Gene expression profiles for B-cell lymphomas and their applications

Description

FIELD OF INVENTION

[0001] The present invention relates to gene expression profiles for B-cell lymphoma. More specifically, the present invention relates to gene expression profiles for diagnosis, prognosis or therapy selection for aggressive B-cell lymphomas.BACKGROUND OF THE INVENTION

[0002] The biological heterogeneity in diffuse large B-cell lymphoma (DLBCL) has prompted significant effort to define distinct molecular subgroups within the disease 1-3< . Accordingly, the most recent revision of the WHO classification divides tumors with DLBCL morphology into cell-of-origin (COO) molecular subtypes: activated B-cell-like (ABC) and germinal center B-cell-like (GCB) subtypes and recognizes high-grade B-cell lymphoma with MYC and BCL2 and / or BCL6 rearrangements (HGBL-DH / TH) 4< , which includes tumors with either DLBCL or high-grade morphology. Approximately 8% of tumors with DLBCL morphology are HGBL-DH / TH and all HGBL-DH / TH with BCL2 translocations (HGBL-DH / TH-BCL2) of DLBCL morphology belong to the GCB molecular subgroup 5,6< . Clinically, despite the generally superior prognosis of GCB-DLBCLs, HGBL-DH / TH-BCL2 patients have poor outcomes 7-12< , prompting treatment of such tumors with dose intensive immunochemotherapy. Genomic studies in DLBCL have identified recurrent mutations and revealed the association of many with COO 13-16< . Genomic landscape studies have defined genetic subgroups based on somatic mutation and structural variants 17-19< .

[0003] Qing Ye et al., Oncotarget, Vol. 7, No. 3, pp. 2401-2416, 2015 describe the characterization of double-hit B-cell lymphomas, being a group of tumors characterized by concurrent translocations of MYC and BCL2, BCL6, or other genes.

[0004] Merron and Davies, Best Practice & Research Clinical Haematology, Vol. 31, No. 3, pp. 233-240, 2018 is a review article summarizing the characterization of double / triple hit lymphoma as a high-grade B-cell lymphoma with rearrangements of MYC and BCL2 and / or BCL6 and treatment options therefor.

[0005] Anupama Reddy et al., Cell, Vol. 171, No. 2, pp. 481-494, 2017, outlines the characterization of diffuse large B cell lymphoma (DLBCL) and the whole exome sequencing and transcriptome sequencing of a cohort of 1001 DLBCL patients to define the landscape of 150 genetic drivers of the disease.SUMMARY OF THE INVENTION

[0006] The present invention is defined by the appended claims.

[0007] In one aspect, the present invention provides a method for classifying an aggressive B-cell lymphoma as defined in claims 10 to 14..

[0008] In some embodiments, the test sample may be a biopsy.

[0009] In some embodiments, the aggressive B-cell lymphoma may be a diffuse large B-cell lymphoma (DLBCL) or high-grade B-cell lymphoma (HGBL).

[0010] In some embodiments, the subject may be a human.

[0011] In an alternative aspect, the present invention provides a kit as defined in claims 1 to 9.

[0012] This summary of the disclosure does not necessarily describe all features of the invention.BRIEF DESCRIPTION OF THE DRAWINGS

[0013] These and other features of the invention will become more apparent from the following description in which reference is made to the appended drawings as follows. FIGURE 1 shows the patient flow for the discovery cohort, two independent validation cohorts and NanoString cohort. ABC, activated B-cell-like subtype; GCB, germinal center B-cell-like subtypes; UNC, unclassified; DHIT, double-hit. FIGURE 2A shows the RNAseq DHITsig scores from 171 GCB-DLBCL used to train and test the DLBCL90 assay. The tumors are arrayed from left to right with increasing DHITsig scores with tumors with a score below 0 being designated DHITsig-neg and above 0 being DHITsig-pos. Selected tumors had digital expression performed using a codeset that contained all 104 genes in the RNAseq model. FIGURE 2B shows the RNAseq DHITsig scores from 171 GCB-DLBCL used to train and test the DLBCL90 assay. The tumors are arrayed from left to right with increasing DHITsig scores with tumors with a score below 0 being designated DHITsig-neg and above 0 being DHITsig-pos. Selected tumors were used to "train" the threshold for the DLBCL90 assay. FIGURE 3 shows the DHITsig score from the RNAseq model (X-axis) against the DHITscore from the DLBCL90 assay in 171 GCB-DLBCL. The 72 biopsies were used to establish the thresholds for the assay. Arrows highlight the 5 (3%) tumors that were frankly misclassified. FIGURE 4A shows comparisons between the linear predictor score (LPS) from the Lymph2Cx (Scott, Mottok et al J Clin Oncol 2015) and the DLBCL90 assay. The figure shows the uncalibrated DLBCL90 LPS scores. Six (6) tumors (2%) were moved from a definitive category to Unclassified (or vice versa). FIGURE 4B shows comparisons between the linear predictor score (LPS) from the Lymph2Cx (Scott, Mottok et al J Clin Oncol 2015) and the DLBCL90 assay. The figures shows the calibrated DLBCL90 LPS scores, where 116.6 points were removed from the uncalibrated scores. Six (6) tumors (2%) were moved from a definitive category to Unclassified (or vice versa). FIGURE 5A shows the gene expression-based model of 104 genes based on HGBL-DH / TH-BCL2 status showing the importance score with 95% confidence interval of the 104 most significantly differentially expressed genes between HGBL-DH / TH-BCL2 and GCB-DLBCL. Genes with dark grey and light grey bars are over- and under-expressed in HGBL-DH / TH-BCL2, respectively. FIGURE 5B shows the mean Z-score of genes over- or under-expressed in HGBL-DH / TH-BCL2 is shown in the form of a heatmap, with the 157 patient biopsies shown as columns. DHITsig groups identified by the signature are shown below the heat map. The status of MYC, BCL2 and BCL6 genetic alterations, HGBL-DH / TH-BCL2, WHO categories and MYC / BCL2 dual protein expresser (DPE) status are displayed beneath the heatmap. FIGURE 6A shows the prognostic association of DHIT signature in DLBCL patients treated with R-CHOP. Kaplan Meier curves of the DHITsig-pos GCB-DLBCL (black) vs DHITsig-neg GCB-DLBCL (light grey) vs ABC-DLBCL (dark grey) for TTP in British Columbia Cancer cohort. HR; hazard ratio. FIGURE 6B shows the prognostic association of DHIT signature in DLBCL patients treated with R-CHOP. Kaplan Meier curves of the DHITsig-pos GCB-DLBCL (black) vs DHITsig-neg GCB-DLBCL (light grey) vs ABC-DLBCL (dark grey) for DSS in British Columbia Cancer cohort. HR; hazard ratio. FIGURE 6C shows the prognostic association of DHIT signature in DLBCL patients treated with R-CHOP. Kaplan Meier curves of the DHITsig-pos GCB-DLBCL (black) vs DHITsig-neg GCB-DLBCL (light grey) vs ABC-DLBCL (dark grey) OS in British Columbia Cancer cohort. HR; hazard ratio. FIGURE 6D shows the prognostic association of DHIT signature in DLBCL patients treated with R-CHOP. Kaplan Meier curves of the DHITsig-pos GCB-DLBCL (black) vs DHITsig-neg GCB-DLBCL (light grey) vs ABC-DLBCL (dark grey ) for OS in the Reddy et al. validation cohort. HR; hazard ratio. FIGURE 7A shows Kaplan Meier curves of the cases with HGBL-DH / TH-BCL2 (black) vs non-HGBL-DH / TH-BCL2 (grey) within DHITsig-pos GCB-DLBCL for TTP. FIGURE 7B shows Kaplan Meier curves of the cases with HGBL-DH / TH-BCL2 (black) vs non-HGBL-DH / TH-BCL2 (grey) within DHITsig-pos GCB-DLBCL for DSS. FIGURE 7C shows Kaplan Meier curves of the cases with HGBL-DH / TH-BCL2 (black) vs non-HGBL-DH / TH-BCL2 (grey) within DHITsig-pos GCB-DLBCL for OS. FIGURE 8A shows Kaplan Meier curves of cases stratified by DHIT signature combined with DPE status in GCB-DLBCL for TTP. FIGURE 8B shows Kaplan Meier curves of cases stratified by DHIT signature combined with DPE status in GCB-DLBCL for DSS. FIGURE 8C shows Kaplan Meier curves of cases stratified by DHIT signature combined with DPE status in GCB-DLBCL for OS. FIGURE 9A shows the genetic, molecular and phenotypic features of DHIT signature comparing Ki67 staining by IHC between DHITsig-pos, DHITsig-neg GCB-DLBCL and ABC-DLBCL. FIGURE 9B shows the genetic, molecular and phenotypic features of DHIT signature comparing linear predictor score (LPS), provided by Lymph2Cx assay, between DHITsig-pos, DHITsig-neg GCB-DLBCL and ABC-DLBCL. Purple dots represent the HGBL-DH / TH-BCL2 tumors. FIGURE 9C shows the genetic, molecular and phenotypic features of DHIT signature comparing IHC staining pattern of CD10 (MME) and MUM1 (IRF4) between DHITsig-pos and DHITsig-neg GCB-DLBCL cases. FIGURE 9D shows the genetic, molecular and phenotypic features of DHIT signature comparing mean Z scores of DZ, IZ and LZ signature gens (20 genes each) between DHITsig-pos and -neg groups. DZ; dark-zone, IZ; intermediate-zone, LZ; light-zone. FIGURE 10 shows the bar plot of the gene set enrichment analysis (GSEA). This analysis include differential expression genes between DHITsig-pos and -neg groups with FDR< 0.1, and log2 Fold Change > abs(0.5). FIGURE 11A shows the genetic, molecular and phenotypic features of DHIT signature comparing fraction of tumor-infiltrating T-cells (CD3 (left), CD4 (center) and CD8 (right) positive T-cells) measured by flow cytometry between DHITsig-pos, DHITsig-neg GCB-DLBCL and ABC-DLBCL. FIGURE 11B shows the genetic, molecular and phenotypic features of DHIT signature comparing frequencies of MHC-I and -II double-negative (purple), isolated MHC-II negative, isolated MHC-I negative and MHC-I and -II double positive cases in DHITsig-pos (left) and DHITsig-neg cases (right). FIGURE 11C shows the genetic, molecular and phenotypic features of DHIT signature by Forest plots summarizing the results of Fisher's exact tests comparing the frequency of mutations affecting individual genes in DHITsig-neg (left) and DHITsig-pos (right) GCB-DLBCL tumors. Significantly enriched genes in either DHITsig-pos or DHITsig-neg cases (FDR<. 10) are represented. Log10 odds ratios and 95% confidence intervals are shown (left panel). Bar plots representing the frequency of mutations in either DHITsig-pos or -neg groups (right panel). FIGURE 12 shows a heatmap of the result of clustering of primary samples with GCB-DLBCL along with 8 GCB-DLBCL cell lines (Pfeiffer, Toledo, SU-DHL-8, WSU-NHL, HT, SU-DHL-5, SU-DHL-4, SU-DHL-10) by DHIT signature. FIGURE 13 shows the gene expression-based model for the DHIT signature in which the DLBCL90 assay is shown in the form of a heatmap, with the 30 informative genes shown as rows, and the cases shown as columns, separated into 220 GCB- and Unclassified DLBCLs. The tumors are arrayed from highest DHIT sig score on the left to lowest DHITsig score on the right. DHITsig groups identified by the signature are shown below the heat map. FIGURE 14A shows the gene expression-based model for the DHIT signature in which the DLBCL90 assay is shown in the form of a heatmap, with the 88 transformed follicular lymphoma (tFL) with DLBCL morphology. The tumors are arrayed from highest DHIT sig score on the left to lowest DHITsig score on the right. DHITsig groups identified by the signature are shown below the heat map. FIGURE 14B shows the gene expression-based model for the DHIT signature in which the DLBCL90 assay is shown in the form of a heatmap, with the 26 high-grade B-cell lymphomas. The tumors are arrayed from highest DHIT sig score on the left to lowest DHITsig score on the right. DHITsig groups identified by the signature are shown below the heat map. The status of MYC, BCL2 and BCL6 genetic alterations, HGBL-DH / TH-BCL2 status and WHO categories are also shown. FIGURE 15A shows the prognostic association of DLBCL90 in DLBCL patients treated with R-CHOP by Kaplan Meier curves of the GCB-DLBCL (light grey) vs DHITsig-pos and -ind (black) vs Unclassified (medium grey) vs ABC-DLBCL (dark grey) for TTP in 322 patients with de novo tumors of DLBCL morphology treated with R-CHOP. FIGURE 15B shows the prognostic association of DLBCL90 in DLBCL patients treated with R-CHOP by Kaplan Meier curves of the GCB-DLBCL (light grey) vs DHITsig-pos and -ind (black) vs Unclassified (medium grey) vs ABC-DLBCL (dark grey) for DSS in 322 patients with de novo tumors of DLBCL morphology treated with R-CHOP. FIGURE 15C shows the prognostic association of DLBCL90 in DLBCL patients treated with R-CHOP by Kaplan Meier curves of the GCB-DLBCL (light grey) vs DHITsig-pos and -ind (black) vs Unclassified (medium grey) vs ABC-DLBCL (dark grey) for PFS in 322 patients with de novo tumors of DLBCL morphology treated with R-CHOP. FIGURE 15D shows the prognostic association of DLBCL90 in DLBCL patients treated with R-CHOP by Kaplan Meier curves of the GCB-DLBCL (light grey) vs DHITsig-pos and -ind (black) vs Unclassified (medium grey) vs ABC-DLBCL (dark grey) for OS in 322 patients with de novo tumors of DLBCL morphology treated with R-CHOP. DETAILED DESCRIPTION

[0014] The present disclosure provides, in part, methods and reagents for classifying and identifying aggressive B-cell lymphomas. In alternative aspects, the present disclosure provides methods and reagents for selecting therapies and / or identifying candidates for therapies for aggressive B-cell lymphomas.

