Purpose: To evaluate the accuracy of magnetic resonance angiography (MRA) to predict vascular involvement in pancreas carcinoma.
Materials and methods: This prospective trial included 42 patients with suspected pancreas carcinoma. All patients underwent MRA, different tomographic sequences (MRI) and surgery. Peripancreatic vessels were analyzed for vascular involvement, with the tumor infiltration graded on a scale of 0 to 4 (grade 0 - 4). The imaging findings were correlated with surgery and histopathology.
Results: The sensitivity of MRA to predict vascular involvement was 61 % (19 of 31 evaluated vessels) and the specificity 97 % (148 of 153). Sensitivity and specificity for axial MRI were both 94 % (29 of 31 and 144 of 153 evaluated vessels). The difference in the sensitivities of MRA and MRI was statistically significant (p < 0.01). Vascular involvement was correctly graded in 58 % (18 of 31) by MRA and in 74 % (23 of 31) by MRI.
Conclusion: MRA alone is not sufficient to diagnose vascular involvement in pancreas carcinoma, but provides an adequate overview of the vascular anatomy. A combined protocol of MRA and MRI is needed for an acceptable evaluation of the peripancreatic vessels.