[Evidence based surgery of cancer of head of pancreas]

Physicians and surgeons who treat patients with gastrointestinal or hepatic disease must prescribe the most appropriate diagnostic tests, together with an accurate prognosis and effective and safe therapy. This paper examines the best modalities of surgical treatment for cancer of the pancreas, in an evidence-based approach. Evidence was classified as follows: Grade A : evidence from large randomized controlled trials (RCT) or systematic reviews (including meta-analyses) of multiple randomized trials which collectively have at least as much data as one single well-defined trial. Grade B: evidence from at least one high-quality study of non-randomized cohorts or evidence from at least one high-quality case-control study or one high-quality case series. Grade C: opinions from experts without references or access to any of the foregoing The data were obtained from Medline and from controlled randomized trials listed in the Cochrane Library up to the end of 2003. Two series (grade B) showed the superiority of Whipple over total pancreatectomy, with respective median survival times of 12.6 months and 9.6 months. Extensive lymphadenectomy (grade A) in patients with positive lymph nodes gave significantly better survival than standard resection in one trial, but this was not confirmed in the other trial. Results of pylorus-preserving pancreaticoduodenectomy (PPPD) were not different from those of the Whipple procedure on postoperative mortality, morbidity or survival (grade A). Portal vein resection increased the resectability rate. Post-operative mortality was not increased: survival was not different in four studies and was shorter in another four studies (grade C). Low-dose postoperative erythromycin accelerates gastric emptying if the right gastric artery is preserved (grade A). One trial suggests that pancreaticogastrostomy reduces the risk of pancreatic fistula. The two other trials are controversial and showed no difference. One prospective non randomized study showed that stenting in pancreaticojejunostomy reduces the risk of pancreatic fistulae and intraabdominal abscess. To prevent this risk of pancreatic fistula, six controlled trials involving patients receiving octreotride were performed Three European trials showed a smaller volume of abdominal drainage fluid and an abnormal amylase concentration; however, two American trials failed to demonstrate a significant difference. Occlusion of the pancreatic duct with fibrin glue did not reduce the risk of pancreatic fistula, but increased the risk of developing diabetes. Intraabdominal drainage after pancreatic resection significantly increased post-operative complications (grade A). Surgical resection and reconstruction procedures for pancreatic cancer must be based on evidence-based studies. However, the most important prognostic factor is the surgeon's experience, not only with regard to the post-operative course, but also survival. Specific teaching and training is thus essential.