Pancreatic carcinoma in Malmö, Sweden: Clinical, epidemiological and biochemical aspects

The objective of this thesis has been to assess whether there are indications suggesting that new methods for diagnosis and treatment of pancreatic carcinoma (PCa) has been followed by an improved prognosis, and to assess the possible role of smoking and gastric resection in the causation of the disease. Of the 1,200 new cases PCa diagnosed in Sweden yearly, the majority are older than 70 years, the case fatality rate being close to 100%. No obvious improvement is noted in prognosis. Case sur- veillence in Malmö is facilitated as one hospital only deals with somatic diseases. Smoking is a risk factor, but dietary habits may play a role, and conditions with increased cholecystokinin (CCK) concentration are of certain interest, as CCK exert a pronounced trophic effect on the gland. The proportion of cases offered attempted radical surgery in Malmö, increased between 1977-1984 and 1985-1991, from 2 to 6%. Following surgery the mean survival time was below one year, and five-year survival rate did not improve between periods (Paper I). Between 1961 and 1990, 1 314 cases of pancreatic carcinoma in Malmö were retrieved from the Regional Tumour Registry and the National Cause of Death Registry. No increase in incidence was found in men. In women, middle-aged and older, there was a relative increase in the incidence rate of about 2% yearly (Paper II). The position of smoking as a risk factor was strengthened by a follow-up of 35,000 Malmö citizens. The incidences of PCa in never-smokers, current smokers and ex-smokers were 1.5, 15.3 and 24.5/100000, respectively. Weight gain of more than 10kg after the age of 30, and a history of epigastric pain were associated with a further increase of the risk (Paper III). A necropsy based study of 21 660 cases left no support for the hypothesis that gastric resection would be associated with an increased risk of PCa. The prevalence of gastric resections in cases dying of pancreatic carcinoma, matched controls, and cases dying of lung cancer, were 3.4%, 4.4%, and 7.6%, respectively (Paper IV). A case-control study, comparing gastric resected and non-resected subjects, did no show any difference, neither in plasma concentrations of CCK, nor in plasma concentrations of pancreas-specific enzymes. Plasma levels of pancreatic enzymes, but not CCK was higher in smokers than in non-smokers (Paper V)