GENETIC EPIDEMIOLOGY OF MALIGNANT GLIOMA

点击翻译按钮获取中文摘要

Malignant glioma, the most common primary malignant brain tumor in adults, is debilitating and rapidly fatal. Very little is known about the causes of malignant glioma; family history and diverse environmental risk factors have been suggested in previous epidemiologic studies. In addition, gliomas may be associated in some families with more common cancers. No study has examined the interrelationships of familial and environmental factors or tested specific genetic models that might, in part, explain the familial associations of these tumors. In this study, cases (n= approximately 450) will be adults newly diagnosed with histologically confirmed malignant glioma from 7/1/1991 through 6/30/1994 in five San Francisco area counties. Population-based controls (n= approximately 450) matched to cases for age, sex, race, and telephone prefix will be obtained through random-digit dialing. This study will collect detailed family medical histories for approximately 5300 first-degree relatives (parents, siblings, and children) of cases and controls, as well as information on occupation; smoking; alcohol use; diet containing nitrites; water sources; extremely-low-frequency electromagnetic field exposures; hair dye use; and personal medical history of cases and controls. Questionnaires to adult relatives and to designated informants of deceased relatives will assess medical conditions and risk factors of interest for the relatives. Pertinent records will be reviewed for further validation of reported medical conditions. Analyses will address five aims: 1) to test hypotheses that cases' relatives are more likely than controls' relatives to have had a) cancer (breast, lung, adrenal, or colo-rectal cancers, sarcoma, and leukemia are of particular interest); b) brain tumors; and/or c) certain nervous system conditions; 2) to evaluate the contributions of shared environmental exposures or cultural practices to familial cancer clustering using logistic regression techniques appropriate for cluster sample data; 3) to test specific genetic models of inherited susceptibility to disease in cases' families using complex segregation analyses; extended pedigrees of multiply affected families will also be evaluated; 4) to compare familial demographic and environmental risk factor in cases and controls; and 5) to examine confounding and effect modification among these factors using logistic regression methods for matched pairs.