[0015] B-cell lymphomas can be diagnostically classified into Hodgkin and non-Hodgkin lymphomas. Most B-cell lymphomas are non-Hodgkin lymphomas and include Burkitt lymphoma, chronic lymphocytic leukemia / small lymphocytic lymphoma (CLL / SLL), diffuse large B-cell lymphoma, follicular lymphoma, mantle cell lymphoma, etc. Diffuse large B-cell lymphoma (DLBCL) is biologically heterogeneous. The WHO classification divides tumors with DLBCL morphology into cell-of-origin (COO) molecular subtypes: activated B-cell-like (ABC) and germinal center B-cell-like (GCB) subtypes and recognizes high-grade B-cell lymphoma with MYC and BCL2 and / or BCL6 rearrangements (HGBL-DH / TH) as including tumors with either DLBCL or high-grade morphology. Approximately 8% of tumors with DLBCL morphology are HGBL-DH / TH and all HGBL-DH / TH with BCL2 translocations (HGBL-DH / TH-BCL2) of DLBCL morphology belong to the GCB molecular subgroup. High grade B cell lymphoma (HGBL) is a heterogeneous entity with morphologic and genetic features intermediate between DLBCL and Burkitt lymphoma (BL) or blastoid morphology. Many patients with HGBL also have concurrent MYC, BCL2 and / or BCL6 rearrangements documented by FISH. HGBL without MYC and BCL2 and / or BCL6 have been termed HGBL-NOS. An "aggressive" B-cell lymphoma, as used herein, is a fast-growing non-Hodgkin lymphoma that is derived from a B lymphocyte.

[0016] In one aspect, the present disclosure provides a method of classifying an aggressive B-cell lymphoma by preparing a gene expression profile for two or more genes listed in any of Tables 1 to 4 from a test sample and classifying the test sample into two molecular subgroups: an aggressive B-cell lymphoma having a positive DHIT signature (DHITsig-pos) or an aggressive B-cell lymphoma having a negative DHIT signature (DHITsig-neg), based on the gene expression profile. Table 1 Gene Name ensembl_gene_id* 1AC104699.1ENSG000002242202ACPPENSG000000142573ADTRPENSG000001118634AFMIDENSG000001830775ALOX5ENSG000000127796ALS2ENSG000000033937ANKRD33BENSG000001642368ARHGAP25ENSG000001632199ARID3BENSG0000017936110ARPC2ENSG0000016346611ASS1P1ENSG0000022051712ATF4ENSG0000012827213BATFENSG0000015612714BCL2A1ENSG0000014037915CAB39ENSG0000013593216CCDC78ENSG0000016200417CCL17ENSG0000010297018CCL22ENSG0000010296219CD24ENSG0000027239820CD80ENSG0000012159421CDK5R1ENSG0000017674922CFLARENSG0000000340223COBLL1ENSG0000008243824CPEB4ENSG0000011374225CR2ENSG0000011732226CTD-3074O7.5ENSG0000025551727DANCRENSG0000022695028DGKGENSG0000005886629DOCK10ENSG0000013590530EBI3ENSG0000010524631EIF4EBP3ENSG0000024305632ETV5ENSG0000024440533FAM216AENSG0000020485634FCRL5ENSG0000014329735FHITENSG0000018928336GALNT6ENSG0000013962937GAMTENSG0000013000538GNG2ENSG0000018646939GPR137BENSG0000007758540HAGHLENSG0000010325341HIVEP1ENSG0000009595142HMSDENSG0000022188743HRKENSG0000013511644IL10RAENSG0000011032445IL21RENSG0000010352246IRF4ENSG0000013726547JCHAINENSG0000013246548LINC00957ENSG0000023531449LRRC75A-AS1ENSG0000017506150LTAENSG0000022697951LY75ENSG0000005421952MACROD1ENSG0000013331553MIR155HGENSG0000023488354MREGENSG0000011824255MVPENSG0000001336456MYCENSG0000013699757MYEOVENSG0000017292758NCOA1ENSG0000008467659NMRAL1ENSG0000015340660OR13A1ENSG0000025657461PARP15ENSG0000017320062PEG10ENSG0000024226563PIK3CD-AS2ENSG0000023178964POU3F1ENSG0000018566865PPP1R14BENSG0000017345766PTPRJENSG0000014917767QRSL1ENSG0000013034868RASGRF1ENSG0000005833569RFFLENSG0000009287170RGCCENSG0000010276071RPL13ENSG0000016752672RPL35ENSG0000013694273RPL6ENSG0000008900974RPL7ENSG0000014760475RPS8ENSG0000014293776SEMA7AENSG0000013862377SFXN4ENSG0000018360578SGCEENSG0000012799079SGPP2ENSG0000016308280SIAH2ENSG0000018178881SIGLEC14ENSG0000025441582SLC25A27ENSG0000015329183SLC29A2ENSG0000017466984SMARCB1ENSG0000009995685SMIM14ENSG0000016368386SNHG11ENSG0000017436587SNHG17ENSG0000019675688SNHG19ENSG0000026026089SNHG7ENSG0000023301690SOX9ENSG0000012539891SPTBN2ENSG0000017389892ST8SIA4ENSG0000011353293STAT3ENSG0000016861094SUGCTENSG0000017560095SYBUENSG0000014764296TACC1ENSG0000014752697TERTENSG0000016436298TLE4ENSG0000010682999TNFSF8ENSG00000106952100UQCRHENSG00000173660101VASPENSG00000125753102VOPP1ENSG00000154978103WDFY1ENSG00000085449104WNK2ENSG00000165238* Zerbino et al. Ensembl 2018. Nucleic Acids Res. 2018 Jan 4;46(D1):D754-D761. Gene annotations used by featureCounts for extracting read counts are from Ensembl gene build 87.

[0017] In an alternative aspect, an aggressive B-cell lymphoma can be classified by preparing or obtaining a gene expression product e.g., a molecule produced as a result of gene transcription, such as a nucleic acid or a protein, from a test sample, preparing or obtaining a gene expression profile for two or more genes listed in any of Tables 1 to 4 from the gene expression product and classifying the test sample into two molecular subgroups: an aggressive B-cell lymphoma having a positive DHIT signature (DHITsig-pos) or an aggressive B-cell lymphoma having a negative DHIT signature (DHITsig-neg), based on the gene expression profile.

[0018] In some embodiments, an aggressive B-cell lymphoma can be classified by determining the expression of two or more genes ("gene expression") listed in any of Tables 1 to 4 from a test sample, such as a cryosection of a fresh frozen biopsy or a formalin-fixed paraffin-embedded tissue (FFPET) biopsy prepared using standard techniques (see, e.g., Keirnan, J. (ed.), Histological and Histochemical Methods: Theory and Practice, 4th edition, Cold Spring Harbor Laboratory Press (2008)). Gene expression can be determined by isolating or otherwise analyzing a nucleic acid (such as RNA or DNA) from the test sample using standard techniques and commercially available reagents such as, without limitation, QIAamp DNA FFPE Tissue Kit, RNAEASY ™< FFPE Kit, AllPREP FFPE Kit (Qiagen, Venlo, Netherlands); and MAGMAX ™< FFPE DNA Isolation Kit (Life Technologies, Carlsbad, CA)).

[0019] In some embodiments, gene expression can be determined by isolating or otherwise analyzing a protein or polypeptide from the test sample using standard techniques and commercially available reagents such as, without limitation, immunohistochemistry techniques, ELISA, western blotting and mass spectrometry.

[0020] By "gene expression profile" or "signature" as used herein, is meant data generated from one or more genes listed in any of Tables 1 to 4 that make up a particular gene expression pattern that may be reflective of level of expression, cell lineage, stage of differentiation, or a particular phenotype or mutation. In some embodiments, a gene expression profile or signature includes data generated from two or more of the genes listed in Table 1 or 3, e.g., 2, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 60, 75, 80, 85, 90, 95, 100, or 104 of the genes listed in Tables 1 or 3. In some embodiments, a gene expression profile or signature includes data generated from two or more of the genes listed in Tables 2 or 4, e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30 of the genes listed in Table 2 or 4. In some embodiments, a gene expression profile or signature includes data generated from all of the genes listed in Table 2 or 4. In some embodiments, a gene expression profile or signature includes data generated from substantially all of the genes listed in Table 2 or 4 e.g. 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30 of the genes listed in Table 2 or 4. In some embodiments, a gene expression profile or signature is "balanced" i.e. includes data generated from similar numbers of genes that are overexpressed and underexpressed as listed in any of Tables 1 to 4. Table 2 Gene Name Accession No. 1AFMIDNM_001010982.42ALOX5NM_000698.23BATFNM_006399.34CD24NM_013230.25CD80NM_005191.36CDK5R1NM_003885.27EBI3NM_005755.28GAMTNM_138924.19GPR137BNM_003272.310IL21RNM_021798.211IRF4NM_002460.112JCHAINNM_144646.313LY75NM_002349.214MIR155HGNR_001458.315MYCNM_002467.316OR13A1NM_001004297.217PEG10NM_001040152.118QRSL1NM_018292.219RFFLNM_001017368.120RGCCXM_011535051.121SEMA7ANM_001146029.122SGPP2NM_152386.223SLC25A27NM_004277.424SMIM14NM_174921.125SNHG19NR_132114.126STAT3NM_003150.327SYBUNM_001099744.128TNFSF8NM_001244.329VASPNM_003370.330VOPP1NM_030796.3

[0021] A "gene expression profile" or "signature" can be prepared by generating data relating to the level of expression of two or more genes listed in in any of Tables 1 to 4, whether absolute or relative to a synthetic control or standard, in a sample, such as a biopsy sample. In some embodiments, the sample may be obtained from a subject prior to, during, or following diagnosis or treatment for an aggressive B-cell lymphoma, or to monitor the progression of an aggressive B-cell lymphoma, or to assess risk for development of an aggressive B-cell lymphoma, or to calculate risk of relapse. In some embodiments, a gene expression profile or signature can be prepared relative to a synthetic control to, for example, standardize lot-to-lot variation. The level of expression of a gene may be determined based on the level of a nucleic acid e.g., RNA, such as mRNA, encoded by the gene. Alternatively, level of expression of a gene may be determined based on the level of a protein or polypeptide or fragment encoded by the gene. In some embodiments, the gene expression data may be "digital," for example, based on the generation of sequence tags. In alternative embodiments, the gene expression data may be "analog," for example, based on hybridization of nucleic acids. Any suitable quantification method as described herein or known in the art can be used, such as without limitation, PCR, quantitative RT-PCR, real-time PCR, digital PCR, RNA amplification, in situ hybridization, immunohistochemistry, immunocytochemistry, FACS, SAGE, RNAseq, etc. In some embodiments, a gene expression profile can be prepared using microarrays, for example, nucleic acid or antibody microarrays. In some embodiments, a gene expression profile can be prepared with RNA gene expression data using the nCounter ®< gene expression assay available from NanoString Technologies, Inc. (Kulkarni, M. M., "Digital Multiplexed Gene Expression Analysis Using the NANOSTRING™ NCOUNTER™ System," Current Protocols in Molecular Biology. 94: 25B.10.1-25B.10.17 (2011); Geiss et al., Nature Biotechnology, 26: 317-325 (2008); or U.S. Patent 7,919,237).

[0022] In some embodiments, a gene expression profile can be prepared by generating data relating to the level of expression of Lymph3x genes, as set forth in Table 6 and described in PCT publication WO / 2018 / 231589, Staudt et al., published December 20, 2018, in addition to the two or more genes listed in in any of Tables 1 to 4. In some embodiments, a gene expression profile" can be prepared by generating data relating to the level of expression of BCL2, FCGR2B and / or PVT1, in addition to the two or more genes listed in in any of Tables 1 to 4 and / or Table 6.

