Carcinogenesis--a synopsis of human experience with external exposure in medicine.

Studies in the 1980s of medically irradiated populations have increased our knowledge of radiation carcinogenesis. (1) Investigations of prenatal x-ray exposures, especially in twins, provide evidence that very low doses of ionizing radiation may cause cancer in humans. (2) Fractionated doses appear as effective as single exposures of the same total dose in causing breast cancer, but seem less effective for lung cancer. (3) Excess breast cancers can occur among women exposed under age 10, indicating that the immature breast is susceptible to the carcinogenic action of radiation. (4) Moderate doses on the order of 1 Gy to the brains of children can cause tumors later in life; moderately high doses to the skin can cause cancer when followed by frequent exposure to ultraviolet light. (5) Radiotherapy for cervical cancer can increase the rate of subsequent leukemia with the best fitting dose-response functions including a negative exponential term to account for cell-killing. (6) Low-dose exposures of about 10 cGy may increase the risk of thyroid cancer. (7) Second cancers following radiotherapy for a variety of cancers occur primarily among long-term survivors. (8) Radiotherapy may not significantly increase the risk of leukemia following childhood cancer, whereas chemotherapy with alkylating agents is a major risk factor. (9) Bone cancer occurs after high-dose radiotherapy for childhood cancer, but children with retinoblastoma are not more susceptible to radiation-induced disease than children with other malignancies. (10) High-dose external beam therapy can cause thyroid cancer, whereas high-dose radioactive 131I may not. (11) Studies of cervical cancer patients indicate that the risk of radiation-induced second malignancies follows a time-response model consistent with a constant multiplication of the underlying background incidence, i.e. a relative risk model seems to hold for projecting risks forward in time.