Abstract: Primary mediastinal large B-cell lymphoma (PMBCL) is recognized as a distinct entity in the World Health Organization classification. Currently, the diagnosis relies on consensus of histopathology, clinical variables, and presentation, giving rise to diagnostic inaccuracy in routine practice. Previous studies have demonstrated that PMBCL can be distinguished from subtypes of diffuse large B-cell lymphoma (DLBCL) based on gene expression signatures. However, requirement of fresh-frozen biopsy material has precluded the transfer of gene expression–based assays to the clinic. Here, we developed a robust and accurate molecular classification assay (Lymph3Cx) for the distinction of PMBCL from DLBCL subtypes based on gene expression measurements in formalin-fixed, paraffin-embedded tissue. A probabilistic model accounting for classification error, comprising 58 gene features, was trained on 68 cases of PMBCL and DLBCL. Performance of the model was subsequently evaluated in an independent validation cohort of 158 cases and showed high agreement of the Lymph3Cx molecular classification with the clinicopathological diagnosis of an expert panel (frank misclassification rate, 3.8%). Furthermore, we demonstrate reproducibility of the assay with 100% concordance of subtype assignments at 2 independent laboratories. Future studies will determine Lymph3Cx’s utility for routine diagnostic purposes and therapeutic decision making.

Abstract: A gene expression signature of tumor proliferation rate in mantle cell lymphoma (MCL) is an overriding molecular predictor of the length of survival following diagnosis. Many strongly proliferative MCL tumors have exceptionally high cyclin D1 mRNA levels and preferentially express short cyclin D1 mRNA isoforms. We demonstrate here that these short mRNAs are cyclin D1a isoforms with truncated 3′UTRs, not alternatively spliced cyclin D1b mRNA isoforms. Among 15 MCL tumors with truncated cyclin D1 mRNAs, 7 had genomic deletions in the CCND1 3′UTR region. In 3 others, CCND1 contained point mutations that created premature polyadenylation signals, giving rise to 1.5-kb mRNAs lacking most of the 3′UTR. Both types of genomic alteration created transcripts lacking mRNA destabilization elements present in the wild-type cyclin D1a mRNA. Premature polyadenylation due to a 3′UTR mutation also was present in the Z-138 MCL cell line, which expressed both truncated and full-length cyclin D1a mRNAs. In these cells, the half-life of the short cyclin D1a mRNA was much longer than that of the full-length mRNA. We conclude that alterations of CCND1 3′UTR structure can significantly increase its oncogenic effect and worsen the clinical course of MCL patients.

Abstract: Gene expression profiling has been used to distinguish two major subtypes of diffuse large B cell lymphoma (DLBCL), termed germinal center B cell-like (GCB) DLBCL and activated B cell-like (ABC) DLBCL. Following CHOP-like chemotherapy, GCB and ABC DLBCLs had distinct 5-year survival rates of ∼60% and ∼30%, respectively. Prognostic gene expression signatures in CHOP-treated DLBCL include the lymph node signature, which reflects a non-malignant host response, the MHC class II signature, both favorable when expressed and the proliferation signature which is adverse when expressed. The addition of rituximab to CHOP chemotherapy (R-CHOP) has significantly improved the outcome for DLBCL patients. We therefore investigated, if gene expression signatures that predicted survival among DLBCL patients treated with CHOP remained predictive for DLBCL patients treated with R-CHOP. Gene expression profiling was performed on 156 samples from previously untreated patients with DLBCL using Affymetrix U133 plus arrays. All patients received rituximab and CHOP-like chemotherapy. Samples were classified as GCB DLBCL, ABC DLBCL, or unclassified, and were assessed for expression of the lymph node and proliferation signatures. A Cox-proportional hazards model was used to determine the association of these gene expression features with overall survival (OS). 71 DLBCL samples were classified as GCB DLBCL, 63 as ABC DLBCL, and 22 were unclassified. The addition of rituximab improved OS for both GCB and ABC DLBCL compared to historical controls treated with CHOP-like chemotherapy alone. After a median follow-up of 2.3 years, GCB DLBCL had a more favorable OS than ABC DLBCL, with 3-year OS rates of 86% vs. 68% (p = 0.014). The 3-year OS rate of unclassified DLBCLs was 69%. The lymph node signature was associated with favorable OS (p = 0.023) and the proliferation signature with inferior OS (p = 0.009), whereas the MHC class II signature was not associated with OS (p = 0.44). In summary, addition of rituximab to CHOP-like chemotherapy improved OS for both GCB and ABC DLBCL but ABC DLBCL remained inferior to GCB DLBCL. The prognostic value of the lymph node and proliferation signatures were maintained in the context of R-CHOP therapy. An understanding of the biological attributes of DLBCL tumors that are reflected in these gene expression signatures remains critical to our ability to improve survival of these patients.