[0023] In some embodiments, a gene expression profile can be prepared and classified as follows. Gene expression levels of two or more of the genes listed in Table 1 or 2 would be obtained from a sample using a suitable technology (for example, RNAseq or the NanoString platform). In one embodiment, using gene expression from RNAseq, the expression of the 104 genes from Table 1 can be inputted into an algorithm, for example: DHITsig Score = ∑ i = 1 m ImportanceScore ∗ log 10 p 1 p 2 where m is the total number of 104 genes that can be matched in a given RNAseq data, p1 is the p value based on t test of a given sample's gene expression value against a normal distribution with mean and standard deviation from DHITsig-pos group, p2 is the p value based on t test of a given sample's gene expression value against a normal distribution with mean and standard deviation from DHITsig-neg group, and the Importance Score are the values in Table 3, to produce a score with an assignment made into the DHIT signature subgroups based on the score obtained, as described herein.

[0024] In another embodiment, using gene expression from the NanoString platform, the gene expression for the genes in Table 2, would be inputted into an algorithm, for example: DHITsig Score = ∑ i = 1 m ImportanceScore ∗ gene expression where m is the total number of genes (in this example, 30), the Importance Score are the values in Table 3, and gene expression is the gene expression of gene m, after the gene expression has been divided by the geometric mean of one or more (or all) of the house keeping genes (DNAJB12, GIT2, GSK3B, IK, ISY1, OPAl, PHF23, R3HDM1, TRIM56, UBXN4, VRK3, WAC and / or WDR55 listed in Table 6 ), multiplied by 1000 and log 2 transformed, to produce a score with an assignment made into the DHIT signature subgroups based on the score obtained, as described herein.

[0025] A "sample" can be a "test sample" and may be any organ, tissue, cell, or cell extract isolated from a subject, such as a sample isolated from a mammal having an aggressive B-cell lymphoma, or a subgroup or subtype of an aggressive B-cell lymphoma, such as a DLBCL, ABC-DLBCL, GCB-DLBCL, HGBL-DH / TH, HGBL-DH / TH-BCL2, HGBL-NOS, etc. For example, a sample can include, without limitation, cells or tissue (e.g., from a biopsy) or any other specimen, or any extract thereof, obtained from a patient (human or animal), test subject, or experimental animal. In some embodiments, it may be desirable to separate cancerous cells from non-cancerous cells in a sample. A sample may be from a cell or tissue known to be cancerous or suspected of being cancerous. Accordingly, a sample can include without limitation a cryosection of a fresh frozen biopsy, a formalin-fixed paraffin-embedded tissue (FFPET) biopsy, a cryopreserved diagnostic cell suspension, or peripheral blood.

[0026] As used herein, a "subject" may be a human, non-human primate, rat, mouse, cow, horse, pig, sheep, goat, dog, cat, etc. The subject may be a clinical patient, a clinical trial volunteer, an experimental animal, etc. The subject may be suspected of having or at risk for having an aggressive B-cell lymphoma or be diagnosed with an aggressive B-cell lymphoma. In some cases, the subject may have relapsed after treatment for a B-cell lymphoma, for example, treatment with rituximab, cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate and prednisone (R-CHOP).

[0027] Gene expression profiles, prepared as described herein, can be used to classify an aggressive B-cell lymphoma into two molecular subgroups: an aggressive B-cell lymphoma having a positive DHIT signature (DHITsig-pos) or an aggressive B-cell lymphoma having a negative DHIT signature (DHITsig-neg). These molecular subgroups can be used for prognosis and / or to determine treatment options.

[0028] Accordingly, in an alternative aspect, the present disclosure provides a method for determining the prognosis of a subject diagnosed with an aggressive B-cell lymphoma by providing a gene expression profile for two or more genes listed in in any of Tables 1 to 4 from a test sample from the subject and classifying the test sample into an aggressive B-cell lymphoma subgroup having a positive DHIT signature (DHITsig-pos) or an aggressive B-cell lymphoma subgroup having a negative DHIT signature (DHITsig-neg) based on said gene expression profile, as described herein, where DHITsig-pos is predictive of a poor prognosis and DHITsig-neg is predictive of a good prognosis.

[0029] In some embodiments, prognosis or outcome may refer to overall or disease-specific survival, event-free survival, progression-free survival or outcome in response to a particular treatment or therapy. In some embodiments, the prognostic methods described herein may be used to predict the likelihood of long-term, disease-free survival i.e., that the subject will not suffer a relapse of the underlying aggressive B-cell lymphoma within a period of at least one year,or at least two years, or at least three years, or at least four years, or at least five years, or at least ten or more years, following initial diagnosis or treatment and / or will survive at least one year,or at least two years, or at least three years, or at least four years, or at least five years, or at least ten or more years, following initial diagnosis or treatment.

[0030] In some embodiments, the methods described herein can be used to screen tumors with DLBCL morphology for FISH testing, for example, for FISH testing for rearrangements involving MYC, BCL2 and / or BCL6.

[0031] In another aspect, the present disclosure provides a method for selecting a therapy, or for predicting a response to a therapy, for an aggressive B-cell lymphoma by determining whether the aggressive B-cell lymphoma has a positive DHIT signature (DHITsig-pos) or a negative DHIT signature (DHITsig-neg) as described herein; and selecting a therapy effective to treat the molecular subgroup thus determined.

[0032] In another aspect, the present disclosure provides a method for identifying a subject with an aggressive B-cell lymphoma for a therapy, or for predicting the response of a subject with an aggressive B-cell lymphoma to a therapy, by determining whether the aggressive B-cell lymphoma has a positive DHIT signature (DHITsig-pos) or a negative DHIT signature (DHITsig-neg) as described herein; and determining whether the candidate is likely to respond to a therapy effective to treat the molecular subgroup thus determined. By "predicting the response of a subject with an aggressive B-cell lymphoma to a therapy" is meant assessing the likelihood that a subject will experience a positive or negative outcome with a particular treatment. As used herein, "indicative of a positive treatment outcome" refers to an increased likelihood that the subject will experience beneficial results from the selected treatment (e.g., complete or partial remission). "Indicative of a negative treatment outcome" is intended to mean an increased likelihood that the patient will not benefit from the selected treatment with respect to the progression and / or relapse of the underlying aggressive B-cell lymphoma.

[0033] Therapies for B-cell lymphoma include, without limitation, rituximab, cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate and prednisone (R-CHOP), as well as alternate therapies, such as a dose intensive immunochemotherapy, a cell-based therapy such as CAR T-cell therapy, a BCL2 inhibitor, an enhancer of zeste homolog 2 (EZH2) inhibitor, a histone deacetylase inhibitor, arachidonate 5-lipoxygenase inhibitor, a Bruton's tyrosine kinase inhibitor (such as ibrutinib), a PIM kinase inhibitor (such as SGI-1776), a histone deacetylase inhibitor (such as belinostat or vorinostat), a PI3K inhibitor (such as copanlisib or buparlisib), a protein kinase C inhibitor (such as sotrastaurin), immunomodulatory drugs (IMiD - such as lenalidomide) newer generation anti-CD20 antibodies, etc.

[0034] In some embodiments, when the molecular subgroup is determined to be DHITsig-neg, the therapy can be rituximab, cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate and prednisone (R-CHOP).

[0035] In some embodiments, when the molecular subgroup is determined to be DHITsig-pos, a therapy other than R-CHOP (an alternate therapy) may be selected.

[0036] In another aspect, the present disclosure provides a kit comprising reagents sufficient for the detection of two or more of the genes listed in any of Tables 1 to 4. In some embodiments, the kits may further include reagents sufficient for the detection or two or more of the genes listed in Tables 5 or 6. The kit may be used for classification of an aggressive B-cell lymphoma and / or for providing prognostic information and / or for providing information to assist in selection of a therapy.

[0037] The kit may include probes and / or primers specific to two or more of the genes listed in any of Tables 1 to 4 as well as reagents sufficient to facilitate detection and / or quantification of the gene expression products. In some embodiments, the kits may further include probes and / or primers specific to one or more of the genes listed in Tables 5 or 6. The kit may further include a computer readable medium.

[0038] The present invention will be further illustrated in the following examples.EXAMPLES METHODS Patient cohort description

[0039] We analyzed RNAseq data from 157 de novo GCB DLBCLs, including 25 HGBL-DH / TH-BCL2, to define gene expression differences between HGBL-DH / TH-BCL2 and other GCB-DLBCLs (discovery cohort). These are GCB-DLBCLs with available MYC and BCL2 FISH results from a cohort of 347 diagnostic biopsies of de novo DLBCL patients treated with R-CHOP who were selected from the BC Cancer population-based registry 6< (Figure 1). This study was reviewed and approved by the University of British Columbia-BC Cancer Research Ethics Board, in accordance with the Declaration of Helsinki.

[0040] We utilized two external cohorts with RNAseq data available (Reddy et al; n= 278 GCB-DLBCL cases, Schmitz et al; n= 162 GCB-DLBCL cases) to explore the prognostic significance and molecular features associated with DHITsig DLBCL 18, 19< FFPE biopsies of 322 of the 347 DLBCLs plus 88 transformed follicular lymphomas (tFL) 20< with DLBCL morphology and 26 high-grade B-cell lymphomas (HGBL) from patients treated in BC were analyzed for the validation of the NanoString assay.Gene expression profiling and mutational analysis

[0041] RNAseq was applied to RNA extracted from fresh frozen biopsies. We compiled mutations from targeted sequencing of the discovery cohort and existing exome data from two validation cohorts, each with matched RNAseq 18, 19< . Sample processing of RNA and DNA, library construction and detailed analytic procedures for RNAseq, targeted resequencing and mutational analysis of exome data were either previously described 6, 21-23< , or are described herein.Phenotypic analysis Sample processing of fresh frozen biopsies

[0042] For genetic analyses performed at BC Cancer, genomic DNA and RNA were extracted using the AllPrep DNA / RNA Mini kit (QIAGEN, Germany) according to the manufacturer's instructions from cryosections of fresh frozen biopsies or from cryopreserved diagnostic cell suspensions. For constitutional DNA, we extracted genomic DNA from peripheral blood using the Gentra Puregene Blood Kit (QIAGEN).IHC and FISH analyses on tissue microarray

[0043] Immunohistochemistry (IHC) and fluorescent in situ hybridization (FISH) was performed on formalin-fixed paraffin-embedded tissue (FFPET) biopsies of 341 DLBCL cases within the cohort as described previously 6, 24< . Briefly, FISH was performed using commercially available dual-color break-apart probes for MYC, BCL2 and BCL6 as previously described 6, 24< . IHC staining on the 4µm slides of TMAs was performed for MYC, BCL2, CD10 (MME), BCL6, MUM1 (IRF4) and Ki67 on the Benchmark XT platform (Ventana, AZ) according to the previously described method 6,24< . For CD10, BCL6 and MUM1 (IRF4), tumor cells with ≥30% positive cells were called as positive. The cut-off points previously described were used for MYC (≥40% positive tumor cells) and BCL2 (≥50% positive tumor cells) 9< .Lymph2Cx assay

[0044] For the determination of COO subtype of BC -Cancer cohort, digital GEP was performed using the Lymph2Cx 20-genes GEP assay on the NanoString platform (NanoString Technologies, WA) 24, 32< . RNA was extracted from 10 µm scrolls using the QIAGEN AllPrep DNA / RNA FFPE kit (Catalogue # 80234, QIAGEN GmbH, Germany) with QIAGEN deparaffinization solution (Catalogue # 19093, QIAGEN GmbH, Germany). Two hundred nanograms of RNA were used to quantitate the 20 genes that contribute to the Lymph2Cx assay. The reactions were processed on an nCounter ™< Prep Station. The COO score was calculated based on the model previously described 32< and assigned to ABC, GCB and Unclassified categories.Flow cytometry analysis

[0045] We performed flow cytometric immunophenotyping on cell suspensions from freshly disaggregated lymph node biopsies using a routine diagnostic panel and stained according to the manufacturer's recommendations with CD3, CD4 and CD8 monoclonal antibodies (Beckman Coulter, USA). Analysis was performed on a Cytomics FC 500 flow cytometer (samples processed between 1985-2009; Beckman Coulter, USA) or BD FACS Canto (samples processed between 2009-2011; BD Biosciences, USA).Gene expression analysis Library preparation and data processing of RNAseq

[0046] RNA-seq data were generated from 322 BC -Cancer DLBCL samples to quantify the gene expression levels. Polyadenylated (polyA+) messenger RNA (mRNA) was purified using the 96-well MultiMACS mRNA isolation kit on the MultiMACS 96 separator (Miltenyi Biotec, Germany) then ethanol-precipitated, and used to synthesize cDNA using the Maxima H Minus First Strand cDNA Synthesis kit (Thermo-Fisher, USA) and random hexamer primers at a concentration of 5 µM along with a final concentration of 1µg / uL Actinomycin D, followed by Ampure XP SPRI bead purification on a Biomek FX robot (Beckman-Coulter, USA). cDNA was fragmented by sonication using a Covaris LE220 (Covaris, USA). Plate-based libraries were prepared using the Biomek FX robot (Beckman-Coulter, USA) according to the British Columbia Cancer, Genome Science Centre paired-end protocol, previously described 33< . The purified libraries with a desired size range were purified and diluted to 8nM, and then pooled at five per lane and sequenced as paired-end 75-bp on the Hiseq 2500 platform. This yielded, on average, 71 million reads per patient (range: 6.5-163.7 million reads).

