Selecting Patients Most Likely to Respond to Therapy

Multiple myeloma (MM) is a heterogeneous disease with a broad range of biological and clinical features.^1,2  Multiple studies have therefore attempted to identify genetic and biologic markers, as well as clinical characteristics to define subgroups of patients. For example, prognostic factors such as beta ²  microglobulin and albumin form the basis of the International Staging System,³  which is predictive of disease course. Conventional cytogenetics can identify patients with adverse outcome to conventional low- and high-dose therapies (i.e., 4;14 translocation or deletion of chromosome 13).⁴  More recent studies have used microarray profiling⁵  and array comparative genomic hybridization⁶  to form the basis for defining mRNA-based and DNA-based prognostic subgroups of myeloma. Importantly, prognostic factors must be defined in a particular clinical context. For example, in patients treated with novel therapies such as bortezomib, chromosome 13 deletion and 4;14 translocation are no longer of adverse prognostic import.⁷  Moreover, recent studies have utilized microarray profiling to define patient populations with gene signatures predictive of response to specific therapies (i.e., bortezomib).⁸ 

Mateo and colleagues have recently carried out a prospective study of 685 newly diagnosed patients with myeloma treated uniformly with six alternating cycles of vincristine, BCNU, melphalan, cytoxan, and prednisone (VBMCP) alternating with vincristine, BCNU, adriamycin, and prednisone (VBAP) therapy, followed by melphalan 200mg/m²  and autologous stem cell transplantation. The median progression-free survival (PFS) and overall survival (OS) were 37 and 67 months, respectively. In order to delineate patient subgroups with differential outcome, CD138 positive bone marrow plasma cells were purified and immunophenotyped prospectively as well. Importantly, CD19+CD28+CD117-phenotype on CD138+BM plasma cells predicted poor outcome, with significantly shorter PFS and OS. An immunophenotype-based staging system, defined on the basis of CD28 and CD117 expression on tumor cells, identified poor-, intermediate-, and good-risk patient subgroups with significantly different PFS and OS. In addition to its clinical utility, understanding the biologic significance of the observed antigen-expression profiles conferring this differential outcome may both delineate mechanisms of sensitivity versus resistance to therapy and also yield insights into MM pathogenesis. Moreover, correlation of these phenotype profiles with microarray profiling and aCGH analysis may identify, and yield, new insights into MM pathogenesis to be exploited in novel, single agent or combination targeted therapeutics.