Thyroid malignancies, the most common endocrine cancer, account for >17,000 new cases and 1,200 cancer deaths per year in the U.S. The bulk (70-95%) of these neoplasms are primary thyroid carcinomas of follicular cell origin, including differentiated papillary (PTC) and follicular (FTC), and undifferentiated anaplastic thyroid cancers. In the U.S. incidence rates are PTC > FTC >> anaplastic carcinomas, while morbidity/mortality rates associated with these cancers are anaplastic carcinomas >> FTC >PTC. Evidence for progression from benign follicular adenoma (Fa) > FTC has been observed, while PTC appears to arise de novo. While numerous studies have attempted to define the molecular genetics of differentiated thyroid cancer, virtually all of these studies have suffered from lack of significant specimen numbers, insufficient pathological criteria, or both. We have shown extensive evidence for frequent loss of heterozygosity (LOH) on chromosomes 3p, 10q, 13q and 17p in FTC , but not FA or PTC, suggesting that tumor suppressor genes (TSGs) may be involved in the genesis of FTC. Known TSGs mapping near regions of LOH on chromosomes 3p (VHL and FHIT) and 17p (p53) do not appear to be involved, since mutations of these sequences are rare in FTC. Thus as yet undefined TSGs appears to be involved in the genesis of FTC. In the current studies we will perform a detail molecular genetic study of at least 30 specimen/tumor type of a well defined and stratified population of thyroid cancers, in which extensive clinical records are available. Tumor types will include PTC (grades 1,2 and 3), FTC (minimally and widely invasive, oxyphilic and non-oxyphilic carcinomas) and FA. With this population of tumors we will: (i) perform a comprehensive LOH analysis of all chromosomes arms at a resolution of approximately 10 cM, (ii) refine mapping of regions of significant LOH at a resolution of less than or qual to 2 cM, (iii) analyze candidate TSGs that resides within the refine map locations and (iv) clone potential tumor suppressor genes that reside in locations for which no known TSG candidates have been identified. The molecular genetic profiles will be correlated with the clinical records to assess the significance of the genetic changes on morbidity and mortality. These studies will offer one of the first comprehensive analyses of LOH in any well defined tumor population that can be reconciled with the clinical record and will provide detailed insight into the pathogenesis of thyroid cancer.