[0047] Paired end RNA-seq FASTQ files were used as input to our gene expression analyses starting with alignment using the STAR aligner (STAR_2.5.1b_modified). The non-default parameters were chosen as recommended by the STAR-Fusion guidelines as follows: --outReadsUnmapped None, --twopassMode Basic,-outSAMunmapped Within. Detailed data analysis was as previously described 21-23< .104 gene DHIT signature

[0048] In order to produce a stable significant gene list, RNAseq count data were normalized in two different ways: voom function in R package limma and vst function in R package DESeq2. DESeq2 was used to normalize the data using variant stabilization. We generated spearman correlation coefficients and Importance Gini Index from a random forest analysis for both data formats to identify genes that discriminated HGBL-DH / TH-BCL2 from other GCB-DLBCLs. For each gene, we derived four "importance scores", namely two correlation coefficients and two Importance Gini Indexes with signs of correlation coefficients. The mean of the four numbers became final Importance Score for each gene. We kept the top 0.1% and down 0.1% genes with the largest absolute Importance Score, removing any genes where the 95% confidence intervals, based on these four importance scores, crossed 0. Additionally, genes with BAC-based names (RP1 and RP11) were removed. This process resulted in identifying the 104 genes (Table 3 ). Table 3. DHITsignature Importance Score No. Gene Name DHITsignature Importance Score 1*OR13A10.6742184282FAM216A0.6662735733*MYC0.6180967684*SLC25A270.5973288825*ALOX50.582284096UQCRH0.5545504117SUGCT0.5447910098SNHG70.5331311069*TNFSF80.48655375110LINC009570.47748213811*PEG100.4756755912PIK3CD-AS20.47136484613*GAMT0.46081880914RPL60.45022222515EIF4EBP30.4495809616*SNHG190.4323041917*QRSL10.42809628118FHIT0.42719022119SLC29A20.42616492920TERT0.42503365921SMARCB10.42500241122*RGCC0.42039377923SNHG170.41538343424*JCHAIN0.41120529925SPTBN20.40516575426ATF40.40426282127*CD240.40243129428RPL350.40100922629HAGHL0.39479781830CTD-3074O7.50.39429680331WNK20.38833052132*AFMID0.38774168133CCDC780.38540686834RPL130.38064750235RPL70.37975941836SFXN40.37827722437SGCE0.37727374738*SMIM140.37675611439LRRC75A-AS10.37463424540HRK0.3733336241DANCR0.36970447242*SYBU0.36849188143RPS80.36645545444SNHG110.36189863345NMRAL10.36133384546PPP1R14B0.36130009247MACROD10.35873597748SOX90.35791079149MYEOV-0.43319519250IL10RA-0.43409960851*GPR137B-0.43664693252TLE4-0.43808895753PARP15-0.43944214454CCL17-0.4408764955HMSD-0.44282181756DOCK10-0.44293364457MVP-0.44456421258ASS1P1-0.44623454459GNG2-0.44625475560*CDK5R1-0.45041720661ETV5-0.45215248962RASGRF1-0.45286422763ACPP-0.45342731664COBLL1-0.46362434365*LY75-0.46539779666ARPC2-0.46544946767CFLAR-0.4696946868AC104699.1-0.47036394869GALNT6-0.47635152270*VASP-0.47820627271ARHGAP25-0.48317427672SIGLEC14-0.48551446773PTPRJ-0.49075617774CR2-0.49280185175CAB39-0.49396459676HIVEP1-0.50348519677*RFFL-0.50984877378ADTRP-0.51518392279*MIR155HG-0.51557665980POU3F1-0.51729636381*VOPP1-0.5179133382*BATF-0.51820083883MREG-0.52059214384*STAT3-0.5280311185TACC1-0.53078222486*IRF4-0.5314413287ST8SIA4-0.5314463788WDFY1-0.53248999889ARID3B-0.53303585290CCL22-0.53621524591SIAH2-0.53721072392*SGPP2-0.57805502193CPEB4-0.58261501494*CD80-0.59198804795*SEMA7A-0.59713292896ANKRD33B-0.60197243297NCOA1-0.60246473598BCL2A1-0.62379397799DGKG-0.633290788100ALS2-0.657454773101LTA-0.673264157102FCRL5-0.750221729103*EBI3-0.776792921104*IL21R-0.778158195* selected for DLCBL90 assay

[0049] To calculate the 104 gene DHITsig score for RNAseq data, we used the following model: DHITsig Score = ∑ i = 1 m ImportanceScore ∗ log 10 p 1 p 2 where m is the total number of 104 genes that we can match in a given RNAseq data, p1 is the p value based on t test of a given sample's gene expression value against a normal distribution with mean and standard deviation from DHITsig-pos group, and p2 is the p value based on t test of a given sample's gene expression value against a normal distribution with mean and standard deviation from DHITsig-neg group,

[0050] When training data with DHITsig information was not available, such as testing on an independent cohort, we used a prior of proportion of DHITsig-pos cases for a given gene to calculate the mean and standard deviation for DHITsig-pos group, with the remaining values used to calculate mean and standard deviation for the DHITsig-negative group.GSEA

[0051] Differentially expressed genes between DHITsig-pos and DHITsig-neg were determined using DESeq2 v.1.20.0 34< . The DESeq pipeline was run using the default parameters, aside from the results, during which the following parameters were set, lfcThreshold=0.5, and alpha=0.1. The resulting differentially expressed genes and their combined test statistics were then used as input for Fast Gene Set Enrichment Analysis v.1.6.0 (FGSEA) 35< . The hallmark gene sets, gene symbols (h.all.v6.2.symbols.gmt) used for FGSEA analysis were obtained from MSigDB / GSEA. FGSEA was then run using 1000 permutations, with the aforementioned gene list, test statistics, and hallmark gene set as input.

[0052] Based on DZ / IZ / LZ gene lists 26< , we selected top 20 genes for each of these lists and extra RNAseq data for these 60 genes for the discovery DLC GCB cohort with 157 samples. For each gene, we calculated z score across all 157 samples. For each sample, we further calculated mean z scores for 20 DZ genes, 20 IZ genes, and 20 LZ genes separately. Then, we separated 157 samples into DHITsig-pos and DHITsig-neg, and compare their median sample mean z score differences between DHITsig POS vs NEG for DZ, IZ and LZ separately based on Wilcoxon rank sum test (also called Mann-Whitney' test for two group comparison). Boxplot showed DZ, IZ, LZ separately with DHITsig-pos and -neg. P values on the boxplot were from Wilcoxon rank sum test.Mutation analysis

[0053] We analyzed the data of targeted re-sequencing, which has been performed using BC Cancer cohort. A gene panel comprising known DLBCL-related genes and novel candidates was sequenced in tumor DNA extracted from FF biopsies in 347 de novo DLBCL patients using a TruSeq Custom Amplicon and custom hybridisation-capture strategy as described previously 6, 21-23< .Statistical analysis

[0054] The Kaplan-Meier method was used to estimate the time-to-progression (TTP; progression / relapse or death from lymphoma or acute treatment toxicity), progression-free survival (PFS; progression / relapse or death from any cause), disease-specific survival (DSS; death from lymphoma or acute treatment toxicity) and overall survival (OS; death from any cause), with log-rank test performed to compare groups. Univariate and multivariate Cox proportional hazard models were used to evaluate proposed prognostic factors.

[0055] Fisher's exact test was used when comparing two categorical data. For the comparison of two continuous variables, data were tested by Wilcoxon rank-sum test, except where noted. Multiple testing correction was performed, where necessary, using the Benjamini-Hochberg procedure. All P values result from two-sided tests and a threshold of 0.05 was used for significance, except where noted. All analyses were performed using R v3.4.1.Digital gene expression profiling

[0056] To translate the signature into an assay applicable to FFPE, we performed digital expression profiling on RNA derived from FFPE biopsies using the NanoString Technology (Seattle, WA) as described herein.Development and testing of the DLBCL90 Digital gene expression

[0057] RNA was extracted from formalin-fixed paraffin-embedded (FFPE) biopsies using the Qiagen AllPrep DNA / RNA FFPE Kit (Qiagen, Hilden, Germany). Digital gene expression was performed on the NanoString technology platform at the highest resolution (555 fields of view).

[0058] Data was normalized for loading and RNA integrity by dividing by the geometric mean of the housekeeping genes for that sample and then multiplying by 1000. The house-keeping genes were the 13 genes used in the Lymph3Cx assay and includes all 5 genes from the Lymph2Cx 27< . The normalized data was then log 2 transformed prior to analysis.Model building Gene selection

[0059] In order to translate the DHITsig from RNAseq to the NanoString platform, digital gene expression was first performed using a code set that included all 104 gene of the RNAseq DHITsig. This was applied to 35 samples that were selected to be representative of the range of scores observed with the RNAseq model (Figure 2A). In the first step, the correlation between gene expression by RNAseq and NanoString in these 35 samples was examined. Genes with R 2< less than 0.6 were excluded leaving 67 genes of interest. These 67 genes were then ranked into two lists ordered according to their Importance Score: A) genes over-expressed in DHITsig-pos tumors and B) genes under-expressed in DHITsig-pos. In order to produce a "balanced" model, that would be less vulnerable to any variability in normalization, the 15 top ranked genes from both lists were selected for the final model (see Table 2 or 4). Model Building

[0060] A NanoString codeset was developed that included the 30 selected genes alongside the genes in the Lymph3Cx - this represented an additional of 29 genes as IRF4 was already included in the Lymph3Cx. The Lymph3Cx included the 20 genes from the Lymph2Cx in addition to 8 further house-keeper genes and 30 genes that discriminate DLBCL from primary mediastinal B-cell lymphoma 12< . The Lymph3Cx genes are listed, for example, in PCT publication WO / 2018 / 231589, Staudt et al., published December 20, 2018. In addition, BCL2, FCGR2B and PVT1 were added for a total of 90 genes, with the assay named "DLBCL90". The probes targeting the 30 selected genes were used in the NanoString assay (Table 4 ). The probes targeting BCL2, FCGR2B and PVT1, used in the NanoString assay, are shown in Table 5. Table 4 Gene Name Accession Position Target Sequence 1AFMIDNM_001010982.4851-9502ALOX5NM_000698.2736-8353BATFNM_006399.3826-9254CD24NM_013230.21860-19595CD80NM_005191.3675-7746CDK5R1NM_003885.21211-13107EBI3NM_005755.2827-9268GAMTNM_138924.1291-3909GPR137BNM_003272.3682-78110IL21RNM_021798.22081-218011IRF4NM_002460.1326-42512JCHAINNM_144646.3436-53513LY75NM_002349.25362-546114MIR155H GNR_001458.3361-46015MYCNM_002467.31611-171016OR13A1NM_001004297.2917-101617PEG10NM_001040152.15001-510018QRSL1NM_018292.21131-123019RFFLNM_001017368.1509-60820RGCCXM_011535051.1381-48021SEMA7ANM_001146029.1661-76022SGPP2NM_152386.2851-95023SLC25A27NM_004277.41481-158024SMIM14NM_174921.1371-47025SNHG19NR_132114.1235-33426STAT3NM_003150.32061-216027SYBUNM_001099744.11493-159228TNFSF8NM_001244.3519-61829VASPNM_003370.31501-160030VOPP1NM_030796.32091-2190 Table 5 Gene Name Accession Position Target Sequence 1BCL2NM 000657.2948-10472FCGR2BNM 001002273.1871-9703PVT1NR 003367.1412-511 Table 6 Gene Accession No. Position Target Sequence 1ASB13NM_024701.31636-17352AUHNM_001698.2591-6903BANK1NM_001083907.11396-14954BATF3NM_018664.2870-9695BTG2NM_006763.21701-18006CARD11NM_032415.21076-11757CCDC50NM_174908.3975-1074 8CCL17NM_002987.2230-3299CREB3L2NM_194071.22556-265510CYB5R2NM_016229.3367-46611DNAJB12NM_017626.41961-206012FAM159ANM_001042693 .2334-43313FSCN1NM_003088.21844-194314GIT2NM_057169.2606-70515GSK3BNM_002093.2926-102516HOMER2NM_004839.21055-115417IFIH1NM_022168.2186-28518IKNM_006083.3557-65619IL13RA1NM_001560.21231-1330 20IRF4NM_002460.1326-42521ISY1NM_020701.287-18622ITPKBNM_002221.34201-430023LIMA1NM_001113547 .12916-301524LIMD1NM_014240.22926-302525MALNM_002371.2706-80526MAML3NM_018717.41351-145027MMENM_000902.25060-515928MOBKL2CNM_145279.41631-173029MSTIRNM_002447.13301-340030MYBL1XM_034274.141441-154031NECAP2NM_018090.4991-1090 32NFIL3NM_005384.2186-28533OPA1NM_130837.11356-145534PDCD1LG2NM_025239.3643-74235PHF23NM_024297.21661-176036PIM2NM_006875.2621-72037PRDX2NM_005809.4651-75038PRKCBNM_212535.11751-185039PRR6NM_181716.2606-70540PTGIRNM_000960.31271-137041QSOX1NM_002826.42566-266542R3HDM1NM_015361.21276-137543RAB7L1NM_001135664.1786-885 44RCL1NM_005772.3696-79545RHOFNM_019034.2142-24146S1PR2NM_004230.2186-28547SERPINA9NM_001042518 .11156-125548SLAMF1NM_003037.2581-68049SNX11NM_013323.21361-146050TFPI2NM_006528.2601-70051TMOD1NM_003275.2771-87052TNFRSF13 BNM_012452.2161-26053TRAF1NM_005658.33736-383554TRIM56NM_030961.12571-267055UBXN4NM_014607.3344-443 56VRK3NM_016440.3821-92057WACNM_100486.2756-85558WDR55NM_017706.4816-915

[0061] The DLBCL90 was applied to 171 GCB-DLBCL including 156 / 157 of the samples whose RNAseq were used define the DHITsig. All 171 GCB-DLBCL were selected from the 347 patient BC Cancer cohort and had RNAseq data available, such that the RNAseq DHITsig score could be calculated and DHITsig categories assigned. Importantly, the 15 additional samples that were not part of the "discovery cohort" had been excluded from that cohort on the basis that they did not have both MYC and BCL2 FISH results available. The QC threshold of the geometric mean of the 13 housekeeping genes being greater than 60 was carried over from the Lymph3Cx.

