Abstract
Multiple myeloma is preceded by monoclonal gammopathy of undetermined significance (MGUS). Serum markers are currently used to stratify MGUS patients into clinical risk groups. A molecular signature predicting MGUS progression has not been produced. We have explored the use of gene expression profiling to risk-stratify MGUS and developed an optimized signature based on large samples with long-term follow-up. Microarrays of plasma cell mRNA from 334 MGUS with stable disease and 40 MGUS that progressed to MM within 10 years, was used to define a molecular signature of MGUS risk. After a three-fold cross-validation analysis, the top thirty-six genes that appeared in each validation and maximized the concordance between risk score and MGUS progression were included in the gene signature (GS36). The GS36 accurately predicted MGUS progression (C-statistic is 0.928). An optimal cut-point for risk of progression by the GS36 score was found to be 0.7, which identified a subset of 61 patients with a 10-year progression probability of 54.1%. The remainder of the 313 patients had a probability of progression of only 2.2%. The sensitivity and specificity were 82.5% and 91.6%. Furthermore, combination of GS36, free light chain ratio and immunoparesis identified a subset of MGUS patients with 82.4% risk of progression to MM within 10 years. A gene expression signature combined with serum markers created a highly robust model for predicting risk of MGUS progression. These findings strongly support the inclusion of genomic analysis in the management of MGUS to identify patients who may benefit from more frequent monitoring.