Benzene exposure and multiple myeloma: a detailed meta-analysis of benzene cohort studies.

Case reports and epidemiological studies of workers exposed to benzene have demonstrated associations with a number of lymphohematopoietic diseases, but the association with multiple myeloma (MM) has been less apparent. Data from all of the "benzene cohort studies" conducted to date have been selected and evaluated for inclusion in a meta-analysis. The analysis demonstrates a significant excess in the relative risk (RR) of MM in relation to benzene exposure. Pooling the data from seven cohort studies, a meta-analysis yields a statistically significant weighted RR estimate of 2.13 (95% CI = 1.31-3.46). In the analysis of cohort data, an understanding of the cohort follow-up period in relation to benzene exposure and RR of MM is important. Exposure-related RRs of disease decline after the median latency periods are exceeded, particularly when exposure has terminated decades earlier. The positive epidemiological evidence for benzene as a cause of MM is supported by biological plausibility for such an effect from benzene exposure. Studies of refinery workers are difficult to interpret in relation to benzene exposure and risk of MM, but are limited in the study design and analysis. Nonetheless, they provide some support for an association between refinery work and MM.