A High-Risk Survival Classifier for Multiple Myeloma

• Published in: Blood 2011-11, Vol.118 (21), p.1800-1800
• Main Authors: Kuiper, Rowan, Broyl, Annemiek, de Knegt, Yvonne, van Vliet, Martin H., van Beers, Erik H., van der Holt, Bronno, Jarari, Laila el, Mulligan, George, Morgan, Gareth, Gregory, Walter M, Goldschmidt, Hartmut, Lokhorst, Henk M., van Duin, Mark, Sonneveld, Pieter
• Format: Article
• Language: English
• ISSN: 0006-4971; 1528-0020
• DOI: 10.1182/blood.V118.21.1800.1800

Abstract 1800▪▪This icon denotes a clinically relevant abstract Survival of patients with newly diagnosed multiple myeloma (MM) is highly variable and currently used clinical prognostic markers such as the international staging system (ISS) and cytogenetic markers are insufficiently adequate for defining individual patient prognosis. We established a prognostic signature based on gene expression profiling. The signature was generated using a training set of 290 newly diagnosed MM patients included in the multicenter, prospective open-label randomized phase 3 HOVON65/GMMG-HD4 trial. Gene expression profiles, obtained from purified plasma cells, were generated using the Affymetrix GeneChip® Human Genome U133 Plus 2.0 platform (GSE19784; Broyl et al.,Blood 2010; 14:2543–2553). The model predictive for survival was built by supervised principal component analysis (Bair et al., J. Amer. Statistical Assoc. 2006;101:119–37) and further optimized by simulated annealing. The generated survival signature was compared to six previously reported MM gene expression signatures (i.e. UAMS-70, UAMS-17 (Shaughnessy et al., Blood. 2007;109:2276–84), gene expression-based proliferation index (GPI, Hose et al., Haematol. 2010; 96: 87–95), MRC-IX-6 gene (Dickens et al., Clin. Cancer Res. 2010;16:1856–1864), Millennium (Mulligan et al., Blood 2007; 109:3177–3188) and IFM (Decaux et al., J. Clin. Oncol. 2008; 26:4798–4805). A signature of 92 probe sets (EMC-92-gene signature) was highly discriminative for high-risk MM patients, defined as overall survival (OS) < 2 yr (21.7%) vs. standard-risk MM. This performance was confirmed in independent validation datasets of newly diagnosed MM patients (UAMS-TT2, n=351, GSE2658; MRC-IX, n=247, GSE15695) and relapse MM patients (APEX, n=264, GSE9782). In the UAMS-TT2 dataset, a high-risk population of 19.1% was identified which had a hazard-ratio of 3.52 (P = 2.5 × 10−8). In the MRC-IX study, 20.2% of patients were identified as high risk with a hazard-ratio of 2·38 (P = 3·6 × 10−6; Figure 1a) The high-risk signature was able to identify patients with significantly shorter survival in both the transplant-eligible and non-transplant-eligible patients included in the MRC-IX study. In non-transplant-eligible patients, 23.8% high risk patients were identified with a hazard-ratio of 2.38 (P = 4.3 × 10−4), whereas 17.5% of transplant-eligible patients were high-risk with a hazard-ratio of 2.54 (P = 1.5 × 10−3). The difference between survival in