Chronic Lymphocytic Leukemia Associated with Autoimmune Hemolytic Anemia Is Characterized by a Distinctive miRNA Signature,

Abstract 3897

Chronic lymphocytic leukemia (CLL) is frequently associated with autoimmune hemolytic anemia (AIHA). However, the mechanisms governing the association between CLL and AIHA are poorly understood. MicroRNAs (miRNAs) are small 18–22 nucleotide long RNA molecules that regulate gene expression and play a key role in several biological processes. Importantly, deregulated miRNA expression has been implicated both in CLL and autoimmunity. This led us to speculate that patients with CLL who develop AIHA might have a different miRNA expression pattern as compared to those in whom this complication is not observed. We report here the first results of this study. We have evaluated the miRNA expression in purified CLL cells (CD19+, CD5+) from 14 patients who developed AIHA over the course of their disease and 19 sex-, age-, and clinical stage-matched controls who, after a comparable follow up time, did not develop this complication. The expression of 377 mature miRNAs was analyzed using TaqMan Human MicroRNA Arrays A v2.0 (Applied Biosystems) in an ABI 7900 HT sequence detection system. miRNA expression data was analyzed by the 2^–ΔΔCt method, using RNU48 as endogenous control. Statistical analyses were performed with TiGR MultiExperiment Viewer, BRB-ArrayTools and R software. The unsupervised hierarchical cluster analysis identified two groups (CLL AIHA+ and CLL AIHA-). The supervised analysis revealed 7 miRNAs that were down-regulated in CLL AIHA+ patients compared to CLL AIHA- patients: miR-19a, miR-20a, miR-29c, miR-146b-5p, miR-324-3p, miR-340, miR-660. Interestingly, miR-146, which has previously been related to autoimmunity, regulates IL-1 receptor-associated kinase 1 and TNF receptor-associated factor 6, two key adapter molecules downstream of Toll-like and cytokine receptors. The prediction analysis of microarrays (PAM) was used to determine a set of miRNAs able to classify the samples into CLL AIHA+ and CLL AIHA-. We obtained a 122-miRNA model that classified CLL AIHA+ patients with a sensitivity of 69.4% and specificity of 68.4%. This model correctly classified 67% of the analyzed samples. In summary, CLL AIHA+ samples were characterized by a distinctive signature of 7 down-regulated miRNAs, one of which (miR-146) has previously been related to autoimmunity. Moreover, we identified 122 miRNAs that are able to predict AIHA in CLL patients. This is the first study establishing a relationship with CLL associated with AIHA and a distinctive miRNA signature, which should be useful for further studies aimed at unfolding the biologic complexity of AIHA in CLL.