Journal Logo

DLBCL Coast to Coast

Tesio, Melania

doi: 10.1097/HS9.0000000000000023

Institut Necker Enfants Malades, Paris, France

Correspondence: Melania Tesio (e-mail:

Citation: Tesio M. DLBCL Coast to Coast. HemaSphere, 2018;2:1.

Funding/support: None.

The authors have indicated they have no potential conflicts of interest to disclose.

This is an open access article distributed under the terms of the Creative Commons Attribution-Non Commercial-No Derivatives License 4.0 (CCBY-NC-ND), where it is permissible to download and share the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Integrating genomic, transcriptomic, functional, and clinical data from 1001 patient samples, a new study provides the first comprehensive characterization of diffuse large B-cell lymphoma.

Diffuse large B-cell lymphoma (DLBCL) is the most common lymphoid malignancy in adults. This tumor, highly heterogeneous from a genetic and clinical point of view, is believed to originate from either activated B cells (ABC subtype) or mature B cells transiting germinal centers within secondary lymphoid organs (GCB subtype).1 Although long-term remission following standard chemotherapy is achieved in nearly half of DLBCL cases, most patients who relapse do not survive. In clinical practice, risk stratification and treatment are based on clinical models for prognostication, such as the International Prognostic Index, which however provide limited insight into the genetic basis of the disease, its pathogenesis, or potential therapeutic targets. Transcriptomic analyses, instead, are more apt to identify molecular subgroups/pathways for therapeutic intervention. Yet, not all transcriptomic studies provided molecular signatures robustly correlating with clinical outcomes.2 Complicating this scenario, survival-predicting genes lack overlap across different studies. In addition, whereas next-generation sequencing studies highlighted the genetic heterogeneity of DLBCL, they could not identify the full mutational spectrum, mostly owing to limited patient samples that hindered the discovery of rare but important transforming lesions.

By integrating genomic, transcriptomic, functional, and clinical data from 1001 DLBCL samples, Reddy et al3 provided the first comprehensive characterization of this disease (Fig. 1). In a Herculean effort, recently published in Cell, the authors first performed whole-exome sequencing of 1001 clinically annotated DLBCL samples and 502 paired germline DNA samples. This exhaustive screening allowed the authors to identify 150 recurrently mutated genes. Remarkably, although the cohort size allowed 90% power for identifying mutations in 5% of the patients, only 27 out of the 150 genes identified had not been previously implicated in the disease. They included KLHL14, MGA, and SPEN, the most frequently altered gene among the newly identified ones. Also known as SHARP, SPEN encodes a transcriptional repressor previously shown to regulate marginal zone B cell development via inhibition of the NOTCH/RBPJ pathway. As such, SPEN mutations have been recently observed in some cases of marginal zone B cell lymphoma,4 where they were predicted to truncate the RBPJ-binding domain of the protein and thus to affect NOTCH2 signaling. Whereas truncating mutations were also observed by Reddy et al, their functional consequences in DLBCL might not strictly relate to aberrant NOTCH2 signaling, given that knocking out NOTCH2 did not impact cellular growth. In DLBCL cells, instead, SPEN lesions might alter the cellular epigenetic state, for instance by affecting the domains that interact with the histone deacetylase HDAC1 or the nuclear receptor corepressor SMRT,5 possibilities worthy of further investigation.

Figure 1

Figure 1

To functionally validate the mutations identified, Reddy and colleagues next complemented their exome sequencing data with a genome-wide CRISPR screen targeting 19,050 protein-coding genes. This analysis revealed that 1956 genes are essential in DLBCL as their silencing impacted cellular growth. They were enriched in the genetic drivers identified by exome sequencing, such as MGA and NCOR1, which were validated as DLBCL suppressors, since their CRISPR-mediated knockout enhanced cellular viability. Conversely, driver genes such as PAX5, HIST1HIE, and SF3B1 whose knockout reduced cellular viability were validated as essential oncogenes. Importantly, among them the authors identified 9 genes already known to be direct therapeutic targets of existing drugs, for example XPO1, BTK, and SF3B1.

To further connect genetic alterations and gene expression signatures, the authors performed RNA sequencing on 775 DLBCL samples. Thanks to a bioinformatic analysis integrating mutations, copy-number alterations, CRISPR scores, and gene expression profiles, the researchers could link distinct mutations to specific gene signatures. Confirming expected connections (such as MYC and RNA metabolic processes), the researchers revealed novel interesting associations (such as a link between NOTCH2 and cholesterol biosynthesis) as well as a lack of association between certain previously identified gene signatures (ie, stromal and immune gene signatures) to the overall mutation burden. To link these scenarios to clinical outcome, the authors next performed a multivariate analysis integrating patient survival data with multiple combinations of genetic drivers and gene expression markers. This strategy enabled Reddy et al to construct a genomic risk model, whereby 39 distinct combinations of genetic alterations and gene expression markers predict a specific death risk score. For example, alterations of the MYC gene combined with MYC expression defined patients with the worst prognosis, whereas a GCB signature combined with CD70 alterations defined patients with the most favorable outcome. Of note, when validated on an independent test set, this genomic risk predictor outperformed other prognostic models, based either exclusively on genetic alterations, exclusively on gene expression profiles or exclusively on clinical features, such as the International Prognostic Index.

While it might be challenging to integrate gene expression profiling and mutational analysis of more than 30 genes in every-day clinical practice, the study by Reddy et al offers an unprecedented coast-to-coast journey across the biology of one of the most common lymphoid malignancies, providing comprehensive data to model disease development and understand the biology of poor-prognosis DLBCL.

Back to Top | Article Outline


1. Alizadeh AA, Eisen MB, Davis RE, et al Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 2000;403:503–511.
2. Monti S, Savage KJ, Kutok JL, et al Molecular profiling of diffuse large B-cell lymphoma identifies robust subtypes including one characterized by host inflammatory response. Blood 2005;105:1851–1861.
3. Reddy A, Zhang J, Davis NS, et al Genetic and functional drivers of diffuse large B-cell lymphoma. Cell 2017;171:481–494.
4. Rossi D, Trifonov V, Fangazio M, et al The coding genome of splenic marginal zone lymphoma: activation of NOTCH2 and other pathways regulating marginal zone development. J Exp Med 2012;209:1537–1551.
5. Shi Y, Downes M, Xie W, et al Sharp, an inducible cofactor that integrates nuclear receptor repression and activation. Genes Dev 2001;15:1140–1151.
Copyright © 2018 The Authors. Published by Wolters Kluwer Health Inc., on behalf of the European Hematology Association.