WO2021087044A1 - Methods for predicting responsiveness of lymphoma to drug and methods for treating lymphoma - Google Patents

Methods for predicting responsiveness of lymphoma to drug and methods for treating lymphoma Download PDF

Info

Publication number
WO2021087044A1
WO2021087044A1 PCT/US2020/057860 US2020057860W WO2021087044A1 WO 2021087044 A1 WO2021087044 A1 WO 2021087044A1 US 2020057860 W US2020057860 W US 2020057860W WO 2021087044 A1 WO2021087044 A1 WO 2021087044A1
Authority
WO
WIPO (PCT)
Prior art keywords
lymphoma
patient
patients
dlbcl
genes
Prior art date
Application number
PCT/US2020/057860
Other languages
French (fr)
Inventor
Fadi George TOWFIC
Maria ORTIZ-ESTEVEZ
Matthew Stokes
Nicholas STONG
Anita GANDHI
Patrick HAGNER
Chris Huang
Original Assignee
Celgene Corporation
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Celgene Corporation filed Critical Celgene Corporation
Publication of WO2021087044A1 publication Critical patent/WO2021087044A1/en

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B40/00—ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
    • G16B40/30—Unsupervised data analysis
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61P—SPECIFIC THERAPEUTIC ACTIVITY OF CHEMICAL COMPOUNDS OR MEDICINAL PREPARATIONS
    • A61P35/00—Antineoplastic agents
    • A61P35/02—Antineoplastic agents specific for leukemia
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
    • C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
    • C12Q1/6876—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes
    • C12Q1/6883—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes for diseases caused by alterations of genetic material
    • C12Q1/6886—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes for diseases caused by alterations of genetic material for cancer
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/106—Pharmacogenomics, i.e. genetic variability in individual responses to drugs and drug metabolism
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/112—Disease subtyping, staging or classification
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/158—Expression markers
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B25/00—ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
    • G16B25/10—Gene or protein expression profiling; Expression-ratio estimation or normalisation

Definitions

  • FIELD Provided herein are methods of predicting the responsiveness of a lymphoma patient to a cancer treatment. Also provided herein are methods of treating a lymphoma patient based on predicting the responsiveness of the lymphoma patient to a cancer treatment. 2.
  • NHLs non-Hodgkin lymphomas
  • Types of NHL vary significantly in their severity, from indolent to very aggressive. Less aggressive non-Hodgkin lymphomas are compatible with a long survival while more aggressive non-Hodgkin lymphomas can be rapidly fatal without treatment. They can be formed from either B-cells or T-cells.
  • B-cell non-Hodgkin lymphomas include Burkitt lymphoma, chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL), diffuse large B-cell lymphoma, follicular lymphoma, immunoblastic large cell lymphoma, precursor B-lymphoblastic lymphoma, and mantle cell lymphoma.
  • T-cell non-Hodgkin lymphomas include mycosis fungoides, anaplastic large cell lymphoma, and precursor T-lymphoblastic lymphoma. Prognosis and treatment depend on the stage and type of disease.
  • DLBCL Diffuse large B-cell lymphoma
  • the diffuse large B-cell lymphomas can be divided into distinct molecular subtypes according to their gene profiling patterns: germinal-center B-cell–like DLBCL (GCB- DLBCL), activated B-cell–like DLBCL (ABC-DLBCL), and primary mediastinal B-cell lymphoma (PMBL) or unclassified type.
  • GCB- DLBCL germinal-center B-cell–like DLBCL
  • ABS-DLBCL activated B-cell–like DLBCL
  • PMBL primary mediastinal B-cell lymphoma
  • neoplastic tissue may not completely remove neoplastic tissue. Radiation therapy is only effective when the neoplastic tissue exhibits a higher sensitivity to radiation than normal tissue. Radiation therapy can also often elicit serious side effects. Hormonal therapy is rarely given as a single agent. Although hormonal therapy can be effective, it is often used to prevent or delay recurrence of cancer after other treatments have removed the majority of cancer cells. Biological therapies and immunotherapies are limited in number and may produce side effects such as rashes or swellings, flu-like symptoms, including fever, chills and fatigue, digestive tract problems or allergic reactions. [0007] In the context of DLBCL, treatment usually includes administration of a combination of chemotherapy and antibody therapy.
  • DLBCL The most widely used treatment of DLBCL is a combination of antibody rituximab (Rituxan) and chemotherapy drugs (cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP), and in some cases etoposide is added (R- EPOCH)).
  • DLBCL also typically requires immediate treatment upon diagnosis due to how quickly the disease can advance. For some patients, DLBCL returns or becomes refractory following treatment.
  • Several alternative treatments, some of which can include use of lenalidomide, are currently being tested in clinical trials for patients with newly diagnosed, relapsed or refractory DLBCL. See Czuczman et al., Clin Cancer Res., 2017, 23:4127-4137.
  • lymphomas are also types of indolent lymphomas.
  • Indolent B- cell lymphomas include, for example, follicular lymphoma, small lymphocytic lymphoma; nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type, and mycosis fungoides.
  • the clinical course of B-cell indolent lymphoma patients is characterized, among other features, by the risk of histologic transformation (HT) to an aggressive lymphoma, mostly DLBCL, and, less commonly, to Burkitt lymphoma (BL) or other types of aggressive lymphomas).
  • HT histologic transformation
  • BL Burkitt lymphoma
  • DLBCL has traditionally been classified by cell of origin (COO) subcategories based on tumor gene expression profiles and includes Activated B-Cell (ABC) and Germinal Center B-Cell (GCB) subtypes.
  • COO cell of origin
  • ABS Activated B-Cell
  • GCB Germinal Center B-Cell
  • a method for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining reference biological samples from each patient in a reference patient group comprising reference patients having a lymphoma; (b) clustering the reference patient group into subgroups of patients using gene expression levels in the reference biological samples; (c) obtaining a biological sample from the lymphoma patient; (d) determining to which subgroup the lymphoma patient belongs using gene expression levels in the biological sample from the lymphoma patient; and (e) predicting the responsiveness of the lymphoma patient to a first cancer treatment.
  • step (b) comprises administering to the lymphoma patient a second cancer treatment.
  • step (b) comprises generating clustering information defining relationships between the expression levels of one or more genes in the reference biological samples; and rearranging heat map representation based on the clustering information.
  • step (b) uses a hierarchical method or a non-hierarchical method.
  • step (b) uses Cluster of Cluster Analysis (COCA) method or iClusterPlus method.
  • step (b) uses COCA method.
  • step (b) uses iClusterPlus method.
  • the reference patients in the reference patient group are clustered into 2-15 subgroups. [0018] In some embodiments, the reference patients in the reference patient group are clustered into 8 subgroups. [0019] In some embodiments, the method provided herein further comprises training a classifier model using expression levels of one or more genes in the reference biological samples. [0020] In some embodiments, expression levels of one, two, three, four, five, or more of the genes identified in Table 1 are used in training the classifier model. [0021] In some embodiments, expression levels of all the genes identified in Table 1 are used in training the classifier model. [0022] In some embodiments, the classifier model is a grouped multinomial generalized linear model (GLM).
  • GLM grouped multinomial generalized linear model
  • the lymphoma is diffuse large B-cell lymphoma (DLBCL).
  • the lymphoma is indolent B cell lymphoma.
  • the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma.
  • the lymphoma is follicular lymphoma.
  • the lymphoma is nodal marginal zone B-cell lymphoma. [0028] In some embodiments, the lymphoma is mantle cell lymphoma. [0029] In some embodiments, the lymphoma is chronic lymphocytic leukemia.
  • the reference patients in the reference patient group are clustered into 8 subgroups; and wherein: (i) subgroup D1 comprises about 55% to 65% patients having germinal center B-cell-like (GCB) DLBCL, about 20% to 30% patients having activated B-cell like (ABC) lymphoma, and about 20% to 30% patients who are DHITsig+ DLBCL patients; (ii) subgroup D2 comprises about 45% to 55% patients having GCB DLBCL, about 20% to 45% patients having ABC DLBCL, and about 20% to 25% patients who are DHITsig+ DLBCL patients; (iii) subgroup D3 comprises about 90% to 95% patients having GCB DLBCL, about 0% to 10% patients having ABC DLBCL, and about 0% to 20% patients who are DHITsig+ DLBCL patients; (iv) subgroup D4 comprises about 0% to 10% patients having GCB DLBCL, about 90% to 100% patients having ABC DLBCL, and about 0% to 10% patients who are
  • the first cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP).
  • R-CHOP rituximab
  • the method comprises predicting that the patient is not likely to be responsive to the first cancer treatment.
  • the second cancer treatment is R-CHOP.
  • the second cancer treatment is not R-CHOP.
  • the second cancer treatment when the lymphoma patient is determined to belong to subgroup D4 or D8, the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor.
  • BET bromodomain and extra-terminal
  • CDK cyclin dependent kinase
  • a method for predicting the responsiveness of lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first patient having a lymphoma; (b) determining the level of expression of one, two, three, four, five or more of the genes identified in Table 1; (c)comparing the level of expression of the one, two, three, four, five or more of the genes identified in Table 1 in the first biological sample with the level of expression of the same genes in a second biological sample(s) from a second patient(s), wherein the second lymphoma patient(s) is responsive to the cancer treatment, and wherein the similar expression of the one, two, three, four, five or more of the genes in the first biological sample relative to the level of expression of the one, two, three, four, five or more of the genes in the second biological sample(s) indicates that the lymphoma in the first patient will be responsive to treatment with the cancer treatment.
  • a method for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first lymphoma patient; (b) determining the expression of the genes or a certain subset of genes set forth in Table 1 or any combination thereof in the first biological sample; and (c) comparing the gene expression profile of the genes or subset of genes in the first biological sample to (i) the gene expression profile of the genes or subset of genes in biological samples from lymphoma patients which are responsive to the drug and (ii) the gene expression of the genes or subset of genes in biological samples from lymphoma patients which are not responsive to the cancer treatment, wherein a gene expression profile for the genes or subset of genes in the first biological sample similar to the gene expression profile for the genes or subset of genes in biological samples from lymphoma patients which are responsive to the cancer treatment indicates that the first lymphoma patient will be responsive to the cancer treatment, and a gene expression profile for the genes or
  • a method of treating a lymphoma patient comprising: (i) identifying a lymphoma patient who is likely to be responsive to the cancer treatment according to the predicting method provided herein, and (ii) administering to the lymphoma patient the cancer treatment.
  • a method of treating a lymphoma patient comprising: (i) identifying a lymphoma patient who is likely to be not responsive to the cancer treatment according to the predicting method provided herein, and (ii) administering to the lymphoma patient an alternative cancer treatment.
  • the cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP).
  • the alternative cancer treatment is a bromodomain and extra- terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor.
  • the lymphoma is diffuse large B-cell lymphoma (DLBCL).
  • the lymphoma is indolent B cell lymphoma.
  • the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma.
  • the lymphoma is follicular lymphoma.
  • the lymphoma is nodal marginal zone B-cell lymphoma.
  • the lymphoma is mantle cell lymphoma.
  • the lymphoma is chronic lymphocytic leukemia.
  • the level of expression of all the genes identified in Table 1 is determined in step (b) and compared in step (c).
  • the biological samples are tumor biopsy samples.
  • the determining step comprises detecting the presence or amount of a complex in the biological sample, wherein the presence or amount of the complex indicates the expression level of the genes in each subset of genes.
  • the complex is a hybridization complex.
  • the hybridization complex is attached to a solid support.
  • the complex is detectably labeled.
  • the determining step comprises detecting the presence or amount of a reaction product in the biological sample, wherein the presence or amount of the reaction product indicates the expression level of the genes.
  • the reaction product is detectably labeled.
  • the reference patients are refractory DLBCL patient.
  • the reference patients are relapsed DLBCL patient.
  • the reference patients are newly diagnosed DLBCL patient.
  • the lymphoma patient is a refractory DLBCL patient.
  • the lymphoma patient is a relapsed DLBCL patient.
  • the lymphoma patient is a newly diagnosed DLBCL patient.
  • the lymphoma patient is a GCB DLBCL patient.
  • the lymphoma patient is an ABC DLBCL patient.
  • the lymphoma patient is a DHITsig+ DLBCL patient.
  • the lymphoma patient is a DHITsig- DLBCL patient. 4. BRIEF DESCRIPTION OF THE FIGURES [0068]
  • FIG.1 illustrates that Principal Component Analysis (PCA) projection of ssNorm datasets shows a high degree of overlap and no systematic stratification by dataset.
  • PCA Principal Component Analysis
  • FIG.2 depicts the flowchart of unsupervised clustering input, clustering evaluation, classifier application to independent data, and exemplary potential biological interpretation of patient subgroups. All four matrices of features (gene expression, Hallmark, C1, and LM23) were used in the unsupervised clustering methods.
  • FIGS.3A-3E show heatmaps of clustering by each different features matrices and heatmap of consensus matrix.
  • FIG.3A heatmap of clustering by gene expression
  • FIG.3B heatmap of clustering by Hallmark pathways
  • FIG.3C heatmap of clustering by the C1 set of positional cytoband features
  • FIG.3D heatmap of clustering by LM23 cell type features
  • FIG.3E heatmap of consensus matrix.
  • FIGS.5A-5E show the gene expression heatmaps for the eight patient clusters identified in the discovery (FIG.5A), ndMER (FIG.5B), Lenz datasets (FIG.5C), rrMER (FIG.5D), and CC-122 (FIG.5E). The heatmaps show the same top 100 most differentially expressed genes for each cluster (50 upregulated, 50 downregulated) in FIGS.5A-5E.
  • FIGS.6A-6B illustrate that the discovery dataset (Commercial) and replication dataset (ndMER) show consistent cluster characteristics. Each pair of bars denotes the discovery dataset (commercial dataset; in blank color) and replication dataset (ndMER dataset; in light grey color).
  • FIG.6A fractions of each cluster that are GCB DLBCL patients in the Commercial and MER datasets;
  • FIG.6B fractions of each cluster that are TME+ DLBCL patients in Commercial and MER datasets. Error bars represent the 95% confidence interval, showing few differences in the COO or TME content of the clusters between datasets.
  • FIG.7 shows the cluster prevalence of the TME+ and TME- patients across the eight subgroups in a combination dataset of MER, Schmitz, and Lenz datasets.
  • FIG.8 illustrates the event-free survival (EFS) of the eight patient clusters in MER dataset treated with R-CHOP as first-line treatment. The EFS of the eight patient clusters show significantly different survival patterns (p ⁇ 0.0001 by longrank test).
  • FIG.9 illustrates that the cluster prevalence of the eight clusters is consistent across ndDLBCL datasets (Commercial, Lenz and ndMER datasets). In rrDLBCL datasets (rrMER and CC-122 datasets) the prevalence of the high risk clusters (D4 and D8) increases.
  • FIG.10 shows that the DESeq2-derived DHITsig score appropriately ranks FISH Double HIT+ samples highly.
  • the selected threshold (circled) identifies all FISH Double HIT+ patients, minimizes false positives, and matches the target prevalence of Double HIT+ patients among GCBs.
  • FIG.11A overall survival (OS) of the DHITsig+ and DHITsig- GCB patients in the Lenz dataset (treated with R-CHOP);
  • FIG.11B overall survival (OS) of the DHITsig+ and DHITsig- GCB patients in the MER dataset (treated with R-CHOP).
  • FIG.12 shows the expression of the BCL6 signature across the eight subgroups in the ndMER dataset.
  • FIGS.13A-13D display the percentage of nucleated cells of CD8, CD4, CD163, and CD20 DLBCL patients in each of the identified eight subgroups (FIG.13A: CD8 cells; FIG. 13B: CD4 cells; FIG.13C: CD163; and FIG.13D: CD20. 5.
  • lymphoma includes, but is not limited to, Hodgkin’s lymphoma, non-Hodgkin’s lymphoma, diffuse large B-Cell lymphoma, indolent B-cell lymphoma, follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, cutaneous T-Cell lymphoma, cutaneous B-Cell lymphoma, mycosis fungoide, mantle cell lymphoma, and chronic lymphocytic leukemia.
  • the terms “treat,” “treating,” and “treatment” refer to an action that occurs while a patient is suffering from the specified cancer (a specific type of lymphoma, e.g., DLBCL), which reduces the severity of the cancer or retards or slows the progression of the cancer.
  • a specific type of lymphoma e.g., DLBCL
  • sensitivity or “sensitive” when made in reference to a cancer treatment is a relative term which refers to the degree of effectiveness of the cancer treatment in lessening or decreasing the progress of a tumor or the cancer being treated.
  • the term “increased sensitivity” when used in reference to treatment of a cell or tumor in connection with a compound refers to an increase of, at least about 5%, or more, in the effectiveness of the cancer treatment.
  • the term “therapeutically effective amount” of a cancer treatment is an amount sufficient to provide a therapeutic benefit in the treatment or management of a cancer, or to delay or minimize one or more symptoms associated with the presence of the cancer.
  • a therapeutically effective amount of a compound means an amount of therapeutic agent, alone or in combination with other therapies, which provides a therapeutic benefit in the treatment or management of the cancer.
  • terapéuticaally effective amount can encompass an amount that improves overall therapy, reduces or avoids symptoms or causes of cancer, or enhances the therapeutic efficacy of another therapeutic agent.
  • the term also refers to the amount of a compound that is sufficient to elicit the biological or medical response of a biological molecule (e.g., a protein, enzyme, RNA, or DNA), cell, tissue, system, animal, or human, which is being sought by a researcher, veterinarian, medical doctor, or clinician.
  • responsiveness or “responsive” when used in reference to a cancer treatment refers to the degree of effectiveness of the treatment in lessening or decreasing the symptoms of a cancer, e.g., DLBCL, being treated.
  • the term “increased responsiveness” when used in reference to a treatment of a cell or a subject refers to an increase in the effectiveness in lessening or decreasing the symptoms of the disease compared to a reference treatment (e.g., of the same cell or subject, or of a different cell or subject) when measured using any methods known in the art.
  • the increase in the effectiveness is at least about 5%, at least about 10%, at least about 20%, at least about 30%, at least about 40%, or at least about 50%.
  • the terms “effective subject response,” “effective patient response,” and “effective patient tumor response” refer to any increase in the therapeutic benefit to the patient.
  • An “effective patient tumor response” can be, for example, about 5%, about 10%, about 25%, about 50%, or about 100% decrease in the rate of progress of the tumor.
  • An “effective patient tumor response” can be, for example, about 5%, about 10%, about 25%, about 50%, or about 100% decrease in the physical symptoms of a cancer.
  • An “effective patient tumor response” can also be, for example, about 5%, about 10%, about 25%, about 50%, about 100%, about 200%, or more increase in the response of the patient, as measured by any suitable means, such as gene expression, cell counts, assay results, tumor size, etc.
  • An improvement in the cancer e.g., DLBCL or a subtype thereof
  • cancer-related disease can be characterized as a complete or partial response.
  • “Complete response” refers to an absence of clinically detectable disease with normalization of any previously abnormal radiographic studies, bone marrow, and cerebrospinal fluid (CSF) or abnormal monoclonal protein measurements. “Partial response” refers to at least about 10%, about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, or about 90% decrease in all measurable tumor burden (i.e., the number of malignant cells present in the subject, or the measured bulk of tumor masses or the quantity of abnormal monoclonal protein) in the absence of new lesions.
  • treatment contemplates both a complete and a partial response.
  • the term “likelihood” generally refers to an increase in the probability of an event.
  • the term “likelihood” when used in reference to the effectiveness of a patient tumor response generally contemplates an increased probability that the rate of tumor progress or tumor cell growth will decrease.
  • the term “likelihood” when used in reference to the effectiveness of a patient tumor response can also generally mean the increase of indicators, such as mRNA or protein expression, that may evidence an increase in the progress in treating the tumor.
  • the term “predict” generally means to determine or tell in advance. When used to “predict” the effectiveness of a cancer treatment, for example, the term “predict” can mean that the likelihood of the outcome of the cancer treatment can be determined at the outset, before the treatment has begun, or before the treatment period has progressed substantially.
  • the term “source” when used in reference to a reference sample refers to the origin of a sample. For example, a sample that is taken from blood would have a reference sample that is also taken from blood. Similarly, a sample that is taken from bone marrow would have a reference sample that is also taken from the bone marrow.
  • the term “refractory” or “resistant” refers to a disorder, disease, or condition that has not responded to prior treatment that can include one or more lines of therapy. In one embodiment, the disorder, disease, or condition has been previously treated one, two, three or four lines of therapy.
  • the disorder, disease, or condition has been previously treated with two or more lines of treatment, and has less than a complete response (CR) to most recent systemic therapy containing regimen.
  • CR complete response
  • the term “relapsed” refers to a disorder, disease, or condition that responded to treatment (e.g., achieved a complete response) then had progression.
  • the treatment can include one or more lines of therapy.
  • the term “expressed” or “expression” as used herein refers to the transcription from a gene to give an RNA nucleic acid molecule at least complementary in part to a region of one of the two nucleic acid strands of the gene.
  • the term “expressed” or “expression” as used herein also refers to the translation from the RNA molecule to give a protein, a polypeptide, or a portion thereof.
  • a “biological marker” or “biomarker” is a substance whose detection indicates a particular biological state, such as, for example, the presence of a type of cancer. In some embodiments, biomarkers can be determined individually. In other embodiments, several biomarkers can be measured simultaneously.
  • the terms “polypeptide” and “protein,” as used interchangeably herein, refer to a polymer of three or more amino acids in a serial array, linked through peptide bonds. The term “polypeptide” includes proteins, protein fragments, protein analogues, oligopeptides, and the like.
  • polypeptide as used herein can also refer to a peptide.
  • the amino acids making up the polypeptide may be naturally derived, or may be synthetic.
  • the polypeptide can be purified from a biological sample.
  • the polypeptide, protein, or peptide also encompasses modified polypeptides, proteins, and peptides, e.g., glycopolypeptides, glycoproteins, or glycopeptides; or lipopolypeptides, lipoproteins, or lipopeptides.
  • antibody immunoglobulin
  • Ig immunoglobulin
  • Antibodies provided herein include, but are not limited to, synthetic antibodies, monoclonal antibodies, polyclonal antibodies, recombinantly produced antibodies, multispecific antibodies (including bi-specific antibodies), human antibodies, humanized antibodies, chimeric antibodies, intrabodies, single-chain Fvs (scFv) (e.g., including monospecific, bispecific, etc.), camelized antibodies, Fab fragments, F(ab’) fragments, disulfide- linked Fvs (sdFv), anti-idiotypic (anti-Id) antibodies, and epitope-binding fragments of any of the above.
  • scFv single-chain Fvs
  • antibodies provided herein include immunoglobulin molecules and immunologically active portions of immunoglobulin molecules, i.e., antigen binding domains or molecules that contain an antigen-binding site that immunospecifically binds to CRBN antigen (e.g., one or more complementarity determining regions (CDRs) of an anti-CRBN antibody).
  • the antibodies provided herein can be of any class (e.g., IgG, IgE, IgM, IgD, and IgA) or any subclass (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2) of immunoglobulin molecule.
  • the anti-CRBN antibodies are fully human, such as fully human monoclonal CRBN antibodies.
  • antibodies provided herein are IgG antibodies, or a subclass thereof (e.g., human IgG1 or IgG4).
  • the terms “antigen binding domain,” “antigen binding region,” “antigen binding fragment,” and similar terms refer to the portion of an antibody that comprises the amino acid residues that interact with an antigen and confer on the binding agent its specificity and affinity for the antigen (e.g., the CDR).
  • the antigen binding region can be derived from any animal species, such as rodents (e.g., rabbit, rat, or hamster) and humans.
  • the antigen-binding region is of human origin.
  • epitope refers to a localized region on the surface of an antigen that is capable of binding to one or more antigen binding regions of an antibody, that has antigenic or immunogenic activity in an animal, such as a mammal (e.g., a human), and that is capable of eliciting an immune response.
  • An epitope having immunogenic activity is a portion of a polypeptide that elicits an antibody response in an animal.
  • An epitope having antigenic activity is a portion of a polypeptide to which an antibody immunospecifically binds as determined by any method well known in the art, for example, by the immunoassays described herein.
  • Antigenic epitopes need not necessarily be immunogenic. Epitopes usually consist of chemically active surface groupings of molecules, such as amino acids or sugar side chains, and have specific three-dimensional structural characteristics as well as specific charge characteristics. A region of a polypeptide contributing to an epitope may be contiguous amino acids of the polypeptide, or the epitope may come together from two or more non-contiguous regions of the polypeptide. The epitope may or may not be a three-dimensional surface feature of the antigen. [0098] The terms “fully human antibody” and “human antibody” are used interchangeably herein and refer to an antibody that comprises a human variable region and, in some embodiments, a human constant region.
  • the terms refer to an antibody that comprises a variable region and a constant region of human origin.
  • the term “fully human antibody” includes antibodies having variable and constant regions corresponding to human germline immunoglobulin sequences as described by Kabat et al., Sequences of Proteins of Immunological Interest, U.S. Department of Health and Human Services, NIH Publication No. 91-3242 (5th ed.1991).
  • recombinant human antibody includes human antibodies that are prepared, expressed, created, or isolated by recombinant means, such as antibodies expressed using a recombinant expression vector transfected into a host cell, antibodies isolated from a recombinant, combinatorial human antibody library, antibodies isolated from an animal (e.g., a mouse or a cow) that is transgenic and/or transchromosomal for human immunoglobulin genes (see, e.g., Taylor et al., Nucl. Acids Res., 1992, 20:6287-6295) or antibodies prepared, expressed, created, or isolated by any other means that involves splicing of human immunoglobulin gene sequences to other DNA sequences.
  • Such recombinant human antibodies can have variable and constant regions derived from human germline immunoglobulin sequences. See Kabat et al., Sequences of Proteins of Immunological Interest, U.S. Department of Health and Human Services, NIH Publication No.91-3242 (5th ed.1991).
  • such recombinant human antibodies are subjected to in vitro mutagenesis (or, when an animal transgenic for human Ig sequences is used, in vivo somatic mutagenesis) and thus the amino acid sequences of the heavy chain variable and light chain variable regions of the recombinant antibodies are sequences that, while derived from and related to human germline heavy chain variable and light chain variable sequences, may not naturally exist within the human antibody germline repertoire in vivo.
  • the term “monoclonal antibody” refers to an antibody obtained from a population of homogenous or substantially homogeneous antibodies, and each monoclonal antibody will typically recognize a single epitope on the antigen.
  • a “monoclonal antibody,” as used herein, is an antibody produced by a single hybridoma or other cell, wherein the antibody immunospecifically binds to only an epitope as determined, e.g., by ELISA or other antigen-binding or competitive binding assay known in the art or in the Examples provided herein.
  • the term “monoclonal” is not limited to any particular method for making the antibody.
  • monoclonal antibodies provided herein may be made by the hybridoma method as described in Kohler et al., Nature, 1975, 256:495-497, or may be isolated from phage libraries using the techniques as described herein.
  • Polyclonal antibodies refers to an antibody population generated in an immunogenic response to a protein having many epitopes and thus includes a variety of different antibodies directed to the same or to different epitopes within the protein. Methods for producing polyclonal antibodies are known in the art.
  • level refers to the amount, accumulation, or rate of a molecule.
  • a level can be represented, for example, by the amount or the rate of synthesis of a messenger RNA (mRNA) encoded by a gene, the amount or the rate of synthesis of a polypeptide or protein encoded by a gene, or the amount or the rate of synthesis of a biological molecule accumulated in a cell or biological fluid.
  • mRNA messenger RNA
  • level refers to an absolute amount of a molecule in a sample or a relative amount of the molecule, determined under steady-state or non-steady-state conditions.
  • An mRNA that is “upregulated” is generally increased upon a given treatment or condition.
  • An mRNA that is “downregulated” generally refers to a decrease in the level of expression of the mRNA in response to a given treatment or condition. In some situations, the mRNA level can remain unchanged upon a given treatment or condition.
  • An mRNA from a patient sample can be “upregulated” when treated with a drug, as compared to a non-treated control.
  • This upregulation can be, for example, an increase of about 5%, about 10%, about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 100%, about 200%, about 300%, about 500%, about 1,000%, about 5,000%, or more of the comparative control mRNA level.
  • an mRNA can be “downregulated”, or expressed at a lower level, in response to administration of certain compounds or other agents.
  • a downregulated mRNA can be, for example, present at a level of about 99%, about 95%, about 90%, about 80%, about 70%, about 60%, about 50%, about 40%, about 30%, about 20%, about 10%, about 1%, or less of the comparative control mRNA level.
  • the level of a polypeptide or protein biomarker from a patient sample can be increased when treated with a drug, as compared to a non-treated control. This increase can be about 5%, about 10%, about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 100%, about 200%, about 300%, about 500%, about 1,000%, about 5,000%, or more of the comparative control protein level.
  • the level of a protein biomarker can be decreased in response to administration of certain compounds or other agents.
  • This decrease can be, for example, present at a level of about 99%, about 95%, about 90%, about 80%, about 70%, about 60%, about 50%, about 40%, about 30%, about 20%, about 10%, about 1%, or less of the comparative control protein level.
  • the terms “determining,” “measuring,” “evaluating,” “assessing,” and “assaying” as used herein generally refer to any form of measurement, and include determining whether an element is present or not. These terms include quantitative and/or qualitative determinations. Assessing may be relative or absolute. “Assessing the presence of” can include determining the amount of something present, as well as determining whether it is present or absent.
  • isolated and purified refer to isolation of a substance (such as mRNA, DNA, or protein) such that the substance comprises a substantial portion of the sample in which it resides, i.e., greater than the portion of the substance that is typically found in its natural or un-isolated state.
  • a substantial portion of the sample comprises, e.g., greater than 1%, greater than 2%, greater than 5%, greater than 10%, greater than 20%, greater than 50%, or more, usually up to about 90%-100% of the sample.
  • a sample of isolated mRNA can typically comprise at least about 1% total mRNA.
  • bound indicates direct or indirect attachment.
  • “bound” may refer to the existence of a chemical bond directly joining two moieties or indirectly joining two moieties (e.g., via a linking group or any other intervening portion of the molecule).
  • the chemical bond may be a covalent bond, an ionic bond, a coordination complex, hydrogen bonding, van der Waals interactions, or hydrophobic stacking, or may exhibit characteristics of multiple types of chemical bonds.
  • sample as used herein relates to a material or mixture of materials, typically, although not necessarily, in fluid form, containing one or more components of interest.
  • a sample can be a biological sample.
  • Biological sample refers to a sample obtained from a biological subject, including a sample of biological tissue or fluid origin, obtained, reached, or collected in vivo or in situ.
  • a biological sample also includes samples from a region of a biological subject containing precancerous or cancer cells or tissues. Such samples can be, but are not limited to, organs, tissues, and cells isolated from a mammal.
  • Exemplary biological samples include but are not limited to cell lysate, a cell culture, a cell line, a tissue, oral tissue, gastrointestinal tissue, an organ, an organelle, a biological fluid, a blood sample, a urine sample, a skin sample, and the like.
  • Preferred biological samples include, but are not limited to, whole blood, partially purified blood, PBMC, tissue biopsies, and the like.
  • the term “analyte” as used herein refers to a known or unknown component of a sample.
  • capture agent refers to an agent that binds an mRNA or protein through an interaction that is sufficient to permit the agent to bind and to concentrate the mRNA or protein from a heterogeneous mixture.
  • pharmaceutically acceptable salt encompasses non-toxic acid and base addition salts of the compound to which the term refers.
  • Acceptable non-toxic acid addition salts include those derived from organic and inorganic acids know in the art, which include, for example, hydrochloric acid, hydrobromic acid, phosphoric acid, sulfuric acid, methanesulphonic acid, acetic acid, tartaric acid, lactic acid, succinic acid, citric acid, malic acid, maleic acid, sorbic acid, aconitic acid, salicylic acid, phthalic acid, embolic acid, enanthic acid, and the like.
  • Compounds that are acidic in nature are capable of forming salts with various pharmaceutically acceptable bases.
  • bases that can be used to prepare pharmaceutically acceptable base addition salts of such acidic compounds are those that form non-toxic base addition salts, i.e., salts containing pharmacologically acceptable cations such as, but not limited to, alkali metal or alkaline earth metal salts (calcium, magnesium, sodium, or potassium salts in particular).
  • Suitable organic bases include, but are not limited to, N,N-dibenzylethylenediamine, chloroprocaine, choline, diethanolamine, ethylenediamine, meglumaine (N-methylglucamine), lysine, and procaine.
  • second active agent refers to any additional treatment that is biologically active.
  • the second active agent can be a hematopoietic growth factor, cytokine, anti-cancer agent, antibiotic, cox-2 inhibitor, immunomodulatory agent, immunosuppressive agent, corticosteroid, therapeutic antibody that specifically binds to a cancer antigen or a pharmacologically active mutant, or derivative thereof.
  • Exemplary second active agents include, but are not limited to, an HDAC inhibitor (e.g., panobinostat, romidepsin, or vorinostat), a BCL2 inhibitor (e.g., venetoclax), a BTK inhibitor (e.g., ibrutinib or acalabrutinib), an mTOR inhibitor (e.g., everolimus), a PI3K inhibitor (e.g., idelalisib), a PKC ⁇ inhibitor (e.g., enzastaurin), a SYK inhibitor (e.g., fostamatinib), a JAK2 inhibitor (e.g., fedratinib, pacritinib, ruxolitinib, baricitinib, gandotinib, lestaurtinib, or momelotinib), an Aurora A kinase inhibitor (e.g., alisertib), an EZH2
  • the term “about” or “approximately” means an acceptable error for a particular value as determined by one of ordinary skill in the art, which depends in part on how the value is measured or determined. In certain embodiments, the term “about” or “approximately” means within 1, 2, 3, or 4 standard deviations. In certain embodiments, the term “about” or “approximately” means within 50%, 20%, 15%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, or 0.05% of a given value or range.
  • the word “a” or “an” when used in conjunction with the term “comprising” in the claims and/or the specification can mean “one,” but it is also consistent with the meaning of “one or more,” “at least one,” and “one or more than one.”
  • pre- treatment refers to prior to administration of a drug.
  • patient and “subject” refer to an animal, such as a mammal. In a specific embodiment, the patient is a human.
  • the patient is a non-human animal, such as a dog, cat, farm animal (e.g., horse, pig, or donkey), chimpanzee, or monkey.
  • the patient is a human with lymphoma (e.g., DLBCL) in need of treatment.
  • lymphoma e.g., DLBCL
  • Methods of Clustering Lymphoma Patients comprising (a) obtain samples from lymphoma patients; (b) measuring the gene expression levels in the samples; and (c) clustering the lymphoma patients into subgroups of patients having lymphoma using gene expression levels in the samples.
  • the lymphoma patients are Diffuse Large B-Cell Lymphoma (DLBCL) patients.
  • the lymphoma is diffuse large B-cell lymphoma (DLBCL).
  • the lymphoma is indolent B cell lymphoma.
  • the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma.
  • the lymphoma is follicular lymphoma. In another embodiment, the lymphoma is nodal marginal zone B-cell lymphoma. In yet another embodiment, the lymphoma is mantle cell lymphoma. In yet another embodiment, the lymphoma is chronic lymphocytic leukemia.
  • the sample is obtained from a tissue of the subject containing DLBCL cells. More detailed description of the sample (or biological sample) is provided in Section 5.7 below. [00120] In some embodiments, the DLBCL patients are newly diagnosed (nd) and relapsed/refractory (r/r) DLBCL patients. In one embodiment, the DLBCL patients are newly diagnosed (nd) DLBCL patients.
  • the DLBCL patients are relapsed/refractory (r/r) DLBCL patients.
  • r/r relapsed/refractory
  • the patients datasets include but not limited to, for example, a set of 267 commercially-sourced newly diagnosed DLBCL (ndDLBCL) patient samples which had molecular profiling but no survival data (“Commercial dataset”), a set of 342 ndDLBCL patients from the Molecular Epidemiology Resource (MER) (“ndMER dataset”; see Cerhan et al., Int. J.
  • the Lenz microarray dataset of 414 ndDLBCL patients was also used as a second replication cohort and had outcome data available (see Lenz et al., N. Engl. J. Med., 2008, 359(22):2313-2323), a set of 86 rrDLBCL patients from the MER cohort (“rrMER dataset”), and a set of 189 rrDLBCL patient samples from two clinical trials (CC-122-ST-001 and CC-122-DLBCL-001) (“CC-122 dataset”; see Clinical Trials Nos. NCT01421524 and NCT02031419).
  • step (a) and step (b) generate a dataset of the lymphoma patients.
  • the lymphoma patients are a newly diagnosed lymphoma patient cohort.
  • the lymphoma patients are a relapsed/refractory lymphoma patient cohort.
  • the clustering step uses a discovery dataset and one or more replication/validation datasets. In one embodiment, the clustering step uses a discovery dataset and a replication/validation dataset. In some embodiments, the discovery dataset is from samples from a discovery cohort. In some embodiments, the replication/validation dataset is from sample from a replication/validation cohort. In some embodiments, the discovery dataset is selected from the group consisting of Commercial dataset, ndMER dataset, Lenz dataset, rrMER dataset, and CC-122 dataset. In some embodiments, the replication/validation dataset is selected from the group consisting of Commercial dataset, ndMER dataset, Lenz dataset, rrMER dataset, and CC-122 dataset. In one embodiment, the discovery dataset is Commercial dataset.
  • the discovery dataset is ndMER dataset. In another embodiment, the discovery dataset is Lenz dataset. In another embodiment, the discovery dataset is rrMER dataset. In another embodiment, the discovery dataset is CC-122 dataset. [00124] In some embodiments, the clustering step comprises (i) normalizing a dataset; (ii) selecting clustering feature(s); and (iii) applying a clustering method using the clustering feature(s). In some embodiments, the clustering step further comprises detecting outliers in the dataset and removing the outliers. In some embodiments, the clustering step further comprises evaluating clustering results of the clustering method. In some embodiments, the clustering is an unsupervised clustering method. In one embodiment, the clustering method is a hierarchical clustering method.
