Benzene As A Cause Of Hematological Cancers: A Recurring Example Of The Importance Of Mechanistic Understanding And Animal Studies

INFANTE P. ²

IARC can be justly proud of decision processes that combine human, animal, and mechanistic information into carcinogen classification. Benzene provides an example of persistent problems. The original identification of benzene as a known cause of acute myelogenous leukemia (AML) was inappropriately delayed until the 1977 cohort study of Infante et al. (IARC 29, 1982) This occurred decades after causality was accepted by clinicians, which was based upon the collected cases of benzene-associated AML among workers around the world and in different occupations; AML occurring often among those with benzene-induced aplastic anemia; and that similarly to ionizing radiation, a known leukemogen, benzene produced chromosomal abnormalities. Authoritative bodies refused to accept the epidemiologic evidence from Turkey or Italy of statistically significant 2 and 20 fold increases in risk because of the absence of well- defined cohorts despite far greater direct information about the cases comprising the numerator of the relative risk calculations. Infante has since shown that death certificates underestimated cases in his cohort study. Unfortunately, the same overdependence on cohort epidemiology was evident in 2009 when IARC again accorded sufficient evidence for benzene as a cause of AML (100F) but not non-Hodgkins lymphomas (NHL). Epidemiologically there were positive and negative studies, but the animal and mechanistic information for NHLs was far superior to that for AML in 1982. Lymphatic cells are now recognized as deriving from the same stem cell mutated in AML; benzene affects multiple hematopoietic precursor chromosomal sites; genotoxicity is readily observable in circulating lymphocytes; and benzene–exposed animals develop NHLs. Listing additional cancers caused by an already known human carcinogen relates directly to cancer prevention. Designation of causality increases the litigation success of affected individuals which affects purveyors of carcinogens. Further, incorporating the additional cancers into risk estimations can increase regulatory priority for the carcinogen.