Introduction: Primary mediastinal B-cell lymphoma (PMBCL) is a distinct subtype of diffuse large B-cell lymphoma (DLBCL) with unique clinical and pathologic features including a female predominance, lack of extra-thoracic disease at presentation and prominent tumor sclerosis. PMBCLs also harbor distinctive chromosomal aberrations including consistent gains in chromosome 9p and 2p not unlike classical Hodgkin’s lymphoma. Notable, is the distinct lack of t(14;18) found in a subset of DLBCLs. In this study we used tiling resolution array CGH to profile a panel of PMBCLs tumor specimens to better delineate genomic changes and provide insight into novel genes that may be critical in disease pathogenesis. Hypotheses: PMBCL pathogenesis occurs via multiple genomic alterations and the critical alterations will be present in multiple patient samples. Objective: Alignment of high resolution genomic profiles from a panel of PMBCL samples will reveal minimal regions of alteration important in the pathogenesis of PMBCL. Experimental Approach. Array competitive genomic hybridization (CGH) was used to generate high resolution segmental copy number profiles of a PMBCL panel. Currently, the whole genome has been arrayed as 32,433 bacterial artificial chromosome (BAC) derived amplified fragment pools spotted in triplicate (97,299 elements) resulting in a tiling resolution with complete coverage of the sequenced human genome. Results: Alignment of these profiles with the human genome map has resulted in the identification of both previously reported and novel regions of recurrent genomic alteration. Previously reported regions identified included gains of 9p, 2p15, 9, 8q24, 12q12-q14, 21q22-qter, Xq21-qter, 21, X and losses of 4q35, 13q12-q13, and 18. Novel regions identified included gains of 5q35, 20q12-13, 2q34-qter, 7, 20p11-p12 and losses of 1p13, 8p21-pter, 15q15, 15q26-qter, 16p13, and 7p. Conclusions: The generation of these high resolution segmental copy number profiles allows the identification of candidate genes involved in the pathogenesis of PMBCL as well as lead to an increased understanding of PMBCL. Acknowledgements: This work was supported by funds from Genome Canada/British Columbia. We would like to thank Spencer Watson for array construction and Lindsey Kimm for technical assistance with array CGH.

[Proc Amer Assoc Cancer Res, Volume 46, 2005]