  • the clustering method is a non-hierarchical method.
  • the clustering method is K means clustering method.
  • the clustering method is a partitioning method.
  • the clustering method is a fuzzy clustering method.
  • the clustering method is a density-based clustering.
  • the clustering method is a model-based clustering.
  • the clustering method is Cluster of Cluster Analysis (COCA) clustering method.
  • the clustering method is iClusterPlus clustering method. [00125]
  • a single sample normalization (ssNorm) method is use to normalize the datasets prior to analysis.
  • ssNorm provides a fixed normalization scheme that can apply to individual samples completely independently. There is no need to re-normalize datasets with the addition of new batches of samples, or when combining datasets for larger analyses.
  • the ssNorm method can also implicitly perform batch correction, thereby removing gene-specific experimental effects that appear as systematic, biased differences in the raw expression space.
  • the ssNorm method can properly align diverse datasets to a common space, showing no meaningful separation by dataset/batch when projecting into PCA space.
  • ssNorm allows simple translatability of any classifier or parameterization from one dataset to another – a classifier built on one ssNorm dataset can be directly applied to any other ssNorm dataset, without any need for reweighting model parameters or setting new thresholds.
  • DLBCL-specific housekeeping genes are used for normalization.
  • ISY1, R3HDM1, TRIM56, UBXN4, and WDR55 are used for normalization.
  • Clustering methods are sensitive to the input data, and algorithm performance can be degraded by introducing noisy or irrelevant features. Both feature selection and feature engineering approaches can be utilized to reduce the dimensionality of the dataset while maintaining a representation of relevant biological activity.
  • one, two, three, four, or more features are selected as the clustering features as input into the clustering method.
  • a subset of the gene expression data is selected as the clustering features.
  • the subset of the gene expression data are gene expression data of the top 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 10%, or 5% most expressed genes.
  • the subset of the gene expression data are gene expression data of the top 25% most expressed genes.
  • Gene Set Variation Analysis is a Gene Set Enrichment (GSE) method that estimates variation of pathway activity over a sample population in an unsupervised manner.
  • GSE Gene Set Enrichment
  • GSVA provides increased power to detect subtle pathway activity changes over a sample population in comparison to corresponding methods.
  • GSVA constitutes a starting point to build pathway-centric models of biology and GSVA can contribute to the current need of GSE methods for RNA-seq data. See Hanzelman et al., BMC Bioinformatics, 2013, 14:7.
  • GSVA can be performed over the 50 hallmark pathway gene sets from MSigDB (Hallmark GSVA scores). See, e.g., Liberzon et al., Cell Syst., 2015, 1:417:425.
  • GSVA can be performed over the 299 C1 set of positional cytoband signatures gene sets from MSigDB (C1 Positional GSVA scores). See, e.g., Alhamdoosh et al., F1000Research, 2017, 6:2010. GSVA can also be performed over the DLBCL-specific LM23 matrix (Cell type LM23 GSVA scores), deriving from the DCQ cellular deconvolution method. See Althoum et al., Mol. Syst. Bio., 2014, 10:720. [00130] In some embodiments, the Hallmark GSVA scores are selected as the clustering features. In some embodiments, the C1 Positional GSVA scores are selected as the clustering features.
  • the Cell type LM23 GSVA scores are selected as the clustering features.
  • a subset of the gene expression data, Hallmark GSVA scores, C1 Positional GSVA scores, Cell type LM23 GSVA scores are selected as the matrices of clustering features.
  • the clustering method performs clustering on each feature matrix independently and aggregates the cluster features. In some embodiments, the clustering method performs clustering aggregates the cluster features from different matrices of features and clustering across all the clustering features.
  • the clustering results are evaluated using metrics selected from the group consisting of silhouette statistic, gap statics, and percentage of variance explained by the clustering. In one embodiments, the clustering results are evaluated by the minium cluster size. In order to have sufficient sample size in each cluster to build a downstream classifier model that could robustly identify each cluster or subgroup, the resulting clusters from each tested clustering need to have a minimum cluster size.
  • the method of clustering lymphoma patients further comprises (i) identifying one or more cluster classifier(s) by training a classifier model using the clustering results of a discovery dataset, (ii) applying the one or more cluster classifier(s) to a replication/validation dataset to classify the replication/validation dataset, (iii) clustering the replication/validation dataset using the clustering method and (iv) comparing the classification results of the replication/validation dataset using the cluster classifier(s) with the clustering results of the replication/validation dataset using the clustering method.
  • the model is a grouped multinomial generalized linear model (GLM).
  • the model is GLM using least absolute shrinkage and selection operator (LASSO).
  • the similar the classification results of the replication/validation dataset using the cluster classifier(s) with the clustering results of the replication/validation dataset using the clustering method indicates that the cluster classifier(s) is effective for classifying the replication/validation dataset.
  • the method of clustering lymphoma patients further comprises (i) identifying one or more cluster classifier(s) by training a classifier model using the clustering results of a discovery dataset, and (ii) applying the one or more cluster classifier(s) to a replication/validation dataset to classify the replication/validation dataset.
  • the model is a grouped multinomial generalized linear model (GLM).
  • the model is GLM using least absolute shrinkage and selection operator (LASSO).
  • LASSO least absolute shrinkage and selection operator
  • expression levels of one or more genes in the discovery dataset are used in training the classifier model.
  • one, two, three, four, five or more of the genes identified in Table 1 are used in training the classifier model.
  • all the genes identified in Table 1 are used in training the classifier model.
  • the age of the lymphoma patient is 30 years or older, 35 years or older, 40 years or older, 45 years or older, 50 years or older, or 55 years or old, or 60 years or older, or 65 years or older, or 70 years or older at baseline. In another embodiment, the age of the lymphoma patient is 70 years or older.
  • the age of the lymphoma patient is 60 years or older. In some embodiments, the age of the lymphoma patient is between 30 to 35, 35 to 40, 40 to 45, 45 to 50, 50 to 55, 55 to 60, 60 to 65, or 65 to 70 years old. [00138] In some embodiments, the classifier model uses one, two, three, four, five, or more of the genes identified in Table 1. In some embodiments, the classifier model uses all the genes identified in Table 1. Table 1. List of Genes Utilized in the Resulting grouped multinomial generalized linear model (GLM) Model
  • the method uses all the genes selected from the group consisting of AARS2, ACPP, ACRC, ACTN4, ADH1B, AGER, ALAS1, AMH, ANKRD20A5P, ANKRD22, APOL3, ARHGEF1, ATM, ATP6AP2, ATP8B1, BMF, CBR3, CCDC9, CCL21, CCR1, CCT7, CD83, CDCA8, CDK12, CECR7, CEP72, CHKA, CILP, CLEC4E, CNKSR2, CORO1C, CPT2, CR2, CTSS, CTSZ, DCAF5, DENND2D, DEPDC5, DGKA, DUSP1, E2F4, EHF, ENTPD1, EPAS1, FAM90A1, FBXO6, FCER1G, FCRL1, FLVCR2, FOXO3, FXYD2, FYB, G6PD, GBP4, GDI1, GIMAP6, GINS1, GMIP, GSTM4, GZMM, H
  • methods for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining reference biological samples from each patient in a reference patient group comprising reference patients having a lymphoma; (b) classifying the reference patient group into subgroups of patients using gene expression levels in the reference biological samples; (c) obtaining a biological sample from the lymphoma patient; (d) determining to which subgroup the lymphoma patient belongs using gene expression levels in the biological sample from the lymphoma patient; and (e) predicting the responsiveness of the lymphoma patient to a first cancer treatment.
  • the method further comprises administering to the lymphoma patient a second cancer treatment.
  • the sample is obtained from a tissue of the subject containing DLBCL cells. More detailed description of the sample (or biological sample) is provided in Section 5.7 below.
  • the lymphoma patients are Diffuse Large B-Cell Lymphoma (DLBCL) patients.
  • the DLBCL patients are newly diagnosed (nd) and relapsed/refractory (r/r) DLBCL patients. In one embodiment, the DLBCL patients are newly diagnosed (nd) DLBCL patients.
  • step (b) comprises generating clustering information defining relationships between the expression levels of one or more genes in the reference biological samples; and rearranging heat map representation based on the clustering information.
  • step (b) uses the clustering method described herein in Section 5.2.
  • the clustering step uses a discovery dataset and one or more replication/validation datasets.
  • the clustering step uses a discovery dataset and a replication/validation dataset.
  • the discovery dataset is from samples from a discovery cohort.
  • the replication/validation dataset is from sample from a replication/validation cohort.
  • the discovery dataset is selected from the group consisting of Commercial dataset, ndMER dataset, Lenz dataset, rrMER dataset, and CC-122 dataset.
  • the replication/validation dataset is selected from the group consisting of Commercial dataset, ndMER dataset, Lenz dataset, rrMER dataset, and CC-122 dataset.
  • the discovery dataset is Commercial dataset.
  • the discovery dataset is ndMER dataset.
  • the discovery dataset is Lenz dataset.
  • the discovery dataset is rrMER dataset.
  • the discovery dataset is CC-122 dataset.
  • the clustering step comprises (i) normalizing a dataset; (ii) selecting clustering feature(s); and (iii) applying a clustering method using the clustering feature(s).
  • the clustering step further comprises detecting outliers in the dataset and removing the outliers.
  • the clustering step further comprises evaluating clustering results of the clustering method.
  • the clustering is an unsupervised clustering method.
  • the clustering method is a hierarchical clustering method.
  • the clustering method is a non-hierarchical method.
  • the clustering method is K means clustering method.
  • the clustering method is a partitioning method.
  • the clustering method is a fuzzy clustering method. In another embodiment, the clustering method is a density-based clustering. In another embodiment, the clustering method is a model-based clustering. In one embodiment, the clustering method is Cluster of Cluster Analysis (COCA) clustering method. In one embodiment, the clustering method is iClusterPlus clustering method. In some embodiments, a single sample normalization (ssNorm) method is use to normalize the datasets prior to analysis. [00149] In certain embodiments, DLBCL-specific housekeeping genes are used for normalization. In one embodiment, ISY1, R3HDM1, TRIM56, UBXN4, and WDR55 are used for normalization.
  • one, two, three, four, or more features are selected as the clustering features as input into the clustering method.
  • a subset of the gene expression data is selected as the clustering features.
  • the subset of the gene expression data are gene expression data of the top 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 10%, or 5% most expressed genes.
  • the subset of the gene expression data are gene expression data of the top 25% most expressed genes.
  • the Hallmark GSVA scores are selected as the clustering features.
  • the C1 Positional GSVA scores are selected as the clustering features.
  • the Cell type LM23 GSVA scores are selected as the clustering features.
  • a subset of the gene expression data, Hallmark GSVA scores, C1 Positional GSVA scores, Cell type LM23 GSVA scores are selected as the matrices of clustering features.
  • the clustering method performs clustering on each feature matrix independently and aggregates the cluster features. In some embodiments, the clustering method performs clustering aggregates the cluster features from different matrices of features and clustering across all the clustering features.
  • the clustering results are evaluated using metrics selected from the group consisting of silhouette statistic, gap statics, and percentage of variance explained by the clustering. In one embodiments, the clustering results are evaluated by the minimum cluster size. In order to have sufficient sample size in each cluster to build a downstream classifier model that could robustly identify each cluster or subgroup, the resulting clusters from each tested clustering need to have a minimum cluster size.
  • the method of for predicting the responsiveness of a lymphoma patient to a cancer treatment further comprises (i) identifying one or more cluster classifier(s) by training a classifier model using the clustering results of a discovery dataset, (ii) applying the one or more cluster classifier(s) to a replication/validation dataset to classify the replication/validation dataset, (iii) clustering the replication/validation dataset using the clustering method and (iv) comparing the classification results of the replication/validation dataset using the cluster classifier(s) with the clustering results of the replication/validation dataset using the clustering method.
  • the model is a grouped multinomial generalized linear model (GLM).
  • the model is GLM using least absolute shrinkage and selection operator (LASSO).
  • LASSO least absolute shrinkage and selection operator
  • the similar the classification results of the replication/validation dataset using the cluster classifier(s) with the clustering results of the replication/validation dataset using the clustering method indicates that the cluster classifier(s) is effective for classifying the replication/validation dataset.
  • the method of for predicting the responsiveness of a lymphoma patient to a cancer treatment further comprises (i) identifying one or more cluster classifier(s) by training a classifier model using the clustering results of a discovery dataset, and (ii) applying the one or more cluster classifier(s) to a replication/validation dataset to classify the replication/validation dataset.
  • the model is a grouped multinomial generalized linear model (GLM). In one specific embodiment, the model is GLM using least absolute shrinkage and selection operator (LASSO). [00159] In some embodiments, expression levels of one or more genes in the discovery dataset are used in training the classifier model. In some embodiments, one, two, three, four, five or more of the genes identified in Table 1 are used in training the classifier model. In one embodiment, all the genes identified in Table 1 are used in training the classifier model. [00160] In some embodiments, the classifier model uses one, two, three, four, five, or more of the genes identified in Table 1. In some embodiments, the classifier model uses all the genes identified in Table 1.
  • the determining step (d) applies the clustering method in step (c) to determine to which subgroup the lymphoma patient belongs to using gene expression levels in the biological sample from the lymphoma patient.
  • the predicting step (e) applies the trained classifier model to predict the responsiveness of the lymphoma patient to a first cancer treatment.
  • the predicting step (e) applies the trained GLM model to predict the responsiveness of the lymphoma patient to a first cancer treatment.
  • the lymphoma is diffuse large B-cell lymphoma (DLBCL). In some embodiments, the lymphoma is indolent B cell lymphoma.
  • the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma.
  • the lymphoma is follicular lymphoma.
  • the lymphoma is nodal marginal zone B-cell lymphoma.
  • the lymphoma is mantle cell lymphoma.
  • the lymphoma is chronic lymphocytic leukemia.
  • the first cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP).
  • the second cancer treatment is R-CHOP. In some embodiments, the second cancer treatment is not R-CHOP.
  • the second cancer treatment is a bromodomain and extra- terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor.
  • BET bromodomain and extra- terminal
  • CDK cyclin dependent kinase
  • Bromodomains (BDs) are protein modules of ⁇ 110 amino acids that recognize acetylated lysine in histones and other proteins.
  • BET bromodomain and extra-terminal
  • BET proteins are composed of four proteins, namely bromodomain-containing protein 2 (BRD2), BRD3, BRD4 and bromodomain testis-specific protein (BRDT).
  • BBD2 bromodomain-containing protein 2
  • BRD3 bromodomain-containing protein 3
  • BRD4 bromodomain testis-specific protein
  • BET inhibitors are acetyl lysine mimetics with a heterocyclic core that occupies the BD pocket15.
  • Deregulation of BET proteins, in particular BRD4 has been implicated in the development of diverse diseases, especially cancers. See Duan et al., MedChemComm, 2018, 9:1779-1802.
  • the BET inhibitor is selected from the group consisting of OTX015, MK-8628, CPI-0610, BMS-986158, ZEN003694, GSK2820151, GSK525762, INCB054329, INCB057643, ODM-207, RO6870810, BAY1238097, CC-90010, AZD5153, FT- 1101, ABBV-075, ABBV-744, SF1126, GS-5829, and CPI-0610.
  • the BET inhibitor is a bromodomain-containing protein 4 (BRD4) inhibitor.
  • the BRD4 inhibitor is selected from the group consisting of OTX015, TEN-010, GSK525762, CPI-0610, I-BET151, PLX51107, INCB0543294, ABBV-075, BI 894999, BMS-986158, and AZD5153.
  • CDKs Cyclin-dependent kinases
  • CKIs endogenous inhibitors
  • CDKs Mammalian CDKs, cyclins and CKIs play important roles in other biological processes including, for example, transcriptional regulation, epigenetics, DNA damage response and repair (DDR), stemness, metabolism and angiogenesis among others.
  • DDR DNA damage response and repair
  • the CDK inhibitor is selected from the group consisting of PD- 0332991 (Ibrance® or Palbociclib), LEE011 (Ribociclib), LY2835219 (Verzenio® or Abemaciclib), G1T28 (Trilaciclib), G1T38 (Lerociclib), SHR-6390, Flavopiridol (Alvocidib), PHA848125 (Milciclib), BCD-115, MM-D37K, PF-06873600, TG-02 (SB-1317 or Zoriraciclib), C7001 (ICEC 0942), BEY-1107, XZP-3287 (Birociclib), BPI-16350, FCN-437, CYC-065, R- Roscovitine (CY-202 or Seliciclib), AT-7519, AGM-130 (Inditinib), FN-1501, SY-1365, AZD- 4573, TP-1287,
  • DHITsig Double Hit Signature Classifier
  • the Ennishi method for calculating the DHITsig score is a variable importance- weighted sum of log likelihood ratios.
  • the likelihood function is calculated by assuming a Gaussian mixture model of gene expression that defines two normal distributions of expression for DHITsig+ and DHITsig- samples for each signature gene.
  • the gene list and variable weights from Ennishi et al. were used, along with the mixture model distribution parameters as derived from their DESeq2-processed dataset. 5.3.2.
  • subgroup D1 comprises about 55% to 65% patients having germinal center B-cell-like (GCB) DLBCL, about 20% to 30% patients having activated B-cell like (ABC) lymphoma, and about 20% to 30% patients who are DHITsig+ DLBCL patients
  • subgroup D2 comprises about 45% to 55% patients having GCB DLBCL, about 20% to 45% patients having ABC DLBCL, and about 20% to 25% patients who are DHITsig+ DLBCL patients
  • subgroup D3 comprises about 90% to 95% patients having GCB DLBCL, about 0% to 10% patients having ABC DLBCL, and about 0% to 20% patients who are DHITsig+ DLBCL patients
  • subgroup D4 comprises about 0% to 10% patients having GCB DLBCL, about 90% to 100% patients having ABC
  • the method comprises predicting that the patient is not likely to be responsive to the first cancer treatment. In some embodiments, when the lymphoma patient is determined to belong to any one of subgroups D1-D3 or D5-D7, the method comprises predicting that the patient is likely to be responsive to the first cancer treatment.
  • the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor.
  • the second cancer treatment is a cyclin dependent kinase (CDK) inhibitor; and (ii) when the lymphoma patient is determined to belong to subgroup D8, the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor.
  • the second cancer treatment is a cyclin dependent kinase (CDK) inhibitor.
  • the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor.
  • the second cancer treatment when the lymphoma patient is determined to belong to subgroup D8, the second cancer treatment is a BCL2 inhibitor. In some embodiments, when the lymphoma patient is determined to belong to subgroup D8, the second cancer treatment is an agent that increases FAS expression. In some embodiments, when the lymphoma patient is determined to belong to subgroup D3 or D7, the second cancer treatment is an inhibitor of human leukocyte antigen (HLA) genes. In one embodiment, the second cancer treatment is an inhibitor of HLA-A. In another embodiment, the second cancer treatment is an inhibitor of HLA-B. In yet another embodiment, the second cancer treatment is an inhibitor of HLA-C. In yet another embodiment, the second cancer treatment is an inhibitor of HLA-E.
  • HLA human leukocyte antigen
  • the second cancer treatment is an inhibitor of HLA-F.
  • the second cancer treatment when the lymphoma patient is determined to belong to subgroup D4 or D8, the second cancer treatment is a CD47 treatment.
  • the second cancer treatment when the lymphoma patient is determined to belong to subgroup D7, the second cancer treatment is an IDO inhibitor or an agent that depletes regulatory T cells.
  • the second cancer treatment is an IDO inhibitor.
  • the second cancer treatment is an agent that depletes regulatory T cells.
  • the second cancer treatment is a histone deacetylase (HDAC) inhibitor.
  • HDAC histone deacetylase
  • the second cancer treatment is a galectin-3 (Gal3) inhibitor.
  • the lymphoma patient is determined to be belong to a subgroup based on the mutational data of one or more features listed in Table 4 or Table 5.
  • kits for predicting the responsiveness of lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first patient having a lymphoma; (b) determining the level of expression of one, two, three, four, five or more of the genes identified in Table 1; (c) comparing the level of expression of the one, two, three, four, five or more of the genes identified in Table 1 in the first biological sample with the level of expression of the same genes in a second biological sample(s) from a second patient(s), wherein the second lymphoma patient(s) is responsive to the cancer treatment, and wherein the similar expression of the one, two, three, four, five or more of the genes in the first biological sample relative to the level of expression of the one, two, three, four, five or more of the genes in the second biological sample(s) indicates that the lymphoma in the first patient will be responsive to treatment with the cancer treatment.
  • a method for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first lymphoma patient; (b) determining the expression of the genes or a certain subset of genes set forth in Table 1 or any combination thereof in the first biological sample; and (c) comparing the gene expression profile of the genes or subset of genes in the first biological sample to (i) the gene expression profile of the genes or subset of genes in biological samples from lymphoma patients which are responsive to the drug and (ii) the gene expression of the genes or subset of genes in biological samples from lymphoma patients which are not responsive to the cancer treatment, wherein a gene expression profile for the genes or subset of genes in the first biological sample similar to the gene expression profile for the genes or subset of genes in biological samples from lymphoma patients which are responsive to the cancer treatment indicates that the first lymphoma patient will be responsive to the cancer treatment, and a gene expression profile for the genes or
  • the determining step of the methods described herein comprising determining the expression of all the genes listed in Table 1.
  • the expression levels of all the genes listed in Table 1 are determined in step (b) and compared in step (c).
  • the determining step of the methods described herein comprising determining the expression of all the genes selected from the group consisting of AARS2, ACPP, ACRC, ACTN4, ADH1B, AGER, ALAS1, AMH, ANKRD20A5P, ANKRD22, APOL3, ARHGEF1, ATM, ATP6AP2, ATP8B1, BMF, CBR3, CCDC9, CCL21, CCR1, CCT7, CD83, CDCA8, CDK12, CECR7, CEP72, CHKA, CILP, CLEC4E, CNKSR2, CORO1C, CPT2, CR2, CTSS, CTSZ, DCAF5, DENND2D, DEPDC5, DGKA, DUSP1, E2F4, EHF, ENTPD1, EPAS1, FAM90A1, FBXO6, FCER1G, FCRL1, FLVCR2, FOXO3, FXYD2, FYB, G6PD, GBP4, GDI1, GIMAP6, GINS1, GM
  • the cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP).
  • R-CHOP prednisone
  • methods of treating a lymphoma patient comprising: (i) identifying a lymphoma patient who is likely to be responsive to the cancer treatment as predicted using the predicting methods described herein above; and (ii) administering to the lymphoma patient the cancer treatment.
  • kits for treating a lymphoma patient comprising: (i) identifying a lymphoma patient who is likely to be not responsive to the cancer treatment as predicted using the predicting methods described herein above; and (ii) administering to the lymphoma patient an alternative cancer treatment.
  • the alternative cancer treatment is a bromodomain and extra- terminal (BET) inhibitor, a cyclin dependent kinase (CDK) inhibitor.
  • BET bromodomain and extra- terminal
  • CDK cyclin dependent kinase
  • all of the genes listed in Table 1 can be used as biomarkers to predict the responsiveness of a lymphoma (e.g., DLBCL) patient to a drug.
  • the subgroups provided herein can be characterized and/or identified based on Bcl6 signature scores as shown in the example section below and in FIG.12. Therefore, Bcl6 sigature score can also be used as a way to classify a patient into one of the 8 subgroups for the purpose of determining a patient’s responsiveness to a treatment.
  • Bcl6 sigature score can also be used as a way to classify a patient into one of the 8 subgroups for the purpose of determining a patient’s responsiveness to a treatment.
  • mutational data were collected and interpreted in the context of the identified subgroups.
  • each subgroup (or cluster) or a subset thereof can also be used to identify a subgroup, or to classify a patient into one of the subgroups for the purpose of determining the patient’s responsiveness to a treatment.
  • the subgroups (or clusters) provided herein were also characterized based on proportions of different T cell populations (e.g., CD8+, CD4+, CD163+, or CD20+ cells) as shown in the example section below and in FIGs.13A-13D.
  • proportions of different T cell populations can be used to identify a subgroup or to classify a patient into one of the subgroups for the purpose of determining the patient’s responsiveness to a treatment.
  • a first cancer treatment compound is administered to a lymphoma patient predicted to be responsive to the first cancer treatment.
  • methods of treating patients who have been previously treated for cancer e.g., DLBCL or a subtype thereof
  • are non-responsive to a first cancer treatment e.g., standard therapies
  • the invention also encompasses methods of treating patients regardless of patient’s age, although some diseases or disorders are more common in certain age groups.
  • the invention further encompasses methods of treating patients who have undergone surgery in an attempt to treat the disease or condition at issue, as well as those who have not. Because patients with cancer have heterogeneous clinical manifestations and varying clinical outcomes, the treatment given to a patient may vary, depending on his/her prognosis. The skilled clinician will be able to readily determine without undue experimentation specific secondary agents, types of surgery, and types of non-drug based standard therapy that can be effectively used to treat an individual patient with cancer (e.g., DLBCL or a subtype thereof).
  • a therapeutically or prophylactically effective amount of the cancer treatment is from about 0.005 to about 1,000 mg per day, from about 0.01 to about 500 mg per day, from about 0.01 to about 250 mg per day, from about 0.01 to about 100 mg per day, from about 0.1 to about 100 mg per day, from about 0.5 to about 100 mg per day, from about 1 to about 100 mg per day, from about 0.01 to about 50 mg per day, from about 0.1 to about 50 mg per day, from about 0.5 to about 50 mg per day, from about 1 to about 50 mg per day, from about 0.02 to about 25 mg per day, or from about 0.05 to about 10 mg per day.
  • the therapeutically or prophylactically effective amount is about 0.1, about 0.2, about 0.5, about 1, about 2, about 5, about 10, about 15, about 20, about 25, about 30, about 40, about 45, about 50, about 60, about 70, about 80, about 90, about 100, or about 150 mg per day.
  • the recommended daily dose range of the cancer treatment for the conditions described herein lie within the range of from about 0.1 mg to about 50 mg per day, preferably given as a single once-a-day dose, or in divided doses throughout a day. In some embodiments, the dosage ranges from about 1 mg to about 50 mg per day. In other embodiments, the dosage ranges from about 0.5 mg to about 5 mg per day.
  • Specific doses per day include 0.1, 0.2, 0.5, 1, 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, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 mg per day.
  • the recommended starting dosage may be 0.5, 1, 2, 3, 4, 5, 10, 15, 20, 25, or 50 mg per day.
  • the recommended starting dosage may be 0.5, 1, 2, 3, 4, or 5 mg per day.
  • the dose may be escalated to 10, 15, 20, 25, 30, 35, 40, 45, or 50 mg per day.
  • the therapeutically or prophylactically effective amount is from about 0.001 to about 100 mg/kg/day, from about 0.01 to about 50 mg/kg/day, from about 0.01 to about 25 mg/kg/day, from about 0.01 to about 10 mg/kg/day, from about 0.01 to about 9 mg/kg/day, 0.01 to about 8 mg/kg/day, from about 0.01 to about 7 mg/kg/day, from about 0.01 to about 6 mg/kg/day, from about 0.01 to about 5 mg/kg/day, from about 0.01 to about 4 mg/kg/day, from about 0.01 to about 3 mg/kg/day, from about 0.01 to about 2 mg/kg/day, or from about 0.01 to about 1 mg/kg/day.
  • the administered dose can also be expressed in units other than mg/kg/day.
  • doses for parenteral administration can be expressed as mg/m 2 /day.
  • doses for parenteral administration can be expressed as mg/m 2 /day.
  • One of ordinary skill in the art would readily know how to convert doses from mg/kg/day to mg/m 2 /day to given either the height or weight of a subject or both (see, www.fda.gov/cder/cancer/animalframe.htm).
  • a dose of 1 mg/kg/day for a 65 kg human is approximately equal to 38 mg/m 2 /day.
  • the amount of the cancer treatment administered is sufficient to provide a plasma concentration of the compound at steady state, ranging from about 0.001 to about 500 ⁇ M, about 0.002 to about 200 ⁇ M, about 0.005 to about 100 ⁇ M, about 0.01 to about 50 ⁇ M, from about 1 to about 50 ⁇ M, about 0.02 to about 25 ⁇ M, from about 0.05 to about 20 ⁇ M, from about 0.1 to about 20 ⁇ M, from about 0.5 to about 20 ⁇ M, or from about 1 to about 20 ⁇ M.
  • the amount of the cancer treatment administered is sufficient to provide a plasma concentration of the compound at steady state, ranging from about 5 to about 100 nM, about 5 to about 50 nM, about 10 to about 100 nM, about 10 to about 50 nM, or from about 50 to about 100 nM.
  • plasma concentration at steady state is the concentration reached after a period of administration of a cancer treatment provided herein. Once steady state is reached, there are minor peaks and troughs on the time-dependent curve of the plasma concentration of the cancer treatment.
  • the amount of the cancer treatment administered is sufficient to provide a maximum plasma concentration (peak concentration) of the compound, ranging from about 0.001 to about 500 ⁇ M, about 0.002 to about 200 ⁇ M, about 0.005 to about 100 ⁇ M, about 0.01 to about 50 ⁇ M, from about 1 to about 50 ⁇ M, about 0.02 to about 25 ⁇ M, from about 0.05 to about 20 ⁇ M, from about 0.1 to about 20 ⁇ M, from about 0.5 to about 20 ⁇ M, or from about 1 to about 20 ⁇ M.
  • peak concentration peak concentration
  • the amount of the cancer treatment administered is sufficient to provide a minimum plasma concentration (trough concentration) of the compound, ranging from about 0.001 to about 500 ⁇ M, about 0.002 to about 200 ⁇ M, about 0.005 to about 100 ⁇ M, about 0.01 to about 50 ⁇ M, from about 1 to about 50 ⁇ M, about 0.01 to about 25 ⁇ M, from about 0.01 to about 20 ⁇ M, from about 0.02 to about 20 ⁇ M, from about 0.02 to about 20 ⁇ M, or from about 0.01 to about 20 ⁇ M.
  • a minimum plasma concentration (trough concentration) of the compound ranging from about 0.001 to about 500 ⁇ M, about 0.002 to about 200 ⁇ M, about 0.005 to about 100 ⁇ M, about 0.01 to about 50 ⁇ M, from about 1 to about 50 ⁇ M, about 0.01 to about 25 ⁇ M, from about 0.01 to about 20 ⁇ M, from about 0.02 to about 20 ⁇ M, from about 0.02 to about 20 ⁇ M, or from about 0.
  • the amount of the cancer treatment administered is sufficient to provide an area under the curve (AUC) of the compound, ranging from about 100 to about 100,000 ng*hr/mL, from about 1,000 to about 50,000 ng*hr/mL, from about 5,000 to about 25,000 ng*hr/mL, or from about 5,000 to about 10,000 ng*hr/mL.
  • AUC area under the curve
  • the lymphoma patient to be treated with one of the methods provided herein has not been treated with anticancer therapy prior to the administration of a standard therapy (e.g., R-CHOP).
  • the lymphoma patient to be treated with one of the methods provided herein has been treated with anticancer therapy (standard therapies, e.g., R-CHOP) prior to the administration of the second treatment.
  • anticancer therapy standard therapies, e.g., R-CHOP
  • the lymphoma patient to be treated with one of the methods provided herein has developed drug resistance to the first cancer treatment.
  • the cancer treatment may be administered by parenteral (e.g., intramuscular, intraperitoneal, intravenous, CIV, intracistemal injection or infusion, subcutaneous injection, or implant), inhalation, nasal, vaginal, rectal, sublingual, or topical (e.g., transdermal or local) routes of administration.
  • parenteral e.g., intramuscular, intraperitoneal, intravenous, CIV, intracistemal injection or infusion, subcutaneous injection, or implant
  • inhalation nasal, vaginal, rectal, sublingual, or topical (e.g., transdermal or local) routes of administration.
  • the cancer treatment may be formulated, alone or together, in suitable dosage unit with pharmaceutically acceptable excipients, carriers, adjuvants, and vehicles, appropriate for each route of administration.
  • the cancer treatment is administered parenterally.
  • the cancer treatment is administered intravenously.
  • the trewatment compound may be administered by oral, parenteral (e.g., intramuscular, intraperitoneal, intravenous, CIV, intracistemal injection or infusion, subcutaneous injection, or implant), inhalation, nasal, vaginal, rectal, sublingual, or topical (e.g., transdermal or local) routes of administration.
  • the treatment compound may be formulated, alone or together, in suitable dosage unit with pharmaceutically acceptable excipients, carriers, adjuvants and vehicles, appropriate for each route of administration.
  • the treatment compound is administered orally.
  • the treatment compound is administered parenterally. In yet another embodiment, the treatment compound is administered intravenously.
  • the treatment compound can be delivered as a single dose such as, e.g., a single bolus injection, or oral capsules, tablets or pills; or over time, such as, e.g., continuous infusion over time or divided bolus doses over time.
  • the cancer treatment as described herein can be administered repeatedly if necessary, for example, until the patient experiences stable disease or regression, or until the patient experiences disease progression or unacceptable toxicity.
  • the treatment compound can be administered once daily (QD), or divided into multiple daily doses such as twice daily (BID), three times daily (TID), and four times daily (QID).
  • the administration can be continuous (i.e., daily for consecutive days or every day), intermittent, e.g., in cycles (i.e., including days, weeks, or months of rest without drug).
  • intermittent e.g., in cycles (i.e., including days, weeks, or months of rest without drug).
  • cycles i.e., including days, weeks, or months of rest without drug.
  • intermittent is intended to mean that a therapeutic compound is administered once or more than once each day, for example, for a period of time.
  • continuous is intended to mean that a therapeutic compound is administered daily for an uninterrupted period of at least 7 days to 52 weeks.
  • the term “intermittent” or “intermittently” as used herein is intended to mean stopping and starting at either regular or irregular intervals.
  • intermittent administration is administration for one to six days per week, administration in cycles (e.g., daily administration for two to eight consecutive weeks, then a rest period with no administration for up to one week), or administration on alternate days.
  • cycling as used herein is intended to mean that a therapeutic compound is administered daily or continuously but with a rest period.
  • the frequency of administration is in the range of about a daily dose to about a monthly dose.
  • the cancer treatment can be delivered as a single dose (e.g., a single bolus injection), or over time (e.g., continuous infusion over time or divided bolus doses over time).
  • stable disease for solid cancers generally means that the perpendicular diameter of measurable lesions has not increased by 25% or more from the last measurement. Therasse et al., J. Natl. Cancer Inst., 2000, 92(3):205-216. Stable disease or lack thereof is determined by methods known in the art such as evaluation of patient symptoms, physical examination, and visualization of the tumor that has been imaged using X-ray, CAT, PET, MRI scan, or other commonly accepted evaluation modalities.
  • the cancer treatment can be administered once daily (QD) or divided into multiple daily doses such as twice daily (BID), three times daily (TID), and four times daily (QID).
  • the administration can be continuous (i.e., daily for consecutive days or every day) or intermittent, e.g., in cycles (i.e., including days, weeks, or months of rest without drug).
  • the term “daily” is intended to mean that a cancer treatment is administered once or more than once each day, for example, for a period of time.
  • continuous is intended to mean that the cancer treatment is administered daily for an uninterrupted period of at least 10 days to 52 weeks.
  • intermittent administration of the cancer treatment is administration for one to six days per week, administration in cycles (e.g., daily administration for two to eight consecutive weeks, then a rest period with no administration for up to one week), or administration on alternate days.
  • cycling as used herein is intended to mean that the cancer treatment is administered daily or continuously but with a rest period.
  • the rest period is the same length as the treatment period.
  • the rest period has different length from the treatment period.
  • the length of cycling is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 weeks.
  • the cancer treatment is administered daily for a period of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, or 30 days, followed by a rest period.
  • the cancer treatment is administered daily for a period of 5 days of a 4-week cycle.
  • the cancer treatment is administered daily for a period of 10 days of a 4-week cycle.
  • the frequency of administration is in the range of about a daily dose to about a monthly dose.
  • administration is once a day, twice a day, three times a day, four times a day, once every other day, twice a week, once every week, once every two weeks, once every three weeks, or once every four weeks.
  • the cancer treatment is administered once a day. In another embodiment, the cancer treatment is administered twice a day. In yet another embodiment, the cancer treatment is administered three times a day. In still another embodiment, the cancer treatment is administered four times a day. [00215] In certain embodiments, the cancer treatment is administered once per day from one day to six months, from one week to three months, from one week to four weeks, from one week to three weeks, or from one week to two weeks. In certain embodiments, the cancer treatment is administered once per day for one week, two weeks, three weeks, or four weeks. In one embodiment, the cancer treatment is administered once per day for one week. In another embodiment, the cancer treatment is administered once per day for two weeks. In yet another embodiment, the cancer treatment is administered once per day for three weeks.
  • the cancer treatment is administered once per day for four weeks.
  • One or more additional therapies such as additional active ingredients or agents, that can be used in combination with the administration of a cancer treatment described herein to treat a lymphoma patient (e.g., a patient having DLBCL).
  • the one or more additional therapies can be administered prior to, concurrently with, or subsequent to the administration of the compound described herein.
  • Administration of a cancer treatment described herein and an additional active agent (“second active agents”) to a patient can occur simultaneously or sequentially by the same or different routes of administration.
  • the suitability of a particular route of administration employed for a particular active agent will depend on the active agent itself (e.g., whether it can be administered orally without decomposing prior to entering the blood stream) and the condition of lymphoma (e.g., DLBCL) being treated.
