PANCREATIC-CARCINOMA FOLLOWING GASTRIC RESECTION - A CASE-CONTROL STUDY BASED ON 21,660 CONSECUTIVE CLINICAL NECROPSIES AT MALMO-UNIVERSITY-HOSPITAL

Conclusion. In this necropsy-based case-control study, there was no re lationship between pancreatic carcinoma and previous gastric resection , Based on the association between lung cancer and gastric resection, it is suggested that the relationship between pancreatic carcinoma and gastric resection shown in other studies may have been confounded by smoking, Background. This case-control study was designed to assess wh ether in patients dying from pancreatic carcinoma, there is a relation ship to previous gastric resection for peptic ulcer disease. Methods. By linking the autopsy data base in Malmo with the national Cause of D eath Register, we identified 439 autopsied individuals who had died of pancreatic carcinoma between 1970 and 1982. The 21,660 individuals in the data base represent 64% of all deaths during that time period. Fo r each of these individuals who died of pancreatic carcinoma, we rando mly chose three controls who were matched for age at death, gender, an d year of death. In order to assess the specificity of the assumed rel ationship, we also used as a control group the 1337 autopsied individu als who had died of lung cancer. Results. The prevalence of previous g astric resections was 3.4% in patients dying from pancreatic carcinoma , 7.6% in patients dying from lung cancer, and 4.4% in the age-and sex -matched control group. The odds for previous gastric resection in pat ients dying from pancreatic carcinoma was 0.9 [95% confidence interval (CI) 0.5-1.7] in comparison with this age-and sex-matched control gro up. The lower odds for previous gastric resection in patients dying of pancreatic carcinoma than in patients dying of lung cancer remained i n the logistic regression analysis after controlling for age at death, gender, and year of death (odds ratio [OR] 0.5; 95% CI 0.3-0.9).