Background: The incidence of carcinoma of the pancreas is increasing in Western societies at great cost. Pancreatic resections have been performed for 60 years and there are hundreds of papers written on the subject, but there is no agreement on the efficacy of resections. Authors have reported on different groups and subsets of patients and have used different statistical methods. Most authors are unable to report a single five-year survivor; others claim a five-year survival rate of 30 to 55 percent.
Study design: I have standardized and compared the results reported in 340 papers that deal with survival rates and where there is apparently adequate confirmation of disease.
Results: Survivors who have been resected may be reported up to six times even from different countries, whereas survivors who were not resected are frequently overlooked. Actuarial statistical methods exaggerate results when data are lost. After corrections for repetitions, approximately 300 survivors were found, of whom 10 percent had not undergone resection, of the estimated 80,000 patients reported. The overall survival rate was less than 0.4 percent. The best overall survival rate in surgical studies reported in detail is only 3.6 percent and for a nonsurgical study it was 1.7 percent. The average excess cost for each resection was at least $150,000. With only one in 30 patients who underwent resection living for five years, the cumulative cost per "successful" resection was therefore approximately $4.5 million.
Conclusions: Pancreatic resections have had minimal impact on survival rates in patients with carcinoma and are wasteful of resources.