| Prostate cancer is a disease of complex genetic mechanisms that has profound effects on the health of the male population in developed countries. This thesis focuses on characterizing the chromosomal aberrations that in part underlie this malignancy. For this work I had access to extraordinary sample sets that often consisted of very few cells. A subset of these remarkable samples included cells that had presumably disseminated out from prostate tumors and made their way to the bone marrow. To detect genomic alterations in these cells, I developed and validated an array comparative genomic hybridization (CGH) method compatible with as few as 10 cells.; Using this method, I demonstrate that disseminated cells from advanced disease patients possess multiple striking deviations. In contrast, disseminated cells from patients with localized disease have fewer and less marked alterations. However, there were concordant changes in disseminated cells from localized and advanced patients, which could point to progression-related factors.; I also identified a myriad of genomic changes in tumors that had acquired an androgen-independent status as a result of treatment by androgen deprivation. Interestingly, when the full collections of genomic changes are considered, tumors from the same patient appear more similar to each other than do tumors from the same organ in different patients. However, some chromosomal alterations appear more often in the tumors of one organ site versus the other organ sites, suggesting that these loci could favor survival in particular microenvironments.; Of the AI tumors, 22/57 had a deletion at 21q22 that points to the fusion of the androgen-regulated TMPRSS2 (21q22.3) and the oncogenic ERG (21q22.2) genes. I confirmed this gene fusion in the AI tumors using fluorescence in situ hybridization (FISH) and found a significant reduction in the frequency of this hybrid gene in early tumors treated with high-doses of an anti-androgen agent versus tumors treated with a low-dose and untreated tumors.; The work presented here provides a foundation for assessing recurrence risk using the genomic characteristics of disseminated cells, for identifying therapeutic targets in late-stage prostate cancer, and for evaluating a putative marker of hormonal escape via sensitization of androgen-response pathways. |