Premium
Cross‐classification and survival characteristics of 5,000 cases of cancer of the pancreas
Author(s) -
Baylor Stephen M.,
Berg John W.
Publication year - 1973
Publication title -
journal of surgical oncology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.201
H-Index - 111
eISSN - 1096-9098
pISSN - 0022-4790
DOI - 10.1002/jso.2930050410
Subject(s) - medicine , pancreas , adenocarcinoma , histology , relative survival , cancer , carcinoma , pancreatic cancer , stage (stratigraphy) , survival rate , gastroenterology , survival analysis , metaplasia , pathology , oncology , cancer registry , paleontology , biology
Over 5,000 histologically confirmed cases of cancer of the pancreas from the End Results Group of cancer registries have been examined for differences in behavior by epidemiologic and pathologic variables. Thirteen percent had a specific histologic diagnosis other than “adenocarcinoma,” “carcinoma,” or “cancer.” The 92 cases of islet cell carcinoma were unique, having a much younger age at diagnosis (two‐thirds instead of one‐third diagnosed under the age of 60), a locational preference for the body and/or tail (only 40% instead of 70% occurring in the head of the pancreas), and a very favorable survival experience (15 year relative survival of 14%). Of cancers of the exocrine pancreas, the 64 cases with papillary histology were most distinct, two‐thirds being found in women, 18% instead of 10% being localized, and having relatively favorable survival (30% one‐year survival, 5% three‐year survival). Patients whose tumors revealed squamous metaplasia occurred in general at older ages and had unusually poor survival. Analysis by anatomic location showed an increasing percentage of carcinomas occurring in the head of the pancreas with increasing age. Analysis of “adenocarcinomas” by histologic grade indicated that grade may carry some prognostic information not explained by difference in stage. A striking gradient of decreasing survival with increasing age did not explain, and was not explained by survival gradients in other variables.