  • Routes of administration for the additional active agents or ingredients are known to those of ordinary skill in the art. See, e.g., Physicians’ Desk Reference.
  • the cancer treatment described herein and an additional active agent are cyclically administered to a patient with lymphoma (e.g., DLBCL). Cycling therapy involves the administration of an active agent for a period of time, followed by a rest for a period of time, and repeating this sequential administration.
  • Second active ingredients or agents can be used in the methods and compositions provided herein.
  • Second active agents can be large molecules (e.g., proteins) or small molecules (e.g., synthetic inorganic, organometallic, or organic molecules).
  • Various agents can be used, such as those described in U.S. Patent Application No.16/390,815 or U.S.
  • Exemplary second active agents include, but are not limited to, an HDAC inhibitor (e.g., panobinostat, romidepsin, or vorinostat), a BCL2 inhibitor (e.g., venetoclax), a BTK inhibitor (e.g., ibrutinib or acalabrutinib), an mTOR inhibitor (e.g., everolimus), a PI3K inhibitor (e.g., idelalisib), a PKC ⁇ inhibitor (e.g., enzastaurin), a SYK inhibitor (e.g., fostamatinib), a JAK2 inhibitor (e.g., fedratinib, pacritinib, ruxolitinib, baricitinib, gandotinib, lestaurtinib, or momelotinib), an Aurora A kinase inhibitor (e.g., alisertib), an EZH2
  • the methods further include administration of one or more of rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone, etoposide, Bendamustine (Treanda), lenalidomide, or gemcitabine.
  • the methods further include administration of one or more of rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone, etoposide, Bendamustine (Treanda), or gemcitabine.
  • the treatment further includes treatment with one or more of R-CHOP (rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone), R EPOCH (etoposide, rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone), stem cell transplant, Bendamustine (Treanda) plus rituximab, rituximab, lenalidomide plus rituximab, or gemcitabine- based combinations.
  • R-CHOP rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone
  • R EPOCH etoposide, rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone
  • stem cell transplant Bendamustine (Treanda) plus
  • the treatment further includes treatment with one or more of R-CHOP (rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone), R EPOCH (etoposide, rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone), stem cell transplant, Bendamustine (Treanda) plus rituximab, rituximab, or gemcitabine-based combinations.
  • the second active agent is rituximab, as provided in U.S. Provisional Application 62/833,432.
  • the second active agent used in the methods provided herein is a histone deacetylase (HDAC) inhibitor.
  • HDAC histone deacetylase
  • the HDAC inhibitor is panobinostat, romidepsin, or vorinostat, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • BCL2 B-cell lymphoma 2
  • the BCL2 inhibitor is venetoclax, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the BCL2 inhibitor is venetoclax.
  • the second active agent used in the methods provided herein is a Bruton’s tyrosine kinase (BTK) inhibitor.
  • the BTK inhibitor is ibrutinib, or acalabrutinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the BTK inhibitor is ibrutinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the BTK inhibitor is ibrutinib.
  • the BTK inhibitor is acalabrutinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the BTK inhibitor is acalabrutinib.
  • the second active agent used in the methods provided herein is a mammalian target of rapamycin (mTOR) inhibitor. In one embodiment, the mTOR inhibitor is rapamycin or an analog thereof (also termed rapalog). In one embodiment, the mTOR inhibitor is everolimus, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the mTOR inhibitor is everolimus.
  • the second active agent used in the methods provided herein is a phosphoinositide 3-kinase (PI3K) inhibitor.
  • the PI3K inhibitor is idelalisib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the PI3K inhibitor is idelalisib.
  • the second active agent used in the methods provided herein is a protein kinase C beta (PKC ⁇ or PKC- ⁇ ) inhibitor.
  • the PKC ⁇ inhibitor is enzastaurin, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the PKC ⁇ inhibitor is enzastaurin.
  • the PKC ⁇ inhibitor is a pharmaceutically acceptable salt of enzastaurin.
  • the PKC ⁇ inhibitor is a hydrochloride salt of enzastaurin.
  • the PKC ⁇ inhibitor is a bis-hydrochloride salt of enzastaurin.
  • the second active agent used in the methods provided herein is a spleen tyrosine kinase (SYK) inhibitor.
  • the SYK inhibitor is fostamatinib, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the SYK inhibitor is fostamatinib. In one embodiment, the SYK inhibitor is a pharmaceutically acceptable salt of fostamatinib. In one embodiment, the SYK inhibitor is fostamatinib disodium hexahydrate. [00227] In one embodiment, the second active agent used in the methods provided herein is a Janus kinase 2 (JAK2) inhibitor.
  • JK2 Janus kinase 2
  • the JAK2 inhibitor is fedratinib, pacritinib, ruxolitinib, baricitinib, gandotinib, lestaurtinib, or momelotinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the JAK2 inhibitor is fedratinib, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the JAK2 inhibitor is fedratinib.
  • the JAK2 inhibitor is pacritinib, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the JAK2 inhibitor is pacritinib. [00230] In one embodiment, the JAK2 inhibitor is ruxolitinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the JAK2 inhibitor is ruxolitinib. In one embodiment, the JAK2 inhibitor is a pharmaceutically acceptable salt of ruxolitinib. In one embodiment, the JAK2 inhibitor is ruxolitinib phosphate.
  • the second active agent used in the methods provided herein is an aurora A kinase inhibitor.
  • the aurora A kinase inhibitor is alisertib, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the aurora A kinase inhibitor is alisertib.
  • the second active agent used in the methods provided herein is an enhancer of zeste homolog 2 (EZH2) inhibitor.
  • the EZH2 inhibitor is tazemetostat, GSK126, CPI-1205, 3-deazaneplanocin A (DZNep), EPZ005687, EI1, UNC1999, or sinefungin, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the EZH2 inhibitor is tazemetostat, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the EZH2 inhibitor is tazemetostat.
  • the EZH2 inhibitor is GSK126, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the EZH2 inhibitor is GSK126 (also known as GSK-2816126).
  • the EZH2 inhibitor is CPI-1205, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the EZH2 inhibitor is CPI-1205.
  • the EZH2 inhibitor is 3-deazaneplanocin A.
  • the EZH2 inhibitor is EPZ005687.
  • the EZH2 inhibitor is EI1. In one embodiment, the EZH2 inhibitor is UNC1999. In one embodiment, the EZH2 inhibitor is sinefungin.
  • the second active agent used in the methods provided herein is a hypomethylating agent. In one embodiment, the hypomethylating agent is 5-azacytidine or decitabine, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. [00238] In one embodiment, the hypomethylating agent is 5-azacytidine, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the hypomethylating agent is 5-azacytidine.
  • the hypomethylating agent is decitabine, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof.
  • the hypomethylating agent is decitabine.
  • the second active agent used in the methods provided herein is a chemotherapy.
  • the chemotherapy is bendamustine, doxorubicin, etoposide, methotrexate, cytarabine, vincristine, ifosfamide, or melphalan, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, prodrug, or pharmaceutically acceptable salt thereof.
  • the second therapeutic agent is administered before, after or simultaneously with a cancer treatment described herein. Administration of a cancer treatment described herein and a second therapeutic agent to a patient can occur simultaneously or sequentially by the same or different routes of administration.
  • a particular route of administration employed for a particular second drug or agent will depend on the second therapeutic agent itself (e.g., whether it can be administered orally or topically without decomposition prior to entering the blood stream) and the subject being treated.
  • Particular routes of administration for the second drug or agents or ingredients are known to those of ordinary skill in the art. See, e.g., The Merck Manual, 448 (17 th ed., 1999).
  • Any combination of the above therapeutic agents, suitable for treatment of the diseases or symptoms thereof, can be administered. Such therapeutic agents can be administered in any combination at the same time or as a separate course of treatment.
  • the term “in combination” does not restrict the order in which therapies (e.g., prophylactic and/or therapeutic agents) are administered to a patient with a disease or disorder.
  • Administration of a second active agent provided herein, to a patient can occur simultaneously or sequentially by the same or different routes of administration.
  • the suitability of a particular route of administration employed for a particular active agent will depend on the active agent itself (e.g., whether it can be administered orally without decomposing prior to entering the blood stream).
  • the cancer treatment provides herein and/or the additional active agent provided herein are formulated in a pharmaceutical composition, and the method provide herein comprises administering a lymphoma (e.g., DLBCL) patient a pharmaceutical composition comprising the cancer treatment.
  • the pharmaceutical compositions provided herein contain therapeutically effective amounts of one or more of the cancer treatment provided herein and a pharmaceutically acceptable carrier, diluents, or excipient.
  • the compounds may be formulated as the sole pharmaceutically active ingredient in the composition or may be combined with other active ingredients.
  • the cancer treatment provided herein can be formulated into suitable pharmaceutical compositions for different routes of administration, such as injection, sublingual and buccal, rectal, vaginal, ocular, otic, nasal, inhalation, nebulization, cutaneous, or transdermal.
  • suitable pharmaceutical compositions for different routes of administration, such as injection, sublingual and buccal, rectal, vaginal, ocular, otic, nasal, inhalation, nebulization, cutaneous, or transdermal.
  • the compounds described above are formulated into pharmaceutical compositions using techniques and procedures well known in the art (see, e.g., Ansel, Introduction to Pharmaceutical Dosage Forms, (7th ed.1999)).
  • effective concentrations of one or more compounds or pharmaceutically acceptable salts are mixed with a suitable pharmaceutical carrier or vehicle.
  • the concentrations of the compounds in the compositions are effective for delivery of an amount, upon administration, that treats, prevents, or ameliorates one or more of the symptoms and/or progression of lymphoma (e.g., DLBCL).
  • the active compound is in an amount sufficient to exert a therapeutically useful effect in the absence of undesirable side effects on the patient treated.
  • the therapeutically effective concentration may be determined empirically by testing the compounds in in vitro and in vivo systems described herein and then extrapolated therefrom for dosages for humans.
  • the concentration of active compound in the pharmaceutical composition will depend on absorption, tissue distribution, inactivation, and excretion rates of the active compound, the physicochemical characteristics of the compound, the dosage schedule, and amount administered as well as other factors known to those of skill in the art.
  • the pharmaceutically therapeutically active compounds and salts thereof are formulated and administered in unit dosage forms or multiple dosage forms.
  • Unit dose forms as used herein refer to physically discrete units suitable for human and animal subjects and packaged individually as is known in the art. Each unit dose contains a predetermined quantity of the therapeutically active compound sufficient to produce the desired therapeutic effect, in association with the required pharmaceutical carriers, vehicles, or diluents. Examples of unit dose forms include ampoules and syringes and individually packaged tablets or capsules.
  • Unit dose forms may be administered in fractions or multiples thereof.
  • a multiple dose form is a plurality of identical unit dosage forms packaged in a single container to be administered in segregated unit dose form. Examples of multiple dose forms include vials, bottles of tablets or capsules, or bottles of pints or gallons. Hence, multiple dose form is a multiple of unit doses which are not segregated in packaging. [00250] It is understood that the precise dosage and duration of treatment is a function of the disease being treated and may be determined empirically using known testing protocols or by extrapolation from in vivo or in vitro test data. It is to be noted that concentrations and dosage values may also vary with the severity of the condition to be alleviated.
  • Solutions or suspensions used for parenteral, intradermal, subcutaneous, or topical application can include any of the following components: a sterile diluents (such as water, saline solution, fixed oil, polyethylene glycol, glycerine, propylene glycol, dimethyl acetamide, or other synthetic solvent), antimicrobial agents (such as benzyl alcohol and methyl parabens), antioxidants (such as ascorbic acid and sodium bisulfate), chelating agents (such as ethylenediaminetetraacetic acid (EDTA)), buffers (such as acetates, citrates, and phosphates), and agents for the adjustment of tonicity (such as sodium chloride or dextrose).
  • a sterile diluents such as water, saline solution, fixed oil, polyethylene glycol, glycerine, propylene glycol, dimethyl acetamide, or other synthetic solvent
  • antimicrobial agents such as benzyl alcohol and methyl para
  • sustained-release preparations can also be prepared. Suitable examples of sustained- release preparations include semipermeable matrices of solid hydrophobic polymers containing the compound provided herein, which matrices are in the form of shaped articles, e.g., films or microcapsule.
  • sustained-release matrices include iontophoresis patches, polyesters, hydrogels (for example, poly(2-hydroxyethyl-methacrylate) or poly(vinylalcohol)), polylactides, copolymers of L-glutamic acid and ethyl-L-glutamate, non-degradable ethylene-vinyl acetate, degradable lactic acid-glycolic acid copolymers such as LUPRON DEPOTTM (injectable microspheres composed of lactic acid-glycolic acid copolymer and leuprolide acetate), and poly-D-(-)-3-hydroxybutyric acid.
  • iontophoresis patches for example, polyesters, hydrogels (for example, poly(2-hydroxyethyl-methacrylate) or poly(vinylalcohol)), polylactides, copolymers of L-glutamic acid and ethyl-L-glutamate, non-degradable ethylene-vinyl
  • anhydrous pharmaceutical compositions and dosage forms containing a compound provided herein can be prepared using anhydrous or low moisture containing ingredients and low moisture or low humidity conditions, as known by those skilled in the art.
  • An anhydrous pharmaceutical composition should be prepared and stored such that its anhydrous nature is maintained.
  • anhydrous compositions are packaged using materials known to prevent exposure to water such that they can be included in suitable formulatory kits.
  • suitable packaging include, but are not limited to, hermetically sealed foils, plastics, unit dose containers (e.g., vials), blister packs, and strip packs.
  • Dosage forms or compositions containing active ingredient in the range of 0.001% to 100% with the balance made up from non-toxic carrier may be prepared.
  • the contemplated compositions contain from about 0.005% to about 95% active ingredient.
  • the contemplated compositions contain from about 0.01% to about 90% active ingredient.
  • the contemplated compositions contain from about 0.1% to about 85% active ingredient.
  • the contemplated compositions contain from about 0.1% to about 75-95% active ingredient.
  • Pharmaceutically acceptable carriers used in parenteral preparations include aqueous vehicles, nonaqueous vehicles, antimicrobial agents, isotonic agents, buffers, antioxidants, local anesthetics, suspending and dispersing agents, emulsifying agents, sequestering or chelating agents, and other pharmaceutically acceptable substances.
  • aqueous vehicles include sodium chloride injection, Ringer’s injection, isotonic dextrose injection, sterile water injection, dextrose and lactated Ringer’s injection.
  • Nonaqueous parenteral vehicles include fixed oils of vegetable origin, such as cottonseed oil, corn oil, sesame oil, and peanut oil.
  • Antimicrobial agents in bacteriostatic or fungistatic concentrations must be added to parenteral preparations packaged in multiple dose containers, which include phenols or cresols, mercurials, benzyl alcohol, chlorobutanol, methyl and propyl-p-hydroxybenzoic acid esters, thimerosal, benzalkonium chloride, and benzethonium chloride.
  • Isotonic agents include sodium chloride and dextrose. Buffers include phosphate and citrate.
  • Antioxidants include sodium bisulfate.
  • Local anesthetics include procaine hydrochloride.
  • Suspending and dispersing agents include sodium carboxymethylcelluose, hydroxypropyl methylcellulose and polyvinylpyrrolidone.
  • Emulsifying agents include Polysorbate 80 (TWEEN® 80).
  • a sequestering or chelating agent of metal ions includes EDTA.
  • Pharmaceutical carriers also include ethyl alcohol, polyethylene glycol and propylene glycol for water miscible vehicles, and sodium hydroxide, hydrochloric acid, citric acid, or lactic acid for pH adjustment. [00257] Injectables are designed for local and systemic administration.
  • a therapeutically effective dosage is formulated to contain a concentration of at least about 0.1% w/w up to about 90% w/w or more, such as more than 1% w/w of the active compound to the treated tissue(s).
  • the active ingredient may be administered at once, or may be divided into a number of smaller doses to be administered at intervals of time. It is understood that the precise dosage and duration of treatment is a function of the tissue being treated and may be determined empirically using known testing protocols or by extrapolation from in vivo or in vitro test data. It is to be noted that concentrations and dosage values may also vary with the age of the individual treated.
  • lyophilized powders which can be reconstituted for administration as solutions, emulsions, and other mixtures. They may also be reconstituted and formulated as solids or gels.
  • the sterile, lyophilized powder is prepared by dissolving a compound provided herein, or a pharmaceutically acceptable salt thereof, in a suitable solvent.
  • the solvent may contain an excipient which improves the stability or other pharmacological component of the powder or reconstituted solution, prepared from the powder.
  • Excipients that may be used include, but are not limited to, dextrose, sorbital, fructose, corn syrup, xylitol, glycerin, glucose, sucrose, or other suitable agent.
  • the solvent may also contain a buffer, such as citrate, phosphate, or other buffers known to those of skill in the art. Subsequent sterile filtration of the solution followed by lyophilization under standard conditions known to those of skill in the art provides the desired formulation. Generally, the resulting solution will be apportioned into vials for lyophilization.
  • the lyophilized powder can be stored under appropriate conditions, such as at about 4 o C to room temperature.
  • the lyophilized formulations are suitable for reconstitution with a suitable diluent to the appropriate concentration prior to administration.
  • the lyophilized formulation is stable at room temperature.
  • the lyophilized formulation is stable at room temperature for up to about 24 months.
  • the lyophilized formulation is stable at room temperature for up to about 24 months, up to about 18 months, up to about 12 months, up to about 6 months, up to about 3 months or up to about 1 month.
  • the lyophilized formulation is stable upon storage under accelerated condition of 40 °C/75% RH for up to about 12 months, up to about 6 months or up to about 3 months.
  • the lyophilized formulation is suitable for reconstitution with an aqueous solution for intravenous administrations.
  • the lyophilized formulation provided herein is suitable for reconstitution with water.
  • the reconstituted aqueous solution is stable at room temperature for up to about 24 hours upon reconsititution.
  • the reconstituted aqueous solution is stable at room temperature from about 1-24, 2-20, 2-15, 2-10 hours upon reconsititution.
  • the reconstituted aqueous solution is stable at room temperature for up to about 20, 15, 12, 10, 8, 6, 4 or 2 hours upon reconsititution.
  • the lyophilized formulations upon reconstitution have a pH of about 4 to 5.
  • Active ingredients provided herein can be administered by controlled release means or by delivery devices that are well known to those of ordinary skill in the art. Examples include, but are not limited to, those described in U.S.
  • Such dosage forms can be used to provide slow or controlled- release of one or more active ingredients using, for example, hydropropylmethyl cellulose, other polymer matrices, gels, permeable membranes, osmotic systems, multilayer coatings, microparticles, liposomes, microspheres, or a combination thereof, to provide the desired release profile in varying proportions.
  • Suitable controlled-release formulations known to those of ordinary skill in the art, including those described herein, can be readily selected for use with the active ingredients provided herein.
  • the various methods provided herein use samples (e.g., biological samples) from lymphoma (e.g., DLBCL) patients.
  • the patient can be male or female, and can be an adult, child or infant.
  • Samples can be analyzed at a time during an active phase of lymphoma (e.g., DLBCL), or when lymphoma (e.g., DLBCL) is inactive.
  • a sample is obtained from a patient prior, concurrently with and/or subsequent to administration of a drug described herein.
  • a sample is obtained from a patient prior to administration of a drug described herein.
  • more than one sample from a patient can be obtained.
  • the sample used in the methods provided herein comprises body fluids from a subject.
  • Non-limiting examples of body fluids include blood (e.g., peripheral whole blood, peripheral blood), blood plasma, amniotic fluid, aqueous humor, bile, cerumen, cowper’s fluid, pre-ejaculatory fluid, chyle, chyme, female ejaculate, interstitial fluid, lymph, menses, breast milk, mucus, pleural fluid, pus, saliva, sebum, semen, serum, sweat, tears, urine, vaginal lubrication, vomit, water, feces, internal body fluids, including cerebrospinal fluid surrounding the brain and the spinal cord, synovial fluid surrounding bone joints, intracellular fluid is the fluid inside cells, and vitreous humour the fluids in the eyeball.
  • blood e.g., peripheral whole blood, peripheral blood
  • blood plasma e.g., amniotic fluid, aqueous humor, bile, cerumen, cowper’s fluid
  • pre-ejaculatory fluid e.g., aque
  • the sample is a blood sample.
  • the blood sample can be obtained using conventional techniques as described in, e.g. Innis et al, editors, PCR Protocols (Academic Press, 1990).
  • White blood cells can be separated from blood samples using convention techniques or commercially available kits, e.g. RosetteSep kit (Stein Cell Technologies, Vancouver, Canada).
  • Sub-populations of white blood cells e.g. mononuclear cells, B cells, T cells, monocytes, granulocytes or lymphocytes, can be further isolated using conventional techniques, e.g.
  • the blood sample is from about 0.1 mL to about 10.0 mL, from about 0.2 mL to about 7 mL, from about 0.3 mL to about 5 mL, from about 0.4 mL to about 3.5 mL, or from about 0.5 mL to about 3 mL.
  • the blood sample is about 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 6.0, 7.0, 8.0, 9.0 or 10.0 mL.
  • the sample used in the present methods comprises a biopsy (e.g., a tumor biopsy).
  • the biopsy can be from any organ or tissue, for example, skin, liver, lung, heart, colon, kidney, bone marrow, teeth, lymph node, hair, spleen, brain, breast, or other organs.
  • the sample used in the methods described herein comprises a tumor biopsy.
  • the sample used in the methods provided herein is obtained from the subject prior to the patient receiving a treatment for lymphoma (e.g., DLBCL).
  • the sample is obtained from the patient during the subject receiving a treatment for the lymphoma (e.g., DLBCL).
  • the sample is obtained from the patient after the patient received a treatment for the lymphoma (e.g., DLBCL).
  • the treatment comprises administering a compound described herein to the subject.
  • the sample used in the methods provided herein comprises a plurality of cells.
  • Such cells can include any type of cells, e.g., stem cells, blood cells (e.g., peripheral blood mononuclear cells), lymphocytes, B cells, T cells, monocytes, granulocytes, immune cells, or tumor or cancer cells.
  • the tumor or cancer cells or a tumor tissue such as a tumor biopsy or a tumor explants.
  • T cells include, for example, helper T cells (effector T cells or Th cells), cytotoxic T cells (CTLs), memory T cells, and regulatory T cells.
  • the cells used in the methods provided herein are CD3 + T cells, e.g., as detected by flow cytometry.
  • the number of T cells used in the methods can range from a single cell to about 10 9 cells.
  • B cells include, for example, plasma B cells, dendritic cells, memory B cells, B1 cells, B2 cells, marginal-zone B cells, and follicular B cells.
  • B cells can express immunoglobulins (antibodies, B cell receptor).
  • Specific cell populations can be obtained using a combination of commercially available antibodies (e.g., Quest Diagnostic (San Juan Capistrano, Calif.); Dako (Denmark)).
  • the sample used in the methods provided herein is from a diseased tissue from a lymphoma (e.g., DLBCL) patient.
  • the number of cells used in the methods provided herein can range from a single cell to about 10 9 cells. In some embodiments, the number of cells used in the methods provided herein is about 1 x 10 4 , 5 x 10 4 , 1 x 10 5 , 5 x 10 5 , 1 x 10 6 , 5 x 10 6 , 1 x 10 7 , 5 x 10 7 , 1 x 10 8 , or 5 x 10 8 .
  • the number and type of cells collected from a subject can be monitored, for example, by measuring changes in morphology and cell surface markers using standard cell detection techniques such as flow cytometry, cell sorting, immunocytochemistry (e.g., staining with tissue specific or cell-marker specific antibodies) fluorescence activated cell sorting (FACS), magnetic activated cell sorting (MACS), by examination of the morphology of cells using light or confocal microscopy, and/or by measuring changes in gene expression using techniques well known in the art, such as PCR and gene expression profiling. These techniques can be used, too, to identify cells that are positive for one or more particular markers.
  • standard cell detection techniques such as flow cytometry, cell sorting, immunocytochemistry (e.g., staining with tissue specific or cell-marker specific antibodies) fluorescence activated cell sorting (FACS), magnetic activated cell sorting (MACS), by examination of the morphology of cells using light or confocal microscopy, and/or by measuring changes in gene expression using techniques well known in
  • Fluorescence activated cell sorting is a well-known method for separating particles, including cells, based on the fluorescent properties of the particles (Kamarch, Methods Enzymol., 1987, 151:150-165). Laser excitation of fluorescent moieties in the individual particles results in a small electrical charge allowing electromagnetic separation of positive and negative particles from a mixture.
  • cell surface marker-specific antibodies or ligands are labeled with distinct fluorescent labels. Cells are processed through the cell sorter, allowing separation of cells based on their ability to bind to the antibodies used. FACS sorted particles may be directly deposited into individual wells of 96-well or 384-well plates to facilitate separation and cloning.
  • subsets of cells are used in the methods provided herein.
  • Methods to sort and isolate specific populations of cells are well-known in the art and can be based on cell size, morphology, or intracellular or extracellular markers.
  • Such methods include, but are not limited to, flow cytometry, flow sorting, FACS, bead based separation such as magnetic cell sorting, size-based separation (e.g., a sieve, an array of obstacles, or a filter), sorting in a microfluidics device, antibody-based separation, sedimentation, affinity adsorption, affinity extraction, density gradient centrifugation, laser capture microdissection, etc.
  • the methods provided here comprise measuring the expression levels of one or more genes identified in Table 1.
  • the expression levels of genes identiefied in Table 1 can be determined by known methods in the art.
  • the expression levels of genes are determined by measuring the mRNA levels of these proteins. Several methods of detecting or quantitating mRNA levels are known in the art. Exemplary methods include but are not limited to northern blots, ribonuclease protection assays, PCR-based methods, and the like.
  • the mRNA sequence can be used to prepare a probe that is at least partially complementary.
  • a nucleic acid assay for testing for immunomodulatory activity in a biological sample can be prepared.
  • An assay typically contains a solid support and at least one nucleic acid contacting the support, where the nucleic acid corresponds to at least a portion of an mRNA.
  • the assay can also have a means for detecting the altered expression of the mRNA in the sample.
  • the assay method can be varied depending on the type of mRNA information desired.
  • Exemplary methods include but are not limited to Northern blots and PCR-based methods (e.g., RT-qPCR). Methods such as RT-qPCR can also accurately quantitate the amount of the mRNA in a sample.
  • Any suitable assay platform can be used to determine the presence of the mRNA in a sample.
  • an assay may be in the form of a dipstick, a membrane, a chip, a disk, a test strip, a filter, a microsphere, a slide, a multiwell plate, or an optical fiber.
  • An assay system may have a solid support on which a nucleic acid corresponding to the mRNA is attached.
  • the solid support may comprise, for example, a plastic, silicon, a metal, a resin, glass, a membrane, a particle, a precipitate, a gel, a polymer, a sheet, a sphere, a polysaccharide, a capillary, a film a plate, or a slide.
  • the assay components can be prepared and packaged together as a kit for detecting an mRNA.
  • the nucleic acid can be labeled, if desired, to make a population of labeled mRNAs.
  • a sample can be labeled using methods that are well known in the art (e.g., using DNA ligase, terminal transferase, or by labeling the RNA backbone, etc.; see, e.g., Ausubel, et al., Short Protocols in Molecular Biology, 3rd ed., Wiley & Sons 1995 and Sambrook et al., Molecular Cloning: A Laboratory Manual, Third Edition, 2001 Cold Spring Harbor, N.Y.).
  • the sample is labeled with fluorescent label.
  • Exemplary fluorescent dyes include but are not limited to xanthene dyes, fluorescein dyes, rhodamine dyes, fluorescein isothiocyanate (FITC), 6 carboxyfluorescein (FAM), 6 carboxy-2',4',7',4,7-hexachlorofluorescein (HEX), 6 carboxy 4', 5' dichloro 2', 7' dimethoxyfluorescein (JOE or J), N,N,N',N' tetramethyl 6 carboxyrhodamine (TAMRA or T), 6 carboxy X rhodamine (ROX or R), 5 carboxyrhodamine 6G (R6G5 or G5), 6 carboxyrhodamine 6G (R6G6 or G6), and rhodamine 110; cyanine dyes, e.g.
  • Cy3, Cy5 and Cy7 dyes include Alexa dyes, e.g. Alexa-fluor-555; coumarin, Diethylaminocoumarin, umbelliferone; benzimide dyes, e.g. Hoechst 33258; phenanthridine dyes, e.g.
  • Texas Red ethidium dyes; acridine dyes; carbazole dyes; phenoxazine dyes; porphyrin dyes; polymethine dyes, BODIPY dyes, quinoline dyes, Pyrene, Fluorescein Chlorotriazinyl, R110, Eosin, JOE, R6G, Tetramethylrhodamine, Lissamine, ROX, Napthofluorescein, and the like.
  • a typical mRNA assay method can contain the steps of 1) obtaining surface-bound subject probes; 2) hybridization of a population of mRNAs to the surface-bound probes under conditions sufficient to provide for specific binding (3) post-hybridization washes to remove nucleic acids not bound in the hybridization; and (4) detection of the hybridized mRNAs.
  • the reagents used in each of these steps and their conditions for use may vary depending on the particular application.
  • Hybridization can be carried out under suitable hybridization conditions, which may vary in stringency as desired. Typical conditions are sufficient to produce probe/target complexes on a solid surface between complementary binding members, i.e., between surface- bound subject probes and complementary mRNAs in a sample.
  • stringent hybridization conditions may be employed.
  • Hybridization is typically performed under stringent hybridization conditions.
  • Standard hybridization techniques e.g. under conditions sufficient to provide for specific binding of target mRNAs in the sample to the probes
  • Several guides to general techniques are available, e.g., Tijssen, Hybridization with Nucleic Acid Probes, Parts I and II (Elsevier, Amsterdam 1993).
  • Tijssen Hybridization with Nucleic Acid Probes, Parts I and II (Elsevier, Amsterdam 1993.
  • the surface bound polynucleotides are typically washed to remove unbound nucleic acids. Washing may be performed using any convenient washing protocol, where the washing conditions are typically stringent, as described above.
  • the hybridization of the target mRNAs to the probes is then detected using standard techniques.
  • Other methods such as PCR-based methods, can also be used to follow the expression of the genes. Examples of PCR methods can be found in the literature. Examples of PCR assays can be found in U.S. Patent No.6,927,024, which is incorporated by reference herein in its entirety. Examples of RT-PCR methods can be found in U.S.
  • RT-qPCR Real-Time Reverse Transcription-PCR
  • RNA targets Bustin, et al., Clin. Sci., 2005, 109:365- 379. Quantitative results obtained by RT-qPCR are generally more informative than qualitative data.
  • RT-qPCR-based assays can be useful to measure mRNA levels during cell-based assays.
  • RT-qPCR is also useful to monitor patient therapy. Examples of RT-qPCR-based methods can be found, for example, in U.S. Patent No.7,101,663, which is incorporated by reference herein in its entirety.
  • real-time PCR gives quantitative results.
  • An additional advantage of real-time PCR is the relative ease and convenience of use. Instruments for real-time PCR, such as the Applied Biosystems 7500, are available commercially, as are the reagents, such as TaqMan Sequence Detection chemistry. For example, TaqMan ® Gene Expression Assays can be used, following the manufacturer’s instructions.
  • kits are pre-formulated gene expression assays for rapid, reliable detection and quantification of human, mouse and rat mRNA transcripts.
  • An exemplary PCR program for example, is 50oC for 2 minutes, 95oC for 10 minutes, 40 cycles of 95oC for 15 seconds, then 60oC for 1 minute.
  • the data can be analyzed, for example, using a 7500 Real-Time PCR System Sequence Detection software v1.3 using the comparative CT relative quantification calculation method. Using this method, the output is expressed as a fold-change of expression levels.
  • the threshold level can be selected to be automatically determined by the software.
  • the threshold level is set to be above the baseline but sufficiently low to be within the exponential growth region of an amplification curve.
  • Techniques known to one skilled in the art may be used to measure the amount of an RNA transcript(s).
  • the amount of one, two, three, four, five, or more RNA transcripts is measured using deep sequencing, such as ILLUMINA® RNASeq, ILLUMINA® next generation sequencing (NGS), ION TORRENT TM RNA next generation sequencing, 454 TM pyrosequencing, or Sequencing by Oligo Ligation Detection (SOLID TM ).
  • the amount of multiple RNA transcripts is measured using a microarray and/or gene chip.
  • the amount of one, two, three, or more RNA transcripts is determined by RT-PCR. In other embodiments, the amount of one, two, three, or more RNA transcripts is measured by RT-qPCR. Techniques for conducting these assays are known to one skilled in the art. In yet other embodiments, NanoString (e.g., nCounter® miRNA Expression Assays provided by NanoString® Technologies) is used for analyzing RNA transcripts. [00289] Several protein detection and quantitation methods can be used to measure the level of proteins. Any suitable protein quantitation method can be used. In some embodiments, antibody-based methods are used.
  • IHC immunohistochemistry
  • the antibodies are usually linked to an enzyme or a fluorescent dye. Typically, when the antibodies bind to the antigen in the tissue sample, the enzyme or dye is activated, and the antigen can then be seen under a microscope.
  • IHC can be used to help diagnose diseases, such as cancer. It may also be used to help tell the difference between different types of cancer.
  • IHC can be used to image discrete components in tissues by using appropriately-labeled antibodies to bind specifically to their target antigens in situ. IHC makes it possible to visualize and document the high-resolution distribution and localization of specific cellular components within cells and within their proper histological context. While there are multiple approaches and permutations in IHC methodology, all of the steps involved can be generally separated into two groups: sample preparation and sample staining.
  • IHC is based on the immunostaining of thin sections of tissues attached to individual glass slides. Multiple small sections can be arranged on a single slide for comparative analysis, a format referred to as a tissue microarray. In other embodiments, IHC is performed by using high-throughput sample preparation and staining. [00291] Samples can be viewed by either light or fluorescence microscopy. In some embodiments, antigen detection in tissue can be performed using an antibody conjugated to an enzyme (horseradish peroxidase) and utilized a colorimetric substrate that could be detected by light microscopy.
  • enzyme horseradish peroxidase
  • the sample e.g., a tissue from the patient
  • the sample has been snap frozen in liquid nitrogen, isopentane or dry ice.
  • the sample e.g., a tissue from the patient
  • FFPE paraffin wax
  • the tissue or sections of the tissue can be mounted on slides prior to staining.
  • the IHC-free-floating technique may be used, where the entire IHC procedure is performed in liquid to increase antibody binding and penetration and slide mounting only takes place upon experimental completion. IHC-free-floating appears to be most popular in neuroscience research.
  • kits predicting the responsiveness of a lymphoma patient to a cancer treatment, comprising agents for measuring the gene expression levels of a biological sample from the lymphoma patient.
  • the kit further comprises an agent (or tool) for taking a sample from a subject.
  • kits further comprises an instruction on how to interpret or use the expression levels determined to predict if a patient has a particular subtype of lymphoma (e.g., DLBCL).
  • a kit comprises a reagent or reagents necessary for carrying out an assay(s) described herein, in one or more other containers.
  • the kit comprises a solid support, and a means for detecting the RNA or protein expression of at least one biomarker in a biological sample.
  • a kit may employ, for example, a dipstick, a membrane, a chip, a disk, a test strip, a filter, a microsphere, a slide, a multiwell plate, or an optical fiber.
  • the solid support of the kit can be, for example, a plastic, silicon, a metal, a resin, glass, a membrane, a particle, a precipitate, a gel, a polymer, a sheet, a sphere, a polysaccharide, a capillary, a film, a plate, or a slide.
  • the kit comprises, in one or more containers, components for conducting RT-PCR, RT-qPCR, deep sequencing, or a microarray such as NanoString assay.
  • the kit comprises a solid support, nucleic acids contacting the support, where the nucleic acids are complementary to at least 10, 20, 50, 100, 200, 350, or more bases of mRNA, and a means for detecting the expression of the mRNA in a biological sample.
  • the kit comprises, in one or more containers, components for conducting assays that can determine one or more protein levels, such flow cytometry, ELISA, or HIC.
  • Such kits may comprise materials and reagents required for measuring RNA or protein.
  • kits include microarrays, wherein the microarray is comprised of oligonucleotides and/or DNA and/or RNA fragments which hybridize to one or more of the genes identified in Table 1.
  • such kits may include primers for PCR of either the RNA product or the cDNA copy of the RNA product of the genes or subset of genes, or both.
  • such kits may include primers for PCR as well as probes for Quantitative PCR.
  • such kits may include multiple primers and multiple probes wherein some of said probes have different fluorophores so as to permit multiplexing of multiple products of a gene product or multiple gene products.
  • kits may further include materials and reagents for creating cDNA from RNA.
  • such kits may include antibodies specific for one or more of the genes identified in Table 1.
  • Such kits may additionally comprise materials and reagents for isolating RNA and/or proteins from a biological sample.
  • such kits may include materials and reagents for synthesizing cDNA from RNA isolated from a biological sample.
  • such kits may include, a computer program product embedded on computer readable media for predicting whether a patient is responsive to a compound as described herein.
  • the kits may include a computer program product embedded on a computer readable media along with instructions.
  • the kit can comprise, for example: (1) a first antibody (which may or may not be attached to a solid support) which binds to a peptide, polypeptide or protein of interest; and, optionally, (2) a second, different antibody which binds to either the peptide, polypeptide or protein, or the first antibody and is conjugated to a detectable label (e.g., a fluorescent label, radioactive isotope or enzyme).
  • a detectable label e.g., a fluorescent label, radioactive isotope or enzyme.
  • the antibody-based kits may also comprise beads for conducting an immunoprecipitation. Each component of the antibody-based kits is generally in its own suitable container. Thus, these kits generally comprise distinct containers suitable for each antibody.
  • kits may comprise instructions for performing the assay and methods for interpreting and analyzing the data resulting from the performance of the assay.
