Evaluating gene expression profiling by quantitative polymerase chain reaction to develop a clinically feasible test for outcome prediction in multiple myeloma

Abstract

The gene expression profiles (GEPs) of 96 selected genes were analysed by real-time quantitative polymerase chain reaction (qPCR) with a TaqMan low-density array card in isolated tumour plasma cells (PCs) from 157 newly diagnosed multiple myeloma (MM) patients. This qPCR-based GEP correctly classified cases following the Translocation-cyclin D classification. Classic prognostic parameters and qPCR-based GEP predicted MM patient outcome and, although multivariate analyses revealed that cytogenetic risk (standard vs. high risk) was the variable that most strongly predicted prognosis, GEP added significant information for risk stratification. Considering only the standard risk cytogenetic patients, multivariate analyses revealed that high β2-microglobulin, low CDKN1A and high SLC19A1 gene expression levels independently predicted a short time-to-progression (TTP), while high International Staging System stage, low CDKN2B and high TBRG4 gene expression predicted poor overall survival (OS). A gene expression risk score enabled the division of standard risk patients into two groups with different TTPs (83% vs. 38% at 3years, P<0·0001) and OS rates (88% vs. 61% at 5years; P=0·003). This study demonstrates that quantitative PCR is a robust, accurate and feasible technique for implementing in the daily routine as a surrogate for GEP-arrays.