The Predictive Value of CT Findings in Invasive Behavior of Pancreatic Neuroendocrine Tumors

Abstract

Objective: To explore the value of CT findings in predicting the invasive behaviors of pancreatic neuroendocrine tumor (pNET). Methods: The clinical data and CT data of 120 patients with pNET confirmed by surgical resection and pathology were retrospectively analyzed. Preoperative CT plain scan and enhanced examination of pancreas were performed. Image analysis included tumor location, shape, capsule, cystic change, calcification, pancreatic duct dilation and enhancement pattern. The maximum and minimum diameter of the tumor were measured. The CT value of the solid part of the tumor was measured in plain scan, arterial phase and portal vein phase. Enhancement difference in arterial phase and portal vein phase were calculated. The tumors were divided into invasive group and non-invasive group according to the pathological results. The difference of CT findings between the two groups was compared. Receivers operating characteristic (ROC) curves were drawn to evaluate the predictive value of tumor maximum diameter, minimum diameter, enhancement difference in arterial phase and portal vein phase on the invasive behavior of pNET. Results: There were no statistical differences in tumor location, pancreatic duct dilation, cystic change, calcification and enhancement pattern between the invasive and non-invasive groups. There were statistically significant differences between the two groups in the quantitative characteristics of the maximum diameter, the minimum diameter, enhancement difference in arterial phase and portal vein phase and the areas under ROC curve were 0.693, 0.69, 0.73 and 0.64, indicating discrimination efficiency. Multivariate Logistic regression analysis of 6 meaningful image features showed that arterial enhancement difference was an independent predictor of pNET invasive behavior, and the optimal critical point of arterial enhancement difference was 90.1HU (sensitivity 0.714, specificity 0.656, positive predictive value 64.5%, negative predictive value 72.4%, accuracy 68.3%). Conclusion: Large volume, irregular shape, incomplete or no capsule of pancreatic neuroendocrine tumor, and low enhancement difference of solid tumor components in arterial and portal phases suggested invasive behavior of tumor, and enhancement difference in arterial phase was an independent predictor of invasive behavior of pNET.

Keywords

• tomography
• x-ray computer
• pancreatic neoplasms
• neuroendocrine tumor