[Surgical palliation for pancreatic cancer. The 25-year experience of a single reference centre]

Abstract

Background: In spite of dynamic development of modern diagnostic and therapeutic methods, the long-term results of surgical therapy in pancreatic cancer are still unsatisfying. The aim of this study was to analyse long-term results of surgical palliation for pancreatic cancer in a pancreatic surgery centre.

Methods: We performed a retrospective analysis of 418 patients who underwent non-resective, palliative procedures for pancreatic cancer between 1975 and 1999. In order to compare two consecutive periods of time, the patients were divided in 2 groups; group I treated from 1975 to 1990 (n = 204), and group II from 1991 to 1999 (n = 214).

Results: Of all patients qualified for surgery, 281 (67.2 %) underwent surgical bypass, 107 (25.6 %) laparotomy, and in 30 cases surgical intervention was limited to implantation of endoprosthesis. A significant tendency towards double (i. e. biliary and gastric) anastomosis was observed (32.3 % vs. 74.8 %; p < 0.01) in patients who underwent bypass procedures. The postoperative morbidity was 16.3 %. The postoperative mortality rate was 5.7 % and significantly (p < 0.01) decreased from 10.3 % (group I) to 1.4 % (group II). No differences neither in mortality nor morbidity related to the type of performed surgery were found. The mean time of hospital stay was 15.5 +/- 6.9 days and showed no differences related to the type of intervention. Jaundice or symptoms of gastric outlet obstruction were observed in 16 % of patients in the follow-up period and concomitantly performed biliary and gastric bypasses were associated with the lowest rate of the late gastrointestinal obstruction (4 %). The median survival time was 169 days and only 4 % of patients survived 12 months. The univariate analysis of prognostic factors showed that location and stage of the tumour, the type of surgical intervention and bypass procedure influenced 1-year survival. The multivariate analysis using Cox proportional hazard model proved that only stage and location of the tumour had independent prognostic value.

Conclusion: Surgical palliation for pancreatic cancer can be performed with acceptable morbidity and mortality rates. For tumours located in the head and body of the pancreas combined biliary and gastric bypass should be preferred. For cancers located in the tail of the pancreas gastric bypass should be performed routinely. Because surgical palliation can prevent gastric outlet obstruction by gastroenterostomy, endoscopic biliary stenting should be only performed in patients with pancreatic head cancers and simultaneous evidence of distal metastases as well as in older patients with high comorbidity.