3D-modeling Capabilities in Assessing Resectability of Pancreatic Head Tumors

Background . Pancreatic head cancer ranks 6–7th among oncologic diseases and 4–5th among causes of mortality, with only 5% of patients achieving 5-year survival rate to date. Despite the successes of modern diagnostics and surgical treatment, the problem of early detection, staging of oncologic process and, as a consequence, combined treatment of pancreatic head cancer remains actual. Objective : increasing the accuracy of diagnostics and estimation of resectability of the pancreatic head cancer on the basis of the complex use of the radiation methods of investigation with pancreaticoduodenal zone 3D-reconstructions. Material and methods . The study included 93 patients (44 (47.31%) males and 49 (52.69%) females) with complicated pancreatoduodenal masses treated from 2019 to 2022 at the Surgical Department of the City Mariinsky Hospital. The patients’ age varied from 44 to 90 years, the mean age was 67±0.74 years. All patients underwent magnetic resonance imaging (MRI) on an Ingenia MR tomograph (Philips Medical Systems, Netherlands) with a magnetic field induction of 3 Tesla. The native examination of the abdomen and retroperitoneum, supplemented with MR-cholangiopancreatography protocol, and dynamic contrast enhancement with data collection in arterial, portal, and delayed phases were carried out. T2-weighted images were then performed using turbospin-echo technology, including fat-suppressed images, to evaluate structural changes and the presence of fluid in fascial spaces. Patients also underwent endoscopic ultrasound of the pancreaticoduodenal zone using the push and pull method, and abdominal multislice computed tomography (MSCT). To build 3D models, we used free 3D-slicer and Mimics programs, which allowed to build semi-automatic model for further evaluation of anatomo-topographic relations. Results . MSCT 3D modeling revealed tumor invasion into the superior mesenteric vein in 6 (23.06%) patients, whereas MRI models showed tumor invasion in 4 (15.38%) patients, intraoperatively the results were confirmed in 5 patients (19.23%). According to both MSCT and MRI modeling data, the invasion of the ventral trunk occurred in 1 (5.2%) case, which was confirmed intraoperatively. Inferior vena cava invasion on MSCT and MRI models was detected in 3 (11.54%) patients, whereas intraoperatively – in 4 (15.38%) patients. The MSCT and MRI 3D models coincided with the data on invasion of the ventral trunk in 1 (3.85%) patient and the superior mesenteric artery in 2