Frequent mutated B2M, EZH2, IRF8, and TNFRSF14 in primary bone diffuse large B-cell lymphoma reflect a GCB phenotype

Abstract

Primary bone diffuse large B-cell lymphoma (PB-DLBCL) is a rare extranodal lymphoma subtype. This retrospective study elucidates the currently unknown genetic background of a large clinically well-annotated cohort of DLBCL with osseous localizations (O-DLBCL), including PB-DLBCL. 103 O-DLBCL patients were included and compared with 63 (extra)nodal non-osseous (NO)-DLBCLs with germinal center B-cell phenotype (NO-DLBCL-GCB). Cell-of-origin (COO) was determined by immunohistochemistry and gene-expression-profiling (GEP) using (extended)-NanoString/Lymph2Cx. Mutational profiles were identified with targeted next-generation deep-sequencing, including 52 B-cell lymphoma-relevant genes. O-DLBCLs, including 34 PB-DLBCL, were predominantly classified as GCB-phenotype based on immunohistochemistry (74%) and NanoString analysis (88%). Unsupervised hierarchical clustering of an extended-NanoString/Lymph2Cx demonstrated significantly different GEP-clusters for PB-DLBCL as opposed to NO-DLBCL-GCB (P<0.001). Expression levels of 23 genes of two different targeted GEP-panels, indicated a centrocyte-like phenotype for PB-DLBCL, whereas NO-DLBCL-GCB showed a centroblast-like constitution. PB-DLBCL had significantly more frequent mutations in four GCB-associated genes, i.e. B2M, EZH2, IRF8, and TNFRSF14, compared to NO-DLBCL-GCB (P=0.031, P=0.010, P=0.047, and P=0.003). PB-DLBCL with its corresponding specific mutational profile were significantly associated with a superior overall survival compared to equivalent Ann Arbor limited-stage I/II NO-DLBCL-GCB (P=0.011). This study is the first to demonstrate that PB-DLBCL is characterized by a GCB-phenotype, with a centrocyte-like GEP-pattern and a GCB-associated mutational profile (both involved in immune surveillance) and a favorable prognosis. These novel biology-associated features provide evidence that PB-DLBCL represents a distinct extranodal DLBCL entity and its specific mutational landscape holds potential for targeted therapies (e.g. EZH2-inhibitors).