Other factors in leukemia.

We appreciate Dr. Aksoy's response to our paper and the specific data it provides regarding the methodology used in his studies to calculate leukemia incidence and to estimate benzene concentrations to which his cohort of Turkish shoe, slipper, and handbag workers was exposed. It is without question that Dr. Aksoy's early epidemiologic observations of the association between benzene exposure and leukemia are to be commended. The major conclusion of our review, however, was that the Aksoy studies, as well as some of the other available benzene epidemiologic studies, are not optimal for risk assessment, due primarily to limitations in quantitative exposure information. Given that individual exposure history data were not apparently available for workers in the Aksoy study, it remains our judgment that it provides a less useful basis for the estimation of a dose-response relationship than the Rinsky et al. study that we selected.