Case referent study of brain tumors The etiology of brain tumors and brain cancer is poorly understood, and recorded incidence rates have increased dramatically over the past several decades. Whether this increase is, in part, real or is entirely an artifact of improved diagnosis is a controversial issue. Nonetheless, concern has arisen that one or more increasingly common environmental exposures might cause brain cancer. Examples include industrial chemicals, pesticides, food additives, and electromagnetic fields. In response to such concerns, and to advance understanding of environmental, behavioral and genetic causes of brain tumors, we are collaborating with investigators at three U.S. hospitals in conducting a case-control study of malignant and benign brain tumors. Factors under consideration include occupational exposures to chemical agents and electromagnetic fields, use of cellular telephones, dietary factors, family history of tumors, genetic determinants of susceptibility, home appliance use, reproductive history and hormonal exposures, viruses, medical and dental exposure to ionizing radiation, and other aspects of medical history. A manuscript concerning occupational exposure to ELF-EMF was published in 2008. In other analyses, excess risks of glioma were found among electricians and farmers, and an elevated risk of meningioma was seen among auto-body painters. Evaluations are in progress to clarify the possible role of solvents, pesticides, lead, EMF and other exposures in explaining these associations. In parallel with these exposure assessments, we have been conducting studies of genes involved in the metabolism of exposures. We observed an association between occupational exposure to lead and meningioma that appeared to be concentrated in a genetically susceptible subgroup. Other analyses indicated associations with polymorphisms in glutathione S-transferase (GST) and cytochrome P450 (CYP) genes involved in metabolism of solvents and polycyclic aromatic hydrocarbons and other substrates. Self-reported family history of cancer showed glioma risk to be elevated only slightly and nonsignificantly among persons with a family history of brain cancer. Persons reporting a family history of prostate, stomach or colon cancer, or Hodgkin lymphoma, had an increased risk of glioma. Risk of meningioma was increased among persons reporting a family history of a benign brain tumor or melanoma. An early age at menarche and early age at first live birth appeared to be protective against glioma, but not meningioma. The risk of glioma was found to be reduced among left-handers relative to right-handers. Risk of glioma and meningioma appeared to vary with season of birth, with higher risks among persons born in the winter and lower risks among those born in the summer. Risk of glioma in women increased with older age at menarche. Early age at first birth was associated with reduced risk. Brain tumor risk did not appear to be related to use of hair dyes. Recent genetic analyses have indicated associations between brain tumor risk and polymorphisms in cytokine, apoptosis and oxidative stress genes. Findings for these pathways are being followed up in larger and more comprehensive studies. UV Dosimetry REB investigators conducted a pilot study of 125 volunteer radiologic technologists in which daily diaries and polysulfone UV dosimeters were used to develop better questionnaire approaches to ascertain environmental UV exposure for skin and other environmental and occupational cancers in this largely female occupational population. The volunteers were queried 6 months later to test the reproducibility of responses to time outdoors. Agreement between reported time on weekdays was significantly higher than for weekends. Improved exposure assessment will enable us to characterize more quantitatively the effects of UV and ionizing radiation on skin and other cancers. We are currently analyzing the validity and reproducibility of hour-based and activity-based questionnaires to capture lifetime sun exposure. PLCO Lung Damage Study Reverse causation bias has been posed as an alternative reason associations have been observed between functional assays and increased cancer risk in published case-control studies. The debate has hinged on the concern that because the tests were performed on biologic specimens collected after cancer diagnosis they may be measuring the consequence, rather than the underlying, causes of cancer (termed reverse causation bias). Several investigators in the Radiation Epidemiology Branch (REB), the Occupational and Environmental Epidemiology Branch (OEEB), and the Genetic Epidemiology Branch (GEB) sought to determine the predictive usefulness of several phenotypic or functional assays in pre-diagnostic samples. Investigators in REB, OEEB, and GEB planned this study using pre-diagnosis samples from the Prostate, Lung, Colon, and Ovary (PLCO) screening trial. The cancer of interest was lung cancer because of the public health impact and because several previous studies had been conducted among lung cancer patients. We sent, 117 lung cancer cases and 117 matched control samples (plus 16 QC samples) to four collaborating investigators at three institutions (Lawrence Livermore National Laboratory, M. D. Anderson Cancer Center, and the University of Minnesota). Laboratory analyses have been completed using the Comet assay and mutagen sensitivity assays (BPDE and bleomycin as the mutagens). A nucleotide excision repair assay is still in progress. Initial data analyses are underway with results available in 2009.
