Use of exon-based transcriptome profiling to identify novel signaling pathways and survival-associated genes in diffuse large B-cell lymphoma.

Abstract

8074 Background: Identification of biological prognostic factors that could be used to define poor risk diffuse large B-cell lymphoma (DLBCL) patients is a main concern. Methods: Study population consisted of 38 de novo high risk DLBCL patients less than 65 years old. The patients were treated in the Nordic phase II protocol with six courses of R-CHOEP14 followed by systemic central nervous system prophylaxis with one course of high dose methotrexate and one course of high dose cytarabine. Exon array-based profiling was used to screen signaling pathways and differentially expressed genes between the clinically high risk patients, who had relapsed or remained in remission in response to dose dense chemoimmunotherapy. At the time of the analysis, median follow up was 34 months, progression free survival (PFS) 78% and overall survival (OS) 78%. Results: The screen between relapsed patients and the patients in remission using criteria of p ≤ 0.05 and fold change ≥ 1.6 revealed 566 differentially expressed genes (131 protein coding), of which 24 were likely to be involved in conventional signaling pathways, including those regulating antigen processing and presentation (CIITA, HLA-DQA2, HLA-DQB1, RFXAP), Jak-STAT signaling (SOCS3), Notch signaling (NOTCH1) and Toll-like receptor signalling (IRF5). In cox univariate analysis, 12 of 24 genes were found to associate with PFS (p<0.05). Of these, high expression of CIITA, DLL4, HLA-DQA2, HLA-DQB1, IRF5, NOTCH1, PER1, RFXAP, SEMA4D and ZFP36 had a favorable impact on PFS, whereas high levels of ENPP3 and PRKAR2B were associated with adverse outcome. Differential expression of four genes was confirmed by quantitative PCR, and prognostic value of six genes validated using Lymphoma/Leukemia Molecular Profiling Project microarray data set. Immunohistochemical validation of the findings in a larger patient cohort is ongoing. Germinal centre B-cell signature did not predict survival in this cohort. Conclusions: The results provide evidence that exon-based transcriptome profiling can identify biologically relevant signaling pathways and genes that discriminate the outcome of homogenously treated young high risk DLBCL patients.