[0062] To prevent over-fitting, the gene coefficients from the RNAseq model, which were the Importance Score for that gene, were carried over to the DLBCL90 model unaltered. The DLBCL90 DHITsig score was calculated as the sum of the gene coefficient (Importance Score) multiplied by the log 2 transformed normalized gene expression. In order to determine the appropriate thresholds for the DLBCL90 score, 72 of the 171 samples were selected on the basis of being equally distributed across the scores for the population ( Figure 2B). To avoid circularity, this cohort included the 35 samples used for gene selection to leave a cohort of samples that had not contributed to gene selection and threshold training. The thresholds were selected according to Bayes rule with 20% and 80% used as the threshold probabilities. This level was used, as opposed to 90%, as it resulted in 10% of the population in an "indeterminate" group where assignment could not be made with sufficient confidence. With these thresholds, 3 (4%) tumors were misclassified with 2 RNAseq DHITsig-neg being called DHITsig-pos by the DLBCL90 (including 1 case that was HGBL-DH / TH-BCL2) and 1 RNAseq DHITsig-pos being called DHITsig-neg by the DLBCL90. Seven (10%) were deemed DHITsig-ind.

[0063] These thresholds were locked and the model was then applied to the remaining 99 samples (blinded to outcome and the DHITsig result from RNAseq) to test the final model, including the thresholds. Nine cases (9%) were assigned to DHITsig-ind. Two cases (2%) were misclassified with one being DHITsig-pos by RNAseq but DHITsig-neg by DLBCL90 and one vice versa. Taken as a total group, the misclassification rate was 3% (5 / 171) ( Figure 3). Applying the DLBCL90 to a population registry-based cohort

[0064] On review of the 347-patient cohort, one tumor from the training cohort (DLC0224) was removed due to a tumor content of <10%. As the thresholds had been "locked" prior to the removal of this sample, the thresholding was not repeated on the data set after removal of the sample. The DLBCL90 was applied to an additional 152 biopsies to complete a total of 322 eligible cases from the 347 patient BC Cancer cohort - RNA was not available for the remaining 24 patients. Note that inclusion of DLC0224 would have strengthened the outcome correlation of the DHITsig-pos group, as the patient was DHITsig-pos and had a poor outcome (death at 0.6 years).Performance of the Lymph2Cx component

[0065] Linear predictor scores (LPS) were available for 320 samples from both the Lymph2Cx assay 2< and the DLBCL90. The correlation between the scores was very high (R 2< = .996) and the slope was 1.007. The bias (the Y-intercept was +116.6 points ( Figure 4A). Therefore, to calibrate the DLBCL90 LPS to the original Lymph2Cx score, 116.6 points were removed from the DLBCL90 LPS ( Figure 4B). In total, six tumors (2%) changed COO, going from definitive COO categories to Unclassified or vice versa - there were no cases that changed from ABC to GCB or vice versa. Thus, the addition of the DHITsig 30 gene module did not impact the performance of the Lymph2Cx component of the assay.The DHITsig across the population registry-based cohort

[0066] The results in the GCB-DLBCL and Unclassified-DLBCL (with COO determined using the DLBCL90 LPS) are shown in Figure 4A. Results in the ABC-DLBCL are not shown. In GCB-DLBCL, 23% were classified as DHITsig-pos, 10% were DHITsig-ind and 66% DHITsig-neg, while in Unclassified-DLBCL, these figures were 6% DHITsig-pos and 94% DHITsig-neg and in ABC-DLBCL 4% were DHITsig-ind and 96% DHITsig-neg. Over the entire cohort, 45 / 322 (14%) were DHITsig-pos, 23 / 322 (7%) were DHITsig-ind and 254 / 322 (79%) were DHITsig-neg.Applying the DLBCL90 to transformed follicular lymphoma and high-grade B-cell lymphomas Transformed follicular lymphoma with DLBCL morphology

[0067] The DLBCL90 was applied to the 88 tFL with DLBCL morphology, previously described in Kridel et al 20< to validate the association between the DHITsig assignment by the DLBCL90 and HGBL-DH / TH-BCL2. The results are shown in Figure 14A, with all HGBL-DH / TH-BCL2 falling with the DHITsig-pos and DHITsig-ind groups.High-grade B-cell lymphoma

[0068] The DLBCL90 was applied to 26 high-grade B-cell lymphomas drawn from the BC Cancer Centre for Lymphoid Cancer Database. These tumors would be categorized as high-grade B-cell lymphoma (n = 4) or HGBL-DH / TH with high-grade morphology (n = 18) with 4 lymphomas having insufficient FISH results to place them in the correct category. The morphology of the tFL cases within this cohort had already been centrally reviewed. The morphology of the remaining 17 cases were reviewed by a panel of expert hematopathologists (PF, GWS, JC and TT) and confirmed to be high-grade as opposed to DLBCL. The results are shown in Figure 14B, with 23 / 26 (88%) being DHITsig-pos and the remaining tumors being DHITsig-ind.

[0069] Following the REMARK guidelines, the assay parameters were locked prior to application to the "validation" cohorts. On review of the assembled data, it would appear that the DHITsig-pos and DHITsig-ind share similar quite outcomes and if considered together they would have detected all HGBL-DH / TH-BCL2 cases within the tFL with DLBCL morphology. For this reason, depending on the application, DHITsig-ind may be considered a positive result, which would maximize specificity thereby enriching for patients with very good outcomes (i.e. DHITsig-neg).RESULTS Development of the DHIT gene expression signature

[0070] We identified 104 genes that were most significantly differentially expressed between HGBL-DH / TH-BCL2 and other GCB-DLBCLs (Figure 5A). We devised a model score using the expression of these 104 genes that separates GCB-DLBCL into two groups. The smaller group, comprising 42 tumors (27%), was termed "double-hit signature"-positive (DHITsig-pos) and included 22 of the 25 HGBL-DH / TH-BCL2 tumors, as determined by FISH. The remaining 115 GCB cases (73%) were considered DHITsig-negative (DHITsig-neg), including 3 HGBL-DH / TH-BCL2 tumors (Figure 5B).Prognostic value of the DHIT signature

[0071] Having developed the DHITsig blinded to patient outcomes, we then explored the prognostic impact of the DHITsig within the 157 uniformly R-CHOP treated cohort of de novo GCB-DLBCL 6, 24< using assignments from the locked RNAseq model. DHITsig was not associated with clinical variables, including the factors of International Prognostic Index (IPI), IPI subgroups, B-symptoms or tumor volume. As expected, MYC and BCL2 translocations and protein expression of MYC and BCL2 were significantly more frequent in DHITsig-pos cases (all, P<.001; Table 10 ).

[0072] DHITsig-pos cases had significantly shorter TTP, DSS and OS when compared with the DHITsig-neg GCB group (log-lank P<.001, P<.001 and P=.012, respectively) exhibiting outcomes comparable to those of ABC-DLBCL from the cohort of 347 patients (Figure 6A-C). Importantly, the non-HGBL-DH / TH-BCL2 cases with the DHITsig-pos group showed comparably poor prognosis to HGBL-DH / TH-BCL2 cases (Figure 7A-C). Although IPI and dual protein expression of MYC and BCL2 (DPE) were also associated with survival in GCB-DLBCL (Table 7), DHITsig remained prognostic of TTP and DSS in multivariate analyses (HR=3.1 [95% CI 1.5-6.4]; P=.002, HR=3.1 [95% CI 1.3-7.1]; P=.008, respectively) independent of these factors (Table 8 ).

[0073] In particular, DPE did not provide statistically significant risk stratification within either the DHITsig-pos or -neg groups ( Figure 8A-C), indicating that the DHITsig designation subsumes the prognostic impact of DPE within GCB-DLBCL. We then applied this gene expression model to GCB-DLBCL from an independent dataset (Reddy et al; n=262 GCB-DLBCLs), in which the DHITsig-pos group also had significantly inferior OS compared with other GCB-DLBCLs (P<0.001) (Figure 6D).Double Hit signature defines a biologically distinct subgroup within GCB-DLBCL

[0074] Exploration of the pathology and gene expression patterns demonstrated that DHITsig-pos tumors form a distinct biological subgroup of GCB-DLBCL characterized by a cell-of-origin from the intermediate- / dark-zone of the germinal center. In a first step, a pathology re-review of the entire 347 DLBCL cases from the BC Cancer cohort was performed by a panel of expert hematopathologists, confirming that DHITsig-pos tumors were indeed of DLBCL morphology. There were no morphological features that distinguished these tumors from DHITsig-neg tumors nor was the proliferation index (Ki67) significantly different between DHITsig groups (Figure 9A).

[0075] In the Lymph2Cx assay, low linear predictor scores (LPS) provide an assignment to the GCB group while high scores result in an ABC assignment. Among the GCB DLBCLs, DHITsig-pos cases had significantly lower LPSs than DHITsig-neg (P<.001, Figure 9B). Moreover, DHITsig-pos tumors were universally positive for CD10 (MME) staining and the vast majority were MUM1 (IRF4) negative. CD10+ / MUM1- cases were significantly more frequent in DHITsig-pos tumors (P<.001; Figure 9C). It has been previously demonstrated that most GCB-DLBCLs have a COO consistent with B-lymphocytes from the light zone (LZ) of the germinal center 25< . Given that the gene features in the Lymph2Cx and these IHC markers are associated with B-cell differentiation states, we considered whether the two DHITsig groups had gene expression patterns implying distinct putative COOs. Gene signatures associated with DZ, LZ and the more recently described intermediate zone (IZ), representing transition stage between these, were explored within the GCB-DLBCLs 26< . Strikingly, DHITsig-pos cases showed significantly lower expression of LZ genes compared to DHITsig-neg tumors (P<.001) ( Figure 9D). The expression of genes in the DZ cluster were not statistically different between the two groups, while genes associated with the IZ had higher expression within the DHITsig-pos tumors. Furthermore, genes characteristic of the IZ are part of the 104-gene DHITsig model. Collectively, these findings demonstrate that while DHITsig-neg tumors have a LZ COO, we postulate that the COO for DHITsig-pos tumors are IZ B-cells transitioning from the LZ to the DZ.

