Meng, Yan

Abstract Introduction: Clinical development of bortezomib in multiple myeloma (MM) included pharmacogenomic gene expression profiling (GEP) of pre-treatment tumor biopsies in order to evaluate the feasibility of developing predictive classifiers in prospective clinical trials and to better define the biological pathways related to drug sensitivity. By partitioning the samples into train (n=101) and test (n=80) datasets, statistically significant classifiers of response and survival were identified and then validated [Mulligan G, et al: Blood. 2007; 109:3177–88]. Here we describe specific genes and biological pathways associated with bortezomib efficacy in relapsed MM. Methods: Using tumor GEP data from 188 patients treated with single-agent bortezomib we identified genes associated with overall survival (OS), as well as genes associated with response (R) or progressive disease (PD) after bortezomib therapy. These clinical endpoints were also analyzed with biological pathway algorithms including GSEA [Subramanian A, et al: Proc Natl Acad Sci2005; 102:15545–50] Ingenuity Pathway Analysis and Gene Ontology/GATHER. These results were then compared to GEP data from a patient subset (n=76) treated with single-agent dexamethasone and also cell lines selected for differential sensitivity to bortezomib. Results: A comparison of R and PD highlighted numerous differentially expressed genes; approximately 200 were over-expressed in R and 500 were over-expressed in PD (t-test <0.01). R-associated genes include signaling molecules like TRADD, CFLAR and TANK, while PD-associated genes included oncogenes (NRAS), apoptotic regulators (DAP3) and cancer antigens (CTAG). Pathway tools helped to integrate and extend these observations, consistently identifying NFkB activation and overexpression of adhesion pathways as R-associated. Overexpression of protein synthesis and mitochondrial pathways were PD-associated. The overexpression of pathways such as NFkB, adhesion and cytokines appear linked specifically with bortezomib and not dexamethasone sensitivity. These approaches were also applied to patient survival. In contrast to analyses of single genes, the pathway tools highlighted links between response and survival, including adhesion and inflammatory cytokine pathway expression. Conclusions: This study highlights the diversity of biological pathways expressed in MM and the utility of clinical genomics to associate pathways with activity of single-agent bortezomib.

Abstract Multiple myeloma is a disease of the antibody producing plasma cells in the bone marrow. It has been characterized by frequent genomic alterations including gains and losses of chromosomes. Some of these alterations result in poor clinical prognosis. Bortezomib, a first in class proteasome inhibitor, is approved for the treatment of relapsed multiple myeloma. To better understand how the genome wide changes observed in myeloma relate to prognosis and treatment response, we have used single nucleotide polymorphism (SNP) array technology to assess genotypes and DNA copy number at up to 100,000 genomic loci. DNA from 147 bone marrow tumor biopsies collected in multi-center phase II and III clinical trials of relapsed multiple myeloma patients prior to treatment with Bortezomib were hybridized on SNP arrays to assess genomic aberrations. Commonly seen genomic alterations in myeloma were observed, including deletions of chromosome 13 (37%), 1p (25%), 6q (23%), and amplifications on 1q (35%). We found that 35% of this relapsed myeloma population exhibited a hyperdiploid genome characterized by gains of chromosomes 3, 5, 7, 9, 11, 15, and 19. Other notable genomic deletions included 17p (22%), 8p (27%) and 10p (16%). A correlation of response to bortezomib therapy in patients with differential gains and losses at the chromosomal level is ongoing. Consistent with previous reports, response to bortezomib was observed in patients with poor prognostic chr 13 deletions. A more in-depth analysis of specific loci associated with clinical outcome will be discussed. In summary, this data will help further define genomic aberrations in multiple myeloma and their association with treatment outcome.