Preoperative evaluation of the resectability of pancreatic cancer with peripancreatic vascular invasion by multi-slice computed tomography

TANG Ping,HUA Yan-qing,WU Wei-lan,WANG Wei,YIN Yu-lei,ZHANG Guo-zhen
2009-01-01
Abstract:Objective To evaluate the MSCT manifestations and the resectability of pancreatic cancer with peripancreatic vascular invasion preoperatively. Methods The preoperative MSCT imagings of sixty-six cases of patients with pancreatic duct cell carcinoma confirmed by surgical pathology were studied. Based on correlation of pancreatic mass with peripancreatic vessels in MSCT, five grades were defined for the circumferential contiguity, and for vascular deformation or stenosis; two grades for the involved length of the vascular long axis. Five criteria were synthesized on the mentioned grades for detecting peripancreatic vascular invasion and predicting tumor resectability. Compared with surgical and pathologic findings, the sensitivity, specificity, accuracy, false negativity, false positivity, positive predictive value and negative predictive value of each criterion were calculated respectively. The differences between these five criteria were analyzed statistically. Results 44 neoplasms were located in the head of the pancreas, 22 involved the body and the tail of the pancreas. Whipple’s operation was performed in30 cases, pancreatic body and tail was resected in 17 cased, internal drainage and laparotomy were performed in 17 cases and 2 cases respectively. 34 cases were diagnosed as free of tumor invasion to peripancreatic vessels, and 32 cases as invaded by tumor. Compared with surgical and pathologic findings, the sensitivity, specificity, accuracy, false negativity, false positivity, positive predictive value and negative predictive value for predicting vascular invasion and tumor resectability respective were 90.63%, 94.12%, 92.42%, 9.40%, 5.88%, 93.55%, 91.43% in the first group. For the second group, third group and fifth group, the rates of the false negativity or false positivity were higher than the first group. The sensitivity and the accuracy of the fourth group were greatly lower than the first group. Conclusion In consideration of the tumor contiguity to the circumference of vessels, the degree of vascular deformation and stenosis, and the involved length of the vascular long axis on MSCT, the assessment accuracy of vascular invasion and respectability may be increased in clinic practice.
What problem does this paper attempt to address?