Abstract: Introduction: Plasmablastic lymphoma (PBL) is an aggressive B-cell lymphoma that predominantly occurs in patients with HIV or other causes of immunodeficiency. Frequent infection by the Epstein-Barr virus (EBV) and MYC translocations have been described as major features contributing to the pathogenesis of PBL. Prior studies examining the genetic landscape of PBL have largely relied on targeted capture-based sequencing approaches. As such, the molecular features of PBL remain to be fully explored. Here, we provide a comprehensive description of the genomic landscape of PBL using whole-genome and whole-exome sequencing to identify commonly perturbed pathways. Method: Archival diagnostic fresh frozen and formalin fixed paraffin embedded tissue biopsies from 58 PBL tumours were accrued from Lymphoma/Leukemia Molecular Profiling Project (LLMPP) sites, including 15 tumours from Ramis-Zaldivar et al., Haematolgica 2021. MYC rearrangements were identified by break-apart fluorescent in situ hybridization (FISH) and rearrangement partner was determined in a subset of tumours from capture or genome sequencing data using structural variant callers Manta, GRIDSS, and Delly. High confidence somatic mutations (SNVs/Indels) were identified in data from whole-genome (n=5) or whole-exome (n=53) sequencing through an ensemble voting approach utilizing four variant callers (Strelka2, Lofreq, Mutect2, SAGE). Mutation frequencies in known lymphoma-related genes were compared to activated B-cell (ABC)-DLBCL (Schmitz et al., NEJM 2018), as the closest tumour entity in terms of putative cell-of-origin differentiation stage, to identify differences in genetic aberrations. Candidate somatic copy number alterations (CNAs) were identified from exome and genome sequencing data, using CopywriteR and ControlFREEC, respectively, and high-confidence CNAs were determined using GISTIC2.0. Results: Within the study cohort, 81% of patients were male with a median age of 59 years (range 11-88). HIV and EBV statuses were available for 47% of patients and 95% of tumours, respectively, with 49% (13/27) of the patients being HIV+ and 69% (38/55) of tumours being EBV+. MYC rearrangement was observed in 60% (35/58) of PBLs, with IGH as the partner gene in 88% (21/24) of tumours. The most frequently mutated genes were STAT3 (38%), TP53 (22%), NRAS (21%), and TET2 (16%), consistent with previous studies, however novel mutations were seen in DUSP2 (21%), KLHL6 (16%), and BHLHE41 (16%). Recurrent CNAs included amplifications in 1q, whole gains of 7, 8q24, 11p12 and deletions affecting 4p16, 5p15, 10q11.22. While the mutational landscapes were similar between samples with and without a MYC translocation, the MYC-translocated PBLs showed more frequent amplification of 1q32.1. When stratifying by EBV status, STAT3 and SOCS1 mutations were more frequent in EBV-positive tumours, whereas TP53, TET2, KRAS, and MMRN2 mutations were associated with EBV-negativity. In comparison to ABC-DLBCL, PBLs were significantly enriched in STAT3 and NRAS mutations, and lacked common mutations affecting the NF-κB pathway (eg. MYD88, CD79B, and NFKBIZ 3' UTR mutations). Mutations in genes that are frequently mutated in ABC-DLBCLs, such as those associated with plasma cell differentiation (eg. PRDM1) or a memory B-cell fate (eg. TBL1XR1), were also not mutated in PBLs. Finally, genetic alterations associated with immune evasion, such as deletion of MHC I and II and mutations in B2M, CIITA, and CD58, were rarely observed. Conclusion: These data present a comprehensive overview of the genomic landscape of PBLs in a large cohort. We show frequent mutations involving the JAK-STAT and MAPK pathways, wherein the genetic landscape can be differentially characterized by EBV status and MYC rearrangement status. We show that PBLs are genetically distinct from ABC-DLBCLs, with absence of mutations in genes affecting the NF-κB pathway, immune evasion, and driving a memory B-cell fate. Disclosures Slack: Seagen: Consultancy, Honoraria. Raess: Scopio Labs: Research Funding. Holte: Gilead: Membership on an entity's Board of Directors or advisory committees; Roche: Membership on an entity's Board of Directors or advisory committees; Nordic: Membership on an entity's Board of Directors or advisory committees; Nanovector: Membership on an entity's Board of Directors or advisory committees, Other: lectures honorarias; Novartis: Membership on an entity's Board of Directors or advisory committees; Takeda: Membership on an entity's Board of Directors or advisory committees. Savage: Servier: Consultancy, Honoraria; Roche: Research Funding; BMS: Consultancy, Honoraria, Other: Institutional clinical trial funding; Astra-Zeneca: Consultancy, Honoraria; Merck: Consultancy, Honoraria, Other: Institutional clinical trial funding; AbbVie: Consultancy, Honoraria; Seattle Genetics: Consultancy, Honoraria; Takeda: Other: Institutional clinical trial funding; Beigene: Other: Institutional clinical trial funding; Genentech: Research Funding. Steidl: Seattle Genetics: Consultancy; Curis Inc.: Consultancy; Bayer: Consultancy; Epizyme: Research Funding; Trillium Therapeutics: Research Funding; AbbVie: Consultancy; Bristol-Myers Squibb: Research Funding. Rimsza: NanoString Technologies: Other: Fee-for-service contract. Morin: Foundation for Burkitt Lymphoma Research: Membership on an entity's Board of Directors or advisory committees; Celgene: Consultancy; Epizyme: Patents & Royalties. Scott: Abbvie: Consultancy; AstraZeneca: Consultancy; Rich/Genentech: Research Funding; BC Cancer: Patents & Royalties: Patent describing assigning DLBCL COO by gene expression profiling--licensed to NanoString Technologies. Patent describing measuring the proliferation signature in MCL using gene expression profiling. ; Incyte: Consultancy; Janssen: Consultancy, Research Funding; Celgene: Consultancy; NanoString Technologies: Patents & Royalties: Patent describing measuring the proliferation signature in MCL using gene expression profiling..

