Cancer Mortality among Nuclear Workers in Belgium

INTRODUCTION Animal experiments and epidemiological data in humans have provided evidence that high doses of ionising radiations can induce cancer (1). However, cancer risk estimates available so far , are based mainly on studies about short term exposure (acute) to high radiation dose, e.g. among radiotherapy patients and atomic bomb survivors (2-7). Cancer risk estimates after chronic exposure to low doses of ionising radiation are, in most cases, obtained from extrapolation to low doses of the data observed after high-dose exposure. Radiation protection regulations and current practices are based on these extrapolations (8). Such types of extrapolation imply of course assumptions, in particular concerning the shape of the initial part of the dose-effect relationship. Therefore, cancer risks after low radiation exposure are difficult to evaluate and current estimates remain controversial. Because it is important to know if the present radiation protection measures are adequate and safe to protect the public as well as the radiation workers, there is a need to measure the cancer risk directly on populations chronically exposed to low doses, such as e.g. radiation workers. The present report is dealing with nuclear workers: this group has been selected because these workers are occupationally exposed to low doses of radiation during several years, and individual data on exposure are available. Most of the studies that have been done in nuclear workers investigate cause specific mortality rates in workers of different dose categories. They include workers of several types of installations such as nuclear research centers, reprocessing plants, power plants, military installations. But they have some common characteristics as well: most of the workers are exposed to low doses of external radiation (X and gamma radiation) and individual yearly dosimetry data have been registered and kept since the years 1940-50. Only a minority of workers has been exposed to neutrons, or has been at considerable risk for internal contamination. When total mortality (all causes of death) or cancer mortality (all types of cancer) are compared for the nuclear workers and the general population, only few studies find a significant higher mortality. No specific cancer type is consistently associated with cumulated dose. On the contrary, many studies describe a significantly lower mortality than in the general population, probably influenced by the so called "healthy worker effect", a selection bias. When looking for association with cumulated dose, some positive trends have been found, some of them significant. Leukemia mortality is most frequently associated with dose (9-13). Concerning internal contamination, most of the studies only flag the workers monitored for internal contamination, but have no individual data. Some research has however been done on particular radionuclides, such as plutonium (14). The many individual studies among nuclear workers done so far often lack statistical power because the effect of low doses of radiation on cancer incidence appears small. Therefore, the International Agency for Research on Cancer (IARC/WHO) in Lyon, France, decided to pool available data. The combined data of several already published studies from Canada, the UK and the USA were analysed to obtain more precise estimates of cancer risk. No association was found between cumulative dose and total cancer mortality. An Excess Relative Risk of 2,2 per sievert cumulated dose for leukemia mortality was observed, but still within a broad confidence interval (0,1-5,7). Therefore it is concluded that so far "there is no evidence that the current radiation protection measures are appreciably in error" (ref. 15-16). Following this combined analysis, IARC set up a multicenter study, coordinated by Dr. E. Cardis, to pool data of 14 different countries. Belgium is one of the participants.