AUTHORS: J.J. Keats1, W-J Chng1, E. Braggio1, V. Shanmugam2, V. Pushparaj2, R. Schop1, A. Baker2, C. Mancini2, T. Price-Troska3, G. Ahmann1, K. Henderson3, P. Greipp3, A. Dispenzieri3, L. Bruhin4, P.L. Bergsagel1, J. Carpten2, R. Fonseca1
1 Mayo Clinic, Comprehensive Cancer Center, Scottsdale, Arizona ; 2 Translational Genomics, Hematological Malignancies Research Unit, Phoenix, Arizona ; 3 Mayo Clinic, Internal Medicine, Rochester, Minnesota; 4 Agilent Technologies, Santa Clara, California
Introduction: To identify novel genetic factors contributing to the pathogenesis and prognosis of multiple myeloma (MM) we have initiated a comprehensive genomic screen using array-based comparative genomic hybridization (aCGH). Materials and Methods : Genomic copy number changes were assessed in 68 MM samples using the Agilent Human Genome 44B microarray (Agilent Technologies). We focused on defining common breakpoint regions and minimally deleted and amplified regions, MDR and MAR respectively, which are not well defined in MM. Genes contained within the MDR and MAR were cross-referenced with a list of ~2100 genes mutated and implicated in neoplasia. Results : A large number of recurrent breakpoints and copy number gains/losses were identified. Known aberrations such as the recurrent trisomies, 1p loss, 1q gain and 13 monosomy were observed and agrees with existing FISH data on these patients. We identified 35 recurrent breakpoints that either occurred within a specific genomic window of ~35 kb or breakpoints that occurred within a specific gene. Six of the identified breakpoints occurred in more than 4 patients and one was present in 6 patients. We identified 99 minimal regions of copy number change that encompassed 352 genes implicated in cancer. Interestingly, we identified nearly equal numbers of MDR and MAR, 49 and 50 respectively. The high-resolution of this aCGH platform allowed the identification of 43 recurrent regions of copy number change below 1.0 Mb. Furthermore, 24 are below 0.5 Mb and 9 are below 0.2 Mb. The number of aberrations per tumor ranged from 2 to 58. The presence of a complex genome (more than 20 aberrations) is associated with a worse prognosis (24 months versus 88 months, log-rank p=0.0016). This effect is not simply a dichotomy of hyper and non-hyperdiploid MM as the incidence of a complex genome was equally distributed between both subgroups. Conclusions : This study has identified a number of recurrent breakpoint regions and copy number changes that were not previously characterized. The high resolution of our aCGH platform has identified 24 recurrent gains or losses smaller than 1.0 Mb that contain genes previously implicated in cancer.