Mir-135a Expression Is Associated with Relapse in Hodgkin Lymphoma.

Abstract

MicroRNAs (miRNA) are small RNAs that act inhibiting the translation of mRNA to protein by complementary binding to his 3′UTR region. In a previous study we analyze the profile of miRNAs in classical Hodgkin lymphoma (HL) and found a 25 miRNAs signature that allows us to discriminate between HL and reactive lymph nodes. And some of them act inhibiting the translation of key genes in B lymphomagenesis. The aim of the study was to analyze if the expression of this miRNAs in the diagnostic samples predict clinical outcome in HL. Eighty nine adult patients (median age, 29 yrs; range, 13–89; males 47%) diagnosed with HL at a single institution between September 1995 and June 2005 have been studied. Seven patients (7.8%) were HIV+. Distribution according to histological subtypes was: nodular sclerosis (79%) and mixed cellularity (21%). Epstein-Barr Virus (EBV) was present in 28% of the samples. First-line treatment consisted of CMOPP/ABV or ABVD. Total RNA was extracted from formalin-fixed paraffin embedded lymph nodes using RecoverAll Total Nucleic Acid Isolation (Ambion). Mature miRNA expression was analyzed by stem loop RT-PCR and Real Time PCR (TaqMan MicroRNA Assay Protocol) in a 7500 Sequence Detection system(AB). Statistical analysis was performed with BRB Array Tools and SPSS 14.0. The characteristics considered were: age (<45 years vs. >45), ECOG, Hasenclever prognostic index (<1 vs 2–4 vs >5), HIV status, Ann Arbor (I–II vs. III–IV), WHO histological classification, bulky disease, EBV (LMP1+ vs LMP1−), ESR (EORTC criteria), and the expression of above-mentioned miRNAs. Clinical outcomes analyzed were relapse rate, disease-free survival (DFS) and overall survival (OS). Out of 89 patients, 75 (84.3%) achieved CR, 7 (7.8%) partial response, 7 (7.8%) were chemoresistant. After a median follow-up of 43 months (1–128), OS was 83% and DFS 74%. Of the 25 miRNAs, miR-135a appears underexpressed in HL samples versus reactive lymph nodes. Moreover, we found two groups of patients relative to miR-135a expression: patients with low expression (Ct<35.4) and patients without miR-135a expression(Ct=>35.4). In the Kaplan Meier analysis the patients without expression of miR-135a had a higher probability of relapse than the patients with miR-135a expression (Log Rank p=0.045). In the multivariate analysis the miRNA was the only prognostic factor for relapse (RR=6.533 p=0.021). In conclusion, miR-135a has as putative target HIF1A, a transcription factor related with hypoxia state that have been related in other tumors with tumoral growth and relapse when it's overexpressed. Maybe through this target miR-135a play a role in relapse as prognostic value, but a higher number of patients and a functional analysis will be required to validate this new prognostic factor for HL.