MATERIAL AND METHODS: The study was prospective. Forty-four patients were included. Ultrasonography with colour Doppler was performed in 30 patients, endoscopic ultrasonography in 43. Prediction of superior mesenteric vein or portal vein involvement was blindly assessed by physicians without knowledge of results of other imaging methods. Resectability or potential resectability (n = 34) and irresectability (n = 10) were assessed in all patients by surgery. Thirty patients underwent a pancreaticoduodenectomy. RESULTS: For endoscopic ultrasonography the sensitivity (0.90), the specificity (0.88), the positive predictive value (0.69) and the negative predictive value (0.97) were better than those observed with ultrasonography and Doppler (050, 0.88, 0.69, 0.97, respectively). The specificity of computed tomography (0.96) was better than that of endoscopic ultrasonography but predictive negative values were similar. CONCLUSIONS: Imaging methods to predict superior mesenteric or portal vein involvement in pancreatic diseases are becoming increasingly numerous, complex, and expensive. Endoscopic ultrasonography has a better diagnostic value for correctly predicting resectability than ultrasonography with Doppler. However, for decision making, usefulness of these methods seems to be limited.