Introduction. Diffuse large B-cell lymphoma (DLBCL) is the most common lymphoid malignancy with a diverse genomic appearance. To date, the coding genomes of almost 2000 DLBCLs have been recognized, and the translation of this genetic information into clinical practice is awaited.

Here, we have investigated translocations and protein expression of the hallmark DLBCL genes BCL2, BCL6 and MYC in a well-characterized sample series. The findings are correlated with gene expression, copy number alterations (CNAs) and exome-wide mutational annotations, and molecular data associated with clinical characteristics and treatment outcome.

Methods. A tissue microarray (TMA) was constructed from formalin fixed paraffin embedded (FFPE) tumor tissue of 152 primary DLBCL patients treated with R-CHOP or R-CHOP-like regimen. Translocation status was determined for BCL2, BCL6 and MYC genes with break-apart fluorescent in situ hybridization (FISH). BCL2, BCL6 and MYC protein expression was determined immunohistochemically. RNA sequencing data of the log2 transformed gene expression values and whole-exome sequencing annotations of the mutations and CNAs were derived for each case from the 1001 DLBCLs study (Reddy et. al. Cell 2017; 171:481-494).

Results. Overall, 55% of the samples were BCL2+. Of the BCL2+ cases, 58% had an underlying structural variation in the BCL2 locus. In the whole study cohort, structural variants of the BCL2 gene were detected in 35% of the cases, and they were associated with high gene expression and immunoreactivity (Figure 1A). BCL2 translocations were detected in 16% of the cases, and they were mutually exclusive with amplifications. Translocations and co-occurring coding mutations of the BCL2 gene were enriched in the GCB subtype and amplifications in the ABC/non-GCB DLBCLs. In the ABC/non-GCB DLBCLs, BCL2 copy number aberrations were associated with poor survival independent of IPI (Figure 1B).

High MYC immunoreactivity (>70% cut-off, 12 % of all cases) was evenly distributed between the subtypes. However, MYC translocations (14% of all cases) were associated with the GCB phenotype and elevated gene expression (Figure 1C). Nonsynonymous mutations of the MYC gene were only seen in the translocated cases. Co-occurring MYC and BCL2/BCL6 translocations (Double hit lymphomas, DHL) were recognized in 5.6% cases, most of which were GCB DLBCLs with poor outcome (Figure 1D). Dual protein expression (DPE) of BCL2 and MYC was detected in 9.5% of the cases. DPE lymphomas were evenly distributed between the subtypes, and they had adverse prognostic significance independent of IPI and molecular subtype (Figure 1D).

BCL6 translocations (15% of the cases) did not associate with elevated gene expression or coding mutations of the BCL6 gene (Figure 1E). Translocations were more prevalent in the non-GCB than the GCB DLBCLs and mutually exclusive with the MYD88 L265P mutations. Furthermore, BCL6 translocation status highlighted a subgroup of patients with a favorable outcome in the non-GCB DLBCLs independent of IPI (Figure 1F).

Conclusions. Our study characterizes the prevalence and clinical impact of hallmark structural variations in DLBCL. The results dignify structural, mutational and expression analysis of BCL2, BCL6 and MYC in the clinical practice, and increase our understanding of different entities within the well-established molecular subtypes.

Figure 1.A, C, E) Interplay between structural variants, gene expression, and nonsynonymous SNVs of BCL2, MYC and BCL6, and with immunohistochemical staining scores for BCL2 and MYC proteins are illustrated. B) CNAs of BCL2 and poor survival in patients with non-GCB DLBCL. D) DHLs and DPEs (other than DHL associated) and poor survival. F) BCL6 translocations and favorable outcome in patients with non-GCB DLBCL.


Leppa:Roche: Consultancy, Honoraria, Research Funding; Takeda: Consultancy, Research Funding; Janssen: Consultancy, Research Funding; Celgene: Consultancy; Bayer: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.