[0076] Gene set enrichment analysis was then used to further uncover additional biological differences between DHITsig-pos and -neg tumors. We found that DHITsig-pos cases demonstrated overexpression of MYC and E2F targets and genes associated with oxidative phosphorylation and MTORC1 signaling (Figure 10). Conversely, DHITsig-pos tumors exhibit lower expression of genes associated with apoptosis, TNF-alpha signaling via NF-kB and decreased IL6 / JAK / STAT3 - processes up-regulated in centrocytes. DHITsig-pos cases also exhibited lower expression of immune and inflammation signatures. Consistently, tumor-infiltrating lymphocytes, especially CD4-positive T-cells, had significantly lower representation in DHITsig-pos cases relative to other GCBs ( Figure 11A). Loss of surface MHC class I and class II protein expression was also more frequent in DHITsig-pos cases (Fisher's exact test for MHC-I and MHC-II; 61% vs 40%; P=.020, 44% vs 14%; P<.001, respectively; Figure 11B) with 68% of DHITsig-pos tumors having loss of either MHC class I or class II expression. Finally, we identified that all representative GCB-DLBCL cell lines tested belonged to the DHITsig-pos subgroup ( Figure 12), consistent with the notion that DHITsig-pos tumors harbor strong cell-autonomous survival and proliferation signals and reduced dependence on the microenvironment.The mutational landscape of DHITsig-pos GCB-DLBCL

[0077] We next sought genetic features associated with DHITsig status within GCB-DLBCL. For this, we used the combined mutation data derived from 569 unique GCB-DLBCL cases in 3 cohorts (BC Cancer, Reddy et al and Schmitz et al). Along with the expected enrichment of mutations in MYC and BCL2 (FDR<.001), mutations affecting CREBBP, EZH2 Y646< , MEF2B and ARID5B were more frequent in DHITsig-pos tumors (all FDR<.10). In contrast, the mutations of TNFAIP3 and NFKBIE were more common among DHITsig-neg GCB tumors (FDR<.01, <.14, respectively; Figure 11C, Table 9). Table 9. The association between mutation and DHIT signatureGene Unmu tated DHITsi g-neg Mut ated DHIT sig-neg Unmu tated DHITsi g-pos Mut ated DHIT sig-pos p.value Odds Ratio 95% CI lower bound 95% CI upper bound FDR MYC_ Nonsyn41913111341.25E-129.8208 254844.8573 5827721.002 688281.49E-10BCL2_ Nonsyn3438980653.75E-083.1243 442492.0493 078364.7676 778931.88E-06CREBBP_ Nonsyn3478582634.70E-083.1292 73412.0451 048014.7918 9021.88E-06EZH2 Codon6463686498478.95E-062.7520 170521.7320 49534.3608 186960.0002 68613CD58_ Nonsyn3914114416.98E-050.0663 789490.0016 303670.3989 740030.0016 7485DDX3X_ Nonsyn41121123220.0001 56693.4910 624391.7661 571576.9240 009840.0031 33805TNFAIP3_ Nonsyn3706213960.0005 314040.2580 508350.0891 758940.6129 637010.0091 09782BCL7A_ Nonsyn38745113320.0006 435242.4310 529871.4236 748324.1196 452710.0096 52867TP53_ Nonsyn36567106390.0029 170172.0017 116911.2387 868553.2085 687440.0388 93564KMT2D_ Nonsyn28914377680.0037 688311.7829 415921.1933 609852.6627 607470.0452 25971KLHL6_ Nonsyn3676513690.0059 454170.3741 628350.1593 341340.7822 554350.0648 59099STAT3_ Nonsyn3904214140.0070 116910.2638 928080.0675 312590.7467 709730.0659 08505NFKBIE_ Nonsyn3953714230.0071 400880.2259 236530.0438 928440.7310 898840.0659 08505TET2_ Nonsyn3775513870.0078 206150.3481 979080.1305 93910.7916 704950.0670 33847BCR_ Nonsyn40626127180.0177 456632.2097 039411.1025 469864.3451 195080.1396 84815RB1_ Nonsyn41814133120.0186 246422.6883 584791.1062 391266.4364 926550.1396 84815MEF2B_ Codon8341319131140.0233 342822.3190 711461.0446 074875.0336 58390.1647 12575PRDM1_ Nonsyn4181414500.0262 75113000.8833 488070.1751 67421C10orf12_ Nonsyn40032125200.0282 803521.9973 314521.0430 175813.7481 637990.1764 9257NFKBIA_ Nonsyn3933914050.0294 154280.3603 892810.1086 989590.9408 143230.1764 9257TMSB4X_ Nonsyn3854713870.0317 852760.4160 285850.1549 190130.9557 533420.1816 3015P2RY8_ Nonsyn211267220.0343 578980.2261 514640.0253 948880.9416 496240.1818 06676UBE2A Nonsyn4131914410.0348 46280.1512 399910.0036 1220.9685 477140.1818 06676CD70_ Nonsyn4023014230.0363 696950.2835 206410.0545 466920.9334 741980.1818 48475GNA13_ Nonsyn34092102430.0420 39291.5567 026130.9912 143742.4244 075780.2017 8859EZH2_ Nonsyn423913780.0457 817362.7387 744980.9004 262748.1722 883320.2113 00318CARD11_ Nonsyn36567132130.0520 420550.5370 353020.2630 388831.0222 512580.2312 98024BCL10 Nonsyn4141814410.0557 651070.1600 237250.0038 114141.0316 285930.2349 99339FOXO1_ Nonsyn39438124210.0576 360071.7539 853970.9405 423693.1992 161930.2349 99339SGK1_ Nonsyn33399123220.0587 498350.6021 173980.3450 510491.0150 745740.2349 99339BTK_ Nonsyn41022131140.0716 320531.9890 310520.9131 5564.2008 621010.2723 06504HLA.B_ Nonsyn113143100.0735 4995001.1791 06250.2723 06504MYD88_ Nonsyn4062614230.0768 339520.3304 072690.0630 533091.1030 966290.2723 06504SOCS1_ Nonsyn332100122230.0783 193180.6263 763510.3625 058011.0480 428710.2723 06504ACTB_ Nonsyn3834913690.0804 398730.5177 691560.2176 665361.1020 93680.2723 06504IRF4_ Nonsyn4112114320.0836 479790.2741 435630.0307 928621.1447 231080.2723 06504CIITA_ Nonsyn3993314050.0839 611720.4323 308940.1292 020441.1449 999920.2723 06504SPEN_ Nonsyn3844813690.1070 594960.5299 200050.2225 342451.1296 670070.3380 82619BTG2_ Nonsyn37953134110.1290 532110.5875 030210.2685 704061.1807 038190.3945 85993CD274_ Nonsyn4181414410.1315 286640.2076 99270.0048 741281.3883 112310.3945 85993HVCN1_ Nonsyn4181413690.1392 215981.9732 578880.7359 117325.0219 919760.4074 77847NOTCH1_ Nonsyn28619102120.1447 714161.7681 694550.7544 530533.9945 604640.4136 32616BCL6_ Nonsyn38349135100.1539 193760.5794 753170.2543 904561.1995 256460.4295 42444NLRCS_ Nonsyn4102214230.1579 918370.3942 454550.0744 300721.3418 097010.4308 86828CD36_ Nonsyn4092314230.1619 332970.3762 058570.0712 372411.2734 792480.4318 22125SETD2_ Nonsyn40725132130.1805 008961.6018 930880.7305 119133.3616 295060.4602 00289NFKBIZ_ 3UTR100104210.1828 627230.2397 215570.0053 648791.7814 48910.4602 00289MEF2B_ Nonsyn39735128170.1843 608951.5053 162510.7635 840892.8697 376930.4602 00289RFXAP_ Nonsyn425714050.1879 151182.1650 251090.5329 675338.0679 295310.4602 00289CD79B_ Nonsyn2941111310.1934 052430.2370 670570.0054 514881.6654 500350.4641 72583B2M_ Nonsyn331101119260.2025 47070.7164 285690.4250 131831.1762 674740.4765 81341BLNK_ Nonsyn23527220.2407 709763.2484 655710.2316 8444245.547 644160.5556 25329HIST1H1C_ Nonsyn38052122230.2540 753741.3768 700910.7707 39452.4007 369480.5623 34348KLHL14_ Nonsyn311119660.2583 171651.7643 474610.5214 887725.3683 584230.5623 34348NOTCH2_ Nonsyn311119660.2583 171651.7643 474610.5214 887725.3683 584230.5623 34348MKI67_ Nonsyn3983413870.2645 024170.5942 518520.2172 305691.4030 068670.5623 34348ZC3H12A_ Nonsyn4161614320.2671 088150.3640 940120.0401 371411.5789 978150.5623 34348OSBPL10_ Nonsyn309139570.2822 325141.7487 975960.5735 500564.8786 321690.5839 29339UNCSD_ Nonsyn2951011310.3021 014130.2616 318750.0059 674911.8768 294520.6079 871ETVE_ Nonsyn3111110110.3083 69780.2805 039690.0064 443811.9735 54320.6079 871MYD88 Codon2734211114410.3112 327340.2661 932790.0061 366291.8607 777350.6079 871TNFSF9_ Nonsyn10824120.3141 266682.6149 396190.1839 6235437.175 941040.6079 871CCND3_ Nonsyn40725133120.3268 806841.4678 141930.6528 848973.1318 193560.6217 34785PPP1R9B_ Nonsyn316610200.3430 31492002.6791 400660.6217 34785TMEM30A_ Nonsyn3993313870.3441 431530.6137 642320.2239 182941.4533 98060.6217 34785GRHPR_ Nonsyn426614500.3450 56931002.5277 642730.6217 34785BRAF_ Nonsyn4112114140.3517 858890.5556 947680.1363 283411.6863 60740.6217 34785XPO1_ Nonsyn423914050.3566 883731.6768 905130.4339 713455.6820 30850.6217 34785PIM1_ Nonsyn36072126190.3574 975010.7543 081810.4125 188371.3251 771960.6217 34785MYOM2_ Nonsyn40824134110.4209 512861.3946 591230.5999 99973.0517 143730.7114 66962S1PR2_ Nonsyn40824134110.4209 512861.3946 591230.5999 99973.0517 143730.7114 66962STAT6_ Nonsyn40527133120.4443 097781.3526 284580.6063 938642.8540 327270.7405 16297PIM2_ Nonsyn424814410.4616 526750.3685 168010.0082 420472.7885 76780.7588 81109MPEG1_ Nonsyn4122014140.4711 823230.5848 552890.1429 208021.7872 689780.7625 3806VPS13B_ Nonsyn304189480.4765 862871.4360 230930.5228 443793.6070 636050.7625 3806HLA.DMB_ Nonsyn116112740.4966 003811.5573 956040.3358 068425.7934 126360.7841 05865FAS_ Nonsyn39042134110.5089 560120.7626 215750.3438 699321.5613 819310.7931 782EP300_ Nonsyn38943128170.5325 459881.2010 529650.6195 216212.2398 162170.8071 31514ARID5B_ Nonsyn3091310020.5380 876760.4760 488720.0513 11092.1582 480020.8071 31514TRRAP_ Nonsyn3091310020.5380 876760.4760 488720.0513 11092.1582 480020.8071 31514CPS1_ Nonsyn4221014050.5454 294831.5059 721940.3969 414154.9356 227260.8080 43678MTOR_ Nonsyn310129750.5700 036251.3306 554030.3581 308514.1842 894730.8341 51647CD83_ Nonsyn3983413690.5870 331330.7749 94480.3184 988221.7035 135210.8416 822HNF1B_ Nonsyn319310020.5979 114742.1222 836520.1749 5924218.793 379140.8416 822IL16_ Nonsyn319310020.5979 114742.1222 836520.1749 5924218.793 379140.8416 822IRF8_ Nonsyn35973124210.6032 055770.8330 914290.4665 783841.4376 163470.8416 822DTX1_ Nonsyn299239390.6668 500281.2573 616120.4941 490472.9425 329490.9197 93142CD79B_ Codon197426614410.6860 102070.4935 297170.0106 50124.1206 133630.9250 78858KLHL21_ Nonsyn315710110.6861 001530.4461 94690.0097 940613.5407 784790.9250 78858PCLO_ Nonsyn36171119260.7007 305541.1106 86860.6486 570371.8585 744590.9323 76878FAT4_ Nonsyn35577117280.7095 533581.1031 438680.6555 587511.8182 772170.9323 76878BIRC6_ Nonsyn39735135100.7229 710170.8404 626420.3611 457671.7926 77240.9323 76878HIST1H1E_ Nonsyn34290117280.7229 742910.9095 393370.5443 956051.4861 533930.9323 76878BCL11A_ Nonsyn31489930.7303 618881.1889 258320.1993 631765.0747 844840.9323 76878TRIP12_ Nonsyn312109840.7508 104571.2727 015020.2849 747044.5368 224230.9411 33237NFKBIZ_ Nonsyn4221014140.7582 779681.1968 150810.2697 101474.2310 573950.9411 33237CXCR4_ Nonsyn4201214230.7713 730340.7398 089970.1321 223432.7951 55570.9411 33237UNCSC_ Nonsyn4191314230.7716 44240.6813 65320.1227 910642.5290 50360.9411 33237SIN3A_ Nonsyn4201214050.7764 349211.2495 090890.3387 788143.8934 375530.9411 33237SETD1B_ Nonsyn305179660.8039 361511.1209 98080.3517 256163.0867 251190.9647 23381TNFRSF14_ Nonsyn324108107380.8253 426261.0653 01190.6728 79321.6650 129330.9806 051POU2F2_ Nonsyn4082413690.8363 042381.1247 503480.4486 973782.5821 212980.9821 98827IL4R_ Nonsyn4042813780.8430 539930.8427 906420.3240 394091.9553 489640.9821 98827ZFP36L1_ Nonsyn38745131140.8750 104310.9192 173870.4508 481361.7724 205131BTG1_ Nonsyn38349130150.8786 244060.9020 44410.4539 023731.7025 254471CHST2_ Nonsyn319310201007.6634 692211USP7_ Nonsyn41319139610.9383 833670.3005 591092.5070 400291ARID1A_ Nonsyn387451301510.9923 193780.4965 403141.8867 509421C1orf186_ Nonsyn1261310100159.37 879921ETS1_ Nonsyn4257143210.8493 848960.0851 643534.5292 153761FOXC1_ Nonsyn3166100211.0532 073420.1024 052946.0099 378681HIST1H2BK _Nonsyn3081498410.8981 823870.2103 768952.9483 92911HIST1H3B_ Nonsyn29114109510.9535 914760.2624 467932.8864 036181KRAS_ Nonsyn42012141410.9929 191290.2296 916473.3443 49411NFKB1_ Nonsyn4239142310.9929 690320.1705 790394.0504 821121NOL9 Nonsyn3157100210.9002 18590.0898 615654.8302 67981PTPN1_ Nonsyn4248143210.7416 272510.0758 778853.7762 435431TAP1_ Nonsyn12343101006.2873 173341TBL1XR1_ Nonsyn41814140511.0662 121340.2951 640013.2043 450681WEE1_ Nonsyn313999311.0537 446890.1800 201344.3302 02251 Translation of the DHIT signature into a clinically relevant assay