  • the kits contain instructions for predicting whether a lymphoma (e.g., DLBCL) patient belongs to a specific subgroup of DLBCL (e.g., a high-risk subgroup of DLBCL).
  • solid phase supports are used for purifying proteins, labeling samples, or carrying out the solid phase assays. Examples of solid phases suitable for carrying out the methods disclosed herein include beads, particles, colloids, single surfaces, tubes, multi-well plates, microtiter plates, slides, membranes, gels, and electrodes.
  • the solid phase is a particulate material (e.g., a bead)
  • it is, in one embodiment, distributed in the wells of multi-well plates to allow for parallel processing of the solid phase supports.
  • the Lenz microarray dataset of 414 ndDLBCL patients was also used as a second replication cohort and had outcome data available (see Lenz et al., N. Engl. J. Med., 2008, 359(22):2313-2323).
  • Two other datasets were used for replication in the relapsed and refractory DLBCL (rrDLBCL) setting, each of which had clinical outcome information.
  • the first was a set of 86 rrDLBCL patients from the MER cohort, and the second was a set of 189 rrDLBCL patient samples from two clinical trials (CC-122-ST-001 and CC-122- DLBCL-001) (see Clinical Trials NCT01421524 and NCT02031419).
  • the ssNorm method properly aligns diverse datasets to a common space, showing no meaningful separation by dataset/batch when projecting into PCA space. From a strategic perspective, ssNorm allows simple translatability of any classifier or parameterization from one dataset to another – a classifier built on one ssNorm dataset can be directly applied to any other ssNorm dataset, without any need for reweighting model parameters or setting new thresholds.
  • the ssNorm method used the Commercial cohort (without the IHC samples) as a reference dataset against which all other samples are normalized. The normalization began with isoform-level TPM data. Gene-level data were derived by summing the expression of all isoforms annotated to a particular gene.
  • DLBCL-specific housekeeping genes (ISY1, R3HDM1, TRIM56, UBXN4, and WDR55) were used for normalization. These five genes were the same as the housekeeping genes from the Nanostring COO assay. For each sample, the geometric mean of the housekeeping genes was computed, and used as a global normalization factor by which all genes’ expressions were divided. For the reference dataset, each gene’s housekeeping-normalized mean and standard deviation were calculated. For each dataset, the housekeeping-normalized gene-level data was scaled and shifted to match the reference distribution by subtracting the reference mean and dividing by the reference standard deviation on a gene-wise basis.
  • the Reddy COO calls derived from ssNorm gene expression data were 80% concordant with the Hans IHC method (when excluding unclassifiable patients), and 91% concordant with the Nanostring method which utilized non-ssNorm gene expression.
  • the DHIT gene expression score from ssNorm data associated strongly with FISH DHITsig positivity, having a classification AUC of 0.82.
  • Cellular deconvolution estimates from ssNorm data were also well correlated with cell marker densities derived from IHC, particularly for abundant markers. Both CD20 and CD3 marker densities showed a Spearman correlation above 0.8 with the deconvolution abundance estimates of their respective cell types (B and T cells).
  • FIG.1 illustrates that PCA projection of ssNorm datasets shows a high degree of overlap and no systematic stratification by dataset.
  • the samples represent a subset of the Commercial cohort that was selected for imaging. 6.1.3 Outlier Detection [00309] Outlier detection was applied to the combined cohort of the ssNorm Commercial+IHC, Lenz, and MER datasets to identify samples that showed unexpected deviation from the rest of the population.
  • the ArrayQualityMetrics R package was applied, using the Kolmogorov-Smirnov, sum and upper quartile methods, each of which detected if the distribution of gene expression on a per-sample basis is significant different from all other samples. Samples which failed 2 of the 3 tests were called as outliers and removed for further analysis. In total, 13 samples were removed from the commercial training data, and 9 more were removed from the MER dataset. 6.1.4 Dimensionality Reduction [00310] Any clustering methodology is sensitive to the input data, and algorithm performance can be degraded by introducing noisy or irrelevant features. Both feature selection and feature engineering approaches were utilized to reduce the dimensionality of the dataset while maintaining a representation of relevant biological activity.
  • the ssNorm expression data was filtered to the top most expressed/most variant genes. This was done by selecting the intersection of the top 25% most expressed and top 25% most variant genes in the ssNorm training data. The selected set of 2479 genes was highly expressed/highly variant in raw TPM space, indicating that the ssNorm method had little effect on the overall gene ranking.
  • the ssNorm filtered gene expression data we calculated enrichment scores over several informative gene sets.
  • the single-sample GSVA score over the set of Hallmark pathways from MSigDB was calculated, as well as for the C1 set of positional cytoband signatures which can be viewed as a proxy for the copy number data (which was unavailable at the time of analysis).
  • the final set of feature scores were derived from the DCQ cellular deconvolution method (Altboum et al., Mol. Syst. Biol., 2014, 10:720-733), using the DLBCL-specific LM23 matrix. 6.1.5 Unsupervised Clustering [00313] The commercial set of samples was used as the discovery cohort for the initial clustering.
  • FIGS 3A-3D show heatmaps of the clustering results by each features matrix.
  • FIGS.3B and 3D the hallmark pathway features and LM23 cell type features are not as specific to a single cluster; and as shown in FIG.3C, the C1 set of positional cytoband features do not stratify by cluster.
  • the consensus matrix is shown in FIG. 3E.
  • Both methods were tested for between 2 and 15 clusters, with each method run 100 times using 20% random feature and sample dropout. The final clusters for each method for each choice of number of clusters (K) was determined as the consensus clustering over the resampling iterations.
  • IPI International Prognostic Index
  • FIG.5A A heatmap representation of the discovery cohort, stratified by the top cluster-differential genes, is shown in FIG.5A.
  • Heatmap representations of the ndMER dataset and the Lez dataset are shown in FIGS.5B and 5C.
  • Heapmap representations of the rr MER dataset and the CC-122 dataset are shown in FIGS.5D and 5E.
  • This clustering method allows for the transcriptomic identification of high-risk patients subgroups which are underserved by the standard R-CHOP therapy.
  • 6.2 Example 2 Cluster Classifier [00319] Having identified 8 novel patient subgroups in the discovery cohort, the next step was to identify those subgroups in another dataset.
  • the cross-validated GLM fitting function was applied and the most parsimonious parameterization (lambda) that had error within one standard error of the minimum was selected, as recommended in Friedman et al., J. Stat. Softw., 2010, 33: 1-22.
  • This practice selects a near- optimal model that minimizes the number of features used in classification while avoiding overfitting (Krstajic et al., J. Cheminformatics, 2014, 6:10.
  • the resulting model which utilized 172 genes, was able to reproduce the cluster labels with zero misclassification, as would be expected when classifying groups derived from an unsupervised clustering.
  • the 172 genes utilized in the results model are listed in Table 1.
  • the GLM cluster classifier was applied to independent ndDLBCL datasets, the ndMER and Lenz cohorts. Because the GLM cluster classifier was trained in the space of ssNorm gene expression, its application to other ssNorm datasets was straightforward, requiring no reweighting of parameters or other translation of the model.
  • DHITsig Double Hit Signature Classifier
  • the Ennishi method for calculating the DHITsig score is a variable importance- weighted sum of log likelihood ratios.
  • the likelihood function is calculated by assuming a Gaussian mixture model of gene expression that defines two normal distributions of expression for DHITsig+ and DHITsig- samples for each signature gene. The gene list and variable weights from Ennishi et al.
  • the Gausssian mixture model parameters in ssNorm space were re-derived by directly observing the mean and standard deviation of the signature genes among the called DHITsig+/- groups in ssNorm space.
  • the classifier threshold was set to 0 in this space, which is the natural cutoff point of a log-likelihood measure such as this one. This new set of parameters and threshold, combined with the original variable importance weights, provided a complete description of the DHITsig classifier that can be directly applied to any RNAseq data in ssNorm space.
  • Integrative clustering identified eight subgroups of ndDLBCL patients (named D1- D8).
  • the resulting clusters are analyzed in the lens of different biological features including gene signatures such as Double HIT gene signature and TMD gen signature.
  • gene signatures such as Double HIT gene signature and TMD gen signature.
  • the prevalence and biological features (such as COO type, EFS24 failure rate, DHITsig positivity, TME gene signature positivity, BCL2/BCL6 translocation, and MYC translocation) of the replication dataset (MER dataset) are summarized in Table 3.
  • the resulting cluster identification are predictive of the likelihood of response to standard treatment (e.g., R-CHOP combination treatment) and can suggest rational targeted therapies based on cluster-specific biological features. Table 3.
  • D4 comprised 21% of the replication cohort (ndMER cohort) with a median event-free survival (mEFS) of 38.2 months and a median overall survival (mOS) of 80.3 months.
  • D8 comprised 5% of the cohort with a mEFS of 7.5 months and a mOS of 12.1 months. The remaining 6 subgroups were standard risk, with mEFS ranging from 82.1 months to not reached, and none reaching mOS.
  • subgroups were not uniquely defined by previously known molecular classification methods such as COO or Double Hit Signature (DHITsig) (Ennishi et al., J. Clin. Oncol., 2018, 37:190-201), nor by clinical risk factors such as age or international prognostic index (IPI).
  • COO Double Hit Signature
  • DHITsig Double Hit Signature
  • IPI international prognostic index
  • D4 was associated with high IPI, with 49% of D4 having IPI>2, compared to 33% of non-D4 with IPI>2 (p ⁇ 0.05).
  • D8 represented a high-risk subset, which was 73% GCB. The mEFS of D8 GCBs was 5.4 months, while mEFS of non-D8 GCBs was not reached (p ⁇ 0.0001). Transcriptomic analysis revealed low expression of immune response and cytokine signaling pathways, consistent with the low abundance of immune cells in D8.
  • Affymetrix HG-U133 Plus 2.0 GeneChipTM microarrays (www.affymetrix.com/) at the Molecular Characterization & Clinical Assay Development Laboratory, SAIC Frederick National Laboratory for Cancer Research, SAIC-Frederick, Frederick, MD. Biopsy samples are flash-frozen at screen (FF), archived having been formalin-fixed and paraffin embedded (FFPE archive), or FFPE treated at screen.
  • Raw Affymetrix image (.cel) files of the patient are imported into the R statistical programming environment v3.0.0 (r-project.org) using functionality of the Affy package of the related Bioconductor suite of open-source bioinformatics software (bioconductor.org).
  • Transcriptional profile QC is performed using the NUSE algorithm, implemented in the Bioconductor package arrayQualityMetrics (Kauffmann et al., 2009), applied to a log2 transformation of raw signal.
  • the RMA (Irizarry et al., 2003) algorithm is applied to background-correct, quantile normalize and summarize profiles that passed QC.
  • Annotation of probe-sets to genes is performed using the R packages annotate (Gentleman, 2013) and genefilter (Gentleman et al., 2013) selecting only one probeset per gene (Entrez Gene ID) and choosing the most variable across profiles according to inter-quartile range in cases wherein multiple probe-sets map to a single gene.
  • the Hallmark GSVA scores, C1 Positional GSVA scores, Cell type LM23 GSVA scores of the patient are calculated.
  • the iClusterPlus method used in Example 1 is applied to the data from the patient using the same subset of gene expression data as well as the calculated Hallmark GSVA scores, C1 Positional GSVA scores, Cell type LM23 GSVA scores as input for clustering features.
  • the patient is determined to belong to, for example, subgroup D4, based on the clustering results from iClusterPlus clustering method.
  • the GLM classifier model trained in Example 1 in connection with subgroup D4 is applied to the patient classify and predict whether the patient will respond to the standard therapy (e.g., R-CHOP). Applying the trained GLM classifier model to patient’s expression level of the 172 genes identified in Table 1 classifies and predicts the patient to be not responsive to R-CHOP treatment and to be a high-risk DLBCL patient.
  • the standard therapy e.g., R-CHOP
  • Applying the trained GLM classifier model to patient’s expression level of the 172 genes identified in Table 1 classifies and predicts the patient to be not responsive to R-CHOP treatment and to be a high-risk DLBCL patient.
  • Example 7 Cell Proportions in Subgroups D1-D8 [00345] The proportions of T-cells were measured in the DLBCL patients from the identified eight subgroups D1-D8.
  • FIGS.13A-13D display the percentage of nucleated cells of CD8, CD4, CD163, and CD20 in the DLBCL patients across the identified eight subgroups D1 through D8.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Organic Chemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Proteomics, Peptides & Aminoacids (AREA)
  • Physics & Mathematics (AREA)
  • Immunology (AREA)
  • Public Health (AREA)
  • Oncology (AREA)
  • Analytical Chemistry (AREA)
  • Wood Science & Technology (AREA)
  • Pathology (AREA)
  • Biotechnology (AREA)
  • Zoology (AREA)
  • Data Mining & Analysis (AREA)
  • Genetics & Genomics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Medical Informatics (AREA)
  • Biophysics (AREA)
  • Databases & Information Systems (AREA)
  • Animal Behavior & Ethology (AREA)
  • Epidemiology (AREA)
  • Evolutionary Computation (AREA)
  • Bioethics (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Veterinary Medicine (AREA)
  • Evolutionary Biology (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Theoretical Computer Science (AREA)
  • Hospice & Palliative Care (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Pharmacology & Pharmacy (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Medicinal Chemistry (AREA)
  • General Chemical & Material Sciences (AREA)
  • Microbiology (AREA)

Abstract

Provided herein are methods of predicting the responsiveness of a lymphoma patient to a cancer treatment comprising clustering patients into subgroups of patients using gene expression levels. Also provided herein are methods of treating a lymphoma patient based on predicting the responsiveness of the lymphoma patient to a cancer treatment.

Description

METHODS FOR PREDICTING RESPONSIVENESS OF LYMPHOMA TO DRUG AND METHODS FOR TREATING LYMPHOMA CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application claims the benefit of U.S. Provisional Patent Application No.62/928,305, filed October 30, 2019, the disclosure of which is incorporated by reference herein in its entirety. 1. FIELD [0002] Provided herein are methods of predicting the responsiveness of a lymphoma patient to a cancer treatment. Also provided herein are methods of treating a lymphoma patient based on predicting the responsiveness of the lymphoma patient to a cancer treatment. 2. BACKGROUND [0003] The non-Hodgkin lymphomas (NHLs) are a diverse group of blood cancers that include any kind of lymphoma except Hodgkin's lymphomas. Types of NHL vary significantly in their severity, from indolent to very aggressive. Less aggressive non-Hodgkin lymphomas are compatible with a long survival while more aggressive non-Hodgkin lymphomas can be rapidly fatal without treatment. They can be formed from either B-cells or T-cells. B-cell non-Hodgkin lymphomas include Burkitt lymphoma, chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL), diffuse large B-cell lymphoma, follicular lymphoma, immunoblastic large cell lymphoma, precursor B-lymphoblastic lymphoma, and mantle cell lymphoma. T-cell non-Hodgkin lymphomas include mycosis fungoides, anaplastic large cell lymphoma, and precursor T-lymphoblastic lymphoma. Prognosis and treatment depend on the stage and type of disease. [0004] Diffuse large B-cell lymphoma (DLBCL) accounts for approximately one-third of non-Hodgkin’s lymphomas. While some DLBCL patients are cured with traditional chemotherapy, the rest die from the disease. Anticancer drugs cause rapid and persistent depletion of lymphocytes, possibly by direct apoptosis induction in mature T and B cells. See Stahnke et al., Blood, 2001, 98:3066-3073. [0005] The diffuse large B-cell lymphomas (DLBCL) can be divided into distinct molecular subtypes according to their gene profiling patterns: germinal-center B-cell–like DLBCL (GCB- DLBCL), activated B-cell–like DLBCL (ABC-DLBCL), and primary mediastinal B-cell lymphoma (PMBL) or unclassified type. These subtypes are characterized by distinct differences in survival, chemo-responsiveness, and signaling pathway dependence, particularly the NF- ^B pathway. See Kim et al., Journal of Clinical Oncology, 2007 ASCO Annual Meeting Proceedings Part I. vol 25, No.18S (June 20 Supplement), 2007: 8082. See Bea et al., Blood, 2005; 106: 3183-3190; Ngo et al., Nature, 2011; 470:115-119. Such differences have prompted the search for more effective and subtype-specific treatment strategies in DLBCL. [0006] Current cancer therapy, in general, can involve surgery, chemotherapy, hormonal therapy and/or radiation treatment to eradicate neoplastic cells in a patient (see, for example, Stockdale, 1998, Medicine, vol.3, Rubenstein and Federman, eds., Chapter 12, Section IV). Recently, cancer therapy could also involve biological therapy or immunotherapy. All of these approaches pose significant drawbacks for the patient. Surgery, for example, may be contraindicated due to the health of a patient or may be unacceptable to the patient. Additionally, surgery may not completely remove neoplastic tissue. Radiation therapy is only effective when the neoplastic tissue exhibits a higher sensitivity to radiation than normal tissue. Radiation therapy can also often elicit serious side effects. Hormonal therapy is rarely given as a single agent. Although hormonal therapy can be effective, it is often used to prevent or delay recurrence of cancer after other treatments have removed the majority of cancer cells. Biological therapies and immunotherapies are limited in number and may produce side effects such as rashes or swellings, flu-like symptoms, including fever, chills and fatigue, digestive tract problems or allergic reactions. [0007] In the context of DLBCL, treatment usually includes administration of a combination of chemotherapy and antibody therapy. The most widely used treatment of DLBCL is a combination of antibody rituximab (Rituxan) and chemotherapy drugs (cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP), and in some cases etoposide is added (R- EPOCH)). DLBCL also typically requires immediate treatment upon diagnosis due to how quickly the disease can advance. For some patients, DLBCL returns or becomes refractory following treatment. Several alternative treatments, some of which can include use of lenalidomide, are currently being tested in clinical trials for patients with newly diagnosed, relapsed or refractory DLBCL. See Czuczman et al., Clin Cancer Res., 2017, 23:4127-4137. [0008] Among non-Hodgkin’s lymphomas are also types of indolent lymphomas. Indolent B- cell lymphomas include, for example, follicular lymphoma, small lymphocytic lymphoma; nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type, and mycosis fungoides. The clinical course of B-cell indolent lymphoma patients is characterized, among other features, by the risk of histologic transformation (HT) to an aggressive lymphoma, mostly DLBCL, and, less commonly, to Burkitt lymphoma (BL) or other types of aggressive lymphomas). See Montoto et al., Journal of Clinical Oncology, 2011, 29:1827-1834. [0009] Due to the clinical and biological heterogeneity of lymphomas (particularly DLBCL), there is a significant need for effective methods of classifying subtypes of DLBCL for administering specific treatments. DLBCL has traditionally been classified by cell of origin (COO) subcategories based on tumor gene expression profiles and includes Activated B-Cell (ABC) and Germinal Center B-Cell (GCB) subtypes. See Alizadeh et al., Nature, 2000, 403:503-511; Wright et al., Proc. Natl. Acad. Sci. U. S. A., 2003, 100:9991-9996; Scott et al., Blood, 2014, 123:1214-1217. The GCB and ABC subtypes have different pathogenic mechanisms that may impact the outcomes of DLBCL patients on targeted therapies. See Nyman et al., Mod. Pathol., 2009, 22:1094-1101; Hans et al., Blood, 2004, 103:275-282; Choi et al., Clin. Cancer Res., 2009, 15:5494-5502; Meyer et al., J. Clin. Oncol., 2011, 29:200-207; Natkunam et al., J. Clin. Oncol., 2008, 26:447-454. [0010] Recently, new classification models have focused on DNA alterations using tumor samples from patients treated with R-CHOP. See, e.g., Schmitz et al., N. Engl. J. Med., 2018, 378(15):1396-1407. However, a comprehensive integrative approach using transcriptomic data across both newly diagnosed (nd) and relapsed/refractory (r/r) DLBCL is yet to be accomplished. A robust clustering will allow for identification of biologically driven DLBCL patient subgroups and may predict patient outcome and inform treatment approaches. By defining the landscape of DLBCL subtypes in a robust and meaningful way, there can be future advances in treatment tailored to those classified subtypes. The identification of patients belonging to particular subtypes, in turn, can lead to the identification of high-risk patient groups who are underserved by current therapies (e.g., R-CHOP). Examination of the biological underpinnings of those groups can also help elucidate the mechanisms underlying the high-risk subtypes. The present invention satisfies this and other needs. 3. SUMMARY OF THE INVENTION [0011] In one aspect, provided herein is a method for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining reference biological samples from each patient in a reference patient group comprising reference patients having a lymphoma; (b) clustering the reference patient group into subgroups of patients using gene expression levels in the reference biological samples; (c) obtaining a biological sample from the lymphoma patient; (d) determining to which subgroup the lymphoma patient belongs using gene expression levels in the biological sample from the lymphoma patient; and (e) predicting the responsiveness of the lymphoma patient to a first cancer treatment. [0012] In some embodiments, the method provided herein further comprises administering to the lymphoma patient a second cancer treatment. [0013] In some embodiments, step (b) comprises generating clustering information defining relationships between the expression levels of one or more genes in the reference biological samples; and rearranging heat map representation based on the clustering information. [0014] In some embodiments, step (b) uses a hierarchical method or a non-hierarchical method. [0015] In some embodiments, step (b) uses Cluster of Cluster Analysis (COCA) method or iClusterPlus method. [0016] In some embodiments, step (b) uses COCA method. In some embodiments, step (b) uses iClusterPlus method. [0017] In some embodiments, the reference patients in the reference patient group are clustered into 2-15 subgroups. [0018] In some embodiments, the reference patients in the reference patient group are clustered into 8 subgroups. [0019] In some embodiments, the method provided herein further comprises training a classifier model using expression levels of one or more genes in the reference biological samples. [0020] In some embodiments, expression levels of one, two, three, four, five, or more of the genes identified in Table 1 are used in training the classifier model. [0021] In some embodiments, expression levels of all the genes identified in Table 1 are used in training the classifier model. [0022] In some embodiments, the classifier model is a grouped multinomial generalized linear model (GLM). [0023] In some embodiments, the lymphoma is diffuse large B-cell lymphoma (DLBCL). [0024] In some embodiments, the lymphoma is indolent B cell lymphoma. [0025] In some embodiments, the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma. [0026] In some embodiments, the lymphoma is follicular lymphoma. [0027] In some embodiments, the lymphoma is nodal marginal zone B-cell lymphoma. [0028] In some embodiments, the lymphoma is mantle cell lymphoma. [0029] In some embodiments, the lymphoma is chronic lymphocytic leukemia. [0030] In some embodiments, the reference patients in the reference patient group are clustered into 8 subgroups; and wherein: (i) subgroup D1 comprises about 55% to 65% patients having germinal center B-cell-like (GCB) DLBCL, about 20% to 30% patients having activated B-cell like (ABC) lymphoma, and about 20% to 30% patients who are DHITsig+ DLBCL patients; (ii) subgroup D2 comprises about 45% to 55% patients having GCB DLBCL, about 20% to 45% patients having ABC DLBCL, and about 20% to 25% patients who are DHITsig+ DLBCL patients; (iii) subgroup D3 comprises about 90% to 95% patients having GCB DLBCL, about 0% to 10% patients having ABC DLBCL, and about 0% to 20% patients who are DHITsig+ DLBCL patients; (iv) subgroup D4 comprises about 0% to 10% patients having GCB DLBCL, about 90% to 100% patients having ABC DLBCL, and about 0% to 10% patients who are DHITsig+ DLBCL patients; (v) subgroup D5 comprises about 0% to 20% patients having GCB DLBCL, about 55% to 65% patients having ABC DLBCL, and about 0% to 20% patients who are DHITsig+ DLBCL patients; (vi) subgroup D6 comprises about 50% to 60% patients having GCB DLBCL, about 20% to 40% patients having ABC DLBCL, and about 15% to 30% patients who are DHITsig+ DLBCL patients; (vii) subgroup D7 comprises about 20% to 35% patients having GCB DLBCL, about 45% to 55% patients having ABC DLBCL, and about 0% to 5% patients who are DHITsig+ DLBCL patients; and (viii) subgroup D8 comprises about 35% to 75% patients having GCB DLBCL, about 15% to 60% patients having ABC DLBCL, and about 25% to 65% patients who are DHITsig+ DLBCL patients. [0031] In some embodiments, the first cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP). [0032] In some embodiments, when the lymphoma patient is determined to belong to subgroup D4 or D8, the method comprises predicting that the patient is not likely to be responsive to the first cancer treatment. [0033] In some embodiments, the second cancer treatment is R-CHOP. [0034] In some embodiments, the second cancer treatment is not R-CHOP. [0035] In some embodiments, when the lymphoma patient is determined to belong to subgroup D4 or D8, the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor. [0036] In some embodiments, (i) when the lymphoma patient is determined to belong to subgroup D4, the second cancer treatment is a cyclin dependent kinase (CDK) inhibitor; and (ii) when the lymphoma patient is determined to belong to subgroup D8, the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor. [0037] In another aspect, provided herein is a method for predicting the responsiveness of lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first patient having a lymphoma; (b) determining the level of expression of one, two, three, four, five or more of the genes identified in Table 1; (c)comparing the level of expression of the one, two, three, four, five or more of the genes identified in Table 1 in the first biological sample with the level of expression of the same genes in a second biological sample(s) from a second patient(s), wherein the second lymphoma patient(s) is responsive to the cancer treatment, and wherein the similar expression of the one, two, three, four, five or more of the genes in the first biological sample relative to the level of expression of the one, two, three, four, five or more of the genes in the second biological sample(s) indicates that the lymphoma in the first patient will be responsive to treatment with the cancer treatment. [0038] In another aspect, provided herein is a method for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first lymphoma patient; (b) determining the expression of the genes or a certain subset of genes set forth in Table 1 or any combination thereof in the first biological sample; and (c) comparing the gene expression profile of the genes or subset of genes in the first biological sample to (i) the gene expression profile of the genes or subset of genes in biological samples from lymphoma patients which are responsive to the drug and (ii) the gene expression of the genes or subset of genes in biological samples from lymphoma patients which are not responsive to the cancer treatment, wherein a gene expression profile for the genes or subset of genes in the first biological sample similar to the gene expression profile for the genes or subset of genes in biological samples from lymphoma patients which are responsive to the cancer treatment indicates that the first lymphoma patient will be responsive to the cancer treatment, and a gene expression profile for the genes or subset of genes in first biological sample similar to the gene expression profile for the genes or subset of genes in biological samples from lymphoma patients which are not responsive to the cancer treatment indicates that the first lymphoma patient will not be responsive to the cancer treatment. [0039] In yet another aspect, provided herein is a method of treating a lymphoma patient, comprising: (i) identifying a lymphoma patient who is likely to be responsive to the cancer treatment according to the predicting method provided herein, and (ii) administering to the lymphoma patient the cancer treatment. [0040] In yet another aspect, provided herein is a method of treating a lymphoma patient, comprising: (i) identifying a lymphoma patient who is likely to be not responsive to the cancer treatment according to the predicting method provided herein, and (ii) administering to the lymphoma patient an alternative cancer treatment. [0041] In some embodiments, the cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP). [0042] In some embodiments, the alternative cancer treatment is a bromodomain and extra- terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor. [0043] In some embodiments, the lymphoma is diffuse large B-cell lymphoma (DLBCL). [0044] In some embodiments, the lymphoma is indolent B cell lymphoma. [0045] In some embodiments, the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma. [0046] In some embodiments, the lymphoma is follicular lymphoma. [0047] In some embodiments, the lymphoma is nodal marginal zone B-cell lymphoma. [0048] In some embodiments, the lymphoma is mantle cell lymphoma. [0049] In some embodiments, the lymphoma is chronic lymphocytic leukemia. [0050] In some embodiments, the level of expression of all the genes identified in Table 1 is determined in step (b) and compared in step (c). [0051] In some embodiments, the biological samples are tumor biopsy samples. [0052] In some embodiments, the determining step comprises detecting the presence or amount of a complex in the biological sample, wherein the presence or amount of the complex indicates the expression level of the genes in each subset of genes. [0053] In some embodiments, the complex is a hybridization complex. [0054] In some embodiments, the hybridization complex is attached to a solid support. [0055] In some embodiments, the complex is detectably labeled. [0056] In some embodiments, the determining step comprises detecting the presence or amount of a reaction product in the biological sample, wherein the presence or amount of the reaction product indicates the expression level of the genes. [0057] In some embodiments, the reaction product is detectably labeled. [0058] In some embodiments, the reference patients are refractory DLBCL patient. [0059] In some embodiments, the reference patients are relapsed DLBCL patient. [0060] In some embodiments, the reference patients are newly diagnosed DLBCL patient. [0061] In some embodiments, the lymphoma patient is a refractory DLBCL patient. [0062] In some embodiments, the lymphoma patient is a relapsed DLBCL patient. [0063] In some embodiments, the lymphoma patient is a newly diagnosed DLBCL patient. [0064] In some embodiments, the lymphoma patient is a GCB DLBCL patient. [0065] In some embodiments, the lymphoma patient is an ABC DLBCL patient. [0066] In some embodiments, the lymphoma patient is a DHITsig+ DLBCL patient. [0067] In some embodiments, the lymphoma patient is a DHITsig- DLBCL patient. 4. BRIEF DESCRIPTION OF THE FIGURES [0068] FIG.1 illustrates that Principal Component Analysis (PCA) projection of ssNorm datasets shows a high degree of overlap and no systematic stratification by dataset. [0069] FIG.2 depicts the flowchart of unsupervised clustering input, clustering evaluation, classifier application to independent data, and exemplary potential biological interpretation of patient subgroups. All four matrices of features (gene expression, Hallmark, C1, and LM23) were used in the unsupervised clustering methods. [0070] FIGS.3A-3E show heatmaps of clustering by each different features matrices and heatmap of consensus matrix. FIG.3A: heatmap of clustering by gene expression; FIG.3B: heatmap of clustering by Hallmark pathways; FIG.3C: heatmap of clustering by the C1 set of positional cytoband features; FIG.3D: heatmap of clustering by LM23 cell type features; and FIG.3E: heatmap of consensus matrix. [0071] FIGS.4A-4B show the silhouette metric of the clustering found by iClusterPlus (FIG. 4A) is superior to the one found by COCA (FIG.4B) for K=8 (total 8 subgroups). [0072] FIGS.5A-5E show the gene expression heatmaps for the eight patient clusters identified in the discovery (FIG.5A), ndMER (FIG.5B), Lenz datasets (FIG.5C), rrMER (FIG.5D), and CC-122 (FIG.5E). The heatmaps show the same top 100 most differentially expressed genes for each cluster (50 upregulated, 50 downregulated) in FIGS.5A-5E. Annotation tracks at the top show clusters’ associations with biological characteristics such as COO, TME, and DHITsig. [0073] FIGS.6A-6B illustrate that the discovery dataset (Commercial) and replication dataset (ndMER) show consistent cluster characteristics. Each pair of bars denotes the discovery dataset (commercial dataset; in blank color) and replication dataset (ndMER dataset; in light grey color). FIG.6A: fractions of each cluster that are GCB DLBCL patients in the Commercial and MER datasets; FIG.6B: fractions of each cluster that are TME+ DLBCL patients in Commercial and MER datasets. Error bars represent the 95% confidence interval, showing few differences in the COO or TME content of the clusters between datasets. [0074] FIG.7 shows the cluster prevalence of the TME+ and TME- patients across the eight subgroups in a combination dataset of MER, Schmitz, and Lenz datasets. [0075] FIG.8 illustrates the event-free survival (EFS) of the eight patient clusters in MER dataset treated with R-CHOP as first-line treatment. The EFS of the eight patient clusters show significantly different survival patterns (p<0.0001 by longrank test). [0076] FIG.9 illustrates that the cluster prevalence of the eight clusters is consistent across ndDLBCL datasets (Commercial, Lenz and ndMER datasets). In rrDLBCL datasets (rrMER and CC-122 datasets) the prevalence of the high risk clusters (D4 and D8) increases. [0077] FIG.10 shows that the DESeq2-derived DHITsig score appropriately ranks FISH Double HIT+ samples highly. The selected threshold (circled) identifies all FISH Double HIT+ patients, minimizes false positives, and matches the target prevalence of Double HIT+ patients among GCBs. [0078] FIGS.11A-11B shows that the application of the ssNorm-adapted DHITsig method to the Lenz and ndMER datasets shows significant prognostic value among R-CHOP-treated GCB patients (p=0.0006). FIG.11A: overall survival (OS) of the DHITsig+ and DHITsig- GCB patients in the Lenz dataset (treated with R-CHOP); FIG.11B: overall survival (OS) of the DHITsig+ and DHITsig- GCB patients in the MER dataset (treated with R-CHOP). [0079] FIG.12 shows the expression of the BCL6 signature across the eight subgroups in the ndMER dataset. [0080] FIGS.13A-13D display the percentage of nucleated cells of CD8, CD4, CD163, and CD20 DLBCL patients in each of the identified eight subgroups (FIG.13A: CD8 cells; FIG. 13B: CD4 cells; FIG.13C: CD163; and FIG.13D: CD20. 