To the Editor.—
Divine and Amanollahi1 argue that the study of Hogstedt and associates2—which confirmed an earlier observation that workers occupationally exposed to ethylene oxide had elevated rates of leukemia mortality—was seriously deficient. They also cite a previous occupational study,3 involving a larger cohort, that found no excess cancer mortality. We address the issue of the carcinogenicity of ethylene oxide by comparing dose-response data from animals4 with human data.2,3Ethylene oxide produced mononuclear cell leukemia in both male and female rats in two separate inhalation bioassays.4,5 Thus, animals and humans of both sexes develop cancer in the same type of tissue after ethylene oxide exposure.
We examined how well the dose-response relationship observed in rats fitted the epidemiologic data from several cohorts.2,3 Our objective was to see if the observed number, of human cancer deaths was in the range of what