Childhood cancer after prenatal exposure to diagnostic X-ray examinations in Britain.

Detailed data were provided by the Oxford Survey of Childhood Cancer OSCC on deaths from childhood cancer in Britain after irradiation of the fetus during diagnostic radiology of the mother. In each age group at death, 0-5, 6-9 and 10-15 years, excess cancer deaths decreased suddenly for births in and after 1958. A major factor was concerted action initiated in 1956 to reduce radiation exposure of fetal gonads for fear of genetic hazards. Dose reduction was achieved during 1957 and early 1958 by reducing the rising rate of obstetric radiography and by virtually abandoning pelvimetry as that had been understood. In the 1970s the rate of X-raying increased again and so did cancer risk but not significantly. Direct evidence that diagnostic X-rays can cause childhood cancer is the similar excess rate per X-ray in twins and singleton births when X-raying rate is 5-6 times higher in twins. In the past a dose-response for cancer in OSCC data based on number of films per X-ray examination was taken to be evidence for causation but dose per film varies with kind of X-ray examination. Fixed values for dose per film were mistakenly assumed by UNSCEAR (1972) and used by it and others when deriving risk co-efficients. In updated OSCC data cancer risk is independent of film number. The odds ratio for childhood cancer deaths after X-raying in birth years 1958-61 (1.23 with 95% confidence intervals CI 1.04-1.48) and the mean fetal whole body dose from obstetric radiography in 1958 (0.6 cGy) can each be derived from nationwide surveys in Britain. The corresponding risk coefficient for irradiation in the third trimester for childhood cancer deaths at ages 0-15 years = 4-5 x 10(-4) per cGy fetal whole body dose (95% CI 0.8-9.5 x 10(-4) per cGy). It is the same for cancer incidence and mortality. A lower risk in bomb survivors exposed in utero is not incompatible since its CI are wide. There is no dependable evidence that radiosensitivity is greater in early pregnancy. A significantly raised cancer rate after diagnostic X-raying supports the hypothesis that carcinogenesis by ionising radiation has no threshold.