[1262] MicroRNA Expression in DLBCL: Identification of Unique miRNAs Signatures Related to Clinical Outcome Prediction in R-CHOP Treated Patients

S Montes-Moreno, N Martinez, A Saez, C Montalban, MH Rodriguez, B Sanchez-Espiridion, SM Rodriguez Pinilla, GB Lemarbre, WM Rehrauer, M Waknitz, KH Young, MA Piris. Spanish National Cancer Center, CNIO, Madrid, Spain; UW Paul P. Carbone Comprehensive Cancer Center, University of Wisconsin School of Medicine and Public Health, Madison, WI

Background: MicroRNAs (miRNAs) can constitute markers of malignant transformation, outcome or sensitivity to specific therapies. In this study we analyzed the miRNA profile of a large series of R-CHOP treated DLBCL patients and correlated with clinical outcome variables.
Design: First, a set of 36 cases was studied by miRNA and Gene Expression profiles by array tech. Both outcome-related candidate miRNAs and Cell of Origin (Wright et al. PNAS USA. 2003) related miRNA candidates were identified. After this, the expression of 14 miRNA candidates was studied by RT-PCR in a large series of 90 samples and results were correlated with IPI score and outcome data (OS, EFS).
Results: Two microRNA expression based models were defined to correlate with outcome. These models summarize the expression of 4 miRNAs each and correlate with OS (Hazard ratio 1.2, 95%CI=1.05-1.38, p=0.004) and EFS (Hazard ratio 1.14, 95%CI=0.99-1.3, p=0.055) in the univariable Cox regression analysis. In the multivariable Cox regression analysis, the miRNA-based models remain independent of IPI, being both variables significant as outcome predictors (figure 1). 12 miRNAs were found to be differentially expressed between DLBCL subtypes (FDR<0.05) (upregulated in the GCB type are miR-331-3p, 151-5p, 28-5p, 454, 210, 183 and miR-138 and uregulated in the ABC subtype are miR-222, 144, 451, 221, 148a).


Conclusions: A unique miRNA signature is identified to associate with poor clinical outcome in R-CHOP treated DLBCL patients. A model based on the expression of 4 miRNAs is able to predict for OS and EFS, which is independent of IPI score. MiRNA expression profiling is able to identify particular miRNA species that are differentially expressed between DLBCL subtypes, providing a valuable technology to identify differentiation stage-related miRNAs and its potential gene targets.
Category: Hematopathology

Monday, March 9, 2009 1:00 PM

Platform Session: Section D, Monday Afternoon