Assessment of resectability in carcinoma pancreas using multi-detector computed tomography with surgical correlation

Abstract

Background: Pancreatic cancers are considered a dreaded disease owing to the grave tumour biology and poor overall survival. The objective of this study was prospectively evaluating the ability of multi detector computed tomography (MDCT) to predict resectability of pancreatic cancer on the basis of surgical correlation. Methods: 72 consecutive patients admitted to the departments of General Surgery and Surgical Gastroenterology with pancreatic cancer underwent preoperative contrast enhanced triple phase 128 slice MDCT- pancreatic protocol. Out of this 31 patients deemed resectable and underwent surgery. The operative resectability of tumour in terms of vascular invasion, local spread and abdominal metastasis was assessed. Surgical findings were reviewed and compared with pre-operative MDCT results. Results: Of 31 tumours, considered to be resectable at MDCT evaluation, 25 were completely resected and six were found to be unresectable at surgery, yielding a positive predictive value of 80.6% with six false-negative results for overall resectability. None of 31 evaluated patients showed any evidence of vascular invasion, local spread or abdominal metastasis at MDCT examination. However out of the six inoperable patients, three had vascular invasion, three had evidence of local spread and one had small liver metastasis detected during surgery. Thus MDCT yielded a negative predictive value of 90.3% (28/31) for identifying the vascular invasion as well as for detection of local spread and a negative predictive value of 96.8% (30/31) at detecting abdominal metastases. Conclusions: MDCT is an effective pre-operative tool for assessing resectability with a good positive predictive value for overall resectability in pancreatic adenocarcinoma.