5. DETAILED DESCRIPTION OF THE INVENTION 5.1 Definitions [0081] As used herein, the term “lymphoma” includes, but is not limited to, Hodgkin’s lymphoma, non-Hodgkin’s lymphoma, diffuse large B-Cell lymphoma, indolent B-cell lymphoma, follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, cutaneous T-Cell lymphoma, cutaneous B-Cell lymphoma, mycosis fungoide, mantle cell lymphoma, and chronic lymphocytic leukemia. [0082] As used herein, and unless otherwise specified, the terms “treat,” “treating,” and “treatment” refer to an action that occurs while a patient is suffering from the specified cancer (a specific type of lymphoma, e.g., DLBCL), which reduces the severity of the cancer or retards or slows the progression of the cancer. [0083] The term “sensitivity” or “sensitive” when made in reference to a cancer treatment is a relative term which refers to the degree of effectiveness of the cancer treatment in lessening or decreasing the progress of a tumor or the cancer being treated. For example, the term “increased sensitivity” when used in reference to treatment of a cell or tumor in connection with a compound refers to an increase of, at least about 5%, or more, in the effectiveness of the cancer treatment. [0084] As used herein, and unless otherwise specified, the term “therapeutically effective amount” of a cancer treatment is an amount sufficient to provide a therapeutic benefit in the treatment or management of a cancer, or to delay or minimize one or more symptoms associated with the presence of the cancer. A therapeutically effective amount of a compound means an amount of therapeutic agent, alone or in combination with other therapies, which provides a therapeutic benefit in the treatment or management of the cancer. The term “therapeutically effective amount” can encompass an amount that improves overall therapy, reduces or avoids symptoms or causes of cancer, or enhances the therapeutic efficacy of another therapeutic agent. The term also refers to the amount of a compound that is sufficient to elicit the biological or medical response of a biological molecule (e.g., a protein, enzyme, RNA, or DNA), cell, tissue, system, animal, or human, which is being sought by a researcher, veterinarian, medical doctor, or clinician. [0085] The term “responsiveness” or “responsive” when used in reference to a cancer treatment refers to the degree of effectiveness of the treatment in lessening or decreasing the symptoms of a cancer, e.g., DLBCL, being treated. For example, the term “increased responsiveness” when used in reference to a treatment of a cell or a subject refers to an increase in the effectiveness in lessening or decreasing the symptoms of the disease compared to a reference treatment (e.g., of the same cell or subject, or of a different cell or subject) when measured using any methods known in the art. In certain embodiments, the increase in the effectiveness is at least about 5%, at least about 10%, at least about 20%, at least about 30%, at least about 40%, or at least about 50%. [0086] As used herein, the terms “effective subject response,” “effective patient response,” and “effective patient tumor response” refer to any increase in the therapeutic benefit to the patient. An “effective patient tumor response” can be, for example, about 5%, about 10%, about 25%, about 50%, or about 100% decrease in the rate of progress of the tumor. An “effective patient tumor response” can be, for example, about 5%, about 10%, about 25%, about 50%, or about 100% decrease in the physical symptoms of a cancer. An “effective patient tumor response” can also be, for example, about 5%, about 10%, about 25%, about 50%, about 100%, about 200%, or more increase in the response of the patient, as measured by any suitable means, such as gene expression, cell counts, assay results, tumor size, etc. [0087] An improvement in the cancer (e.g., DLBCL or a subtype thereof) or cancer-related disease can be characterized as a complete or partial response. “Complete response” refers to an absence of clinically detectable disease with normalization of any previously abnormal radiographic studies, bone marrow, and cerebrospinal fluid (CSF) or abnormal monoclonal protein measurements. “Partial response” refers to at least about 10%, about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, or about 90% decrease in all measurable tumor burden (i.e., the number of malignant cells present in the subject, or the measured bulk of tumor masses or the quantity of abnormal monoclonal protein) in the absence of new lesions. The term “treatment” contemplates both a complete and a partial response. [0088] The term “likelihood” generally refers to an increase in the probability of an event. The term “likelihood” when used in reference to the effectiveness of a patient tumor response generally contemplates an increased probability that the rate of tumor progress or tumor cell growth will decrease. The term “likelihood” when used in reference to the effectiveness of a patient tumor response can also generally mean the increase of indicators, such as mRNA or protein expression, that may evidence an increase in the progress in treating the tumor. [0089] The term “predict” generally means to determine or tell in advance. When used to “predict” the effectiveness of a cancer treatment, for example, the term “predict” can mean that the likelihood of the outcome of the cancer treatment can be determined at the outset, before the treatment has begun, or before the treatment period has progressed substantially. [0090] As used herein, the term “source” when used in reference to a reference sample refers to the origin of a sample. For example, a sample that is taken from blood would have a reference sample that is also taken from blood. Similarly, a sample that is taken from bone marrow would have a reference sample that is also taken from the bone marrow. [0091] As used herein, the term “refractory” or “resistant” refers to a disorder, disease, or condition that has not responded to prior treatment that can include one or more lines of therapy. In one embodiment, the disorder, disease, or condition has been previously treated one, two, three or four lines of therapy. In one embodiment, the disorder, disease, or condition has been previously treated with two or more lines of treatment, and has less than a complete response (CR) to most recent systemic therapy containing regimen. [0092] As used herein, the term “relapsed” refers to a disorder, disease, or condition that responded to treatment (e.g., achieved a complete response) then had progression. The treatment can include one or more lines of therapy. [0093] The term “expressed” or “expression” as used herein refers to the transcription from a gene to give an RNA nucleic acid molecule at least complementary in part to a region of one of the two nucleic acid strands of the gene. The term “expressed” or “expression” as used herein also refers to the translation from the RNA molecule to give a protein, a polypeptide, or a portion thereof.A “biological marker” or “biomarker” is a substance whose detection indicates a particular biological state, such as, for example, the presence of a type of cancer. In some embodiments, biomarkers can be determined individually. In other embodiments, several biomarkers can be measured simultaneously. [0094] The terms “polypeptide” and “protein,” as used interchangeably herein, refer to a polymer of three or more amino acids in a serial array, linked through peptide bonds. The term “polypeptide” includes proteins, protein fragments, protein analogues, oligopeptides, and the like. The term “polypeptide” as used herein can also refer to a peptide. The amino acids making up the polypeptide may be naturally derived, or may be synthetic. The polypeptide can be purified from a biological sample. The polypeptide, protein, or peptide also encompasses modified polypeptides, proteins, and peptides, e.g., glycopolypeptides, glycoproteins, or glycopeptides; or lipopolypeptides, lipoproteins, or lipopeptides. [0095] The term “antibody,” “immunoglobulin,” or “Ig” as used interchangeably herein, encompasses fully assembled antibodies and antibody fragments that retain the ability to specifically bind to the antigen. Antibodies provided herein include, but are not limited to, synthetic antibodies, monoclonal antibodies, polyclonal antibodies, recombinantly produced antibodies, multispecific antibodies (including bi-specific antibodies), human antibodies, humanized antibodies, chimeric antibodies, intrabodies, single-chain Fvs (scFv) (e.g., including monospecific, bispecific, etc.), camelized antibodies, Fab fragments, F(ab’) fragments, disulfide- linked Fvs (sdFv), anti-idiotypic (anti-Id) antibodies, and epitope-binding fragments of any of the above. In particular, antibodies provided herein include immunoglobulin molecules and immunologically active portions of immunoglobulin molecules, i.e., antigen binding domains or molecules that contain an antigen-binding site that immunospecifically binds to CRBN antigen (e.g., one or more complementarity determining regions (CDRs) of an anti-CRBN antibody). The antibodies provided herein can be of any class (e.g., IgG, IgE, IgM, IgD, and IgA) or any subclass (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2) of immunoglobulin molecule. In some embodiments, the anti-CRBN antibodies are fully human, such as fully human monoclonal CRBN antibodies. In certain embodiments, antibodies provided herein are IgG antibodies, or a subclass thereof (e.g., human IgG1 or IgG4). [0096] The terms “antigen binding domain,” “antigen binding region,” “antigen binding fragment,” and similar terms refer to the portion of an antibody that comprises the amino acid residues that interact with an antigen and confer on the binding agent its specificity and affinity for the antigen (e.g., the CDR). The antigen binding region can be derived from any animal species, such as rodents (e.g., rabbit, rat, or hamster) and humans. In some embodiments, the antigen-binding region is of human origin. [0097] The term “epitope” as used herein refers to a localized region on the surface of an antigen that is capable of binding to one or more antigen binding regions of an antibody, that has antigenic or immunogenic activity in an animal, such as a mammal (e.g., a human), and that is capable of eliciting an immune response. An epitope having immunogenic activity is a portion of a polypeptide that elicits an antibody response in an animal. An epitope having antigenic activity is a portion of a polypeptide to which an antibody immunospecifically binds as determined by any method well known in the art, for example, by the immunoassays described herein. Antigenic epitopes need not necessarily be immunogenic. Epitopes usually consist of chemically active surface groupings of molecules, such as amino acids or sugar side chains, and have specific three-dimensional structural characteristics as well as specific charge characteristics. A region of a polypeptide contributing to an epitope may be contiguous amino acids of the polypeptide, or the epitope may come together from two or more non-contiguous regions of the polypeptide. The epitope may or may not be a three-dimensional surface feature of the antigen. [0098] The terms “fully human antibody” and “human antibody” are used interchangeably herein and refer to an antibody that comprises a human variable region and, in some embodiments, a human constant region. In specific embodiments, the terms refer to an antibody that comprises a variable region and a constant region of human origin. The term “fully human antibody” includes antibodies having variable and constant regions corresponding to human germline immunoglobulin sequences as described by Kabat et al., Sequences of Proteins of Immunological Interest, U.S. Department of Health and Human Services, NIH Publication No. 91-3242 (5th ed.1991). [0099] The phrase “recombinant human antibody” includes human antibodies that are prepared, expressed, created, or isolated by recombinant means, such as antibodies expressed using a recombinant expression vector transfected into a host cell, antibodies isolated from a recombinant, combinatorial human antibody library, antibodies isolated from an animal (e.g., a mouse or a cow) that is transgenic and/or transchromosomal for human immunoglobulin genes (see, e.g., Taylor et al., Nucl. Acids Res., 1992, 20:6287-6295) or antibodies prepared, expressed, created, or isolated by any other means that involves splicing of human immunoglobulin gene sequences to other DNA sequences. Such recombinant human antibodies can have variable and constant regions derived from human germline immunoglobulin sequences. See Kabat et al., Sequences of Proteins of Immunological Interest, U.S. Department of Health and Human Services, NIH Publication No.91-3242 (5th ed.1991). In certain embodiments, however, such recombinant human antibodies are subjected to in vitro mutagenesis (or, when an animal transgenic for human Ig sequences is used, in vivo somatic mutagenesis) and thus the amino acid sequences of the heavy chain variable and light chain variable regions of the recombinant antibodies are sequences that, while derived from and related to human germline heavy chain variable and light chain variable sequences, may not naturally exist within the human antibody germline repertoire in vivo. [00100] The term “monoclonal antibody” refers to an antibody obtained from a population of homogenous or substantially homogeneous antibodies, and each monoclonal antibody will typically recognize a single epitope on the antigen. In some embodiments, a “monoclonal antibody,” as used herein, is an antibody produced by a single hybridoma or other cell, wherein the antibody immunospecifically binds to only an epitope as determined, e.g., by ELISA or other antigen-binding or competitive binding assay known in the art or in the Examples provided herein. The term “monoclonal” is not limited to any particular method for making the antibody. For example, monoclonal antibodies provided herein may be made by the hybridoma method as described in Kohler et al., Nature, 1975, 256:495-497, or may be isolated from phage libraries using the techniques as described herein. Other methods for the preparation of clonal cell lines and of monoclonal antibodies expressed thereby are well known in the art. See, e.g., Short Protocols in Molecular Biology, Chapter 11 (Ausubel et al., eds., John Wiley and Sons, New York, 5th ed.2002). Other exemplary methods of producing other monoclonal antibodies are provided in the Examples herein. [00101] “Polyclonal antibodies” as used herein refers to an antibody population generated in an immunogenic response to a protein having many epitopes and thus includes a variety of different antibodies directed to the same or to different epitopes within the protein. Methods for producing polyclonal antibodies are known in the art. See, e.g., Short Protocols in Molecular Biology, Chapter 11 (Ausubel et al., eds., John Wiley and Sons, New York, 5th ed.2002). [00102] The term “level” refers to the amount, accumulation, or rate of a molecule. A level can be represented, for example, by the amount or the rate of synthesis of a messenger RNA (mRNA) encoded by a gene, the amount or the rate of synthesis of a polypeptide or protein encoded by a gene, or the amount or the rate of synthesis of a biological molecule accumulated in a cell or biological fluid. The term "level" refers to an absolute amount of a molecule in a sample or a relative amount of the molecule, determined under steady-state or non-steady-state conditions. [00103] An mRNA that is “upregulated” is generally increased upon a given treatment or condition. An mRNA that is “downregulated” generally refers to a decrease in the level of expression of the mRNA in response to a given treatment or condition. In some situations, the mRNA level can remain unchanged upon a given treatment or condition. An mRNA from a patient sample can be “upregulated” when treated with a drug, as compared to a non-treated control. This upregulation can be, for example, an increase of about 5%, about 10%, about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 100%, about 200%, about 300%, about 500%, about 1,000%, about 5,000%, or more of the comparative control mRNA level. Alternatively, an mRNA can be “downregulated”, or expressed at a lower level, in response to administration of certain compounds or other agents. A downregulated mRNA can be, for example, present at a level of about 99%, about 95%, about 90%, about 80%, about 70%, about 60%, about 50%, about 40%, about 30%, about 20%, about 10%, about 1%, or less of the comparative control mRNA level. [00104] Similarly, the level of a polypeptide or protein biomarker from a patient sample can be increased when treated with a drug, as compared to a non-treated control. This increase can be about 5%, about 10%, about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 100%, about 200%, about 300%, about 500%, about 1,000%, about 5,000%, or more of the comparative control protein level. Alternatively, the level of a protein biomarker can be decreased in response to administration of certain compounds or other agents. This decrease can be, for example, present at a level of about 99%, about 95%, about 90%, about 80%, about 70%, about 60%, about 50%, about 40%, about 30%, about 20%, about 10%, about 1%, or less of the comparative control protein level. [00105] The terms “determining,” “measuring,” “evaluating,” “assessing,” and “assaying” as used herein generally refer to any form of measurement, and include determining whether an element is present or not. These terms include quantitative and/or qualitative determinations. Assessing may be relative or absolute. “Assessing the presence of” can include determining the amount of something present, as well as determining whether it is present or absent. [00106] The terms “isolated” and “purified” refer to isolation of a substance (such as mRNA, DNA, or protein) such that the substance comprises a substantial portion of the sample in which it resides, i.e., greater than the portion of the substance that is typically found in its natural or un-isolated state. Typically, a substantial portion of the sample comprises, e.g., greater than 1%, greater than 2%, greater than 5%, greater than 10%, greater than 20%, greater than 50%, or more, usually up to about 90%-100% of the sample. For example, a sample of isolated mRNA can typically comprise at least about 1% total mRNA. Techniques for purifying polynucleotides are well known in the art and include, for example, gel electrophoresis, ion-exchange chromatography, affinity chromatography, flow sorting, and sedimentation according to density. [00107] As used herein, the term “bound” indicates direct or indirect attachment. In the context of chemical structures, “bound” (or “bonded”) may refer to the existence of a chemical bond directly joining two moieties or indirectly joining two moieties (e.g., via a linking group or any other intervening portion of the molecule). The chemical bond may be a covalent bond, an ionic bond, a coordination complex, hydrogen bonding, van der Waals interactions, or hydrophobic stacking, or may exhibit characteristics of multiple types of chemical bonds. In certain instances, “bound” includes embodiments where the attachment is direct and embodiments where the attachment is indirect. [00108] The term “sample” as used herein relates to a material or mixture of materials, typically, although not necessarily, in fluid form, containing one or more components of interest. In some embodiments, a sample can be a biological sample. “Biological sample” as used herein refers to a sample obtained from a biological subject, including a sample of biological tissue or fluid origin, obtained, reached, or collected in vivo or in situ. A biological sample also includes samples from a region of a biological subject containing precancerous or cancer cells or tissues. Such samples can be, but are not limited to, organs, tissues, and cells isolated from a mammal. Exemplary biological samples include but are not limited to cell lysate, a cell culture, a cell line, a tissue, oral tissue, gastrointestinal tissue, an organ, an organelle, a biological fluid, a blood sample, a urine sample, a skin sample, and the like. Preferred biological samples include, but are not limited to, whole blood, partially purified blood, PBMC, tissue biopsies, and the like. [00109] The term “analyte” as used herein refers to a known or unknown component of a sample. [00110] The term “capture agent” as used herein refers to an agent that binds an mRNA or protein through an interaction that is sufficient to permit the agent to bind and to concentrate the mRNA or protein from a heterogeneous mixture. [00111] As used herein and unless otherwise indicated, the term “pharmaceutically acceptable salt” encompasses non-toxic acid and base addition salts of the compound to which the term refers. Acceptable non-toxic acid addition salts include those derived from organic and inorganic acids know in the art, which include, for example, hydrochloric acid, hydrobromic acid, phosphoric acid, sulfuric acid, methanesulphonic acid, acetic acid, tartaric acid, lactic acid, succinic acid, citric acid, malic acid, maleic acid, sorbic acid, aconitic acid, salicylic acid, phthalic acid, embolic acid, enanthic acid, and the like. Compounds that are acidic in nature are capable of forming salts with various pharmaceutically acceptable bases. The bases that can be used to prepare pharmaceutically acceptable base addition salts of such acidic compounds are those that form non-toxic base addition salts, i.e., salts containing pharmacologically acceptable cations such as, but not limited to, alkali metal or alkaline earth metal salts (calcium, magnesium, sodium, or potassium salts in particular). Suitable organic bases include, but are not limited to, N,N-dibenzylethylenediamine, chloroprocaine, choline, diethanolamine, ethylenediamine, meglumaine (N-methylglucamine), lysine, and procaine. [00112] As described herein, the term “second active agent” refers to any additional treatment that is biologically active. It is understood that the second active agent can be a hematopoietic growth factor, cytokine, anti-cancer agent, antibiotic, cox-2 inhibitor, immunomodulatory agent, immunosuppressive agent, corticosteroid, therapeutic antibody that specifically binds to a cancer antigen or a pharmacologically active mutant, or derivative thereof. Exemplary second active agents include, but are not limited to, an HDAC inhibitor (e.g., panobinostat, romidepsin, or vorinostat), a BCL2 inhibitor (e.g., venetoclax), a BTK inhibitor (e.g., ibrutinib or acalabrutinib), an mTOR inhibitor (e.g., everolimus), a PI3K inhibitor (e.g., idelalisib), a PKCβ inhibitor (e.g., enzastaurin), a SYK inhibitor (e.g., fostamatinib), a JAK2 inhibitor (e.g., fedratinib, pacritinib, ruxolitinib, baricitinib, gandotinib, lestaurtinib, or momelotinib), an Aurora A kinase inhibitor (e.g., alisertib), an EZH2 inhibitor (e.g., tazemetostat, GSK126, CPI-1205, 3-deazaneplanocin A, EPZ005687, EI1, UNC1999, or sinefungin), a BET inhibitor (e.g., birabresib or 4 [2- (cyclopropylmethoxy)-5-(methanesulfonyl)phenyl]-2-methylisoquinolin-1(2H)-one), a hypomethylating agent (e.g., 5-azacytidine or decitabine), a chemotherapy (e.g., bendamustine, doxorubicin, etoposide, methotrexate, cytarabine, vincristine, ifosfamide, or melphalan), or an epigenetic compound (e.g., a DOT1L inhibitor such as pinometostat, a HAT inhibitor such as C646, a WDR5 inhibitor such as OICR-9429, a HDAC6 inhibitor such as ACY-241, a DNMT1 selective inhibitor such as GSK3484862, a LSD-1 inhibitor such as Compound C or seclidemstat, a G9A inhibitor such as UNC 0631, a PRMT5 inhibitor such as GSK3326595, a BRPF1B/2 inhibitor such as OF-1, a BRD9/7 inhibitor such as LP99, a SUV420H1/H2 inhibitor such as A-196, a Menin-MLL inhibitor such as MI-503, a CARM1 inhibitor such as EZM2302, a BRD9 such as an inhibitor dBrd9), aiolos/ikaros degrading cereblon E3 ligase modulator (CELMoDs), CREBBp2 inhibitors, anti-CD79b antibody, CD19 CAR-T, inhibitors of p53 (nutlins), Bcl6 inhibitors, CREBBp2 CELMoDs, CD79b CELMoDs, CD19 CELMoDs, p53 (nutlins) CELMoDs, Bcl6 CELMoDs, inhibitors of ligand directed degradation (LDD) of CREBBP2, inhibitors of LDD of CD79b, inhibitors of LDD of CD19, inhibitors of LDD of p53(nutlins), inhibitors of LDD of Bcl6, inhibitors of LDD of CK1a, inhibitors of LDD of IRAK4 (e.g., in MYD88 L265p lymphoma), MALT1 inhibitors such as JNJ-67856633, MAT2A inhibitors (e.g., for 9p21 deletions), anti-CD3 x anti-CD19 bispecific antibodies, and anti-CD3 x anti-CD20 bispecific antibodies. [00113] The term “about” or “approximately” means an acceptable error for a particular value as determined by one of ordinary skill in the art, which depends in part on how the value is measured or determined. In certain embodiments, the term “about” or “approximately” means within 1, 2, 3, or 4 standard deviations. In certain embodiments, the term “about” or “approximately” means within 50%, 20%, 15%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, or 0.05% of a given value or range. [00114] The use of the word “a” or “an” when used in conjunction with the term “comprising” in the claims and/or the specification can mean “one,” but it is also consistent with the meaning of “one or more,” “at least one,” and “one or more than one.” [00115] As used herein, unless otherwise specified or indicated from context, the term “pre- treatment” as used in accordance with the methods described herein refers to prior to administration of a drug. [00116] As used herein, the terms “patient” and “subject” refer to an animal, such as a mammal. In a specific embodiment, the patient is a human. In other embodiments, the patient is a non-human animal, such as a dog, cat, farm animal (e.g., horse, pig, or donkey), chimpanzee, or monkey. In specific embodiments, the patient is a human with lymphoma (e.g., DLBCL) in need of treatment. 5.2 Methods of Clustering Lymphoma Patients [00117] In one aspect, provided herein are method of classifying lymphoma patients, comprising (a) obtain samples from lymphoma patients; (b) measuring the gene expression levels in the samples; and (c) clustering the lymphoma patients into subgroups of patients having lymphoma using gene expression levels in the samples. In some embodiments, the lymphoma patients are Diffuse Large B-Cell Lymphoma (DLBCL) patients. [00118] In some embodiments, the lymphoma is diffuse large B-cell lymphoma (DLBCL). In some embodiments, the lymphoma is indolent B cell lymphoma. In some embodiments, the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma. In one embodiment, the lymphoma is follicular lymphoma. In another embodiment, the lymphoma is nodal marginal zone B-cell lymphoma. In yet another embodiment, the lymphoma is mantle cell lymphoma. In yet another embodiment, the lymphoma is chronic lymphocytic leukemia. [00119] In some embodiments, the sample is obtained from a tissue of the subject containing DLBCL cells. More detailed description of the sample (or biological sample) is provided in Section 5.7 below. [00120] In some embodiments, the DLBCL patients are newly diagnosed (nd) and relapsed/refractory (r/r) DLBCL patients. In one embodiment, the DLBCL patients are newly diagnosed (nd) DLBCL patients. In another embodiment, the DLBCL patients are relapsed/refractory (r/r) DLBCL patients. [00121] As exemplified in Example 1 provide in Section 6.1.1 below, multiple patients datasets can be utilized in the classifying/clustering method. The patients datasets include but not limited to, for example, a set of 267 commercially-sourced newly diagnosed DLBCL (ndDLBCL) patient samples which had molecular profiling but no survival data (“Commercial dataset”), a set of 342 ndDLBCL patients from the Molecular Epidemiology Resource (MER) (“ndMER dataset”; see Cerhan et al., Int. J. Epidermiol., 2017, 46(6):1753-1754i), the Lenz microarray dataset of 414 ndDLBCL patients was also used as a second replication cohort and had outcome data available (see Lenz et al., N. Engl. J. Med., 2008, 359(22):2313-2323), a set of 86 rrDLBCL patients from the MER cohort (“rrMER dataset”), and a set of 189 rrDLBCL patient samples from two clinical trials (CC-122-ST-001 and CC-122-DLBCL-001) (“CC-122 dataset”; see Clinical Trials Nos. NCT01421524 and NCT02031419). A subset of the Commercial dataset also had IHC imaging data, so subset of the Commercial dataset is referred to the Commercial+IHC dataset. The cohort in the MER dataset is well characterized in terms of clinical outcome and treatment. The Lenz microarray dataset has outcome data available. The two other datasets (rrMER dataset and CC-122 dataset) had clinical outcome information. [00122] In some embodiments, step (a) and step (b) generate a dataset of the lymphoma patients. In one embodiment, the lymphoma patients are a newly diagnosed lymphoma patient cohort. In another embodiment, the lymphoma patients are a relapsed/refractory lymphoma patient cohort. [00123] In some embodiments, the clustering step uses a discovery dataset and one or more replication/validation datasets. In one embodiment, the clustering step uses a discovery dataset and a replication/validation dataset. In some embodiments, the discovery dataset is from samples from a discovery cohort. In some embodiments, the replication/validation dataset is from sample from a replication/validation cohort. In some embodiments, the discovery dataset is selected from the group consisting of Commercial dataset, ndMER dataset, Lenz dataset, rrMER dataset, and CC-122 dataset. In some embodiments, the replication/validation dataset is selected from the group consisting of Commercial dataset, ndMER dataset, Lenz dataset, rrMER dataset, and CC-122 dataset. In one embodiment, the discovery dataset is Commercial dataset. In another embodiment, the discovery dataset is ndMER dataset. In another embodiment, the discovery dataset is Lenz dataset. In another embodiment, the discovery dataset is rrMER dataset. In another embodiment, the discovery dataset is CC-122 dataset. [00124] In some embodiments, the clustering step comprises (i) normalizing a dataset; (ii) selecting clustering feature(s); and (iii) applying a clustering method using the clustering feature(s). In some embodiments, the clustering step further comprises detecting outliers in the dataset and removing the outliers. In some embodiments, the clustering step further comprises evaluating clustering results of the clustering method. In some embodiments, the clustering is an unsupervised clustering method. In one embodiment, the clustering method is a hierarchical clustering method. In another embodiment, the clustering method is a non-hierarchical method. In one embodiment, the clustering method is K means clustering method. In another embodiment, the clustering method is a partitioning method. In another embodiment, the clustering method is a fuzzy clustering method. In another embodiment, the clustering method is a density-based clustering. In another embodiment, the clustering method is a model-based clustering. In one embodiment, the clustering method is Cluster of Cluster Analysis (COCA) clustering method. In one embodiment, the clustering method is iClusterPlus clustering method. [00125] In some embodiments, a single sample normalization (ssNorm) method is use to normalize the datasets prior to analysis. The practice of performing ssNorm can benefit both immediate short-term analyses and long-term translatability of findings. In the short term, ssNorm provides a fixed normalization scheme that can apply to individual samples completely independently. There is no need to re-normalize datasets with the addition of new batches of samples, or when combining datasets for larger analyses. The ssNorm method can also implicitly perform batch correction, thereby removing gene-specific experimental effects that appear as systematic, biased differences in the raw expression space. In practice, the ssNorm method can properly align diverse datasets to a common space, showing no meaningful separation by dataset/batch when projecting into PCA space. From a strategic perspective, ssNorm allows simple translatability of any classifier or parameterization from one dataset to another – a classifier built on one ssNorm dataset can be directly applied to any other ssNorm dataset, without any need for reweighting model parameters or setting new thresholds. [00126] In certain embodiments, DLBCL-specific housekeeping genes are used for normalization. In one embodiment, ISY1, R3HDM1, TRIM56, UBXN4, and WDR55 are used for normalization. [00127] Clustering methods are sensitive to the input data, and algorithm performance can be degraded by introducing noisy or irrelevant features. Both feature selection and feature engineering approaches can be utilized to reduce the dimensionality of the dataset while maintaining a representation of relevant biological activity. [00128] In some embodiments, one, two, three, four, or more features are selected as the clustering features as input into the clustering method. In some embodiments, a subset of the gene expression data is selected as the clustering features. In some embodiments, the subset of the gene expression data are gene expression data of the top 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 10%, or 5% most expressed genes. In one embodiment, the subset of the gene expression data are gene expression data of the top 25% most expressed genes. [00129] Some types of derived features scores that aggregate multiple biologically related genes into a single feature can also be used as clustering features. Gene Set Variation Analysis (GSVA) is a Gene Set Enrichment (GSE) method that estimates variation of pathway activity over a sample population in an unsupervised manner. GSVA provides increased power to detect subtle pathway activity changes over a sample population in comparison to corresponding methods. GSVA constitutes a starting point to build pathway-centric models of biology and GSVA can contribute to the current need of GSE methods for RNA-seq data. See Hanzelman et al., BMC Bioinformatics, 2013, 14:7. GSVA can be performed over the 50 hallmark pathway gene sets from MSigDB (Hallmark GSVA scores). See, e.g., Liberzon et al., Cell Syst., 2015, 1:417:425. GSVA can be performed over the 299 C1 set of positional cytoband signatures gene sets from MSigDB (C1 Positional GSVA scores). See, e.g., Alhamdoosh et al., F1000Research, 2017, 6:2010. GSVA can also be performed over the DLBCL-specific LM23 matrix (Cell type LM23 GSVA scores), deriving from the DCQ cellular deconvolution method. See Althoum et al., Mol. Syst. Bio., 2014, 10:720. [00130] In some embodiments, the Hallmark GSVA scores are selected as the clustering features. In some embodiments, the C1 Positional GSVA scores are selected as the clustering features. In some embodiments, the Cell type LM23 GSVA scores are selected as the clustering features. In one embodiments, a subset of the gene expression data, Hallmark GSVA scores, C1 Positional GSVA scores, Cell type LM23 GSVA scores are selected as the matrices of clustering features. [00131] In some embodiments, the clustering method performs clustering on each feature matrix independently and aggregates the cluster features. In some embodiments, the clustering method performs clustering aggregates the cluster features from different matrices of features and clustering across all the clustering features. [00132] In some embodiments, the dataset is clustered into 2-20 clusters (K= 2-20). In some embodiment, the dataset is clustered into 2-15 clusters (K = 2-15). In some embodiments, the dataset is clustered into 2-10 clusters (K = 2-10). In some embodiments, the dataset is clustered into 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 clusters. In some embodiments, the optimal number of clusters is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20. In one embodiment, the dataset is clustered into 8 clusters (K = 8). In one embodiment, the optimal number of cluster is 8. [00133] In some embodiments, the clustering step further comprises evaluating clustering results of the clustering method. In some embodiments, the clustering results of each number of clusters (K) are evaluated. In one embodiment, the clustering results are evaluated using nbClust R package. In one embodiments, the clustering results are evaluated using metrics selected from the group consisting of silhouette statistic, gap statics, and percentage of variance explained by the clustering. In one embodiments, the clustering results are evaluated by the minium cluster size. In order to have sufficient sample size in each cluster to build a downstream classifier model that could robustly identify each cluster or subgroup, the resulting clusters from each tested clustering need to have a minimum cluster size. [00134] In some embodiments, the method of clustering lymphoma patients further comprises (i) identifying one or more cluster classifier(s) by training a classifier model using the clustering results of a discovery dataset, (ii) applying the one or more cluster classifier(s) to a replication/validation dataset to classify the replication/validation dataset, (iii) clustering the replication/validation dataset using the clustering method and (iv) comparing the classification results of the replication/validation dataset using the cluster classifier(s) with the clustering results of the replication/validation dataset using the clustering method. In one embodiment, the model is a grouped multinomial generalized linear model (GLM). In one specific embodiment, the model is GLM using least absolute shrinkage and selection operator (LASSO). In some embodiments, the similar the classification results of the replication/validation dataset using the cluster classifier(s) with the clustering results of the replication/validation dataset using the clustering method indicates that the cluster classifier(s) is effective for classifying the replication/validation dataset. [00135] In some embodiments, the method of clustering lymphoma patients further comprises (i) identifying one or more cluster classifier(s) by training a classifier model using the clustering results of a discovery dataset, and (ii) applying the one or more cluster classifier(s) to a replication/validation dataset to classify the replication/validation dataset. In one embodiment, the model is a grouped multinomial generalized linear model (GLM). In one specific embodiment, the model is GLM using least absolute shrinkage and selection operator (LASSO). [00136] In some embodiments, expression levels of one or more genes in the discovery dataset are used in training the classifier model. In some embodiments, one, two, three, four, five or more of the genes identified in Table 1 are used in training the classifier model. In one embodiment, all the genes identified in Table 1 are used in training the classifier model. [00137] In some embodiments, the age of the lymphoma patient is 30 years or older, 35 years or older, 40 years or older, 45 years or older, 50 years or older, or 55 years or old, or 60 years or older, or 65 years or older, or 70 years or older at baseline. In another embodiment, the age of the lymphoma patient is 70 years or older. In another embodiment, the age of the lymphoma patient is 60 years or older. In some embodiments, the age of the lymphoma patient is between 30 to 35, 35 to 40, 40 to 45, 45 to 50, 50 to 55, 55 to 60, 60 to 65, or 65 to 70 years old. [00138] In some embodiments, the classifier model uses one, two, three, four, five, or more of the genes identified in Table 1. In some embodiments, the classifier model uses all the genes identified in Table 1. Table 1. List of Genes Utilized in the Resulting grouped multinomial generalized linear model (GLM) Model
Figure imgf000027_0001
Figure imgf000028_0001
[00139] In some embodiments, the method uses all the genes selected from the group consisting of AARS2, ACPP, ACRC, ACTN4, ADH1B, AGER, ALAS1, AMH, ANKRD20A5P, ANKRD22, APOL3, ARHGEF1, ATM, ATP6AP2, ATP8B1, BMF, CBR3, CCDC9, CCL21, CCR1, CCT7, CD83, CDCA8, CDK12, CECR7, CEP72, CHKA, CILP, CLEC4E, CNKSR2, CORO1C, CPT2, CR2, CTSS, CTSZ, DCAF5, DENND2D, DEPDC5, DGKA, DUSP1, E2F4, EHF, ENTPD1, EPAS1, FAM90A1, FBXO6, FCER1G, FCRL1, FLVCR2, FOXO3, FXYD2, FYB, G6PD, GBP4, GDI1, GIMAP6, GINS1, GMIP, GSTM4, GZMM, HAT1, HDAC1, HES4, HLA-DQA1, HMBOX1, HSD11B1, HSP90AB1, HTN3, IL4I1, IL9R, INO80D, IRAK3, ITPKB, IVNS1ABP, KDM8, KIAA0226L, KLF12, KLHL9, LAMC3, LCOR, LGALS3, LGI2, LIF, LIPE, LY9, MAD2L2, MAN1B1, MAST3, MCM8, MPC2, MPEG1, MRPL3, MSH6, MXD1, MYOCD, NDST1, NDUFB2, NFKBIA, NLRP14, NRF1, NUDT1, ODF3B, PABPC1L, PDXK, PHLDB3, PIK3CG, PILRA, PKM, PLXNA3, POP1, PRM1, PRMT1, PSEN2, PTPN1, QTRTD1, RAD51D, RANBP6, RASGRP1, RASGRP2, RELB, RGL1, RRP1, S1PR4, SEC61B, SERPINB3, SERPINE1, SETBP1, SH3BP1, SH3BP5, SLAMF8, SLC7A5, SLC9A9, SNX10, ST6GALNAC4, SYNGAP1, TAPT1-AS1, TBC1D27, TGDS, TGFBRAP1, THBD, TLR7, TM2D1, TMEM165, TMEM175, TMEM256, TMEM63B, TMLHE, TNFRSF13B, TRAV12-2, TRAV8-3, TRIB2, TRMU, TXK, U2AF2, UNC13D, UNK, VAV1, WARS, WDR19, WSB1, ZC3H3, ZMYM2, ZNF134, ZNF165, ZNF324, ZNF333, ZNF347, ZNF524, ZNF623, ZNF765, ZNF850, and ZNF91. 5.3 Methods of Predicting Responsiveness of Lymphoma Patient and Methods of Treating [00140] In another aspect, provided herein are methods for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining reference biological samples from each patient in a reference patient group comprising reference patients having a lymphoma; (b) classifying the reference patient group into subgroups of patients using gene expression levels in the reference biological samples; (c) obtaining a biological sample from the lymphoma patient; (d) determining to which subgroup the lymphoma patient belongs using gene expression levels in the biological sample from the lymphoma patient; and (e) predicting the responsiveness of the lymphoma patient to a first cancer treatment. [00141] In some embodiments, the method further comprises administering to the lymphoma patient a second cancer treatment. [00142] In some embodiments, the sample is obtained from a tissue of the subject containing DLBCL cells. More detailed description of the sample (or biological sample) is provided in Section 5.7 below. [00143] In some embodiments, the lymphoma patients are Diffuse Large B-Cell Lymphoma (DLBCL) patients. [00144] In some embodiments, the DLBCL patients are newly diagnosed (nd) and relapsed/refractory (r/r) DLBCL patients. In one embodiment, the DLBCL patients are newly diagnosed (nd) DLBCL patients. In another embodiment, the DLBCL patients are relapsed/refractory (r/r) DLBCL patients. [00145] In some embodiments, step (b) comprises generating clustering information defining relationships between the expression levels of one or more genes in the reference biological samples; and rearranging heat map representation based on the clustering information. [00146] In some embodiments, step (b) uses the clustering method described herein in Section 5.2. [00147] In some embodiments, the clustering step uses a discovery dataset and one or more replication/validation datasets. In one embodiment, the clustering step uses a discovery dataset and a replication/validation dataset. In some embodiments, the discovery dataset is from samples from a discovery cohort. In some embodiments, the replication/validation dataset is from sample from a replication/validation cohort. In some embodiments, the discovery dataset is selected from the group consisting of Commercial dataset, ndMER dataset, Lenz dataset, rrMER dataset, and CC-122 dataset. In some embodiments, the replication/validation dataset is selected from the group consisting of Commercial dataset, ndMER dataset, Lenz dataset, rrMER dataset, and CC-122 dataset. In one embodiment, the discovery dataset is Commercial dataset. In another embodiment, the discovery dataset is ndMER dataset. In another embodiment, the discovery dataset is Lenz dataset. In another embodiment, the discovery dataset is rrMER dataset. In another embodiment, the discovery dataset is CC-122 dataset. [00148] In some embodiments, the clustering step comprises (i) normalizing a dataset; (ii) selecting clustering feature(s); and (iii) applying a clustering method using the clustering feature(s). In some embodiments, the clustering step further comprises detecting outliers in the dataset and removing the outliers. In some embodiments, the clustering step further comprises evaluating clustering results of the clustering method. In some embodiments, the clustering is an unsupervised clustering method. In one embodiment, the clustering method is a hierarchical clustering method. In another embodiment, the clustering method is a non-hierarchical method. In one embodiment, the clustering method is K means clustering method. In another embodiment, the clustering method is a partitioning method. In another embodiment, the clustering method is a fuzzy clustering method. In another embodiment, the clustering method is a density-based clustering. In another embodiment, the clustering method is a model-based clustering. In one embodiment, the clustering method is Cluster of Cluster Analysis (COCA) clustering method. In one embodiment, the clustering method is iClusterPlus clustering method. In some embodiments, a single sample normalization (ssNorm) method is use to normalize the datasets prior to analysis. [00149] In certain embodiments, DLBCL-specific housekeeping genes are used for normalization. In one embodiment, ISY1, R3HDM1, TRIM56, UBXN4, and WDR55 are used for normalization. [00150] In some embodiments, one, two, three, four, or more features are selected as the clustering features as input into the clustering method. In some embodiments, a subset of the gene expression data is selected as the clustering features. In some embodiments, the subset of the gene expression data are gene expression data of the top 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 10%, or 5% most expressed genes. In one embodiment, the subset of the gene expression data are gene expression data of the top 25% most expressed genes. [00151] In some embodiments, the Hallmark GSVA scores are selected as the clustering features. In some embodiments, the C1 Positional GSVA scores are selected as the clustering features. In some embodiments, the Cell type LM23 GSVA scores are selected as the clustering features. In one embodiments, a subset of the gene expression data, Hallmark GSVA scores, C1 Positional GSVA scores, Cell type LM23 GSVA scores are selected as the matrices of clustering features. [00152] In some embodiments, the clustering method performs clustering on each feature matrix independently and aggregates the cluster features. In some embodiments, the clustering method performs clustering aggregates the cluster features from different matrices of features and clustering across all the clustering features. [00153] In some embodiments, the dataset is clustered into 2-20 clusters (K= 2-20). In some embodiment, the dataset is clustered into 2-15 clusters (K = 2-15). In some embodiments, the dataset is clustered into 2-10 clusters (K = 2-10). In some embodiments, the dataset is clustered into 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 clusters. In some embodiments, the optimal number of clusters is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20. In one embodiment, the dataset is clustered into 8 clusters (K = 8). In one embodiment, the optimal number of cluster is 8. [00154] The terms “clusters” and “subgroups” are herein used interchangeably. [00155] In some embodiments, the reference patients in the reference patient group are clustered into 2-20 subgroups (K= 2-20). In some embodiments, the reference patients in the reference patient group are clustered into 2-15 subgroups (K = 2-15). In some embodiments, the reference patients in the reference patient group are clustered into 2-10 subgroups (K = 2-10). In some embodiments, the reference patients in the reference patient group are clustered into 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 subgroups. In one embodiment, the reference patients in the reference patient group are clustered into 8 groups (K = 8). [00156] In some embodiments, the clustering step further comprises evaluating clustering results of the clustering method. In some embodiments, the clustering results of each number of clusters (K) are evaluated. In one embodiment, the clustering results are evaluated using nbClust R package. In one embodiments, the clustering results are evaluated using metrics selected from the group consisting of silhouette statistic, gap statics, and percentage of variance explained by the clustering. In one embodiments, the clustering results are evaluated by the minimum cluster size. In