LAUSR

Background Diffuse large B-cell lymphoma (DLBCL) is a lymphoproliferative disorder originating in B-lymphocytes, and accounting for about 30% of all non-Hodgkin Lymphoma (NHL) . Since most patients show a lack of apparent symptoms in early stage, they are diagnosed at a late stage and therefore lose the best therapeutic opportunity. An increasing number of studies have indicated indicates that exosomal miRNAs extracted from peripheral blood might be usable as noninvasive biomarkers for the early diagnosis of tumors, therapy response monitoring, and prognosis. Although circulating exosomal miRNA has been studied in many malignant cancers, there are currently no reports examining their use in terms of DLBCL diagnosis. In light of previous research findings, we hypothesized that exosomal miRNA from peripheral blood can work as a candidate for DLBCL diagnosis. Materials and methods Blood samples were collected from 99 newly diagnosed DLBCL patients, 21 natural killer/T-cell lymphoma (NKTCL ) patients and 94 healthy people. NKTCL is another common NHL type which with high incidence in Asia and was chosen as case-control in our study. In the initial of the study, exosomes were extracted and identified by nanoparticle tracking analysis (NTA), Transmission Electron Microscope (TEM) and western blot. Next, Exiqon microarray was performed in 10 DLBCL patients and 5 HCs to determine serum exosomal miRNA profiles, which we called the screening stage. Identified miRNAs were confirmed through testing stage (24 DLBCL patients and 24 HCs) using quantitative reverse transcription polymerase chain reaction (qRT-PCR). According to the criterion that Ct(miRNA)<34,Ct(negative control)>40,fold change >1.5 and two-tailed P<0.05, miRNAs tested in testing stage were further chosen and confirmed through validation stages (65 DLBCL patients and 65 HCs) using qRT-PCR. Next, we carried out ROC (Receiver operating characteristic) curves to assess the performance of identified miRNAs in diagnosis DLBCL. Finally, we test the specificity and sensitivity of the identified miRNA by 21 NKTCL patients. Results Through the three-stage process, five miRNAs, including three high expression (miR-379-5p, miR-135a-3p ,miR-4476) and two low expression (miR-483-3p, miR-451a) were identified between DLBCLs and HCs. the AUCs (area under the curve , AUC)of miR-379-5p, miR-135a-3p, miR-4476, miR-483-3p, and miR-451a in testing stage were 0.757 (95% CI 0.612-0.869), 0.712 (95% CI 0.563-0.833), 0.673 (95%CI 0.522-0.801), 0.674 (95% CI 0.524-0.803), and 0.696 (95% CI 0.544-0.818), respectively, and the AUC of the panel was 0.951 (95% CI0.847-0.993). the AUCs of miR-379-5p, miR-135a-3p, miR-4476, miR-483-3p, and miR-451a in validation stage were 0.731(95% CI 0.646-0.805), 0.616(95% CI 0.527-0.700), 0.660 (95%CI 0.572-0.741), 0.620 (95% CI 0.530-0.703), and 0.648 (95% CI 0.560-0.730), respectively, and the AUC of panel was 0.841 (95% CI 0.767-0.900). When combined the testing and validation stage, revealing AUCs for miR-379-5p, miR-135a-3p, miR-4476, miR-483-3p, and miR-451a were 0.734 (95% CI 0.663-0.798), 0.639 (95% CI 0.564-0.710), 0.661 (95%CI 0.586-0.730), 0.637 (95% CI 0.561-0.707), and 0.663 (95% CI 0.588-0.732),respectively; the AUC of the combined panel was 0.824 (95% CI 0.760-0.877) .The panel also showed good resolving power in distinguishing DLBCL patients from NKTCL patients with the AUC of 0.848(95%CI 0.756-0.902).Moreover,miR-451a showed fairish performance in distinguishing NKTCL and HC, as well as discriminate late stage (IIB-IV) from early stage DLBCL, A tendency for the level of miR-135a-3p to declined gradually with increases in IPI was noted. Conclusions The panel composed by subsets of circulating exosomal miRNAs, including hsa-miR-379-5p, hsa-miR-135a-3p, hsa-miR-4476, hsa-miR-483-3p and hsa-miR-451a,can be used as noninvasive biomarkers for DLBCL diagnosis. Circulating exosomal hsa-miR-379-5p,hsa-miR-135a-3p,hsa-miR-4476 and hsa-miR-451a formed a panel which can be well distinguished DLBCL and NKTCL. Circulating exosomal hsa-miR-451a can be used as noninvasive biomarkers for NKTCL diagnosis. Certain circulating exosoaml miRNAs were related to clinical characteristics, specifically, circulating exosomal hsa-miR-451a was related to stage, while hsa-miR-135a-3p was related to IPI. No relevant conflicts of interest to declare.