The investigation of a geographical cluster of bladder cancer

Abstract

Previous analyses of the Limburg Cancer registry (LIKAR) indicated the existence of a geographical cluster of bladder cancer incidence, particularly transitional cell carcinomas, amongst males in the surrounding area of the Belgian cities Hasselt and Alken. In subsequent ecologic analyses no risk factors were identified which could explain the existence of this cluster. Therefore, an epidemiological case-control study has been performed in the province of Limburg to explore the possible determinants of the cluster. The following cluster determinants were investigated: socio-demographic characteristics, life style factors, occupational and environmental exposures, and genetic predisposition. A weight was assigned to all the study subjects. Secondly, we used unconditional logistic regression to calculate crude odd ratios (ORs) and 95% confidence intervals (CIs) for each source of exposure and bladder cancer. Thirdly, we used a multivariate logistic regression analysis which included all the parameters found to be significant at the 0.1 significance level in the univariate analysis. Although some of the specific odds ratios decreased compared with the crude results, the overall excess did not change specifically. In conclusion, we were not able to identify a clear-cut explanation for the existence of the geographical cluster. We therefore recommend an increased surveillance of bladder cancer incidence for 3 to 5 years. If the "excess" of cancer cases has occurred due to random variation, the rate will return to the "expected" range in the next years. If the rate remains elevated, further studies may need to be performed using investigative tools that are not available currently.