Serum miR-22 as potential non-invasive predictor of poor clinical outcome in newly diagnosed, uniformly treated patients with diffuse large B-cell lymphoma: an explorative pilot study

Francesco Marchesi, Giulia Regazzo, Francesca Palombi, Irene Terrenato, Andrea Sacconi, Manuela Spagnuolo, Sara Donzelli, Mirella Marino, Cristiana Ercolani, Anna Di Benedetto, Giovanni Blandino, Gennaro Ciliberto, Andrea Mengarelli, Maria Giulia Rizzo

Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous group of tumors, with aggressive clinical course that renders prognostication and choice of treatment strategy difficult. Chemo-immunotherapy with rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone (R-CHOP) is the current first-line treatment. MicroRNAs (miRNAs) are under investigation as novel diagnostic and prognostic biomarkers in several malignancies, including malignant lymphomas. While tissue miRNAs in DLBCL patients have been extensively studied as biomarkers, only few reports to date have evaluated the role of circulating/serum miRNAs as potential prognostic factors. Here circulating/serum miRNAs, including miR-22, were investigated as potential non-invasive biomarkers, with the aim of a better prognostic stratification of DLBCL patients. MiRNAs were selected by global expression profile of serum miRNAs of DLBCL patients, The Cancer Genome Atlas (TCGA) analysis and literature research. Serum and tissues miRNA expression profile in de novo DLBCL patients, consecutively enrolled for this study, were detected by quantitative real-time polymerase chain reaction. Relative expression was calculated using the comparative Ct method. Statistical significance was determined using the Mann-Whitney rank sum and Fisher's exact test. Survival analysis was conducted through the use of Kaplan-Meier method. Spearman's Rho was applied to study the correlation between miRNA distributions and days to first relapse. Experimentally validated miRNA-target interactions were assessed by miRTarBase database. Negative miRNA-mRNA correlation was evaluated in TCGA DLBCL dataset. Pathway analysis was performed by the functional annotation clustering DAVID tool. We showed a significant modulation of serum miR-22 after R-CHOP treatment compared with basal values but no difference between baseline serum miRNAs values of DLBCL patients and healthy controls. High expression level of serum miR-22 in DLBCL at diagnosis (n = 36) is associated with a worse PFS and is independent of the currently used clinical prognostic index. Integrative and pathways analysis of miR-22 identified target genes involved in different important pathways such as p53 signaling. Our data suggest that miR-22 is of potential interest as non-invasive biomarker to predict clinical outcome in DLBCL patients. Characterization of miR-22 pathways can pave the way to the development of targeted therapy approaches for specific subgroups of DLBCL patients.