[0078] To provide an assay applicable to routinely available biopsies, the 104-gene RNAseq model was reduced to a 30-gene module. This module was added to the Lymph3Cx 27< , which in turn is an extension of Lymph2Cx containing a module to distinguish primary mediastinal B-cell lymphomas. This NanoString-based assay, named DLBCL90, assigns tumors into DHITsig-pos and DHITsig-neg groups using a Bayes rule with 20% and 80% probability thresholds, with an "Indeterminate" group (DHITsig-ind) where the tumor could not be assigned with sufficient confidence. This was applied to 171 GCB-DLBCL tumors from the 347-patient cohort (including 156 from the discovery cohort), giving 26% DHITsig-pos, 64% DHITsig-neg and 10% DHITsig-ind, with a frank misclassification rate of 3% against the RNAseq comparator (Figure 3). The integrity of the Lymph2Cx assay was maintained (Figures 8A-B). The assay was then applied to the remaining available 322 FFPE biopsies from the 347 de novo DLBCL cohort, showing that the DHITsig was not seen in ABC-DLBCL with 4 / 102 (4%) being DHITsig-ind (Figure 13, ABC-DLBCL results not shown). The prognostic significance for TTP, DSS, PFS and OS of DHITsig was maintained (all, P<.001). As the DHITsig-ind group had similar outcomes to DHITsig-pos, these two groups are shown together in Figure 15A-D. Importantly, the assay identified a group with very good prognosis with DHITsig-neg GCB-DLBCLs exhibiting a DSS of 90% at five years. Although small numbers preclude a definitive statement, the patients with rare HGBL-DH / TH-BCL2 and DHITsig-neg status experienced good outcomes with all three patients in remission at 9.2 years.

[0079] To validate the association between the DHITsig and HGBL-DH / TH-BCL2, DLBCL90 was applied to 88 tFL with DLBCL morphology. Within these 88 tFL cases, 11 of the 25 DHITsig-pos tumors were HGBL-DH / TH-BCL2 compared with 0 / 50 in the DHITsig-neg group. Within the DHITsig-ind group, 4 / 13 tumors were HGBL-DH / TH-BCL2 (Figure 4B). Finally, the DLBCL90 assay was applied to 26 HGBL tumors, including 7 classified as high-grade B-cell lymphoma NOS and 18 classified as HGBL-DH / TH with high-grade morphology - one case could not be assigned due to an unknown MYC rearrangement status. Among these tumors, the vast majority were assigned to the DHITsig-pos group (23 (88%)) with 3 (12%) being DHITsig-ind (Figure 4C). REFERENCES

[0080] 1. Lenz G, Wright G, Dave SS, et al: Stromal gene signatures in large-B-cell lymphomas. N Engl J Med 359:2313-23, 2008 2. Shipp MA, Ross KN, Tamayo P, et al: Diffuse large B-cell lymphoma outcome prediction by gene-expression profiling and supervised machine learning. Nat Med 8:68-74, 2002 3. Alizadeh AA, Eisen MB, Davis RE, et al: Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 403 :503-11, 2000 4. Swerdlow SH, Campo E, Pileri SA, et al: The 2016 revision of the World Health Organization (WHO) classification of lymphoid neoplasms. Blood 127:2375-2390,2016 5. Scott DW, King RL, Staiger AM, et al: High-grade B-cell lymphoma with MYC and BCL2 and / or BCL6 rearrangements with diffuse large B-cell lymphoma morphology. Blood 131 :2060-2064, 2018 6. Ennishi D, Mottok A, Ben-Neriah S, et al: Genetic profiling of MYC and BCL2 in diffuse large B-cell lymphoma determines cell-of-origin-specific clinical impact. Blood 129:2760-2770, 2017 7. Ott G, Rosenwald A, Campo E: Understanding MYC-driven aggressive B-cell lymphomas: pathogenesis and classification. Blood 122:575-583, 2015 8. Sarkozy C, Traverse-Glehen A, Coiffier B: Double-hit and double-protein-expression lymphomas: aggressive and refractory lymphomas. Lancet Oncol 16:e555-e567, 2015 9. Johnson NA, Slack GW, Savage KJ, et al: Concurrent expression of MYC and BCL2 in diffuse large B-cell lymphoma treated with rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone. J Clin Oncol 30:3452-3459, 2012 10. Green TM, Young KH, Visco C, et al: Immunohistochemical double-hit score is a strong predictor of outcome in patients with diffuse large B-cell lymphoma treated with rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone. J Clin Oncol 30:3460-3467, 2012 11. Johnson NA, Savage KJ, Ludkovski O, et al: Lymphomas with concurrent BCL2 and MYC translocations: the critical factors associated with survival. Blood 114:2273-2279, 2009 12. Savage KJ, Johnson N a, Ben-neriah S, et al: MYC gene rearrangements are associated with a poor prognosis in diffuse large B-cell lymphoma patients treated with R-CHOP chemotherapy. Blood 114:3533-3537, 2009 13. Pasqualucci L, Trifonov V, Fabbri G, et al: Analysis of the coding genome of diffuse large B-cell lymphoma. Nat Genet 43:830-7, 2011 14. Morin RD, Mendez-Lago M, Mungall AJ, et al: Frequent mutation of histone-modifying genes in non-Hodgkin lymphoma. Nature 476:298-303, 2011 15. Morin RD, Mungall K, Pleasance E, et al: Mutational and structural analysis of diffuse large B-cell lymphoma using whole-genome sequencing. Blood 122:1256-65, 2013 16. Lohr JG, Stojanov P, Lawrence MS, et al: Discovery and prioritization of somatic mutations in diffuse large B-cell lymphoma (DLBCL) by whole-exome sequencing. Proc Natl Acad Sci U S A 109:3879-3884, 2012 17. Schmitz R, Wright GW, Huang DW, et al: Genetics and pathogenesis of diffuse large B-cell lymphoma. N Engl J Med 378:1396-1407, 2018 18. Chapuy B, Stewart C, Dunford AJ, et al: Molecular subtypes of diffuse large B cell lymphoma are associated with distinct pathogenic mechanisms and outcomes. Nat Med 24:679-690, 2018 19. Reddy A, Zhang J, Davis NS, et al: Genetic and functional drivers of diffuse large B cell lymphoma. Cell 171:481-494.e15, 2017 20. Kridel R, Mottok A, Farinha P, et al: Cell-of-origin of transformed follicular lymphoma. Blood 126:2118-2127, 2015 21. Arthur S, Jiang A, Grande B, et al: Genome-wide discovery of somatic coding and regulatory variants in Diffuse Large B-cell Lymphoma. Nat Commun 9: 4001, 2018. 22. Ortega-Molina A, Boss IW, Canela A, et al: The histone lysine methyltransferase KMT2D sustains a gene expression program that represses B cell lymphoma development. Nat Med 21:1199-1208, 2015 23. Jiang Y, Ortega-Molina A, Geng H, et al: CREBBP Inactivation Promotes the Development of HDAC3-Dependent Lymphomas. Cancer Discov 7:38-53, 2017 24. Scott DW, Mottok A, Ennishi D, et al: Prognostic significance of diffuse large B-cell lymphoma cell of origin determined by digital gene expression in formalin-fixed paraffin-embedded tissue biopsies. J Clin Oncol 33:2848-2856, 2015 25. Victora GD, Dominguez-Sola D, Holmes AB, et al: Identification of human germinal center light and dark zone cells and their relationship to human B-cell lymphomas. Blood 120:2240-8, 2012 26. Milpied P, Cervera-Marzal I, Mollichella M-L, et al: Human germinal center transcriptional programs are de-synchronized in B cell lymphoma. Nat Immunol 19:1013-1024, 2018 27. Mottok A, Wright G, Rosenwald A, et al: Molecular classification of primary mediastinal large B-cell lymphoma using routinely available tissue specimens. Blood 132:2401-2405, epub Sep 26th, 2018 28. Dominguez-Sola D, Victora GD, Ying CY, et al: The proto-oncogene MYC is required for selection in the germinal center and cyclic reentry. Nat Immunol 13:1083-1091, 2012 29. Calado DP, Sasaki Y, Godinho SA, et al: The cell-cycle regulator c-Myc is essential for the formation and maintenance of germinal centers. Nat Immunol 13:1092-1100, 2012 30. Green MR, Kihira S, Liu CL, et al: Mutations in early follicular lymphoma progenitors are associated with suppressed antigen presentation. Proc Natl Acad Sci U S A 112:E1116-25, 2015 31. Davids MS, Roberts AW, Seymour JF, et al: Phase i first-in-human study of venetoclax in patients with relapsed or refractory non-Hodgkin lymphoma. J Clin Oncol 35:826-833, 2017 32. Scott DW, Wright GW, Williams PM, et al: Determining cell-of-origin subtypes of diffuse large B-cell lymphoma using gene expression in formalin-fixed paraffin-embedded tissue. Blood 123:1214-1217, 2014 33. Cancer Genome Atlas Research Network: Comprehensive molecular characterization of gastric adenocarcinoma. Nature 513:202-9, 2014 34. Love MI, Huber W, Anders S: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15:1-21, 2014 35. Sergushichev A: An algorithm for fast preranked gene set enrichment analysis using cumulative statistic calculation. bioRxiv 60012, 2016

Claims

1. A kit consisting of probes and / or primers sufficient for the detection of all of the following genes, as well as regents sufficient to facilitate detection and / or quantification of the gene expression products, to determine the molecular subgroup of an aggressive B-cell lymphoma, wherein the molecular subgroup is a positive double-hit signature for high-grade B-cell lymphomas with MYC and BCL2 and / or BCL6 rearrangements ("DHIT signature") (DHITsig-pos) or a negative DHIT signature (DHITsig-neg) lymphoma: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1, and optionally: more than one of AC104699.1, ACPP, ADTRP, ALS2, ANKRD33B, ARHGAP25, ARID3B, ARPC2, ASS1P1, ATF4, BCL2A1, CAB39, CCDC78, CCL17, CCL22, CFLAR, COBLL1, CPEB4, CR2, CTD-3074O7.5, DANCR, DGKG, DOCK10, El F4EBP3, ETV5, FAM216A, FCRL5, FHIT, GALNT6, GNG2, HAGHL, HIVEP1, HMSD, HRK, IL10RA, LINC00957, LRRC75A-AS1, LTA, MACROD1, MREG, MVP, MYEOV, NCOA1, NMRAL1, PARP15, PIK3CD-AS2, POU3F1, PPP1R14B, PTPRJ, RASGRF1, RPL13, RPL35, RPL6, RPL7, RPS8, SFXN4, SGCE, SIAH2, SIGLEC14, SLC29A2, SMARCB1, SNHG11, SNHG17, SNHG7, SOX9, SPTBN2, ST8SIA4, SUGCT, TACC1, TERT, TLE4, UQCRH, WDFY1, or WNK2, and optionally: more than one of BCL2, FCGR2B or PVT1, and optionally: more than one of ASB13, AUH, BANK1, BATF3, BTG2, CARD11, CCDC50, CCL17, CREB3L2, CYB5R2, DNAJB12, FAM159A, FSCN1, GIT2, GSK3B, HOMER2, IFIH1, IK, IL13RA1, IRF4, ISY1, ITPKB, LIMA1, LIMD1, MAL, MAML3, MME, MOBKL2C, MST1R, MYBL1, NECAP2, NFIL3, OPA1, PDCD1LG2, PHF23, PIM2, PRDX2, PRKCB, PRR6, PTGIR, QSOX1, R3HDM1, RAB7L1, RCL1, RHOF, S1PR2, SERPINA9, SLAMF1, SNX11, TFPI2, TMOD1, TNFRSF13 B, TRAF1, TRIM56, UBXN4, VRK3, WAC, or WDR55.