order to have sufficient sample size in each cluster to build a downstream classifier model that could robustly identify each cluster or subgroup, the resulting clusters from each tested clustering need to have a minimum cluster size. [00157] In some embodiments, the method of for predicting the responsiveness of a lymphoma patient to a cancer treatment further comprises (i) identifying one or more cluster classifier(s) by training a classifier model using the clustering results of a discovery dataset, (ii) applying the one or more cluster classifier(s) to a replication/validation dataset to classify the replication/validation dataset, (iii) clustering the replication/validation dataset using the clustering method and (iv) comparing the classification results of the replication/validation dataset using the cluster classifier(s) with the clustering results of the replication/validation dataset using the clustering method. In one embodiment, the model is a grouped multinomial generalized linear model (GLM). In one specific embodiment, the model is GLM using least absolute shrinkage and selection operator (LASSO). In some embodiments, the similar the classification results of the replication/validation dataset using the cluster classifier(s) with the clustering results of the replication/validation dataset using the clustering method indicates that the cluster classifier(s) is effective for classifying the replication/validation dataset. [00158] In some embodiments, the method of for predicting the responsiveness of a lymphoma patient to a cancer treatment further comprises (i) identifying one or more cluster classifier(s) by training a classifier model using the clustering results of a discovery dataset, and (ii) applying the one or more cluster classifier(s) to a replication/validation dataset to classify the replication/validation dataset. In one embodiment, the model is a grouped multinomial generalized linear model (GLM). In one specific embodiment, the model is GLM using least absolute shrinkage and selection operator (LASSO). [00159] In some embodiments, expression levels of one or more genes in the discovery dataset are used in training the classifier model. In some embodiments, one, two, three, four, five or more of the genes identified in Table 1 are used in training the classifier model. In one embodiment, all the genes identified in Table 1 are used in training the classifier model. [00160] In some embodiments, the classifier model uses one, two, three, four, five, or more of the genes identified in Table 1. In some embodiments, the classifier model uses all the genes identified in Table 1. [00161] In some embodiments, the determining step (d) applies the clustering method in step (c) to determine to which subgroup the lymphoma patient belongs to using gene expression levels in the biological sample from the lymphoma patient. [00162] In some embodiments, the predicting step (e) applies the trained classifier model to predict the responsiveness of the lymphoma patient to a first cancer treatment. In one embodiment, the predicting step (e) applies the trained GLM model to predict the responsiveness of the lymphoma patient to a first cancer treatment. [00163] In some embodiments, the lymphoma is diffuse large B-cell lymphoma (DLBCL). In some embodiments, the lymphoma is indolent B cell lymphoma. In some embodiments, the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma. In one embodiment, the lymphoma is follicular lymphoma. In another embodiment, the lymphoma is nodal marginal zone B-cell lymphoma. In yet another embodiment, the lymphoma is mantle cell lymphoma. In yet another embodiment, the lymphoma is chronic lymphocytic leukemia. [00164] In some embodiments, the first cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP). [00165] In some embodiments, the second cancer treatment is R-CHOP. In some embodiments, the second cancer treatment is not R-CHOP. [00166] In some embodiments, the second cancer treatment is a bromodomain and extra- terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor. [00167] Bromodomains (BDs) are protein modules of ~110 amino acids that recognize acetylated lysine in histones and other proteins. The bromodomain and extra-terminal (BET) family is a subset of 46 bromodomain-containing proteins found only in the human genome. BET proteins are composed of four proteins, namely bromodomain-containing protein 2 (BRD2), BRD3, BRD4 and bromodomain testis-specific protein (BRDT). See Cochran et al., Nature Reviews Drug Discovery, 2019, 18:609-628; Duan et al., MedChemComm, 2018, 9:1779- 1802. Tool compounds showing robust preclinical activity have led to advances in the development of BET inhibitors, primarily in oncology indications including, for example, DLBCL. As the chemical probes described for other BD targets, clinical BET inhibitors are acetyl lysine mimetics with a heterocyclic core that occupies the BD pocket15. Deregulation of BET proteins, in particular BRD4, has been implicated in the development of diverse diseases, especially cancers. See Duan et al., MedChemComm, 2018, 9:1779-1802. [00168] In some embodiments, the BET inhibitor is selected from the group consisting of OTX015, MK-8628, CPI-0610, BMS-986158, ZEN003694, GSK2820151, GSK525762, INCB054329, INCB057643, ODM-207, RO6870810, BAY1238097, CC-90010, AZD5153, FT- 1101, ABBV-075, ABBV-744, SF1126, GS-5829, and CPI-0610. [00169] In some embodiments, the BET inhibitor is a bromodomain-containing protein 4 (BRD4) inhibitor. In some embodiments, the BRD4 inhibitor is selected from the group consisting of OTX015, TEN-010, GSK525762, CPI-0610, I-BET151, PLX51107, INCB0543294, ABBV-075, BI 894999, BMS-986158, and AZD5153. [00170] Cyclin-dependent kinases (CDKs) are a family of serine-threonine kinases that are identified as gene products involved in cell cycle control. A close cooperation between CDKs, cyclins and cells contain endogenous inhibitors (CKIs) is necessary for orderly cell cycle progressions under regulation. Mammalian CDKs, cyclins and CKIs play important roles in other biological processes including, for example, transcriptional regulation, epigenetics, DNA damage response and repair (DDR), stemness, metabolism and angiogenesis among others. See Sanchez-Martinez et al., Bioorg. Med. Chem. Lett., 2019, 29:126637. [00171] In some embodiments, the CDK inhibitor is selected from the group consisting of PD- 0332991 (Ibrance® or Palbociclib), LEE011 (Ribociclib), LY2835219 (Verzenio® or Abemaciclib), G1T28 (Trilaciclib), G1T38 (Lerociclib), SHR-6390, Flavopiridol (Alvocidib), PHA848125 (Milciclib), BCD-115, MM-D37K, PF-06873600, TG-02 (SB-1317 or Zoriraciclib), C7001 (ICEC 0942), BEY-1107, XZP-3287 (Birociclib), BPI-16350, FCN-437, CYC-065, R- Roscovitine (CY-202 or Seliciclib), AT-7519, AGM-130 (Inditinib), FN-1501, SY-1365, AZD- 4573, TP-1287, P-1446A-05 (Voruciclib), BAY-1251152, SCH-727965 (MK-765 or Dinaciclib), BEBT-209, TQB-3616, BAY-1000394 (Roniciclib), BAY-1143572 (Atuveciclib), and AGM- 925 (FLX-925). 5.3.1. Double Hit Signature (DHITsig) Classifier [00172] It has been well described in literature that there exists a subset of high-risk DLBCL patients who are hallmarked by a so-called “double hit” (DHIT) translocation in MYC and either BCL2 or BCL6. Patients who are DHIT+ tend to have worse survival than DHIT- patients, and these patients are typically GCB with MYC+BCL2 translocations. DHIT status has commonly been measured using DNA sequencing or FISH probes to determine translocation status. [00173] Ennishi et al. derived a gene expression signature that reflects DHIT status, capturing a broader segment of the population that has differential clinical outcome. See Ennishi et al., J. Clin. Oncol., 2018, 37:190-201. Using the 104 genes, parameterization, and methodology described in the manuscript, the DHITsig method was adapted and implemented for use in the single-sample normalized RNAseq space. [00174] The Ennishi method for calculating the DHITsig score is a variable importance- weighted sum of log likelihood ratios. The likelihood function is calculated by assuming a Gaussian mixture model of gene expression that defines two normal distributions of expression for DHITsig+ and DHITsig- samples for each signature gene. The gene list and variable weights from Ennishi et al. were used, along with the mixture model distribution parameters as derived from their DESeq2-processed dataset. 5.3.2. Identifying Biological Features of Clusters of Patients and Treating [00175] In some embodiments, the reference patients in the reference patient group are clustered into 8 subgroups; and wherein: (i) subgroup D1 comprises about 55% to 65% patients having germinal center B-cell-like (GCB) DLBCL, about 20% to 30% patients having activated B-cell like (ABC) lymphoma, and about 20% to 30% patients who are DHITsig+ DLBCL patients; (ii) subgroup D2 comprises about 45% to 55% patients having GCB DLBCL, about 20% to 45% patients having ABC DLBCL, and about 20% to 25% patients who are DHITsig+ DLBCL patients; (iii) subgroup D3 comprises about 90% to 95% patients having GCB DLBCL, about 0% to 10% patients having ABC DLBCL, and about 0% to 20% patients who are DHITsig+ DLBCL patients; (iv) subgroup D4 comprises about 0% to 10% patients having GCB DLBCL, about 90% to 100% patients having ABC DLBCL, and about 0% to 10% patients who are DHITsig+ DLBCL patients; (v) subgroup D5 comprises about 0% to 20% patients having GCB DLBCL, about 55% to 65% patients having ABC DLBCL, and about 0% to 20% patients who are DHITsig+ DLBCL patients; (vi) subgroup D6 comprises about 50% to 60% patients having GCB DLBCL, about 20% to 40% patients having ABC DLBCL, and about 15% to 30% patients who are DHITsig+ DLBCL patients; (vii) subgroup D7 comprises about 20% to 35% patients having GCB DLBCL, about 45% to 55% patients having ABC DLBCL, and about 0% to 10% patients who are DHITsig+ DLBCL patients; and (viii) subgroup D8 comprises about 35% to 75% patients having GCB DLBCL, about 15% to 60% patients having ABC DLBCL, and about 25% to 65% patients who are DHITsig+ DLBCL patients. [00176] In some embodiments, when the lymphoma patient is determined to belong to subgroup D4 or D8, the method comprises predicting that the patient is not likely to be responsive to the first cancer treatment. In some embodiments, when the lymphoma patient is determined to belong to any one of subgroups D1-D3 or D5-D7, the method comprises predicting that the patient is likely to be responsive to the first cancer treatment. [00177] In some embodiments, when the lymphoma patient is determined to belong to subgroup D4 or D8, the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor. In some embodiments, (i) when the lymphoma patient is determined to belong to subgroup D4, the second cancer treatment is a cyclin dependent kinase (CDK) inhibitor; and (ii) when the lymphoma patient is determined to belong to subgroup D8, the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor. In one embodiment, when the lymphoma patient is determined to belong to subgroup D4, the second cancer treatment is a cyclin dependent kinase (CDK) inhibitor. In another embodiment, when the lymphoma patient is determined to belong to subgroup D8, the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor. [00178] In some embodiments, when the lymphoma patient is determined to belong to subgroup D8, the second cancer treatment is a BCL2 inhibitor. In some embodiments, when the lymphoma patient is determined to belong to subgroup D8, the second cancer treatment is an agent that increases FAS expression. In some embodiments, when the lymphoma patient is determined to belong to subgroup D3 or D7, the second cancer treatment is an inhibitor of human leukocyte antigen (HLA) genes. In one embodiment, the second cancer treatment is an inhibitor of HLA-A. In another embodiment, the second cancer treatment is an inhibitor of HLA-B. In yet another embodiment, the second cancer treatment is an inhibitor of HLA-C. In yet another embodiment, the second cancer treatment is an inhibitor of HLA-E. In yet another emdodiment, the second cancer treatment is an inhibitor of HLA-F. In some embodiments, when the lymphoma patient is determined to belong to subgroup D4 or D8, the second cancer treatment is a CD47 treatment. In some embodiments, when the lymphoma patient is determined to belong to subgroup D7, the second cancer treatment is an IDO inhibitor or an agent that depletes regulatory T cells. In one embodiment, the second cancer treatment is an IDO inhibitor. In another embodiment, the second cancer treatment is an agent that depletes regulatory T cells. In some embodiments, when the lymphoma patient is determined to belong to subgroup D1 or D5, the second cancer treatment is a histone deacetylase (HDAC) inhibitor. In some embodiments, when the lymphoma patient is determined to belong to subgroup D2 or D7, the second cancer treatment is a galectin-3 (Gal3) inhibitor. [00179] In some embodiments, the lymphoma patient is determined to be belong to a subgroup based on the mutational data of one or more features listed in Table 4 or Table 5. [00180] In another aspect, provided herein are methods of predicting the responsiveness of lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first patient having a lymphoma; (b) determining the level of expression of one, two, three, four, five or more of the genes identified in Table 1; (c) comparing the level of expression of the one, two, three, four, five or more of the genes identified in Table 1 in the first biological sample with the level of expression of the same genes in a second biological sample(s) from a second patient(s), wherein the second lymphoma patient(s) is responsive to the cancer treatment, and wherein the similar expression of the one, two, three, four, five or more of the genes in the first biological sample relative to the level of expression of the one, two, three, four, five or more of the genes in the second biological sample(s) indicates that the lymphoma in the first patient will be responsive to treatment with the cancer treatment. [00181] In another aspect, provided herein is a method for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first lymphoma patient; (b) determining the expression of the genes or a certain subset of genes set forth in Table 1 or any combination thereof in the first biological sample; and (c) comparing the gene expression profile of the genes or subset of genes in the first biological sample to (i) the gene expression profile of the genes or subset of genes in biological samples from lymphoma patients which are responsive to the drug and (ii) the gene expression of the genes or subset of genes in biological samples from lymphoma patients which are not responsive to the cancer treatment, wherein a gene expression profile for the genes or subset of genes in the first biological sample similar to the gene expression profile for the genes or subset of genes in biological samples from lymphoma patients which are responsive to the cancer treatment indicates that the first lymphoma patient will be responsive to the cancer treatment, and a gene expression profile for the genes or subset of genes in first biological sample similar to the gene expression profile for the genes or subset of genes in biological samples from lymphoma patients which are not responsive to the cancer treatment indicates that the first lymphoma patient will not be responsive to the cancer treatment. [00182] In some embodiments, the determining step of the methods described herein comprising determining the expression of all the genes listed in Table 1. In some embodiments, the expression levels of all the genes listed in Table 1 are determined in step (b) and compared in step (c). In some embodiments, the determining step of the methods described herein comprising determining the expression of all the genes selected from the group consisting of AARS2, ACPP, ACRC, ACTN4, ADH1B, AGER, ALAS1, AMH, ANKRD20A5P, ANKRD22, APOL3, ARHGEF1, ATM, ATP6AP2, ATP8B1, BMF, CBR3, CCDC9, CCL21, CCR1, CCT7, CD83, CDCA8, CDK12, CECR7, CEP72, CHKA, CILP, CLEC4E, CNKSR2, CORO1C, CPT2, CR2, CTSS, CTSZ, DCAF5, DENND2D, DEPDC5, DGKA, DUSP1, E2F4, EHF, ENTPD1, EPAS1, FAM90A1, FBXO6, FCER1G, FCRL1, FLVCR2, FOXO3, FXYD2, FYB, G6PD, GBP4, GDI1, GIMAP6, GINS1, GMIP, GSTM4, GZMM, HAT1, HDAC1, HES4, HLA-DQA1, HMBOX1, HSD11B1, HSP90AB1, HTN3, IL4I1, IL9R, INO80D, IRAK3, ITPKB, IVNS1ABP, KDM8, KIAA0226L, KLF12, KLHL9, LAMC3, LCOR, LGALS3, LGI2, LIF, LIPE, LY9, MAD2L2, MAN1B1, MAST3, MCM8, MPC2, MPEG1, MRPL3, MSH6, MXD1, MYOCD, NDST1, NDUFB2, NFKBIA, NLRP14, NRF1, NUDT1, ODF3B, PABPC1L, PDXK, PHLDB3, PIK3CG, PILRA, PKM, PLXNA3, POP1, PRM1, PRMT1, PSEN2, PTPN1, QTRTD1, RAD51D, RANBP6, RASGRP1, RASGRP2, RELB, RGL1, RRP1, S1PR4, SEC61B, SERPINB3, SERPINE1, SETBP1, SH3BP1, SH3BP5, SLAMF8, SLC7A5, SLC9A9, SNX10, ST6GALNAC4, SYNGAP1, TAPT1-AS1, TBC1D27, TGDS, TGFBRAP1, THBD, TLR7, TM2D1, TMEM165, TMEM175, TMEM256, TMEM63B, TMLHE, TNFRSF13B, TRAV12-2, TRAV8-3, TRIB2, TRMU, TXK, U2AF2, UNC13D, UNK, VAV1, WARS, WDR19, WSB1, ZC3H3, ZMYM2, ZNF134, ZNF165, ZNF324, ZNF333, ZNF347, ZNF524, ZNF623, ZNF765, ZNF850, and ZNF91. [00183] In some embodiments, the cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP). [00184] In another aspect, provided herein are methods of treating a lymphoma patient, comprising: (i) identifying a lymphoma patient who is likely to be responsive to the cancer treatment as predicted using the predicting methods described herein above; and (ii) administering to the lymphoma patient the cancer treatment. [00185] In another aspect, provided herein are methods of treating a lymphoma patient, comprising: (i) identifying a lymphoma patient who is likely to be not responsive to the cancer treatment as predicted using the predicting methods described herein above; and (ii) administering to the lymphoma patient an alternative cancer treatment. [00186] In some embodiments, the alternative cancer treatment is a bromodomain and extra- terminal (BET) inhibitor, a cyclin dependent kinase (CDK) inhibitor. [00187] Additionally, in some embodiments, all of the genes listed in Table 1 can be used as biomarkers to predict the responsiveness of a lymphoma (e.g., DLBCL) patient to a drug. [00188] In another aspect, the subgroups provided herein (e.g., D1-D8) can be characterized and/or identified based on Bcl6 signature scores as shown in the example section below and in FIG.12. Therefore, Bcl6 sigature score can also be used as a way to classify a patient into one of the 8 subgroups for the purpose of determining a patient’s responsiveness to a treatment. [00189] In another aspect, as shown in the example section below and in Tables 4-5, mutational data were collected and interpreted in the context of the identified subgroups. Thus, in some embodiments, the mutation profile of each subgroup (or cluster) or a subset thereof can also be used to identify a subgroup, or to classify a patient into one of the subgroups for the purpose of determining the patient’s responsiveness to a treatment. [00190] The subgroups (or clusters) provided herein were also characterized based on proportions of different T cell populations (e.g., CD8+, CD4+, CD163+, or CD20+ cells) as shown in the example section below and in FIGs.13A-13D. Therefore, in yet another aspect, proportions of different T cell populations (e.g., CD8+, CD4+, CD163+, or CD20+ cells) can be used to identify a subgroup or to classify a patient into one of the subgroups for the purpose of determining the patient’s responsiveness to a treatment. 5.4 Administration Methods [00191] In some embodiments of the various methods provided herein, a first cancer treatment compound is administered to a lymphoma patient predicted to be responsive to the first cancer treatment. Also provided herein are methods of treating patients who have been previously treated for cancer (e.g., DLBCL or a subtype thereof) but are non-responsive to a first cancer treatment (e.g., standard therapies), as well as those who have not previously been treated. The invention also encompasses methods of treating patients regardless of patient’s age, although some diseases or disorders are more common in certain age groups. The invention further encompasses methods of treating patients who have undergone surgery in an attempt to treat the disease or condition at issue, as well as those who have not. Because patients with cancer have heterogeneous clinical manifestations and varying clinical outcomes, the treatment given to a patient may vary, depending on his/her prognosis. The skilled clinician will be able to readily determine without undue experimentation specific secondary agents, types of surgery, and types of non-drug based standard therapy that can be effectively used to treat an individual patient with cancer (e.g., DLBCL or a subtype thereof). [00192] In certain embodiments, a therapeutically or prophylactically effective amount of the cancer treatment is from about 0.005 to about 1,000 mg per day, from about 0.01 to about 500 mg per day, from about 0.01 to about 250 mg per day, from about 0.01 to about 100 mg per day, from about 0.1 to about 100 mg per day, from about 0.5 to about 100 mg per day, from about 1 to about 100 mg per day, from about 0.01 to about 50 mg per day, from about 0.1 to about 50 mg per day, from about 0.5 to about 50 mg per day, from about 1 to about 50 mg per day, from about 0.02 to about 25 mg per day, or from about 0.05 to about 10 mg per day. [00193] In certain embodiments, the therapeutically or prophylactically effective amount is about 0.1, about 0.2, about 0.5, about 1, about 2, about 5, about 10, about 15, about 20, about 25, about 30, about 40, about 45, about 50, about 60, about 70, about 80, about 90, about 100, or about 150 mg per day. [00194] In one embodiment, the recommended daily dose range of the cancer treatment for the conditions described herein lie within the range of from about 0.1 mg to about 50 mg per day, preferably given as a single once-a-day dose, or in divided doses throughout a day. In some embodiments, the dosage ranges from about 1 mg to about 50 mg per day. In other embodiments, the dosage ranges from about 0.5 mg to about 5 mg per day. Specific doses per day include 0.1, 0.2, 0.5, 1, 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, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 mg per day. [00195] In a specific embodiment, the recommended starting dosage may be 0.5, 1, 2, 3, 4, 5, 10, 15, 20, 25, or 50 mg per day. In another embodiment, the recommended starting dosage may be 0.5, 1, 2, 3, 4, or 5 mg per day. The dose may be escalated to 10, 15, 20, 25, 30, 35, 40, 45, or 50 mg per day. [00196] In certain embodiments, the therapeutically or prophylactically effective amount is from about 0.001 to about 100 mg/kg/day, from about 0.01 to about 50 mg/kg/day, from about 0.01 to about 25 mg/kg/day, from about 0.01 to about 10 mg/kg/day, from about 0.01 to about 9 mg/kg/day, 0.01 to about 8 mg/kg/day, from about 0.01 to about 7 mg/kg/day, from about 0.01 to about 6 mg/kg/day, from about 0.01 to about 5 mg/kg/day, from about 0.01 to about 4 mg/kg/day, from about 0.01 to about 3 mg/kg/day, from about 0.01 to about 2 mg/kg/day, or from about 0.01 to about 1 mg/kg/day. [00197] The administered dose can also be expressed in units other than mg/kg/day. For example, doses for parenteral administration can be expressed as mg/m2/day. One of ordinary skill in the art would readily know how to convert doses from mg/kg/day to mg/m2/day to given either the height or weight of a subject or both (see, www.fda.gov/cder/cancer/animalframe.htm). For example, a dose of 1 mg/kg/day for a 65 kg human is approximately equal to 38 mg/m2/day. [00198] In certain embodiments, the amount of the cancer treatment administered is sufficient to provide a plasma concentration of the compound at steady state, ranging from about 0.001 to about 500 μM, about 0.002 to about 200 μM, about 0.005 to about 100 μM, about 0.01 to about 50 μM, from about 1 to about 50 μM, about 0.02 to about 25 μM, from about 0.05 to about 20 μM, from about 0.1 to about 20 μM, from about 0.5 to about 20 μM, or from about 1 to about 20 μM. [00199] In other embodiments, the amount of the cancer treatment administered is sufficient to provide a plasma concentration of the compound at steady state, ranging from about 5 to about 100 nM, about 5 to about 50 nM, about 10 to about 100 nM, about 10 to about 50 nM, or from about 50 to about 100 nM. [00200] As used herein, the term “plasma concentration at steady state” is the concentration reached after a period of administration of a cancer treatment provided herein. Once steady state is reached, there are minor peaks and troughs on the time-dependent curve of the plasma concentration of the cancer treatment. [00201] In certain embodiments, the amount of the cancer treatment administered is sufficient to provide a maximum plasma concentration (peak concentration) of the compound, ranging from about 0.001 to about 500 μM, about 0.002 to about 200 μM, about 0.005 to about 100 μM, about 0.01 to about 50 μM, from about 1 to about 50 μM, about 0.02 to about 25 μM, from about 0.05 to about 20 μM, from about 0.1 to about 20 μM, from about 0.5 to about 20 μM, or from about 1 to about 20 μM. [00202] In certain embodiments, the amount of the cancer treatment administered is sufficient to provide a minimum plasma concentration (trough concentration) of the compound, ranging from about 0.001 to about 500 μM, about 0.002 to about 200 μM, about 0.005 to about 100 μM, about 0.01 to about 50 μM, from about 1 to about 50 μM, about 0.01 to about 25 μM, from about 0.01 to about 20 μM, from about 0.02 to about 20 μM, from about 0.02 to about 20 μM, or from about 0.01 to about 20 μM. [00203] In certain embodiments, the amount of the cancer treatment administered is sufficient to provide an area under the curve (AUC) of the compound, ranging from about 100 to about 100,000 ng*hr/mL, from about 1,000 to about 50,000 ng*hr/mL, from about 5,000 to about 25,000 ng*hr/mL, or from about 5,000 to about 10,000 ng*hr/mL. [00204] In certain embodiments, the lymphoma patient to be treated with one of the methods provided herein has not been treated with anticancer therapy prior to the administration of a standard therapy (e.g., R-CHOP). In certain embodiments, the lymphoma patient to be treated with one of the methods provided herein has been treated with anticancer therapy (standard therapies, e.g., R-CHOP) prior to the administration of the second treatment. In certain embodiments, the lymphoma patient to be treated with one of the methods provided herein has developed drug resistance to the first cancer treatment. [00205] Depending on the subtype of lymphoma (e.g., DLBCL) to be treated and the subject’s condition, the cancer treatment may be administered by parenteral (e.g., intramuscular, intraperitoneal, intravenous, CIV, intracistemal injection or infusion, subcutaneous injection, or implant), inhalation, nasal, vaginal, rectal, sublingual, or topical (e.g., transdermal or local) routes of administration. The cancer treatment may be formulated, alone or together, in suitable dosage unit with pharmaceutically acceptable excipients, carriers, adjuvants, and vehicles, appropriate for each route of administration. [00206] In one embodiment, the cancer treatment is administered parenterally. In another embodiment, the cancer treatment is administered intravenously. [00207] Depending on the state of the lymphoma to be treated and the subject’s condition, in one embodiment, the trewatment compound may be administered by oral, parenteral (e.g., intramuscular, intraperitoneal, intravenous, CIV, intracistemal injection or infusion, subcutaneous injection, or implant), inhalation, nasal, vaginal, rectal, sublingual, or topical (e.g., transdermal or local) routes of administration. In one embodiment, the treatment compound may be formulated, alone or together, in suitable dosage unit with pharmaceutically acceptable excipients, carriers, adjuvants and vehicles, appropriate for each route of administration. [00208] In one embodiment, the treatment compound is administered orally. In another embodiment, the treatment compound is administered parenterally. In yet another embodiment, the treatment compound is administered intravenously. [00209] In one embodiment, the treatment compound can be delivered as a single dose such as, e.g., a single bolus injection, or oral capsules, tablets or pills; or over time, such as, e.g., continuous infusion over time or divided bolus doses over time. The cancer treatment as described herein can be administered repeatedly if necessary, for example, until the patient experiences stable disease or regression, or until the patient experiences disease progression or unacceptable toxicity. [00210] In one embodiment, the treatment compound can be administered once daily (QD), or divided into multiple daily doses such as twice daily (BID), three times daily (TID), and four times daily (QID). In addition, the administration can be continuous (i.e., daily for consecutive days or every day), intermittent, e.g., in cycles (i.e., including days, weeks, or months of rest without drug). As used herein, the term “daily” is intended to mean that a therapeutic compound is administered once or more than once each day, for example, for a period of time. The term “continuous” is intended to mean that a therapeutic compound is administered daily for an uninterrupted period of at least 7 days to 52 weeks. The term “intermittent” or “intermittently” as used herein is intended to mean stopping and starting at either regular or irregular intervals. For example, intermittent administration is administration for one to six days per week, administration in cycles (e.g., daily administration for two to eight consecutive weeks, then a rest period with no administration for up to one week), or administration on alternate days. The term “cycling” as used herein is intended to mean that a therapeutic compound is administered daily or continuously but with a rest period. [00211] In some embodiments, the frequency of administration is in the range of about a daily dose to about a monthly dose. [00212] The cancer treatment can be delivered as a single dose (e.g., a single bolus injection), or over time (e.g., continuous infusion over time or divided bolus doses over time). The compound can be administered repeatedly if necessary, for example, until the patient experiences stable disease or regression, or until the patient experiences disease progression or unacceptable toxicity. For example, stable disease for solid cancers generally means that the perpendicular diameter of measurable lesions has not increased by 25% or more from the last measurement. Therasse et al., J. Natl. Cancer Inst., 2000, 92(3):205-216. Stable disease or lack thereof is determined by methods known in the art such as evaluation of patient symptoms, physical examination, and visualization of the tumor that has been imaged using X-ray, CAT, PET, MRI scan, or other commonly accepted evaluation modalities. [00213] The cancer treatment can be administered once daily (QD) or divided into multiple daily doses such as twice daily (BID), three times daily (TID), and four times daily (QID). In addition, the administration can be continuous (i.e., daily for consecutive days or every day) or intermittent, e.g., in cycles (i.e., including days, weeks, or months of rest without drug). As used herein, the term “daily” is intended to mean that a cancer treatment is administered once or more than once each day, for example, for a period of time. The term “continuous” is intended to mean that the cancer treatment is administered daily for an uninterrupted period of at least 10 days to 52 weeks. The term “intermittent” or “intermittently” as used herein is intended to mean stopping and starting at either regular or irregular intervals. For example, intermittent administration of the cancer treatment is administration for one to six days per week, administration in cycles (e.g., daily administration for two to eight consecutive weeks, then a rest period with no administration for up to one week), or administration on alternate days. The term “cycling” as used herein is intended to mean that the cancer treatment is administered daily or continuously but with a rest period. In certain embodiments, the rest period is the same length as the treatment period. In other embodiments, the rest period has different length from the treatment period. In some embodiments, the length of cycling is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 weeks. In some embodiments of cycling, the cancer treatment is administered daily for a period of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, or 30 days, followed by a rest period. In a particular embodiment, the cancer treatment is administered daily for a period of 5 days of a 4-week cycle. In another particular embodiment, the cancer treatment is administered daily for a period of 10 days of a 4-week cycle. [00214] In some embodiments, the frequency of administration is in the range of about a daily dose to about a monthly dose. In certain embodiments, administration is once a day, twice a day, three times a day, four times a day, once every other day, twice a week, once every week, once every two weeks, once every three weeks, or once every four weeks. In one embodiment, the cancer treatment is administered once a day. In another embodiment, the cancer treatment is administered twice a day. In yet another embodiment, the cancer treatment is administered three times a day. In still another embodiment, the cancer treatment is administered four times a day. [00215] In certain embodiments, the cancer treatment is administered once per day from one day to six months, from one week to three months, from one week to four weeks, from one week to three weeks, or from one week to two weeks. In certain embodiments, the cancer treatment is administered once per day for one week, two weeks, three weeks, or four weeks. In one embodiment, the cancer treatment is administered once per day for one week. In another embodiment, the cancer treatment is administered once per day for two weeks. In yet another embodiment, the cancer treatment is administered once per day for three weeks. In still another embodiment, the cancer treatment is administered once per day for four weeks. 5.5 Combination Therapy [00216] One or more additional therapies, such as additional active ingredients or agents, that can be used in combination with the administration of a cancer treatment described herein to treat a lymphoma patient (e.g., a patient having DLBCL). The one or more additional therapies can be administered prior to, concurrently with, or subsequent to the administration of the compound described herein. Administration of a cancer treatment described herein and an additional active agent (“second active agents”) to a patient can occur simultaneously or sequentially by the same or different routes of administration. The suitability of a particular route of administration employed for a particular active agent will depend on the active agent itself (e.g., whether it can be administered orally without decomposing prior to entering the blood stream) and the condition of lymphoma (e.g., DLBCL) being treated. Routes of administration for the additional active agents or ingredients are known to those of ordinary skill in the art. See, e.g., Physicians’ Desk Reference. [00217] In certain embodiments, the cancer treatment described herein and an additional active agent are cyclically administered to a patient with lymphoma (e.g., DLBCL). Cycling therapy involves the administration of an active agent for a period of time, followed by a rest for a period of time, and repeating this sequential administration. Cycling therapy can reduce the development of resistance to one or more of the therapies, avoid or reduce the side effects of one of the therapies, and/or improves the efficacy of the treatment. [00218] One or more second active ingredients or agents can be used in the methods and compositions provided herein. Second active agents can be large molecules (e.g., proteins) or small molecules (e.g., synthetic inorganic, organometallic, or organic molecules). Various agents can be used, such as those described in U.S. Patent Application No.16/390,815 or U.S. Provisional Application entitled, "SUBSTITUTED 4-AMINOISOINDOLINE-1,3-DIONE COMPOUNDS AND SECOND ACTIVE AGENTS FOR COMBINED USE," filed on even date herewith (Attorney Docket No.14247-390-888), each of which is incorporated herein by reference in their entirety. Exemplary second active agents include, but are not limited to, an HDAC inhibitor (e.g., panobinostat, romidepsin, or vorinostat), a BCL2 inhibitor (e.g., venetoclax), a BTK inhibitor (e.g., ibrutinib or acalabrutinib), an mTOR inhibitor (e.g., everolimus), a PI3K inhibitor (e.g., idelalisib), a PKCβ inhibitor (e.g., enzastaurin), a SYK inhibitor (e.g., fostamatinib), a JAK2 inhibitor (e.g., fedratinib, pacritinib, ruxolitinib, baricitinib, gandotinib, lestaurtinib, or momelotinib), an Aurora A kinase inhibitor (e.g., alisertib), an EZH2 inhibitor (e.g., tazemetostat, GSK126, CPI-1205, 3-deazaneplanocin A, EPZ005687, EI1, UNC1999, or sinefungin), a BET inhibitor (e.g., birabresib or 4-[2-(cyclopropylmethoxy)-5- (methanesulfonyl)phenyl]-2-methylisoquinolin-1(2H)-one), a hypomethylating agent (e.g., 5- azacytidine or decitabine), a chemotherapy (e.g., bendamustine, doxorubicin, etoposide, methotrexate, cytarabine, vincristine, ifosfamide, or melphalan), or an epigenetic compound (e.g., a DOT1L inhibitor such as pinometostat, a HAT inhibitor such as C646, a WDR5 inhibitor such as OICR-9429, a HDAC6 inhibitor such as ACY-241, a DNMT1 selective inhibitor such as GSK3484862, a LSD-1 inhibitor such as Compound C or seclidemstat, a G9A inhibitor such as UNC 0631, a PRMT5 inhibitor such as GSK3326595, a BRPF1B/2 inhibitor such as OF-1, a BRD9/7 inhibitor such as LP99, a SUV420H1/H2 inhibitor such as A-196, a Menin-MLL inhibitor such as MI-503, a CARM1 inhibitor such as EZM2302, a BRD9 such as an inhibitor dBrd9), aiolos/ikaros degrading cereblon E3 ligase modulator (CELMoDs), CREBBp2 inhibitors, anti-CD79b antibody, CD19 CAR-T, inhibitors of p53 (nutlins), Bcl6 inhibitors, CREBBp2 CELMoDs, CD79b CELMoDs, CD19 CELMoDs, p53 (nutlins) CELMoDs, Bcl6 CELMoDs, inhibitors of ligand directed degradation (LDD) of CREBBP2, inhibitors of LDD of CD79b, inhibitors of LDD of CD19, inhibitors of LDD of p53(nutlins), inhibitors of LDD of Bcl6, inhibitors of LDD of CK1a, inhibitors of LDD of IRAK4 (e.g., in MYD88 L265p lymphoma), MALT1 inhibitors such as JNJ-67856633, MAT2A inhibitors (e.g., for 9p21 deletions), anti-CD3 x anti-CD19 bispecific antibodies, and anti-CD3 x anti-CD20 bispecific antibodies. [00219] In some embodiments of the methods described herein the methods further include administration of one or more of rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone, etoposide, Bendamustine (Treanda), lenalidomide, or gemcitabine. In some embodiments of the methods described herein the methods further include administration of one or more of rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone, etoposide, Bendamustine (Treanda), or gemcitabine. In some embodiments of the methods described herein, the treatment further includes treatment with one or more of R-CHOP (rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone), R EPOCH (etoposide, rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone), stem cell transplant, Bendamustine (Treanda) plus rituximab, rituximab, lenalidomide plus rituximab, or gemcitabine- based combinations. In some embodiments of the methods described herein, the treatment further includes treatment with one or more of R-CHOP (rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone), R EPOCH (etoposide, rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone), stem cell transplant, Bendamustine (Treanda) plus rituximab, rituximab, or gemcitabine-based combinations. In certain embodiments, the second active agent is rituximab, as provided in U.S. Provisional Application 62/833,432. [00220] In one embodiment, the second active agent used in the methods provided herein is a histone deacetylase (HDAC) inhibitor. In one embodiment, the HDAC inhibitor is panobinostat, romidepsin, or vorinostat, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. [00221] In one embodiment, the second active agent used in the methods provided herein is a B-cell lymphoma 2 (BCL2) inhibitor. In one embodiment, the BCL2 inhibitor is venetoclax, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the BCL2 inhibitor is venetoclax. [00222] In one embodiment, the second active agent used in the methods provided herein is a Bruton’s tyrosine kinase (BTK) inhibitor. In one embodiment, the BTK inhibitor is ibrutinib, or acalabrutinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the BTK inhibitor is ibrutinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the BTK inhibitor is ibrutinib. In one embodiment, the BTK inhibitor is acalabrutinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the BTK inhibitor is acalabrutinib. [00223] In one embodiment, the second active agent used in the methods provided herein is a mammalian target of rapamycin (mTOR) inhibitor. In one embodiment, the mTOR inhibitor is rapamycin or an analog thereof (also termed rapalog). In one embodiment, the mTOR inhibitor is everolimus, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the mTOR inhibitor is everolimus. [00224] In one embodiment, the second active agent used in the methods provided herein is a phosphoinositide 3-kinase (PI3K) inhibitor. In one embodiment, the PI3K inhibitor is idelalisib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the PI3K inhibitor is idelalisib. [00225] In one embodiment, the second active agent used in the methods provided herein is a protein kinase C beta (PKCβ or PKC-β) inhibitor. In one embodiment, the PKCβ inhibitor is enzastaurin, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the PKCβ inhibitor is enzastaurin. In one embodiment, the PKCβ inhibitor is a pharmaceutically acceptable salt of enzastaurin. In one embodiment, the PKCβ inhibitor is a hydrochloride salt of enzastaurin. In one embodiment, the PKCβ inhibitor is a bis-hydrochloride salt of enzastaurin. [00226] In one embodiment, the second active agent used in the methods provided herein is a spleen tyrosine kinase (SYK) inhibitor. In one embodiment, the SYK inhibitor is fostamatinib, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the SYK inhibitor is fostamatinib. In one embodiment, the SYK inhibitor is a pharmaceutically acceptable salt of fostamatinib. In one embodiment, the SYK inhibitor is fostamatinib disodium hexahydrate. [00227] In one embodiment, the second active agent used in the methods provided herein is a Janus kinase 2 (JAK2) inhibitor. In one embodiment, the JAK2 inhibitor is fedratinib, pacritinib, ruxolitinib, baricitinib, gandotinib, lestaurtinib, or momelotinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. [00228] In one embodiment, the JAK2 inhibitor is fedratinib, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the JAK2 inhibitor is fedratinib. [00229] In one embodiment, the JAK2 inhibitor is pacritinib, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the JAK2 inhibitor is pacritinib. [00230] In one embodiment, the JAK2 inhibitor is ruxolitinib, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the JAK2 inhibitor is ruxolitinib. In one embodiment, the JAK2 inhibitor is a pharmaceutically acceptable salt of ruxolitinib. In one embodiment, the JAK2 inhibitor is ruxolitinib phosphate. [00231] In one embodiment, the second active agent used in the methods provided herein is an aurora A kinase inhibitor. In one embodiment, the aurora A kinase inhibitor is alisertib, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the aurora A kinase inhibitor is alisertib. [00232] In one embodiment, the second active agent used in the methods provided herein is an enhancer of zeste homolog 2 (EZH2) inhibitor. In one embodiment, the EZH2 inhibitor is tazemetostat, GSK126, CPI-1205, 3-deazaneplanocin A (DZNep), EPZ005687, EI1, UNC1999, or sinefungin, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. [00233] In one embodiment, the EZH2 inhibitor is tazemetostat, or a tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the EZH2 inhibitor is tazemetostat. [00234] In one embodiment, the EZH2 inhibitor is GSK126, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the EZH2 inhibitor is GSK126 (also known as GSK-2816126). [00235] In one embodiment, the EZH2 inhibitor is CPI-1205, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the EZH2 inhibitor is CPI-1205. [00236] In one embodiment, the EZH2 inhibitor is 3-deazaneplanocin A. In one embodiment, the EZH2 inhibitor is EPZ005687. In one embodiment, the EZH2 inhibitor is EI1. In one embodiment, the EZH2 inhibitor is UNC1999. In one embodiment, the EZH2 inhibitor is sinefungin. [00237] In one embodiment, the second active agent used in the methods provided herein is a hypomethylating agent. In one embodiment, the hypomethylating agent is 5-azacytidine or decitabine, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. [00238] In one embodiment, the hypomethylating agent is 5-azacytidine, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the hypomethylating agent is 5-azacytidine. [00239] In one embodiment, the hypomethylating agent is decitabine, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, or pharmaceutically acceptable salt thereof. In one embodiment, the hypomethylating agent is decitabine. [00240] In one embodiment, the second active agent used in the methods provided herein is a chemotherapy. In one embodiment, the chemotherapy is bendamustine, doxorubicin, etoposide, methotrexate, cytarabine, vincristine, ifosfamide, or melphalan, or a stereoisomer, mixture of stereoisomers, tautomer, isotopolog, prodrug, or pharmaceutically acceptable salt thereof. [00241] In certain embodiments, the second therapeutic agent is administered before, after or simultaneously with a cancer treatment described herein. Administration of a cancer treatment described herein and a second therapeutic agent to a patient can occur simultaneously or sequentially by the same or different routes of administration. The suitability of a particular route of administration employed for a particular second drug or agent will depend on the second therapeutic agent itself (e.g., whether it can be administered orally or topically without decomposition prior to entering the blood stream) and the subject being treated. Particular routes of administration for the second drug or agents or ingredients are known to those of ordinary skill in the art. See, e.g., The Merck Manual, 448 (17th ed., 1999). [00242] Any combination of the above therapeutic agents, suitable for treatment of the diseases or symptoms thereof, can be administered. Such therapeutic agents can be administered in any combination at the same time or as a separate course of treatment. [00243] As used herein, the term “in combination” does not restrict the order in which therapies (e.g., prophylactic and/or therapeutic agents) are administered to a patient with a disease or disorder. Administration of a second active agent provided herein, to a patient can occur simultaneously or sequentially by the same or different routes of administration. The suitability of a particular route of administration employed for a particular active agent will depend on the active agent itself (e.g., whether it can be administered orally without decomposing prior to entering the blood stream). 