CANCER OF THE EXOCRINE PANCREAS - A PLEA FOR RESECTION

Background/Aims: The aim of this retrospective multicentric study was to compare the results of resections with those of surgical palliative procedures. Material and Methods: The 3231 patients included had hist ologically pl oven adenocarcinoma of the pancreas and were operated on between 1982 and 1988. Seven hundred eighty-seven underwent surgical resection, and 2444 a palliative procedure. Step by step logistic regr ession was used to determine variables having the greatest impact on p ost operative mortality. Survivals were compared with the logRank test . A semi parametric Cox model was applied to estimate adjusted relativ e risk of death. Results: After resection and bypass, postoperative mo rtality was 10% and 15% (P<0.001), morbidity 35% and 27% (P<0.001), an d mean survival times 19.5 months (SE=1.1), and 8.8 months (SE=0.3) (P <0.001) respectively. In patients without metastases, survival was bet ter after resection than. after bypass, even in patients with involvem ent of lymph nodes adjacent to or distant from the tumor (P=0.001). Co nclusions: In spite of the retrospective nature of such comparisons, t hese results acredit the idea that in. patients with pancreatic cancer without metastases, resection should be attempted whenever possible.