Tissue-Specific microRNA Expression Profiling to Derive Novel Biomarkers for the Diagnosis and Subtyping of Small B-Cell Lymphomas

SIMPLE SUMMARY: It is highly challenging for pathologists to distinguish small B-cell lymphomas from reactive lymphoid tissue and to accurately diagnose common histological subtypes of such lymphomas. This is due to overlapping morphological features and limitations of current ancillary testing, which itself involves many further tests. Hence, there is a pressing need for better biomarkers for accurate diagnosis and subtyping of small B-cell lymphomas as better diagnosis can lead to better treatments and clinical outcomes for patients. In this study, we identified and validated two sets of microRNA biomarkers that can distinguish small B-cell lymphomas from reactive lymphoid tissue and distinguish between four subtypes of such lymphomas, respectively. This study suggests that miRNA expression profiling may serve as a promising tool to aid in the diagnosis of small B-cell lymphomas. ABSTRACT: Accurate diagnosis of the most common histological subtypes of small B-cell lymphomas is challenging due to overlapping morphological features and limitations of ancillary testing, which involves a large number of immunostains and molecular investigations. In addition, a common diagnostic challenge is to distinguish reactive lymphoid hyperplasia that do not require additional stains from such lymphomas that need ancillary investigations. We investigated if tissue-specific microRNA (miRNA) expression may provide potential biomarkers to improve the pathology diagnostic workflow. This study seeks to distinguish reactive lymphoid proliferation (RL) from small B-cell lymphomas, and to further distinguish the four main subtypes of small B-cell lymphomas. Two datasets were included: a discovery cohort (n = 100) to screen for differentially expressed miRNAs and a validation cohort (n = 282) to develop classification models. The models were evaluated for accuracy in subtype prediction. MiRNA gene set enrichment was also performed to identify differentially regulated pathways. 306 miRNAs were detected and quantified, resulting in 90-miRNA classification models from which smaller panels of miRNAs biomarkers with good accuracy were derived. Bioinformatic analysis revealed the upregulation of known and other potentially relevant signaling pathways in such lymphomas. In conclusion, this study suggests that miRNA expression profiling may serve as a promising tool to aid the diagnosis of common lymphoid lesions.