Application Of Multi-slice Spiral CT On Evaluation Of Respectability Of Vascular Invasion Of Pancreatic Carcinoma

Objective The study was conducted to evaluate the vascular invasion degree of pancreatic carcinoma by analyzing the MSCT images of patients with pancreatic carcinoma,to compare the judgement differences of different standards on vascular invasion of pancreatic carcinoma by referring to the operation results,also to compare the evaluations of correctness,sensitivity,specificity and other indicators of the resectability of vascular invasion of pancreatic carcinoma by using different standards, and figure out their consistency with the actual situation of the surgery, and to test the feasibility of the standard made in this study on the evaluation of vascular invasion of pancreatic carcinoma.Materials and Methods(1) Both two senior physicians from the pancreas and spleen professional groups in the radiological department and the general surgery together made an evaluation standard for the resectability of vascular invasion of pancreatic carcinoma as follows:①When there is normal fat space between tumor and SMA;CA and HA are not involved;SMV and PV both are patent and not involved;then this is resectable.②When the contact surface between tumor margin and SMA is no more than 180°(50%or less);the tumor is close to or enveloping a short HA that is less than 2 cm long,without involving CA,and there is possibility to resect and reconstruct blood vessels(with or without vascular graft);SMV or its merging part with PV is involved,and the contact surface between tumor margin and veins is no more than 180°,at the same time,the vascular wall is smooth;or a shorter part of vein contacts with rumor by more than 180°,even is occlusive,but there is upper or lower terminal suitable for vascular anastomosis reconstruction. Above describes the critical state when it is resectable.③SMA contacts with the tumor by more than 180°,or HA and CA blood vessels are enveloped,impossible for vascular reconstruction;SMV and PV are involved,and the contact surface between tumor and vein is no more than 180°,but the vascrlar wall is rough,highly indicating that the vascular wall is invaded;or the contact surface between tumor and vein is more than 180°,even there is occlusion,impossible for vascular reconstruction. This is not resectable.(2) Retrospective study:56 cases diagnosed to have pancreatic carcinoma in our hospital were selected from January 2001 to December 2005.Apart from vascular invasion,there were no other factors that might limit the resectability.They all underwent multi-slice spiral CT enhanced inspection with Siemens Somatom Plus 4 Volume Zoom spiral CT machine before operation.In addition to plain scan,three scan were carried out at the arterial period,the pancreas period and the portal period after enhancement.By reviewing CT images,the vascular invasion degrees of 5 large vessels around the pancreas(arteries including superior mesenteric artery,hepatic artery,abdominal artery;veins including superior mesenteric vein,portal vein) were evaluated in accordance with Loyer classification standard,Lu classification standard and the standard made in this study,to figure out whether the pancreatic carcinoma was resectable or not.There were 280 vessels in total from the 56 cases. Choosing the surgical result as the gold standard,statistics was taken on the accuracy of the evaluation of each vessel;and the correctness, sensitivity,specificity,positive predictive value and negative predictive value of the evaluations of resectability with different standards,as well as the Kappa coefficient were all calculated.(3) Prospective study:From June 2005 to January 2007,21 cases first diagnosed to have pancreatic carcinoma had MSCT inspection with Siemens Sensation Cardiac 64 spiral CT machine in our hospital,and were proved to have pancreatic carcinoma by surgical treatment within a short term.Similarly,three dynamic enhanced scan at the arterial period, the pancreas period and the portal period were carried out after plain scan. In addition to the conventional construction of axial image,thin layer images of 0.75 mm and 0.5 mm thick were reconstructed at the arterial period and the pancreas period.They were taken as the original images to realize VR,MPR and MIP three-dimensional reconstruction to better display the vessels around the pancreas.Before operation,the vascular invasion degrees of 5 large vessels around the pancreas were evaluated in accordance with the standard made in this study,to determine the resectability of pancreatic carcinoma.Furthermore, compared to the actual surgical situation,all the 105 vessels from 21 cases were evaluated.Choosing the surgical result as the gold standard, the correctness,sensitivity,specificity,positive predictive value and negative predictive value of the evaluation with the standard in this study, as well as the Kappa coefficient were all calculatedResults Choosing the surgical result as the gold standard,the retrospective study on the 280 vessels from 56 patients showed that,there were 206 resectable(A-C type) vessels according to Loyer classification standard,including 199 vessels resectable and 7 vessets not resectable; and there were 74 vessels not resectable(D-F type),including 44 vessels not resectable and 30 vessels resectable,with the Kappa coefficient of 0.623 and the correctness,sensitivity,specificity,positive predictive value and negative predictive value of the resectability of tumor being 86.79%,86.27%,86.90%,59.46%and 96.60%respectively.According to Lu classification standard,there were 226 resectable(0-2 grade) vessels,including 218 vessels resectable and 8 vessels not resectable;and there were 54 vessels not resectable,including 43 vessels not resectable and 11 vessels resectable,with the Kappa coefficient of 0.777 and the correctness,sensitivity,specificity,positive predictive value and negative predictive value of the resectability of tumor being 93.21%,84.31%, 95.20%,79.63%and 96.46%respectively.But with the standard made in this study,it was found that there were 232 resectable vessels,including 224 vessels resectable and 8 vessels not resectable;and there were 48 vessels not resectable,including 43 vessels not resectable and 5 vessels resectable,with the Kappa coefficient of 0.841 and the correctness, sensitivity,specificity,positive predictive value and negative predictive value of the resectability of tumor being 95.36%,84.31%,97.82%, 89.58%and 96.55%respectively.In the prospective study on 105 vessels from 21 cases,according to the standard made in this study,there were 86 resectable vessels,including 83 vessels resectable and 3 vessels not resectable;and there were 19 vessels not resectable,including 17 vessels not resectable and 2 vessels resectable,with the Kappa coefficient of 0.841 and the correctness,sensitivity,specificity,positive predictive value and negative predictive value of the resectability of tumor being 95.36%,84.31%,97.82%,89.58%and 96.55%respectively.Conclusion The MSCT preoperative evaluation of the resectability of vascular invasion of pancreatic carcinoma is not a topic within a single subject.It is necessary to combine with the knowledge and technology of doctors from the radiological department and the surgery to make the standard,so as to increase the accuracy of evaluations.The standard made in this study is feasible to evaluate the vascular invasion of pancreatic carcinoma.CT three-dimensional reconstruction technology plays an important supplemental role in the MSCT evaluation of the resectability of vascular invasion of pancreatic carcinoma.It should be applied together with axial images.Because of the difference of the structure and flexibility of arterial and venous wall,the evaluation standard for arterial invasion and venous invasion should thereby be different,in order to increase the accuracy of evaluations.For the arterial invasion,MSCT may show excessive assessment.That is one of the reasons why there is false positive in evaluation of vascular invasion of pancreatic carcinoma. For venous invasion,MSCT may show deficient assessment;and that is also an important reason why there is false positive in evaluation of vascular invasion of pancreatic carcinoma.