No abstract

The prognosis of patients who undergo resection for pancreatic ductal adenocarcinoma with curative intention is generally poor unless they have early-stage disease. Based on our 25-year experience, the results of 194 patients after a standardized Kausch-Whipple resection for adenocarcinoma of the pancreatic head were analyzed and the prognostic factors were evaluated. Between 1972 and 1998 a total of 221 patients were diagnosed for ductal adenocarcinoma of the pancreatic head, and 194 of them subsequently underwent a standardized Kausch-Whipple resection. Long-term results and prognostic factors were examined by multivariate and univariate analyses. The overall postoperative mortality was 3.09%, and the morbidity was 29.9%. By multivariate analysis only curative resection (R0) was significantly related to a favorable prognosis ( p < 0.0001). Furthermore, in case of a curative resection, the presence of lymph node metastases showed prognostic significance in the multivariate analysis ( p = 0.005). Cumulative survival analysis revealed a 5-year survival rate of 25.4%, a 7-year survival rate of 12.3%, and a 10-year survival rate of 8.2% for patients who underwent curative resection (R0) for adenocarcinoma of the pancreatic head. We demonstrated that the R0 status is the only independent prognostic factor after surgery for adenocarcinoma of the pancreatic head. In the case of a curative resection, the presence of lymph node metastases is of prognostic relevance. In view of considerable surgical morbidity and mortality, resection for cancer of the pancreatic head is the only option if the lesion is resectable. We concluded that surgical treatment is "as good as it gets," as extended techniques have not proved to produce better results.

ObjectiveThe authors reviewed the hemorrhagic complications of patients who underwent pancreatoduodenectomies between 1972 and 1996. Summary Background DataAlthough recent studies have demonstrated a reduction in the mortality of pancreatic resection, morbidity is still high. Bleeding is a close second to anastomotic dehiscence in the list of dangerous postoperative complications. MethodsThe medical records from a prospective data bank of 559 patients who underwent pancreatic resection at the Surgical Clinic of Mannheim (Heidelberg University) were analyzed in regard to postoperative hemorrhagic complications. Differences were evaluated with the Fisher exact test.

This paper analyses the early postoperative complications after 285 pancreaticoduodenectomies performed during the past 15 years in the Surgical University Clinic, Mannheim. There were 235 partial (Whipple) and 52 total pancreatectomies performed for pancreatic and periampullary tumors (181 patients) and complicated chronic pancreatitis (104 patients). A total of 92 complications requiring relaparotomy in 42 patients ended fatally in nine patients. The overall operative and hospital mortality rate was 3.1%. The most frequent and most dangerous were complications at or around the pancreaticojejunal anastomosis, which occurred 25 times with five deaths. Postoperative hemorrhage was seen in 16 patients; endoscopic treatment in four patients and operation in 12 patients was successful in stopping the bleeding in all but one patient. Eight biliary fistulae either ceased spontaneously (3 patients) or after operative reintervention (5 patients) without any mortality. Control of these complications depends on four lines of approach: (1) before operation: optimal preparation of the jaundiced patient including endoscopic transpapillary decompression of the common duct; (2) during operation: a meticulous and standardized technique is mandatory; (3) after operation: continuous observation in the surgical intensive care unit is essential for the timely detection of possible complications; and (4) early reintervention can salvage the great majority of these patients with deleterious complications.

Twenty-one years ago, Howard published a paper entitled "Forty-one Consecutive Whipple Resections Without an Operative Mortality." That paper stimulated the present analysis of the last 118 consecutive pancreatoduodenectomies (107 Whipple and 11 total resections) performed at the Surgical University Clinic Mannheim from November 1985 to the present day with no deaths. Ninety-one resections were performed for neoplasms and 27 were for complicated chronic pancreatitis. The preoperative evaluation, operative technique, and postoperative care of these cases is discussed in detail and compared to the experience of Howard. While there was general agreement on operative technique, there were differences concerning preoperative evaluation (modern imaging methods) and postoperative care (simplification). In this series 21 postoperative complications required seven relaparotomies. Long-term survival after resection for carcinoma was analyzed for 133 consecutive patients who were shown to have true ductal adenocarcinoma. In 76 patients, who had radical (R0-) resections, the actuarial 5-year-survival rate was 36%. In 44 patients, whose R0-resections for pancreatic cancer occurred more than 5 years ago, the actual survival rate was 25%.

There is no reason to exclude patients with suspected venous infiltration from radical pancreaticoduodenectomy including venous resection.

Significant complications continue to occur in a minority of patients undergoing pancreaticoduodenectomy; these frequently have fatal consequences. Analysis of 458 patients undergoing the Whipple procedure in this institution from 1972 to 1994 revealed that 16 patients with malignant periampullary tumours (ampullary, eight; pancreatic, seven; bile duct, one), and one patient with chronic pancreatitis subsequently required completion pancreatectomy. Postoperative difficulties after pancreaticoduodenectomy and indications for re-exploration were multifactorial: leakage (n = 8), pancreatitis (n = 7), bleeding (n = 1), and a delayed report of cancer at the margin of the pancreatic transection (n = 1). Completion pancreatectomy was often difficult (mean operating time 2.7 h, mean estimated blood loss 1897 ml). There was considerable significant postoperative morbidity (41 per cent) and mortality (24 per cent) after completion pancreatectomy. Patients who survived completion pancreatectomy lived a mean of nearly 4 years (range 4 months to 9.7 years, median 2.6 years). Tumour recurrence led to death in ten of 13 patients. Three patients remain alive and free of recurrence each more than 8 years after resection. Re-exploration and subsequent completion pancreatectomy after pancreaticoduodenectomy is rarely necessary, but if clinical manifestations occur secondary to failure of the pancreaticojejunostomy, early surgical intervention may maximize survival.

Adenocarcinoma of the pancreatic body and tail often presents late and is widely regarded as incurable by surgical resection; long-term survivors are rare. Thirteen patients underwent left resection (n = 7) or total pancreatectomy (n = 6) in a consecutive series of 105 patients with carcinoma of the body or tail of the pancreas. Comparison was made with 17 patients with locally advanced or metastatic disease. Preoperative computed tomography predicted irresectable disease when a large perivascular lymph node mass was demonstrated. Preoperative angiography predicted irresectable disease when there was encasement or obliteration of the coeliac axis or its major branches, or of the superior mesenteric artery or vein. Splenic vessel involvement was sometimes compatible with resection. After resection, median survival was 13 (range 3-50) months, with a minimum follow-up of 2 years. Five patients survived more than 2 years, and three are still alive 30, 43 and 50 months after resection. Resection of carcinoma of the body or tail of the pancreas was possible in 12 per cent of patients and long-term survival was observed in some of these.