5.6 Pharmaceutical Compositions [00244] In certain embodiments, the cancer treatment provides herein and/or the additional active agent provided herein are formulated in a pharmaceutical composition, and the method provide herein comprises administering a lymphoma (e.g., DLBCL) patient a pharmaceutical composition comprising the cancer treatment. [00245] In some embodiments, the pharmaceutical compositions provided herein contain therapeutically effective amounts of one or more of the cancer treatment provided herein and a pharmaceutically acceptable carrier, diluents, or excipient. In some embodiments, the compounds may be formulated as the sole pharmaceutically active ingredient in the composition or may be combined with other active ingredients. [00246] The cancer treatment provided herein can be formulated into suitable pharmaceutical compositions for different routes of administration, such as injection, sublingual and buccal, rectal, vaginal, ocular, otic, nasal, inhalation, nebulization, cutaneous, or transdermal. Typically, the compounds described above are formulated into pharmaceutical compositions using techniques and procedures well known in the art (see, e.g., Ansel, Introduction to Pharmaceutical Dosage Forms, (7th ed.1999)). [00247] In the compositions, effective concentrations of one or more compounds or pharmaceutically acceptable salts are mixed with a suitable pharmaceutical carrier or vehicle. In certain embodiments, the concentrations of the compounds in the compositions are effective for delivery of an amount, upon administration, that treats, prevents, or ameliorates one or more of the symptoms and/or progression of lymphoma (e.g., DLBCL). [00248] The active compound is in an amount sufficient to exert a therapeutically useful effect in the absence of undesirable side effects on the patient treated. The therapeutically effective concentration may be determined empirically by testing the compounds in in vitro and in vivo systems described herein and then extrapolated therefrom for dosages for humans. The concentration of active compound in the pharmaceutical composition will depend on absorption, tissue distribution, inactivation, and excretion rates of the active compound, the physicochemical characteristics of the compound, the dosage schedule, and amount administered as well as other factors known to those of skill in the art. [00249] The pharmaceutically therapeutically active compounds and salts thereof are formulated and administered in unit dosage forms or multiple dosage forms. Unit dose forms as used herein refer to physically discrete units suitable for human and animal subjects and packaged individually as is known in the art. Each unit dose contains a predetermined quantity of the therapeutically active compound sufficient to produce the desired therapeutic effect, in association with the required pharmaceutical carriers, vehicles, or diluents. Examples of unit dose forms include ampoules and syringes and individually packaged tablets or capsules. Unit dose forms may be administered in fractions or multiples thereof. A multiple dose form is a plurality of identical unit dosage forms packaged in a single container to be administered in segregated unit dose form. Examples of multiple dose forms include vials, bottles of tablets or capsules, or bottles of pints or gallons. Hence, multiple dose form is a multiple of unit doses which are not segregated in packaging. [00250] It is understood that the precise dosage and duration of treatment is a function of the disease being treated and may be determined empirically using known testing protocols or by extrapolation from in vivo or in vitro test data. It is to be noted that concentrations and dosage values may also vary with the severity of the condition to be alleviated. It is to be further understood that for any particular subject, specific dosage regimens should be adjusted over time according to the individual need and the professional judgment of the person administering or supervising the administration of the compositions, and that the concentration ranges set forth herein are exemplary only and are not intended to limit the scope or practice of the claimed compositions. [00251] Solutions or suspensions used for parenteral, intradermal, subcutaneous, or topical application can include any of the following components: a sterile diluents (such as water, saline solution, fixed oil, polyethylene glycol, glycerine, propylene glycol, dimethyl acetamide, or other synthetic solvent), antimicrobial agents (such as benzyl alcohol and methyl parabens), antioxidants (such as ascorbic acid and sodium bisulfate), chelating agents (such as ethylenediaminetetraacetic acid (EDTA)), buffers (such as acetates, citrates, and phosphates), and agents for the adjustment of tonicity (such as sodium chloride or dextrose). Parenteral preparations can be enclosed in ampoules, pens, disposable syringes, or single or multiple dose vials made of glass, plastic, or other suitable material. [00252] Sustained-release preparations can also be prepared. Suitable examples of sustained- release preparations include semipermeable matrices of solid hydrophobic polymers containing the compound provided herein, which matrices are in the form of shaped articles, e.g., films or microcapsule. Examples of sustained-release matrices include iontophoresis patches, polyesters, hydrogels (for example, poly(2-hydroxyethyl-methacrylate) or poly(vinylalcohol)), polylactides, copolymers of L-glutamic acid and ethyl-L-glutamate, non-degradable ethylene-vinyl acetate, degradable lactic acid-glycolic acid copolymers such as LUPRON DEPOT™ (injectable microspheres composed of lactic acid-glycolic acid copolymer and leuprolide acetate), and poly-D-(-)-3-hydroxybutyric acid. While polymers such as ethylene-vinyl acetate and lactic acid-glycolic acid enable release of molecules for over 100 days, certain hydrogels release proteins for shorter time periods. When encapsulated compound remain in the body for a long time, they may denature or aggregate as a result of exposure to moisture at 37 oC, resulting in a loss of biological activity and possible changes in their structure. Rational strategies can be devised for stabilization depending on the mechanism of action involved. For example, if the aggregation mechanism is discovered to be intermolecular di-sulfate bond formation through thio-disulfide interchange, stabilization may be achieved by modifying sulfhydryl residues, lyophilizing from acidic solutions, controlling moisture content, using appropriate additives, and developing specific polymer matrix compositions. [00253] Further encompassed are anhydrous pharmaceutical compositions and dosage forms containing a compound provided herein. Anhydrous pharmaceutical compositions and dosage forms provided herein can be prepared using anhydrous or low moisture containing ingredients and low moisture or low humidity conditions, as known by those skilled in the art. An anhydrous pharmaceutical composition should be prepared and stored such that its anhydrous nature is maintained. Accordingly, anhydrous compositions are packaged using materials known to prevent exposure to water such that they can be included in suitable formulatory kits. Examples of suitable packaging include, but are not limited to, hermetically sealed foils, plastics, unit dose containers (e.g., vials), blister packs, and strip packs. [00254] Dosage forms or compositions containing active ingredient in the range of 0.001% to 100% with the balance made up from non-toxic carrier may be prepared. In some embodiments, the contemplated compositions contain from about 0.005% to about 95% active ingredient. In other embodiments, the contemplated compositions contain from about 0.01% to about 90% active ingredient. In certain embodiments, the contemplated compositions contain from about 0.1% to about 85% active ingredient. In other embodiments, the contemplated compositions contain from about 0.1% to about 75-95% active ingredient. [00255] Pharmaceutically acceptable carriers used in parenteral preparations include aqueous vehicles, nonaqueous vehicles, antimicrobial agents, isotonic agents, buffers, antioxidants, local anesthetics, suspending and dispersing agents, emulsifying agents, sequestering or chelating agents, and other pharmaceutically acceptable substances. [00256] Examples of aqueous vehicles include sodium chloride injection, Ringer’s injection, isotonic dextrose injection, sterile water injection, dextrose and lactated Ringer’s injection. Nonaqueous parenteral vehicles include fixed oils of vegetable origin, such as cottonseed oil, corn oil, sesame oil, and peanut oil. Antimicrobial agents in bacteriostatic or fungistatic concentrations must be added to parenteral preparations packaged in multiple dose containers, which include phenols or cresols, mercurials, benzyl alcohol, chlorobutanol, methyl and propyl-p-hydroxybenzoic acid esters, thimerosal, benzalkonium chloride, and benzethonium chloride. Isotonic agents include sodium chloride and dextrose. Buffers include phosphate and citrate. Antioxidants include sodium bisulfate. Local anesthetics include procaine hydrochloride. Suspending and dispersing agents include sodium carboxymethylcelluose, hydroxypropyl methylcellulose and polyvinylpyrrolidone. Emulsifying agents include Polysorbate 80 (TWEEN® 80). A sequestering or chelating agent of metal ions includes EDTA. Pharmaceutical carriers also include ethyl alcohol, polyethylene glycol and propylene glycol for water miscible vehicles, and sodium hydroxide, hydrochloric acid, citric acid, or lactic acid for pH adjustment. [00257] Injectables are designed for local and systemic administration. Typically, a therapeutically effective dosage is formulated to contain a concentration of at least about 0.1% w/w up to about 90% w/w or more, such as more than 1% w/w of the active compound to the treated tissue(s). The active ingredient may be administered at once, or may be divided into a number of smaller doses to be administered at intervals of time. It is understood that the precise dosage and duration of treatment is a function of the tissue being treated and may be determined empirically using known testing protocols or by extrapolation from in vivo or in vitro test data. It is to be noted that concentrations and dosage values may also vary with the age of the individual treated. It is to be further understood that for any particular subject, specific dosage regimens should be adjusted over time according to the individual need and the professional judgment of the person administering or supervising the administration of the formulations, and that the concentration ranges set forth herein are exemplary only and are not intended to limit the scope or practice of the claimed formulations. [00258] Of interest herein are also lyophilized powders, which can be reconstituted for administration as solutions, emulsions, and other mixtures. They may also be reconstituted and formulated as solids or gels. [00259] The sterile, lyophilized powder is prepared by dissolving a compound provided herein, or a pharmaceutically acceptable salt thereof, in a suitable solvent. The solvent may contain an excipient which improves the stability or other pharmacological component of the powder or reconstituted solution, prepared from the powder. Excipients that may be used include, but are not limited to, dextrose, sorbital, fructose, corn syrup, xylitol, glycerin, glucose, sucrose, or other suitable agent. The solvent may also contain a buffer, such as citrate, phosphate, or other buffers known to those of skill in the art. Subsequent sterile filtration of the solution followed by lyophilization under standard conditions known to those of skill in the art provides the desired formulation. Generally, the resulting solution will be apportioned into vials for lyophilization. Each vial will contain a single dosage or multiple dosages of the compound. The lyophilized powder can be stored under appropriate conditions, such as at about 4 oC to room temperature. [00260] In one aspect, the lyophilized formulations are suitable for reconstitution with a suitable diluent to the appropriate concentration prior to administration. In one embodiment, the lyophilized formulation is stable at room temperature. In one embodiment, the lyophilized formulation is stable at room temperature for up to about 24 months. In one embodiment, the lyophilized formulation is stable at room temperature for up to about 24 months, up to about 18 months, up to about 12 months, up to about 6 months, up to about 3 months or up to about 1 month. In one embodiment, the lyophilized formulation is stable upon storage under accelerated condition of 40 °C/75% RH for up to about 12 months, up to about 6 months or up to about 3 months. [00261] In some embodiments, the lyophilized formulation is suitable for reconstitution with an aqueous solution for intravenous administrations. In certain embodiments, the lyophilized formulation provided herein is suitable for reconstitution with water. In one embodiment, the reconstituted aqueous solution is stable at room temperature for up to about 24 hours upon reconsititution. In one embodiment, the reconstituted aqueous solution is stable at room temperature from about 1-24, 2-20, 2-15, 2-10 hours upon reconsititution. In one embodiment, the reconstituted aqueous solution is stable at room temperature for up to about 20, 15, 12, 10, 8, 6, 4 or 2 hours upon reconsititution. In certain embodiments, the lyophilized formulations upon reconstitution have a pH of about 4 to 5. [00262] Active ingredients provided herein can be administered by controlled release means or by delivery devices that are well known to those of ordinary skill in the art. Examples include, but are not limited to, those described in U.S. Patent Nos.: 3,845,770, 3,916,899, 3,536,809, 3,598,123, 4,008,719, 5,674,533, 5,059,595, 5,591,767, 5,120,548, 5,073,543, 5,639,476, 5,354,556, 5,639,480, 5,733,566, 5,739,108, 5,891,474, 5,922,356, 5,972,891, 5,980,945, 5,993,855, 6,045,830, 6,087,324, 6,113,943, 6,197,350, 6,248,363, 6,264,970, 6,267,981, 6,376,461, 6,419,961, 6,589,548, 6,613,358, 6,699,500, and 6,740,634, each of which is incorporated herein by reference. Such dosage forms can be used to provide slow or controlled- release of one or more active ingredients using, for example, hydropropylmethyl cellulose, other polymer matrices, gels, permeable membranes, osmotic systems, multilayer coatings, microparticles, liposomes, microspheres, or a combination thereof, to provide the desired release profile in varying proportions. Suitable controlled-release formulations known to those of ordinary skill in the art, including those described herein, can be readily selected for use with the active ingredients provided herein. 5.7 Biological Samples [00263] In certain embodiments, the various methods provided herein use samples (e.g., biological samples) from lymphoma (e.g., DLBCL) patients. The patient can be male or female, and can be an adult, child or infant. Samples can be analyzed at a time during an active phase of lymphoma (e.g., DLBCL), or when lymphoma (e.g., DLBCL) is inactive. In one embodiment, a sample is obtained from a patient prior, concurrently with and/or subsequent to administration of a drug described herein. In a specific embodiment, a sample is obtained from a patient prior to administration of a drug described herein. In certain embodiments, more than one sample from a patient can be obtained. [00264] In certain embodiments, the sample used in the methods provided herein comprises body fluids from a subject. Non-limiting examples of body fluids include blood (e.g., peripheral whole blood, peripheral blood), blood plasma, amniotic fluid, aqueous humor, bile, cerumen, cowper’s fluid, pre-ejaculatory fluid, chyle, chyme, female ejaculate, interstitial fluid, lymph, menses, breast milk, mucus, pleural fluid, pus, saliva, sebum, semen, serum, sweat, tears, urine, vaginal lubrication, vomit, water, feces, internal body fluids, including cerebrospinal fluid surrounding the brain and the spinal cord, synovial fluid surrounding bone joints, intracellular fluid is the fluid inside cells, and vitreous humour the fluids in the eyeball. In some embodiments, the sample is a blood sample. The blood sample can be obtained using conventional techniques as described in, e.g. Innis et al, editors, PCR Protocols (Academic Press, 1990). White blood cells can be separated from blood samples using convention techniques or commercially available kits, e.g. RosetteSep kit (Stein Cell Technologies, Vancouver, Canada). Sub-populations of white blood cells, e.g. mononuclear cells, B cells, T cells, monocytes, granulocytes or lymphocytes, can be further isolated using conventional techniques, e.g. magnetically activated cell sorting (MACS) (Miltenyi Biotec, Auburn, California) or fluorescently activated cell sorting (FACS) (Becton Dickinson, San Jose, California). [00265] In one embodiment, the blood sample is from about 0.1 mL to about 10.0 mL, from about 0.2 mL to about 7 mL, from about 0.3 mL to about 5 mL, from about 0.4 mL to about 3.5 mL, or from about 0.5 mL to about 3 mL. In another embodiment, the blood sample is about 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 6.0, 7.0, 8.0, 9.0 or 10.0 mL. [00266] In some embodiments, the sample used in the present methods comprises a biopsy (e.g., a tumor biopsy). The biopsy can be from any organ or tissue, for example, skin, liver, lung, heart, colon, kidney, bone marrow, teeth, lymph node, hair, spleen, brain, breast, or other organs. In a specific embodiment, the sample used in the methods described herein comprises a tumor biopsy. Any biopsy technique known by those skilled in the art can be used for isolating a sample from a subject, for instance, open biopsy, close biopsy, core biopsy, incisional biopsy, excisional biopsy, or fine needle aspiration biopsy. [00267] In one embodiment, the sample used in the methods provided herein is obtained from the subject prior to the patient receiving a treatment for lymphoma (e.g., DLBCL). In another embodiment, the sample is obtained from the patient during the subject receiving a treatment for the lymphoma (e.g., DLBCL). In another embodiment, the sample is obtained from the patient after the patient received a treatment for the lymphoma (e.g., DLBCL). In various embodiments, the treatment comprises administering a compound described herein to the subject. [00268] In certain embodiments, the sample used in the methods provided herein comprises a plurality of cells. Such cells can include any type of cells, e.g., stem cells, blood cells (e.g., peripheral blood mononuclear cells), lymphocytes, B cells, T cells, monocytes, granulocytes, immune cells, or tumor or cancer cells. The tumor or cancer cells or a tumor tissue, such as a tumor biopsy or a tumor explants. T cells (T lymphocytes) include, for example, helper T cells (effector T cells or Th cells), cytotoxic T cells (CTLs), memory T cells, and regulatory T cells. In one embodiment, the cells used in the methods provided herein are CD3+ T cells, e.g., as detected by flow cytometry. The number of T cells used in the methods can range from a single cell to about 109 cells. B cells (B lymphocytes) include, for example, plasma B cells, dendritic cells, memory B cells, B1 cells, B2 cells, marginal-zone B cells, and follicular B cells. B cells can express immunoglobulins (antibodies, B cell receptor). [00269] Specific cell populations can be obtained using a combination of commercially available antibodies (e.g., Quest Diagnostic (San Juan Capistrano, Calif.); Dako (Denmark)). [00270] In certain embodiments, the sample used in the methods provided herein is from a diseased tissue from a lymphoma (e.g., DLBCL) patient. In certain embodiments, the number of cells used in the methods provided herein can range from a single cell to about 109 cells. In some embodiments, the number of cells used in the methods provided herein is about 1 x 104, 5 x 104, 1 x 105, 5 x 105, 1 x 106, 5 x 106, 1 x 107, 5 x 107, 1 x 108, or 5 x 108. [00271] The number and type of cells collected from a subject can be monitored, for example, by measuring changes in morphology and cell surface markers using standard cell detection techniques such as flow cytometry, cell sorting, immunocytochemistry (e.g., staining with tissue specific or cell-marker specific antibodies) fluorescence activated cell sorting (FACS), magnetic activated cell sorting (MACS), by examination of the morphology of cells using light or confocal microscopy, and/or by measuring changes in gene expression using techniques well known in the art, such as PCR and gene expression profiling. These techniques can be used, too, to identify cells that are positive for one or more particular markers. Fluorescence activated cell sorting (FACS) is a well-known method for separating particles, including cells, based on the fluorescent properties of the particles (Kamarch, Methods Enzymol., 1987, 151:150-165). Laser excitation of fluorescent moieties in the individual particles results in a small electrical charge allowing electromagnetic separation of positive and negative particles from a mixture. In one embodiment, cell surface marker-specific antibodies or ligands are labeled with distinct fluorescent labels. Cells are processed through the cell sorter, allowing separation of cells based on their ability to bind to the antibodies used. FACS sorted particles may be directly deposited into individual wells of 96-well or 384-well plates to facilitate separation and cloning. [00272] In certain embodiments, subsets of cells are used in the methods provided herein. Methods to sort and isolate specific populations of cells are well-known in the art and can be based on cell size, morphology, or intracellular or extracellular markers. Such methods include, but are not limited to, flow cytometry, flow sorting, FACS, bead based separation such as magnetic cell sorting, size-based separation (e.g., a sieve, an array of obstacles, or a filter), sorting in a microfluidics device, antibody-based separation, sedimentation, affinity adsorption, affinity extraction, density gradient centrifugation, laser capture microdissection, etc. 5.8 Methods for Detecting Expression Levels [00273] In some embodiments, the methods provided here comprise measuring the expression levels of one or more genes identified in Table 1. The expression levels of genes identiefied in Table 1 can be determined by known methods in the art. [00274] In some embodiments, the expression levels of genes are determined by measuring the mRNA levels of these proteins. Several methods of detecting or quantitating mRNA levels are known in the art. Exemplary methods include but are not limited to northern blots, ribonuclease protection assays, PCR-based methods, and the like. The mRNA sequence can be used to prepare a probe that is at least partially complementary. The probe can then be used to detect the mRNA sequence in a sample, using any suitable assay, such as PCR-based methods, digital PCR (dPCR), Northern blotting, a dipstick assay, and the like. [00275] In other embodiments, a nucleic acid assay for testing for immunomodulatory activity in a biological sample can be prepared. An assay typically contains a solid support and at least one nucleic acid contacting the support, where the nucleic acid corresponds to at least a portion of an mRNA. The assay can also have a means for detecting the altered expression of the mRNA in the sample. [00276] The assay method can be varied depending on the type of mRNA information desired. Exemplary methods include but are not limited to Northern blots and PCR-based methods (e.g., RT-qPCR). Methods such as RT-qPCR can also accurately quantitate the amount of the mRNA in a sample. [00277] Any suitable assay platform can be used to determine the presence of the mRNA in a sample. For example, an assay may be in the form of a dipstick, a membrane, a chip, a disk, a test strip, a filter, a microsphere, a slide, a multiwell plate, or an optical fiber. An assay system may have a solid support on which a nucleic acid corresponding to the mRNA is attached. The solid support may comprise, for example, a plastic, silicon, a metal, a resin, glass, a membrane, a particle, a precipitate, a gel, a polymer, a sheet, a sphere, a polysaccharide, a capillary, a film a plate, or a slide. The assay components can be prepared and packaged together as a kit for detecting an mRNA. [00278] The nucleic acid can be labeled, if desired, to make a population of labeled mRNAs. In general, a sample can be labeled using methods that are well known in the art (e.g., using DNA ligase, terminal transferase, or by labeling the RNA backbone, etc.; see, e.g., Ausubel, et al., Short Protocols in Molecular Biology, 3rd ed., Wiley & Sons 1995 and Sambrook et al., Molecular Cloning: A Laboratory Manual, Third Edition, 2001 Cold Spring Harbor, N.Y.). In some embodiments, the sample is labeled with fluorescent label. Exemplary fluorescent dyes include but are not limited to xanthene dyes, fluorescein dyes, rhodamine dyes, fluorescein isothiocyanate (FITC), 6 carboxyfluorescein (FAM), 6 carboxy-2',4',7',4,7-hexachlorofluorescein (HEX), 6 carboxy 4', 5' dichloro 2', 7' dimethoxyfluorescein (JOE or J), N,N,N',N' tetramethyl 6 carboxyrhodamine (TAMRA or T), 6 carboxy X rhodamine (ROX or R), 5 carboxyrhodamine 6G (R6G5 or G5), 6 carboxyrhodamine 6G (R6G6 or G6), and rhodamine 110; cyanine dyes, e.g. Cy3, Cy5 and Cy7 dyes; Alexa dyes, e.g. Alexa-fluor-555; coumarin, Diethylaminocoumarin, umbelliferone; benzimide dyes, e.g. Hoechst 33258; phenanthridine dyes, e.g. Texas Red; ethidium dyes; acridine dyes; carbazole dyes; phenoxazine dyes; porphyrin dyes; polymethine dyes, BODIPY dyes, quinoline dyes, Pyrene, Fluorescein Chlorotriazinyl, R110, Eosin, JOE, R6G, Tetramethylrhodamine, Lissamine, ROX, Napthofluorescein, and the like. [00279] A typical mRNA assay method can contain the steps of 1) obtaining surface-bound subject probes; 2) hybridization of a population of mRNAs to the surface-bound probes under conditions sufficient to provide for specific binding (3) post-hybridization washes to remove nucleic acids not bound in the hybridization; and (4) detection of the hybridized mRNAs. The reagents used in each of these steps and their conditions for use may vary depending on the particular application. [00280] Hybridization can be carried out under suitable hybridization conditions, which may vary in stringency as desired. Typical conditions are sufficient to produce probe/target complexes on a solid surface between complementary binding members, i.e., between surface- bound subject probes and complementary mRNAs in a sample. In certain embodiments, stringent hybridization conditions may be employed. [00281] Hybridization is typically performed under stringent hybridization conditions. Standard hybridization techniques (e.g. under conditions sufficient to provide for specific binding of target mRNAs in the sample to the probes) are described in Kallioniemi et al., Science, 258:818-821 (1992) and WO 93/18186. Several guides to general techniques are available, e.g., Tijssen, Hybridization with Nucleic Acid Probes, Parts I and II (Elsevier, Amsterdam 1993). For descriptions of techniques suitable for in situ hybridizations, see Gall et al. Meth. Enzymol., 1981, 21:470-480; and Angerer et al. in Genetic Engineering: Principles and Methods (Setlow and Hollaender, Eds.) Vol.7, pgs 43-65 (Plenum Press, New York 1985). Selection of appropriate conditions, including temperature, salt concentration, polynucleotide concentration, hybridization time, stringency of washing conditions, and the like will depend on experimental design, including source of sample, identity of capture agents, degree of complementarity expected, etc., and may be determined as a matter of routine experimentation for those of ordinary skill in the art. [00282] Those of ordinary skill will readily recognize that alternative but comparable hybridization and wash conditions can be utilized to provide conditions of similar stringency. [00283] After the mRNA hybridization procedure, the surface bound polynucleotides are typically washed to remove unbound nucleic acids. Washing may be performed using any convenient washing protocol, where the washing conditions are typically stringent, as described above. The hybridization of the target mRNAs to the probes is then detected using standard techniques. [00284] Other methods, such as PCR-based methods, can also be used to follow the expression of the genes. Examples of PCR methods can be found in the literature. Examples of PCR assays can be found in U.S. Patent No.6,927,024, which is incorporated by reference herein in its entirety. Examples of RT-PCR methods can be found in U.S. Patent No.7,122,799, which is incorporated by reference herein in its entirety. A method of fluorescent in situ PCR is described in U.S. Patent No.7,186,507, which is incorporated by reference herein in its entirety. [00285] In some embodiments, Real-Time Reverse Transcription-PCR (RT-qPCR) can be used for both the detection and quantification of RNA targets (Bustin, et al., Clin. Sci., 2005, 109:365- 379). Quantitative results obtained by RT-qPCR are generally more informative than qualitative data. Thus, in some embodiments, RT-qPCR-based assays can be useful to measure mRNA levels during cell-based assays. The RT-qPCR method is also useful to monitor patient therapy. Examples of RT-qPCR-based methods can be found, for example, in U.S. Patent No.7,101,663, which is incorporated by reference herein in its entirety. [00286] In contrast to regular reverse transcriptase-PCR and analysis by agarose gels, real-time PCR gives quantitative results. An additional advantage of real-time PCR is the relative ease and convenience of use. Instruments for real-time PCR, such as the Applied Biosystems 7500, are available commercially, as are the reagents, such as TaqMan Sequence Detection chemistry. For example, TaqMan® Gene Expression Assays can be used, following the manufacturer’s instructions. These kits are pre-formulated gene expression assays for rapid, reliable detection and quantification of human, mouse and rat mRNA transcripts. An exemplary PCR program, for example, is 50ºC for 2 minutes, 95ºC for 10 minutes, 40 cycles of 95ºC for 15 seconds, then 60ºC for 1 minute. [00287] To determine the cycle number at which the fluorescence signal associated with a particular amplicon accumulation crosses the threshold (referred to as the CT), the data can be analyzed, for example, using a 7500 Real-Time PCR System Sequence Detection software v1.3 using the comparative CT relative quantification calculation method. Using this method, the output is expressed as a fold-change of expression levels. In some embodiments, the threshold level can be selected to be automatically determined by the software. In some embodiments, the threshold level is set to be above the baseline but sufficiently low to be within the exponential growth region of an amplification curve. [00288] Techniques known to one skilled in the art may be used to measure the amount of an RNA transcript(s). In some embodiments, the amount of one, two, three, four, five, or more RNA transcripts is measured using deep sequencing, such as ILLUMINA® RNASeq, ILLUMINA® next generation sequencing (NGS), ION TORRENTTM RNA next generation sequencing, 454TM pyrosequencing, or Sequencing by Oligo Ligation Detection (SOLIDTM). In other embodiments, the amount of multiple RNA transcripts is measured using a microarray and/or gene chip. In certain embodiments, the amount of one, two, three, or more RNA transcripts is determined by RT-PCR. In other embodiments, the amount of one, two, three, or more RNA transcripts is measured by RT-qPCR. Techniques for conducting these assays are known to one skilled in the art. In yet other embodiments, NanoString (e.g., nCounter® miRNA Expression Assays provided by NanoString® Technologies) is used for analyzing RNA transcripts. [00289] Several protein detection and quantitation methods can be used to measure the level of proteins. Any suitable protein quantitation method can be used. In some embodiments, antibody-based methods are used. Exemplary methods that can be used include but are not limited to immunoblotting (western blot), enzyme-linked immunosorbent assay (ELISA), immunohistochemistry, flow cytometry, cytometric bead array, mass spectroscopy, and the like. Several types of ELISA are commonly used, including direct ELISA, indirect ELISA, and sandwich ELISA. [00290] In a specific embodiment, the protein level is determined by immunohistochemistry (IHC). IHC refers to a lab test that uses antibodies to test for certain antigens (markers) in a sample of tissue, and is a process of detecting antigens (e.g., proteins) in cells of a tissue section by exploiting the principle of antibodies binding specifically to antigens in biological tissues. The antibodies are usually linked to an enzyme or a fluorescent dye. Typically, when the antibodies bind to the antigen in the tissue sample, the enzyme or dye is activated, and the antigen can then be seen under a microscope. IHC can be used to help diagnose diseases, such as cancer. It may also be used to help tell the difference between different types of cancer. IHC can be used to image discrete components in tissues by using appropriately-labeled antibodies to bind specifically to their target antigens in situ. IHC makes it possible to visualize and document the high-resolution distribution and localization of specific cellular components within cells and within their proper histological context. While there are multiple approaches and permutations in IHC methodology, all of the steps involved can be generally separated into two groups: sample preparation and sample staining. In some embodiments, IHC is based on the immunostaining of thin sections of tissues attached to individual glass slides. Multiple small sections can be arranged on a single slide for comparative analysis, a format referred to as a tissue microarray. In other embodiments, IHC is performed by using high-throughput sample preparation and staining. [00291] Samples can be viewed by either light or fluorescence microscopy. In some embodiments, antigen detection in tissue can be performed using an antibody conjugated to an enzyme (horseradish peroxidase) and utilized a colorimetric substrate that could be detected by light microscopy. [00292] In some embodiments, the sample (e.g., a tissue from the patient) has been snap frozen in liquid nitrogen, isopentane or dry ice. In other embodiments, the sample (e.g., a tissue from the patient) has been fixed in formaldehyde and embedded in paraffin wax (FFPE). In both of the above-mentioned methods, the tissue or sections of the tissue can be mounted on slides prior to staining. In yet other embodiments, the IHC-free-floating technique may be used, where the entire IHC procedure is performed in liquid to increase antibody binding and penetration and slide mounting only takes place upon experimental completion. IHC-free-floating appears to be most popular in neuroscience research. When analysis of the tissue by electron microscopy is desired, the tissue can be embedded in acrylate resins such as glycol methacrylate (GMA), a technique referred to as IHC-resin. [00293] In a specific embodiment, IHC can be performed using the method described in the Examples section below. 5.9 Kits [00294] In one aspect, provided herein is a kit predicting the responsiveness of a lymphoma patient to a cancer treatment, comprising agents for measuring the gene expression levels of a biological sample from the lymphoma patient. In some embodiments, the kit further comprises an agent (or tool) for taking a sample from a subject. In some embodiments, the kit further comprises an instruction on how to interpret or use the expression levels determined to predict if a patient has a particular subtype of lymphoma (e.g., DLBCL). [00295] In certain embodiments, a kit comprises a reagent or reagents necessary for carrying out an assay(s) described herein, in one or more other containers. In certain embodiments, the kit comprises a solid support, and a means for detecting the RNA or protein expression of at least one biomarker in a biological sample. Such a kit may employ, for example, a dipstick, a membrane, a chip, a disk, a test strip, a filter, a microsphere, a slide, a multiwell plate, or an optical fiber. The solid support of the kit can be, for example, a plastic, silicon, a metal, a resin, glass, a membrane, a particle, a precipitate, a gel, a polymer, a sheet, a sphere, a polysaccharide, a capillary, a film, a plate, or a slide. [00296] In a specific embodiment, the kit comprises, in one or more containers, components for conducting RT-PCR, RT-qPCR, deep sequencing, or a microarray such as NanoString assay. In some embodiments, the kit comprises a solid support, nucleic acids contacting the support, where the nucleic acids are complementary to at least 10, 20, 50, 100, 200, 350, or more bases of mRNA, and a means for detecting the expression of the mRNA in a biological sample. [00297] In another specific embodiment, the kit comprises, in one or more containers, components for conducting assays that can determine one or more protein levels, such flow cytometry, ELISA, or HIC. [00298] Such kits may comprise materials and reagents required for measuring RNA or protein. In some embodiments, such kits include microarrays, wherein the microarray is comprised of oligonucleotides and/or DNA and/or RNA fragments which hybridize to one or more of the genes identified in Table 1. In some embodiments, such kits may include primers for PCR of either the RNA product or the cDNA copy of the RNA product of the genes or subset of genes, or both. In some embodiments, such kits may include primers for PCR as well as probes for Quantitative PCR. In some embodiments, such kits may include multiple primers and multiple probes wherein some of said probes have different fluorophores so as to permit multiplexing of multiple products of a gene product or multiple gene products. In some embodiments, such kits may further include materials and reagents for creating cDNA from RNA. In some embodiments, such kits may include antibodies specific for one or more of the genes identified in Table 1. Such kits may additionally comprise materials and reagents for isolating RNA and/or proteins from a biological sample. In addition, such kits may include materials and reagents for synthesizing cDNA from RNA isolated from a biological sample. In some embodiments, such kits may include, a computer program product embedded on computer readable media for predicting whether a patient is responsive to a compound as described herein. In some embodiments, the kits may include a computer program product embedded on a computer readable media along with instructions. [00299] For antibody based kits, the kit can comprise, for example: (1) a first antibody (which may or may not be attached to a solid support) which binds to a peptide, polypeptide or protein of interest; and, optionally, (2) a second, different antibody which binds to either the peptide, polypeptide or protein, or the first antibody and is conjugated to a detectable label (e.g., a fluorescent label, radioactive isotope or enzyme). The antibody-based kits may also comprise beads for conducting an immunoprecipitation. Each component of the antibody-based kits is generally in its own suitable container. Thus, these kits generally comprise distinct containers suitable for each antibody. Further, the antibody-based kits may comprise instructions for performing the assay and methods for interpreting and analyzing the data resulting from the performance of the assay. In a specific embodiment, the kits contain instructions for predicting whether a lymphoma (e.g., DLBCL) patient belongs to a specific subgroup of DLBCL (e.g., a high-risk subgroup of DLBCL). [00300] In certain embodiments of the methods and kits provided herein, solid phase supports are used for purifying proteins, labeling samples, or carrying out the solid phase assays. Examples of solid phases suitable for carrying out the methods disclosed herein include beads, particles, colloids, single surfaces, tubes, multi-well plates, microtiter plates, slides, membranes, gels, and electrodes. When the solid phase is a particulate material (e.g., a bead), it is, in one embodiment, distributed in the wells of multi-well plates to allow for parallel processing of the solid phase supports. [00301] The practice of the embodiments provided herein will employ, unless otherwise indicated, conventional techniques of molecular biology, microbiology, and immunology, which are within the skill of those working in the art. Such techniques are explained fully in the literature. Examples of particularly suitable texts for consultation include the following: Sambrook et al., Molecular Cloning: A Laboratory Manual (2d ed.1989); Glover, ed., DNA Cloning, Volumes I and II (1985); Gait, ed., Oligonucleotide Synthesis (1984); Hames & Higgins, eds., Nucleic Acid Hybridization (1984); Hames & Higgins, eds., Transcription and Translation (1984); Freshney, ed., Animal Cell Culture: Immobilized Cells and Enzymes (IRL Press, 1986); Immunochemical Methods in Cell and Molecular Biology (Academic Press, London); Scopes, Protein Purification: Principles and Practice (Springer Verlag, N.Y., 2d ed. 1987); and Weir & Blackwell, eds., Handbook of Experimental Immunology, Volumes I-IV (1986). [00302] From the foregoing, it will be appreciated that, although specific embodiments have been described herein for the purpose of illustration, various modifications may be made without deviating from the spirit and scope of what is provided herein. All of the references referred to above are incorporated herein by reference in their entireties. [00303] Certain embodiments of the invention are illustrated by the following non-limiting examples. 6. EXAMPLES [00304] The examples below are carried out using standard techniques, which are well known and routine to those of skill in the art, except where otherwise described in detail. The examples are intended to be merely illustrative. 6.1 Example 1 Unsupervised Clustering of DLBCL Patients 6.1.1 Data Overview [00305] Multiple DLBCL patient datasets were utilized. The initial discovery dataset was a set of 267 commercially-sourced newly diagnosed DLBCL (ndDLBCL) patient samples that had molecular profiling but no survival data. A subset of these patients also had IHC imaging data, so this initial discovery set is referred to the Commercial+IHC dataset. As a replication cohort, a set of 342 ndDLBCL patients from the Molecular Epidemiology Resource (MER) which were well-characterized in terms of clinical outcome and treatment (see Cerhan et al., Int. J. Epidermiol., 2017, 46(6):1753-1754i). The Lenz microarray dataset of 414 ndDLBCL patients was also used as a second replication cohort and had outcome data available (see Lenz et al., N. Engl. J. Med., 2008, 359(22):2313-2323). Two other datasets were used for replication in the relapsed and refractory DLBCL (rrDLBCL) setting, each of which had clinical outcome information. The first was a set of 86 rrDLBCL patients from the MER cohort, and the second was a set of 189 rrDLBCL patient samples from two clinical trials (CC-122-ST-001 and CC-122- DLBCL-001) (see Clinical Trials NCT01421524 and NCT02031419). An overview of the dataset is provided in Table 2 below. Table 2. Dataset Overview