2. A kit according to claim 1, consisting of probes and / or primers sufficient for the detection of all of the following genes, as well as regents sufficient to facilitate detection and / or quantification of the gene expression products: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1.

3. A kit according to claim 1, consisting of probes and / or primers sufficient for the detection of all of the following genes, as well as regents sufficient to facilitate detection and / or quantification of the gene expression products: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1, and more than one of AC104699.1, ACPP, ADTRP, ALS2, ANKRD33B, ARHGAP25, ARID3B, ARPC2, ASS1P1, ATF4, BCL2A1, CAB39, CCDC78, CCL17, CCL22, CFLAR, COBLL1, CPEB4, CR2, CTD-3074O7.5, DANCR, DGKG, DOCK10, El F4EBP3, ETV5, FAM216A, FCRL5, FHIT, GALNT6, GNG2, HAGHL, HIVEP1, HMSD, HRK, IL10RA, LINC00957, LRRC75A-AS1, LTA, MACROD1, MREG, MVP, MYEOV, NCOA1, NMRAL1, PARP15, PIK3CD-AS2, POU3F1, PPP1R14B, PTPRJ, RASGRF1, RPL13, RPL35, RPL6, RPL7, RPS8, SFXN4, SGCE, SIAH2, SIGLEC14, SLC29A2, SMARCB1, SNHG11, SNHG17, SNHG7, SOX9, SPTBN2, ST8SIA4, SUGCT, TACC1, TERT, TLE4, UQCRH, WDFY1, or WNK2.

4. A kit according to claim 1, consisting of probes and / or primers sufficient for the detection of all of the following genes, as well as regents sufficient to facilitate detection and / or quantification of the gene expression products: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1, and more than one of ASB13, AUH, BANK1, BATF3, BTG2, CARD11, CCDC50, CCL17, CREB3L2, CYB5R2, DNAJB12, FAM159A, FSCN1, GIT2, GSK3B, HOMER2, IFIH1, IK, IL13RA1, IRF4, ISY1, ITPKB, LIMA1, LIMD1, MAL, MAML3, MME, MOBKL2C, MST1R, MYBL1, NECAP2, NFIL3, OPA1, PDCD1LG2, PHF23, PIM2, PRDX2, PRKCB, PRR6, PTGIR, QSOX1, R3HDM1, RAB7L1, RCL1, RHOF, S1PR2, SERPINA9, SLAMF1, SNX11, TFPI2, TMOD1, TNFRSF13 B, TRAF1, TRIM56, UBXN4, VRK3, WAC, or WDR55.

5. A kit according to claim 1, consisting of probes and / or primers sufficient for the detection of all of the following genes, as well as regents sufficient to facilitate detection and / or quantification of the gene expression products: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1, and more than one of BCL2, FCGR2B or PVT1.

6. A kit according to claim 1, consisting of probes and / or primers sufficient for the detection of all of the following genes, as well as regents sufficient to facilitate detection and / or quantification of the gene expression products: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1, and more than one of AC104699.1, ACPP, ADTRP, ALS2, ANKRD33B, ARHGAP25, ARID3B, ARPC2, ASS1P1, ATF4, BCL2A1, CAB39, CCDC78, CCL17, CCL22, CFLAR, COBLL1, CPEB4, CR2, CTD-3074O7.5, DANCR, DGKG, DOCK10, El F4EBP3, ETV5, FAM216A, FCRL5, FHIT, GALNT6, GNG2, HAGHL, HIVEP1, HMSD, HRK, IL10RA, LINC00957, LRRC75A-AS1, LTA, MACROD1, MREG, MVP, MYEOV, NCOA1, NMRAL1, PARP15, PIK3CD-AS2, POU3F1, PPP1R14B, PTPRJ, RASGRF1, RPL13, RPL35, RPL6, RPL7, RPS8, SFXN4, SGCE, SIAH2, SIGLEC14, SLC29A2, SMARCB1, SNHG11, SNHG17, SNHG7, SOX9, SPTBN2, ST8SIA4, SUGCT, TACC1, TERT, TLE4, UQCRH, WDFY1, or WNK2, and more than one of BCL2, FCGR2B or PVT1.

7. A kit according to claim 1, consisting of probes and / or primers sufficient for the detection of all of the following genes, as well as regents sufficient to facilitate detection and / or quantification of the gene expression products: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1, and more than one of ASB13, AUH, BANK1, BATF3, BTG2, CARD11, CCDC50, CCL17, CREB3L2, CYB5R2, DNAJB12, FAM159A, FSCN1, GIT2, GSK3B, HOMER2, IFIH1, IK, IL13RA1, IRF4, ISY1, ITPKB, LIMA1, LIMD1, MAL, MAML3, MME, MOBKL2C, MST1R, MYBL1, NECAP2, NFIL3, OPA1, PDCD1LG2, PHF23, PIM2, PRDX2, PRKCB, PRR6, PTGIR, QSOX1, R3HDM1, RAB7L1, RCL1, RHOF, S1PR2, SERPINA9, SLAMF1, SNX11, TFPI2, TMOD1, TNFRSF13 B, TRAF1, TRIM56, UBXN4, VRK3, WAC, or WDR55, and more than one of BCL2, FCGR2B or PVT1.

8. A kit according to claim 1, consisting of probes and / or primers sufficient for the detection of all of the following genes, as well as regents sufficient to facilitate detection and / or quantification of the gene expression products: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1, and more than one of AC104699.1, ACPP, ADTRP, ALS2, ANKRD33B, ARHGAP25, ARID3B, ARPC2, ASS1P1, ATF4, BCL2A1, CAB39, CCDC78, CCL17, CCL22, CFLAR, COBLL1, CPEB4, CR2, CTD-3074O7.5, DANCR, DGKG, DOCK10, El F4EBP3, ETV5, FAM216A, FCRL5, FHIT, GALNT6, GNG2, HAGHL, HIVEP1, HMSD, HRK, IL10RA, LINC00957, LRRC75A-AS1, LTA, MACROD1, MREG, MVP, MYEOV, NCOA1, NMRAL1, PARP15, PIK3CD-AS2, POU3F1, PPP1R14B, PTPRJ, RASGRF1, RPL13, RPL35, RPL6, RPL7, RPS8, SFXN4, SGCE, SIAH2, SIGLEC14, SLC29A2, SMARCB1, SNHG11, SNHG17, SNHG7, SOX9, SPTBN2, ST8SIA4, SUGCT, TACC1, TERT, TLE4, UQCRH, WDFY1, or WNK2, and more than one of ASB13, AUH, BANK1, BATF3, BTG2, CARD11, CCDC50, CCL17, CREB3L2, CYB5R2, DNAJB12, FAM159A, FSCN1, GIT2, GSK3B, HOMER2, IFIH1, IK, IL13RA1, IRF4, ISY1, ITPKB, LIMA1, LIMD1, MAL, MAML3, MME, MOBKL2C, MST1R, MYBL1, NECAP2, NFIL3, OPA1, PDCD1LG2, PHF23, PIM2, PRDX2, PRKCB, PRR6, PTGIR, QSOX1, R3HDM1, RAB7L1, RCL1, RHOF, S1PR2, SERPINA9, SLAMF1, SNX11, TFPI2, TMOD1, TNFRSF13 B, TRAF1, TRIM56, UBXN4, VRK3, WAC, or WDR55.

9. A kit according to claim 1, consisting of probes and / or primers sufficient for the detection of all of the following genes, as well as regents sufficient to facilitate detection and / or quantification of the gene expression products: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1, and more than one of AC104699.1, ACPP, ADTRP, ALS2, ANKRD33B, ARHGAP25, ARID3B, ARPC2, ASS1P1, ATF4, BCL2A1, CAB39, CCDC78, CCL17, CCL22, CFLAR, COBLL1, CPEB4, CR2, CTD-3074O7.5, DANCR, DGKG, DOCK10, El F4EBP3, ETV5, FAM216A, FCRL5, FHIT, GALNT6, GNG2, HAGHL, HIVEP1, HMSD, HRK, IL10RA, LINC00957, LRRC75A-AS1, LTA, MACROD1, MREG, MVP, MYEOV, NCOA1, NMRAL1, PARP15, PIK3CD-AS2, POU3F1, PPP1R14B, PTPRJ, RASGRF1, RPL13, RPL35, RPL6, RPL7, RPS8, SFXN4, SGCE, SIAH2, SIGLEC14, SLC29A2, SMARCB1, SNHG11, SNHG17, SNHG7, SOX9, SPTBN2, ST8SIA4, SUGCT, TACC1, TERT, TLE4, UQCRH, WDFY1, or WNK2, and more than one of ASB13, AUH, BANK1, BATF3, BTG2, CARD11, CCDC50, CCL17, CREB3L2, CYB5R2, DNAJB12, FAM159A, FSCN1, GIT2, GSK3B, HOMER2, IFIH1, IK, IL13RA1, IRF4, ISY1, ITPKB, LIMA1, LIMD1, MAL, MAML3, MME, MOBKL2C, MST1R, MYBL1, NECAP2, NFIL3, OPA1, PDCD1LG2, PHF23, PIM2, PRDX2, PRKCB, PRR6, PTGIR, QSOX1, R3HDM1, RAB7L1, RCL1, RHOF, S1PR2, SERPINA9, SLAMF1, SNX11, TFPI2, TMOD1, TNFRSF13 B, TRAF1, TRIM56, UBXN4, VRK3, WAC, or WDR55, and more than one of BCL2, FCGR2B or PVT1.

10. A method of classifying an aggressive B-cell lymphoma comprising: i) preparing a gene expression profile for all of the following genes in a test sample: AFMID, ALOX5, BATF, CD24, CD80, CDK5R1, EBI3, GAMT, GPR137B, IL21R, IRF4, JCHAIN, LY75, MIR155HG, MYC, OR13A1, PEG10, QRSL1, RFFL, RGCC, SEMA7A, SGPP2, SLC25A27, SMIM14, SNHG19, STAT3, SYBU, TNFSF8, VASP, and VOPP1 from said test sample; and iii) classifying said test sample into an aggressive B-cell lymphoma having a positive double-hit signature for high-grade B-cell lymphomas with MYC and BCL2 and / or BCL6 rearrangements ("DHIT signature") (DHITsig-pos) or an aggressive B-cell lymphoma having a negative DHIT signature (DHITsig-neg) based on said gene expression profile.

11. The method of claim 10 further comprising preparing a gene expression profile for two or more genes in the test sample selected from: AC104699.1, ACPP, ADTRP, ALS2, ANKRD33B, ARHGAP25, ARID3B, ARPC2, ASS1P1, ATF4, BCL2A1, CAB39, CCDC78, CCL17, CCL22, CFLAR, COBLL1, CPEB4, CR2, CTD-3074O7.5, DANCR, DGKG, DOCK10, El F4EBP3, ETV5, FAM216A, FCRL5, FHIT, GALNT6, GNG2, HAGHL, HIVEP1, HMSD, HRK, IL10RA, LINC00957, LRRC75A-AS1, LTA, MACROD1, MREG, MVP, MYEOV, NCOA1, NMRAL1, PARP15, PIK3CD-AS2, POU3F1, PPP1R14B, PTPRJ, RASGRF1, RPL13, RPL35, RPL6, RPL7, RPS8, SFXN4, SGCE, SIAH2, SIGLEC14, SLC29A2, SMARCB1, SNHG11, SNHG17, SNHG7, SOX9, SPTBN2, ST8SIA4, SUGCT, TACC1, TERT, TLE4, UQCRH, WDFY1, or WNK2.

12. The method of any one of claims 10 or 11 wherein the genes further comprise one or more of the following genes: ASB13, AUH, BANK1, BATF3, BTG2, CARD11, CCDC50, CCL17, CREB3L2, CYB5R2, DNAJB12, FAM159A, FSCN1, GIT2, GSK3B, HOMER2, IFIH1, IK, IL13RA1, IRF4, ISY1, ITPKB, LIMA1, LIMD1, MAL, MAML3, MME, MOBKL2C, MST1R, MYBL1, NECAP2, NFIL3, OPA1, PDCD1LG2, PHF23, PIM2, PRDX2, PRKCB, PRR6, PTGIR, QSOX1, R3HDM1, RAB7L1, RCL1, RHOF, S1PR2, SERPINA9, SLAMF1, SNX11, TFPI2, TMOD1, TNFRSF13 B, TRAF1, TRIM56, UBXN4, VRK3, WAC, WDR55.

13. The method of any one of claims 10 to 12 wherein the genes further comprise one or more of BCL2, FCGR2B and PVT1.

14. The method of any one of claims 10 to 13 wherein the aggressive B-cell lymphoma is a diffuse large B-cell lymphoma (DLBCL) or high-grade B-cell lymphoma (HGBL) and / or wherein the subject is a human.