Colorectal cancer incidence among atomic bomb survivors, 1950-80.

Colorectal cancer incidence in the LSS sample during 1950-80 was investigated. A total of 730 incidence cases of colorectal cancer were confirmed from a variety of sources. Sixty-two percent of the cancers were microscopically verified and 12% were ascertained through death certificate only. The risk of colon cancer increased significantly with intestinal dose, but no definite increase of risk was observed for rectal cancer. Relative risk at 1 Sv and excess risk per 10(4) PY-Sv for colon cancer are 1.80 (90% confidence internal 1.37-2.36) and 0.36 (90% confidence interval 0.06-0.77) respectively. City and sex did not significantly modify the dose-response of colon cancer, but the risk decreased with age at the time of bombings (ATB). The relative risk of colon cancer does not vary substantially over time following exposure. A non-linear dose response did not significantly improve the fit. Further, the anatomic location of the tumors indicate that the cecum and ascending, transverse and descending, and sigmoid colon seem equally sensitive to radiation. No difference in the distribution of tumor histological types could be observed by radiation dose.