Abstract: Most MCLs harbor the t(11;14) translocation resulting in overexpression of cyclin D1 that promotes tumor proliferation. Gene expression profiling of 92 cyclin D1 positive MCL biopsy samples defined a gene expression signature of tumor proliferation that provided strong prognostic information: the quartile of patients with the highest proliferation gene expression signature had a 0.8 year median survival whereas the quartile with the lowest expression had a 6.7 year median survival (Cancer Cell 2003, 3:185). Sixteen of the 23 MCL samples in the highest proliferation quartile displayed exceptionally high cyclin D1 levels and preferentially expressed short cyclin D1 mRNA forms that lacked parts of the 3′UTR known to harbor RNA-destabilizing elements. To test whether expression of short cyclin D1 mRNA is due to genomic deletions, we screened DNA from all samples in the highest proliferation quartile by quantitative genomic PCR. We found genomic deletions of the 3′UTR region in 6 of the 10 samples that expressed short mRNAs. We next sequenced the proximal 400bp region of the 3′UTR in those samples with high expression of short mRNAs that did not have genomic deletions. In 2 samples, we found 2 distinct mutations: a 3bp deletion and an A insertion. Both of these mutations created a premature polyadenylation signal (AATAAA) about 300bp 3′ of the stop codon, giving rise to a predicted mRNA of 1.5kb that lacks the destabilizing elements present in the full-length form. We identified two MCL cell lines, Jeko-1 and Z138, that also showed high-level expression of the short cyclin D1 form. In Jeko-1 we found a deletion of the 3′UTR similar to that observed in the MCL patient samples, whereas the mechanism responsible for expression of the short cyclin D1 mRNA in Z138 remains unclear. In both lines, we confirmed that the half-life of the short cyclin D1 mRNA is longer than that of the full length mRNA. Our results demonstrate that MCLs increase cyclin D1 expression not only through the t(11;14), but also by the additional mechanisms of genomic deletions and mutations that create alternative polyadenylation sites that yield more stable mRNAs. Higher cyclin D1 expression increases proliferation and decreases survival, suggesting that strategies aimed at reducing cyclin D1 expression in MCL could have therapeutic promise.

