Familial Associations of Colon and Rectal Cancers With Other Cancers

BACKGROUND: Many studies have indicated that colon and rectal cancers differ in etiology and histology. OBJECTIVE: The aim of this study was to investigate whether the associations of colon and rectal cancers with any other (discordant) cancer were site specific. DESIGN: A novel approach was implemented in which cancer risks were analyzed in families with increasing numbers of family members diagnosed with defined cancers. The novel assumption was that, for a true familial association, the risk should increase by the number of affected family members. In separate analyses, familial risks were calculated after the exclusion of putative families with hereditary nonpolyposis colorectal cancer. SETTINGS: The study was conducted using the Swedish Family-Cancer Database. MAIN OUTCOME MEASURES: The outcome measure was relative risk. RESULTS: Relative risks of colorectal cancer and colon cancer were higher when family members were diagnosed with colon cancer than when family members were diagnosed with rectal cancer (incidence rate ratio for colorectal: 1.82 (95% Cl, 1.74-1.90) vs 1.61 (95% Cl, 1.51-1.71); incidence rate ratio for colon: 1.92 (95% Cl, 1.83-2.02) vs 1.56 (95% CI, 1.45-1.69)). Relative risks for 10 discordant cancers were increased in colon or rectal cancer families, whereas none of the relative risks differed significantly between colon and rectal cancers. After deleting hereditary nonpolyposis colorectal cancer families, the relative risks of endometrial and ovarian cancers were no longer significant. LIMITATIONS: Genetic data are unavailable in the database. CONCLUSIONS: Our results suggested that familial risks for colon cancer were higher than risks for rectal cancer in families of patients with colorectal cancer and colon cancer. The relationships of lung cancer and nervous system cancer with colorectal cancer were site specific. The associations of colon and rectal cancers with lung cancer, myeloma, and cancer of unknown primary appeared not to point out known syndromes and may suggest involvement of a novel predisposition.