[Evaluation of CT diagnostic criteria for peri-pancreatic artery and vein invasion in pancreatic carcinoma].

Unlabelled: OBJECTIVE; To evaluate the clinical value of different CT diagnostic criteria for peripancreatic artery and vein invasion in pancreatic carcinoma through comparison with the findings on surgical exploration.

Methods: Of 72 patients of having suspected pancreatic carcinoma were examined by multiplane spiral CT. Among 43 confirmed by surgical pathology; 15 underwent pancreaticoduodenectomy; 28 were found to have unresectable tumors. The peri-pancreatic major vessels including the superior mesenteric artery, celiac artery, hepatic artery, superior mesenteric vein and portal vein were explored carefully during surgical exploration.

Results: The criteria for peri-pancreatic artery invasion was the presence of one of the following signs: artery embeded in tumor, or more than half of the artery circumference involved by tumor with wall irregularity or stenosis. The sensitivity of the above described criteria was 75.0% (12/16). If the criteria of tumor involvement exceeding half of the vessel circumference were adhered to, the sensitivity was 87.5% (14/16), which was high than the former, but the specificity was lower than that of the former one (90.2% versus 95.1%). The criteria for peri-pancreatic vein invasion was presence of any of the following signs: vein obliteration, more than half of the vein circumference involved by tumor, vein wall irregularity, vein stenosis, tear-drop sign of superior mesenteric artery. The sensitivity of the above described criteria was 92.9% (39/42), higher than that of the criteria that more than half of the vessel circumference was involved by the tumor (69.0%, 29/42), but the specificity of both criteria was the same (97.4%, 37/38).

Conclusion: For assessing peri-pancreatic artery and vein invasion, using the combination of different CT diagnostic criteria has higher accuracy than when using only criteria of more than half of vessel circumference involved by tumor.