Abstract: Mantle cell lymphoma is an aggressive B-cell neoplasm that displays heterogeneous outcomes after treatment. In 2003, the Lymphoma/Leukemia Molecular Profiling Project described a powerful biomarker—the proliferation signature—using gene expression in fresh frozen material. Herein, we describe the training and validation of a new assay that measures the proliferation signature in RNA derived from routinely available formalin-fixed paraffin-embedded (FFPE) biopsies. Methods Forty-seven FFPE biopsies were used to train an assay on the NanoString platform, using microarray gene expression data of matched fresh frozen biopsies as a gold standard. The locked assay was applied to pretreatment FFPE lymph node biopsies from an independent cohort of 110 patients uniformly treated with rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone. Seventeen biopsies were tested across three laboratories to assess assay reproducibility. Results The MCL35 assay, which contained a 17-gene proliferation signature, yielded gene expression of sufficient quality to assign an assay score and risk group in 108 (98%) of 110 archival FFPE biopsies. The MCL35 assay assigned patients to high-risk (26%), standard-risk (29%), and low-risk (45%) groups, with different lengths of overall survival (OS): a median of 1.1, 2.6, and 8.6 years, respectively (log-rank for trend, P 〈 .001). In multivariable analysis, these risk groups and the Mantle Cell Lymphoma International Prognostic Index were independently associated with OS ( P 〈 .001 for both variables). Concordance of risk assignment across the three independent laboratories was 100%. Conclusion The newly developed and validated MCL35 assay for FFPE biopsies uses the proliferation signature to define groups of patients with significantly different OS independent of the Mantle Cell Lymphoma International Prognostic Index. Importantly, the analytic and clinical validity of this assay defines it as a reliable biomarker to support risk-adapted clinical trials.

Abstract: Whole-genome integrative analyses in FL reveal that genes strongly influenced by copy number are highly enriched for NF-kB pathway regulators. Subsignatures of the NF-kB targets predict transformation in FL.

Abstract: Using current diagnostic criteria, primary mediastinal B cell lymphoma (PMBL) cannot be distinguished from other types of diffuse large B cell lymphoma (DLBCL) reliably. We used gene expression profiling to develop a more precise molecular diagnosis of PMBL. PMBL patients were considerably younger than other DLBCL patients, and their lymphomas frequently involved other thoracic structures but not extrathoracic sites typical of other DLBCLs. PMBL patients had a relatively favorable clinical outcome, with a 5-yr survival rate of 64% compared with 46% for other DLBCL patients. Gene expression profiling strongly supported a relationship between PMBL and Hodgkin lymphoma: over one third of the genes that were more highly expressed in PMBL than in other DLBCLs were also characteristically expressed in Hodgkin lymphoma cells. PDL2, which encodes a regulator of T cell activation, was the gene that best discriminated PMBL from other DLBCLs and was also highly expressed in Hodgkin lymphoma cells. The genomic loci for PDL2 and several neighboring genes were amplified in over half of the PMBLs and in Hodgkin lymphoma cell lines. The molecular diagnosis of PMBL should significantly aid in the development of therapies tailored to this clinically and pathogenetically distinctive subgroup of DLBCL.