Figure imgf000068_0001
6.1.2 Data Pre-Processing [00306] A single sample normalization (ssNorm) method was applied to all RNAseq datasets prior to analysis. The practice of performing ssNorm has benefits in both immediate short-term analyses, as well as long-term translatability of findings. In the short term, ssNorm provides a fixed normalization scheme that can apply to individual samples completely independently. There is no need to re-normalize datasets with the addition of new batches of samples, or when combining datasets for larger analyses. The ssNorm method also implicitly performs batch correction, removing gene-specific experimental effects that appear as systematic, biased differences in the raw expression space. In practice, the ssNorm method properly aligns diverse datasets to a common space, showing no meaningful separation by dataset/batch when projecting into PCA space. From a strategic perspective, ssNorm allows simple translatability of any classifier or parameterization from one dataset to another – a classifier built on one ssNorm dataset can be directly applied to any other ssNorm dataset, without any need for reweighting model parameters or setting new thresholds. [00307] The ssNorm method used the Commercial cohort (without the IHC samples) as a reference dataset against which all other samples are normalized. The normalization began with isoform-level TPM data. Gene-level data were derived by summing the expression of all isoforms annotated to a particular gene. Then five DLBCL-specific housekeeping genes (ISY1, R3HDM1, TRIM56, UBXN4, and WDR55) were used for normalization. These five genes were the same as the housekeeping genes from the Nanostring COO assay. For each sample, the geometric mean of the housekeeping genes was computed, and used as a global normalization factor by which all genes’ expressions were divided. For the reference dataset, each gene’s housekeeping-normalized mean and standard deviation were calculated. For each dataset, the housekeeping-normalized gene-level data was scaled and shifted to match the reference distribution by subtracting the reference mean and dividing by the reference standard deviation on a gene-wise basis. In the reference data, all genes had a mean of 0 and a standard deviation of 1, and other datasets normalized to this space matched that distribution closely. [00308] The ssNorm method removed systematic differences between datasets while leaving relevant biological signal intact. Molecularly derived features such as COO, DHITsig, or cellular deconvolution estimates can be computed from the ssNorm RNAseq data and showed good concordance with measures that are not derived from RNAseq data. In the ndMER (replication) dataset, for example, the Reddy COO calls derived from ssNorm gene expression data were 80% concordant with the Hans IHC method (when excluding unclassifiable patients), and 91% concordant with the Nanostring method which utilized non-ssNorm gene expression. Similarly, the DHIT gene expression score from ssNorm data associated strongly with FISH DHITsig positivity, having a classification AUC of 0.82. Cellular deconvolution estimates from ssNorm data were also well correlated with cell marker densities derived from IHC, particularly for abundant markers. Both CD20 and CD3 marker densities showed a Spearman correlation above 0.8 with the deconvolution abundance estimates of their respective cell types (B and T cells). Finally, sample rankings were also relatively unchanged by the ssNorm method, particularly for highly expressed genes – for the highly expressed/highly variant genes used in the iClusterPlus approach, the mean gene-wise correlation between sample TPMs and ssNorm data was greater than 0.9. FIG.1 illustrates that PCA projection of ssNorm datasets shows a high degree of overlap and no systematic stratification by dataset. The samples represent a subset of the Commercial cohort that was selected for imaging. 6.1.3 Outlier Detection [00309] Outlier detection was applied to the combined cohort of the ssNorm Commercial+IHC, Lenz, and MER datasets to identify samples that showed unexpected deviation from the rest of the population. To achieve this, the ArrayQualityMetrics R package was applied, using the Kolmogorov-Smirnov, sum and upper quartile methods, each of which detected if the distribution of gene expression on a per-sample basis is significant different from all other samples. Samples which failed 2 of the 3 tests were called as outliers and removed for further analysis. In total, 13 samples were removed from the commercial training data, and 9 more were removed from the MER dataset. 6.1.4 Dimensionality Reduction [00310] Any clustering methodology is sensitive to the input data, and algorithm performance can be degraded by introducing noisy or irrelevant features. Both feature selection and feature engineering approaches were utilized to reduce the dimensionality of the dataset while maintaining a representation of relevant biological activity. As input data for the clustering methods, a subset of the gene expression data and several types of derived feature scores that aggregate multiple biologically related genes into a single feature were used. [00311] To perform feature selection on the genes, the ssNorm expression data was filtered to the top most expressed/most variant genes. This was done by selecting the intersection of the top 25% most expressed and top 25% most variant genes in the ssNorm training data. The selected set of 2479 genes was highly expressed/highly variant in raw TPM space, indicating that the ssNorm method had little effect on the overall gene ranking. [00312] In addition to the ssNorm filtered gene expression data, we calculated enrichment scores over several informative gene sets. In particular, the single-sample GSVA score over the set of Hallmark pathways from MSigDB was calculated, as well as for the C1 set of positional cytoband signatures which can be viewed as a proxy for the copy number data (which was unavailable at the time of analysis). The final set of feature scores were derived from the DCQ cellular deconvolution method (Altboum et al., Mol. Syst. Biol., 2014, 10:720-733), using the DLBCL-specific LM23 matrix. 6.1.5 Unsupervised Clustering [00313] The commercial set of samples was used as the discovery cohort for the initial clustering. All four matrices of features (gene expression, Hallmark, C1, and LM23) were used as input to two clustering methods, Cluster of Clusters Analysis (COCA) (Cabassi, Cluster of Clusters Analysis, R Package 2019) and iClusterPlus (Mo & Shen, iClusterPlus: Intergrative clustering of multi-type genomics data, R Package, 2019). The COCA method performs clustering on each feature matrix independently and aggregates the cluster labels, while the iClusterPlus method is a latent variable model that aggregates the feature matrices and finds a cluster labeling that is informative across all the inputs simultaneously. An overall schematic of the clustering methodology and downstream validation approach is shown in FIG.2. [00314] No specific features selection was made among each of the feature matrix (gene expression, Hallmark, C1, and LM23). FIGS 3A-3D show heatmaps of the clustering results by each features matrix. As shown in FIGS.3B and 3D, the hallmark pathway features and LM23 cell type features are not as specific to a single cluster; and as shown in FIG.3C, the C1 set of positional cytoband features do not stratify by cluster. The consensus matrix is shown in FIG. 3E. [00315] Both methods were tested for between 2 and 15 clusters, with each method run 100 times using 20% random feature and sample dropout. The final clusters for each method for each choice of number of clusters (K) was determined as the consensus clustering over the resampling iterations. We performed parameter fitting by internal cross-validation routine, parameters were optimized for K between 2 and 5 and after noting that the hyper parameter converged, this fixed parameterization was carried forward to higher K. [00316] Each clustering method produced one final clustering for each value of K, and these cluster assignments were evaluated to determine the optimal number of clusters and best algorithm. The nbClust R package was used for cluster evaluation, which used a variety of metrics to evaluate internal cluster stability such as the silhouette statistic (FIGS.4A-4B), gap statistic, and percentage of variance explained by the clustering. The clusters were also evaluated in terms of the minimum cluster size, in order to have sufficient sample size in each cluster to be able to build a downstream classifier that could robustly identify each subgroup. 6.1.6 Clustering Results [00317] The optimal clustering was found using the iClusterPlus method for K=8 clusters. In the discovery dataset, these clusters ranged in prevalence from 7% to 19% of the cohort, with a minimum cluster size of N=18 patients. The clusters were associated with, but not uniquely determined by, relevant biological characteristics such as COO, TME, and DHITsig. Clusters were not globally associated with clinical characteristics such as age or sex group in external replication datasets, although cluster D4 specifically was associated with a higher International Prognostic Index (IPI) clinical risk score. These findings indicate that the discovered subgroups represent a novel method of classifying DLBCL patients that is distinct from previous molecular or clinical classification methods. A heatmap representation of the discovery cohort, stratified by the top cluster-differential genes, is shown in FIG.5A. Heatmap representations of the ndMER dataset and the Lez dataset are shown in FIGS.5B and 5C. Heapmap representations of the rr MER dataset and the CC-122 dataset are shown in FIGS.5D and 5E. [00318] This clustering method allows for the transcriptomic identification of high-risk patients subgroups which are underserved by the standard R-CHOP therapy. 6.2 Example 2 Cluster Classifier [00319] Having identified 8 novel patient subgroups in the discovery cohort, the next step was to identify those subgroups in another dataset. To accomplish this, a grouped multinomial generalized linear model (GLM) was trained using LASSO (alpha=1, least absolute shrinkage and selection operator) to predict the cluster labels in the discovery cohort, using the ssNorm gene expression data filtered to the top most expressed/most variant genes used for clustering. The cross-validated GLM fitting function was applied and the most parsimonious parameterization (lambda) that had error within one standard error of the minimum was selected, as recommended in Friedman et al., J. Stat. Softw., 2010, 33: 1-22. This practice selects a near- optimal model that minimizes the number of features used in classification while avoiding overfitting (Krstajic et al., J. Cheminformatics, 2014, 6:10. The resulting model, which utilized 172 genes, was able to reproduce the cluster labels with zero misclassification, as would be expected when classifying groups derived from an unsupervised clustering. The 172 genes utilized in the results model are listed in Table 1. [00320] Next, the GLM cluster classifier was applied to independent ndDLBCL datasets, the ndMER and Lenz cohorts. Because the GLM cluster classifier was trained in the space of ssNorm gene expression, its application to other ssNorm datasets was straightforward, requiring no reweighting of parameters or other translation of the model. Direct application of the classifier yielded consistent cluster features between the discovery and replication cohorts, including cluster prevalence, the cluster-defining gene sets, and higher-level biological features like TME or COO content of each cluster (see FIGS.6A-6B). The cluster prevalence of the TME+ and TME- patients across the eight subgroups in a combination dataset of MER, Schmitz, and Lenz datasets is shown in FIG.7. See Schmitz et al., N. Engl. J. Med., 2018, 378(15):1396- 1407; Lenz et al., Proc. Natl. Acad. Sci. U. S. A., 2008, 105:13520-13525. [00321] Examination of the clinical outcome stratified by cluster label revealed that among R- CHOP-treated patients in the ndMER cohort, two clusters had significantly worse outcomes than the others (FIG.8). One of the high-risk clusters, called D4, comprised ~20% of ndDLBCL patients and was a subset of high-risk ABC patients with low immune infiltration, hallmarked by upregulation of MYC-related pathways. The other high-risk cluster, called D8, represented ~5% of the ndDLBCL population, had very low immune infiltration, and was associated with a double-hit gene signature (DHITsig) positivity. [00322] Application of the cluster classifier to the rrDLBCL cohorts yielded similar results, with consistent cluster-defining genes and biological characteristics of the clusters, although at different prevalence (FIG.9). In particular, the prevalence of the high-risk clusters (D4 and D8) was increased in rrDLBCL compared to ndDLBCL, with a prevalence of ~30% and ~15%, respectively. As the patients in the high-risk clusters progress into a relapse setting at a higher rate than the other patients, they are over-represented in rrDLBCL cohort. 6.3 Example 3 Double Hit Signature (DHITsig) Classifier [00323] It has been well described in literature that there exists a subset of high-risk DLBCL patients who are hallmarked by a so-called “double hit” (DHIT) translocation in MYC and either BCL2 or BCL6. Patients who are DHIT+ tend to have worse survival than DHIT- patients, and these patients are typically GCB with MYC+BCL2 translocations. DHIT status has commonly been measured using DNA sequencing or FISH probes to determine translocation status. [00324] Ennishi et al. derived a gene expression signature that reflects DHIT status, capturing a broader segment of the population that has differential clinical outcome. See Ennishi et al., J. Clin. Oncol., 2018, 37:190-201. Using the 104 genes, parameterization, and methodology described in the manuscript, the DHITsig method was adapted and implemented for use in the single-sample normalized RNAseq space. [00325] The Ennishi method for calculating the DHITsig score is a variable importance- weighted sum of log likelihood ratios. The likelihood function is calculated by assuming a Gaussian mixture model of gene expression that defines two normal distributions of expression for DHITsig+ and DHITsig- samples for each signature gene. The gene list and variable weights from Ennishi et al. were used, along with the mixture model distribution parameters as derived from their DESeq2-processed dataset. [00326] In order to adapt the parameters to the ssNorm setting, it began with raw RNAseq counts from the MER 1L dataset. These counts were processed using a DESeq2 pipeline, following the Ennishi processing steps as closely as possible given the description in the manuscript. The classifier was then applied directly using the importance scores, mixture model parameters, and classification threshold from Ennishi et al. It was found that the given threshold to be too conservative in calling DHITsig+ on our data, finding only 4% of GCB patients as DHITsig+ and missing many patients confirmed to be DHIT+ by FISH. The overall sample ranking by DHITsig score was largely correct, however, as evidenced by the high AUC (0.82) in predicting FISH DHIT+ patients from the DHITsig score (FIG.10). Given this finding, a new threshold that identified all FISH DHIT+ patients (sensitivity = 1) was set while minimizing false positives (specificity = 0.4). This choice of threshold was further supported by the prevalence of DHITsig+ patients among GCBs, which matched the prevalence of 27% reported by Ennishi et al. [00327] Using the new classification threshold, the DHITsig+/- calls for the MER 1L samples were fixed. Then the Gausssian mixture model parameters in ssNorm space were re-derived by directly observing the mean and standard deviation of the signature genes among the called DHITsig+/- groups in ssNorm space. The classifier threshold was set to 0 in this space, which is the natural cutoff point of a log-likelihood measure such as this one. This new set of parameters and threshold, combined with the original variable importance weights, provided a complete description of the DHITsig classifier that can be directly applied to any RNAseq data in ssNorm space. [00328] The ssNorm-adapted DHITsig method was applied to the Lenz and ndMER datasets and showed significant prognostic value among GCB patients (p=0.0006), but not among ABC or Unclassified patients, as expected (FIGS.11A and 11B). The proportion of GCB patients called DHITsig+ was 28%, almost perfectly replicating the prevalence observed by Ennishi et al. and training dataset. [00329] The expression of the BCL6 signature across the eight subgroups in the ndMER cohort are measured and shown in FIG.12. (see Polo et al., Proc. Natl. Acad. Sci. U. S. A., 2007, 104(9):3207-3212). 6.4 Example 4 High Risk Subgroups D4 and D8 [00330] Integrative clustering identified eight subgroups of ndDLBCL patients (named D1- D8). The resulting clusters are analyzed in the lens of different biological features including gene signatures such as Double HIT gene signature and TMD gen signature. The prevalence and biological features (such as COO type, EFS24 failure rate, DHITsig positivity, TME gene signature positivity, BCL2/BCL6 translocation, and MYC translocation) of the replication dataset (MER dataset) are summarized in Table 3. The resulting cluster identification are predictive of the likelihood of response to standard treatment (e.g., R-CHOP combination treatment) and can suggest rational targeted therapies based on cluster-specific biological features. Table 3. Prevalence and biological features of replication dataset (MER dataset)
Figure imgf000075_0001
[00331] Among R-CHOP-treated patients, subgroups D4 (p<0.01) and D8 (p<0.0001) had significantly worse survival outcomes than the rest of the population. D4 comprised 21% of the replication cohort (ndMER cohort) with a median event-free survival (mEFS) of 38.2 months and a median overall survival (mOS) of 80.3 months. D8 comprised 5% of the cohort with a mEFS of 7.5 months and a mOS of 12.1 months. The remaining 6 subgroups were standard risk, with mEFS ranging from 82.1 months to not reached, and none reaching mOS. The subgroups were not uniquely defined by previously known molecular classification methods such as COO or Double Hit Signature (DHITsig) (Ennishi et al., J. Clin. Oncol., 2018, 37:190-201), nor by clinical risk factors such as age or international prognostic index (IPI). [00332] Within subgroup D4, 92% of patients were ABC, representing a high-risk subset of ABC patients. The mEFS in D4 ABCs was 38.2 months, while mEFS of non-D4 ABCs was not reached (p<0.005). Transcriptomic analysis revealed upregulation of MYC signaling, cell cycle, metabolism, and DNA recombination pathways, and a lower abundance of immune infiltration. D4 was associated with high IPI, with 49% of D4 having IPI>2, compared to 33% of non-D4 with IPI>2 (p<0.05). [00333] D8 represented a high-risk subset, which was 73% GCB. The mEFS of D8 GCBs was 5.4 months, while mEFS of non-D8 GCBs was not reached (p<0.0001). Transcriptomic analysis revealed low expression of immune response and cytokine signaling pathways, consistent with the low abundance of immune cells in D8. This subgroup consisted of 63% DHITsig+ patients, and although only 20% of all DHITsig+ patients were in D8, these D8 DHITsig+ patients showed significantly worse survival than non-D8 DHITsig+ patients (mEFS 11.3 months vs. not reached, p<0.0001). [00334] In the relapsed/refractory (r/r) DLBCL setting, D1-D8 were all present, with an increased prevalence of D4 and D8 in CC-122 cohort (30% and 17%, respectively) and rrMER cohort (30% and 14%) compared to the newly diagnosed setting. As these high-risk patients progress at higher rates than others, they appear overrepresented in the r/r population. [00335] This clustering method allows for the transcriptomic identification of high-risk patients subgroups (e.g., subgroups D4 and D8) which are underserved by the standard R-CHOP therapy. 6.5 Example 5 Association of DLBCL Driver Mutations with Clustered Subgroups [00336] Mutational data for the studied cohorts were collected and interpreted in the context of the discovered subgroups. As shown Table 4, all DLBCL driver mutations are listed in and analyzed by counts of mutations or wildtypes in each of the identified subgroups (inCount = number of mutated smaples in the corresponding subgroup; inNoCount = number of wild type samples in the corresponding subgroup; outCount = number of mutated samples in other subgroups; and outNotCount – number of wild type samples in other subgroups). Table 4. List of DLBCL Driver Mutations and Their Association of Subgroups
Figure imgf000077_0001
Figure imgf000078_0001
Figure imgf000079_0001
Figure imgf000080_0001
Figure imgf000081_0001
Figure imgf000082_0001
Figure imgf000083_0001
Figure imgf000084_0001
Figure imgf000085_0001
Figure imgf000086_0001
Figure imgf000087_0001
Figure imgf000088_0001
Figure imgf000089_0001
Figure imgf000090_0001
Figure imgf000091_0001
Figure imgf000092_0001
Figure imgf000093_0001
Figure imgf000094_0001
Figure imgf000095_0001
Figure imgf000096_0001
Figure imgf000097_0001
Figure imgf000098_0001
Figure imgf000099_0001
Figure imgf000100_0001
Figure imgf000101_0001
Figure imgf000102_0001
Figure imgf000103_0001
Figure imgf000104_0001
Figure imgf000105_0001
Figure imgf000106_0001
Figure imgf000107_0001
Figure imgf000108_0001
Figure imgf000109_0001
Figure imgf000110_0001
Figure imgf000111_0001
Figure imgf000112_0001
Figure imgf000113_0001
Figure imgf000114_0001
Figure imgf000115_0001
Figure imgf000116_0001
Figure imgf000117_0001
Figure imgf000118_0001
Figure imgf000119_0001
Figure imgf000120_0001
Figure imgf000121_0001
Figure imgf000122_0001
Figure imgf000123_0001
Figure imgf000124_0001
Figure imgf000125_0001
Figure imgf000126_0001
Figure imgf000127_0001
Figure imgf000128_0001
Figure imgf000129_0001
Figure imgf000130_0001
Figure imgf000131_0001
Figure imgf000132_0001
Figure imgf000133_0001
Figure imgf000134_0001
Figure imgf000135_0001
Figure imgf000136_0001
Figure imgf000137_0001
Figure imgf000138_0001
[00337] Similar mutational data analysis was done among the ABC subtype patient in the ndMER patients by comparing subgroup D4 with other subgroups (D1-D3 and D5-D8) The mutational data analysis is illusted in Table 5 (inCount = number of mutated smaples in the corresponding subgroup; inNoCount = number of wild type samples in the corresponding subgroup; outCount = number of mutated samples in other subgroups; and outNotCount – number of wild type samples in other subgroups). Table 5 List of DLBCL Driver Mutations and Their Association in ABC Subtype Patients in ndMER cohort (Subgroup D4 vs other Subgroups)
Figure imgf000139_0001
Figure imgf000140_0001
Figure imgf000141_0001
Figure imgf000142_0001
Figure imgf000143_0001
Figure imgf000144_0001
Figure imgf000145_0001
Figure imgf000146_0001
6.6 Example 6 Predicting Responsiveness of a DLBCL patient to R-CHOP [00338] Samples from biopsies of a patient that has been diagnosed to have DLBCL are taken prior to receiving any therapy. Such patient may be selected from, for example, a clinical trial. Such patient may be classified as presenting GCB type DLBCL, or may be classified as presenting ABC subtype DLBCL. [00339] The samples from the patient biopsies are hybridized to Affymetrix HG-U133 Plus 2.0 GeneChip™ microarrays (www.affymetrix.com/) at the Molecular Characterization & Clinical Assay Development Laboratory, SAIC Frederick National Laboratory for Cancer Research, SAIC-Frederick, Frederick, MD. Biopsy samples are flash-frozen at screen (FF), archived having been formalin-fixed and paraffin embedded (FFPE archive), or FFPE treated at screen. [00340] Raw Affymetrix image (.cel) files of the patient are imported into the R statistical programming environment v3.0.0 (r-project.org) using functionality of the Affy package of the related Bioconductor suite of open-source bioinformatics software (bioconductor.org). Transcriptional profile QC is performed using the NUSE algorithm, implemented in the Bioconductor package arrayQualityMetrics (Kauffmann et al., 2009), applied to a log2 transformation of raw signal. [00341] The RMA (Irizarry et al., 2003) algorithm is applied to background-correct, quantile normalize and summarize profiles that passed QC. Annotation of probe-sets to genes is performed using the R packages annotate (Gentleman, 2013) and genefilter (Gentleman et al., 2013) selecting only one probeset per gene (Entrez Gene ID) and choosing the most variable across profiles according to inter-quartile range in cases wherein multiple probe-sets map to a single gene. [00342] After normalizing the raw data of the patient using the single sample normalization (ssNorm) method, the Hallmark GSVA scores, C1 Positional GSVA scores, Cell type LM23 GSVA scores of the patient are calculated. [00343] Next, the iClusterPlus method used in Example 1 is applied to the data from the patient using the same subset of gene expression data as well as the calculated Hallmark GSVA scores, C1 Positional GSVA scores, Cell type LM23 GSVA scores as input for clustering features. The patient is determined to belong to, for example, subgroup D4, based on the clustering results from iClusterPlus clustering method. [00344] Then, the GLM classifier model trained in Example 1 in connection with subgroup D4 is applied to the patient classify and predict whether the patient will respond to the standard therapy (e.g., R-CHOP). Applying the trained GLM classifier model to patient’s expression level of the 172 genes identified in Table 1 classifies and predicts the patient to be not responsive to R-CHOP treatment and to be a high-risk DLBCL patient. 6.7 Example 7 Cell Proportions in Subgroups D1-D8 [00345] The proportions of T-cells were measured in the DLBCL patients from the identified eight subgroups D1-D8. FIGS.13A-13D display the percentage of nucleated cells of CD8, CD4, CD163, and CD20 in the DLBCL patients across the identified eight subgroups D1 through D8. [00346] From the foregoing, it will be appreciated that, although specific embodiments have been described herein for the purpose of illustration, various modifications may be made without deviating from the spirit and scope of what is provided herein. All of the references referred to above are incorporated herein by reference in their entireties.

Claims

WHAT IS CLAIMED IS: 1. A method for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining reference biological samples from each patient in a reference patient group comprising reference patients having a lymphoma; (b) clustering the reference patient group into subgroups of patients using gene expression levels in the reference biological samples; (c) obtaining a biological sample from the lymphoma patient; (d) determining to which subgroup the lymphoma patient belongs using gene expression levels in the biological sample from the lymphoma patient; and (e) predicting the responsiveness of the lymphoma patient to a first cancer treatment.
2. The method of claim 1, further comprising administering to the lymphoma patient a second cancer treatment.
3. The method of claim 1 or 2, wherein step (b) comprises generating clustering information defining relationships between the expression levels of one or more genes in the reference biological samples; and rearranging heat map representation based on the clustering information.
4. The method of any one of claims 1-3, wherein step (b) uses a hierarchical method or a non-hierarchical method.
5. The method of any one of claims 1-3, wherein step (b) uses Cluster of Cluster Analysis (COCA) method or iClusterPlus method.
6. The method of claim 5, wherein step (b) uses COCA method.
7. The method of claim 5, wherein step (b) uses iClusterPlus method.
8. The method of any one of claims 1-7, wherein the reference patients in the reference patient group are clustered into 2-15 subgroups.
9. The method of claim 8, wherein the reference patients in the reference patient group are clustered into 8 subgroups.
10. The method of any one of claim 1-9, wherein the method further comprises training a classifier model using expression levels of one or more genes in the reference biological samples.
11. The method of claim 10, wherein expression levels of one, two, three, four, five or more of the genes identified in Table 1 are used in training the classifier model.
12. The method of claim 11, wherein expression levels of all the genes identified in Table 1 are used in training the classifier model.
13. The method of any one of claims 10-12, wherein the classifier model is a grouped multinomial generalized linear model (GLM).
14. The method of any one of claims 1-13, wherein the lymphoma is diffuse large B- cell lymphoma (DLBCL).
15. The method of any one of claims 1-13, wherein the lymphoma is indolent B cell lymphoma.
16. The method of any one of claims 1-13, wherein the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma.
17. The method of 16, wherein the lymphoma is follicular lymphoma.
18. The method of 16, wherein the lymphoma is nodal marginal zone B-cell lymphoma.
19. The method of 16, wherein the lymphoma is mantle cell lymphoma.
20. The method of 16, wherein the lymphoma is chronic lymphocytic leukemia.
21. The method of any one of claims 1-14, wherein the reference patients in the reference patient group are clustered into 8 subgroups; and wherein: (i) subgroup D1 comprises about 55% to 65% patients having germinal center B- cell-like (GCB) DLBCL, about 20% to 30% patients having activated B-cell like (ABC) lymphoma, and about 20% to 30% patients who are DHITsig+ DLBCL patients; (ii) subgroup D2 comprises about 45% to 55% patients having GCB DLBCL, about 20% to 45% patients having ABC DLBCL, and about 20% to 25% patients who are DHITsig+ DLBCL patients; (iii) subgroup D3 comprises about 90% to 95% patients having GCB DLBCL, about 0% to 10% patients having ABC DLBCL, and about 0% to 20% patients who are DHITsig+ DLBCL patients; (iv) subgroup D4 comprises about 0% to 10% patients having GCB DLBCL, about 90% to 100% patients having ABC DLBCL, and about 0% to 10% patients who are DHITsig+ DLBCL patients; (v) subgroup D5 comprises about 0% to 20% patients having GCB DLBCL, about 55% to 65% patients having ABC DLBCL, and about 0% to 20% patients who are DHITsig+ DLBCL patients; (vi) subgroup D6 comprises about 50% to 60% patients having GCB DLBCL, about 20% to 40% patients having ABC DLBCL, and about 15% to 30% patients who are DHITsig+ DLBCL patients; (vii) subgroup D7 comprises about 20% to 35% patients having GCB DLBCL, about 45% to 55% patients having ABC DLBCL, and about 0% to 10% patients who are DHITsig+ DLBCL patients; and (viii) subgroup D8 comprises about 35% to 75% patients having GCB DLBCL, about 15% to 60% patients having ABC DLBCL, and about 25% to 65% patients who are DHITsig+ DLBCL patients.
22. The method of any one of claims 1-21, wherein the first cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP).
23. The method of any one of claims 1-22, wherein when the lymphoma patient is determined to belong to subgroup D4 or D8, the method comprises predicting that the patient is not likely to be responsive to the first cancer treatment.
24. The method of any one of claims 2-23, wherein the second cancer treatment is R- CHOP.
25. The method of any one of claims 2-23, wherein the second cancer treatment is not R-CHOP.
26. The method of any one of claims 2-25, wherein when the lymphoma patient is determined to belong to subgroup D4 or D8, the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor.
27. The method of any one of claims 2-25, wherein: (i) when the lymphoma patient is determined to belong to subgroup D4, the second cancer treatment is a cyclin dependent kinase (CDK) inhibitor; and (ii) when the lymphoma patient is determined to belong to subgroup D8, the second cancer treatment is a bromodomain and extra-terminal (BET) inhibitor.
28. A method for predicting the responsiveness of lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first patient having a lymphoma; (b) determining the level of expression of one, two, three, four, five or more of the genes identified in Table 1; (c) comparing the level of expression of the one, two, three, four, five or more of the genes identified in Table 1 in the first biological sample with the level of expression of the same genes in a second biological sample(s) from a second patient(s), wherein the second lymphoma patient(s) is responsive to the cancer treatment, and wherein the similar expression of the one, two, three, four, five or more of the genes in the first biological sample relative to the level of expression of the one, two, three, four, five or more of the genes in the second biological sample(s) indicates that the lymphoma in the first patient will be responsive to treatment with the cancer treatment.
29. A method for predicting the responsiveness of a lymphoma patient to a cancer treatment comprising: (a) obtaining a first biological sample from a first lymphoma patient; (b) determining the expression of the genes or a certain subset of genes set forth in Table 1 or any combination thereof in the first biological sample; and (c) comparing the gene expression profile of the genes or subset of genes in the first biological sample to (i) the gene expression profile of the genes or subset of genes in biological samples from lymphoma patients which are responsive to the drug and (ii) the gene expression of the genes or subset of genes in biological samples from lymphoma patients which are not responsive to the cancer treatment, wherein a gene expression profile for the genes or subset of genes in the first biological sample similar to the gene expression profile for the genes or subset of genes in biological samples from lymphoma patients which are responsive to the cancer treatment indicates that the first lymphoma patient will be responsive to the cancer treatment, and a gene expression profile for the genes or subset of genes in first biological sample similar to the gene expression profile for the genes or subset of genes in biological samples from lymphoma patients which are not responsive to the cancer treatment indicates that the first lymphoma patient will not be responsive to the cancer treatment.
30. A method of treating a lymphoma patient, comprising: (i) identifying a lymphoma patient who is likely to be responsive to the cancer treatment according to the method of any one of claims 28-29; and (ii) administering to the lymphoma patient the cancer treatment.
31. A method of treating a lymphoma patient, comprising: (i) identifying a lymphoma patient who is likely to be not responsive to the cancer treatment according to the method of any one of claims 28-29; and (ii) administering to the lymphoma patient an alternative cancer treatment.
32. The method of claim 30, wherein the cancer treatment is a combination treatment with rituximab (Rituxan), cyclophosphamide, doxorubicin, vincristine, and prednisone (R- CHOP).
33. The method of claim 31, wherein the altnerative cancer treatment is a bromodomain and extra-terminal (BET) inhibitor, or a cyclin dependent kinase (CDK) inhibitor.
34. The method of any one of claims 28-33, wherein the lymphoma is diffuse large B- cell lymphoma (DLBCL).
35. The method of any one of claims 28-33, wherein the lymphoma is indolent B cell lymphoma.
36. The method of any one of claims 28-33, wherein the lymphoma is selected from the group consisting of follicular lymphoma, small lymphocytic lymphoma, nodal marginal zone B-cell lymphoma, lymphoplasmacytic lymphoma, anaplastic large cell lymphoma, primary cutaneous type lymphoma, mycosis fungoides, chronic lymphocytic leukemia, and mantle cell Lymphoma.
37. The method of 36, wherein the lymphoma is follicular lymphoma.
38. The method of 36, wherein the lymphoma is nodal marginal zone B-cell lymphoma.
39. The method of 36, wherein the lymphoma is mantle cell lymphoma.
40. The method of 36, wherein the lymphoma is chronic lymphocytic leukemia.
41. The method of any one of claims 28-39, wherein the level of expression of all the genes identified in Table 1 is determined in step (b) and compared in step (c).
42. The method of any one of claims 1-41, wherein the biological samples are tumor biopsy samples.
43. The method of any one of claims 1-42, wherein the determining step comprises detecting the presence or amount of a complex in the biological sample, wherein the presence or amount of the complex indicates the expression level of the genes in each subset of genes.
44. The method of claim 43, wherein the complex is a hybridization complex.
45. The method of claim 44, wherein the hybridization complex is attached to a solid support.
46. The method of claim 43, wherein the complex is detectably labeled.
47. The method of any one of claims 1-35, wherein the determining step comprises detecting the presence or amount of a reaction product in the biological sample, wherein the presence or amount of the reaction product indicates the expression level of the genes.
48. The method of claim 47, wherein the reaction product is detectably labeled.
49. The method of any one of claims 14, 21-27, and 41-48, wherein the reference patients are refractory DLBCL patient.
50. The method of any one of claims 14, 21-27, and 41-48, wherein the reference patients are relapsed DLBCL patient.
51. The method of any one of claims 14, 21-27, and 41-48, wherein the reference patients are newly diagnosed DLBCL patient.
52. The method of any one of claims 1-14, 21-27, 33 and 40-51, wherein the lymphoma patient is a refractory DLBCL patient.
53. The method of any one of claims 14, 21-27, 33 and 40-51, wherein the lymphoma patient is a relapsed DLBCL patient.
54. The method of any one of claims 14, 21-27, 33 and 40-51, wherein the lymphoma patient is a newly diagnosed DLBCL patient.
55. The method of any one of claims 1-14, 21-34, and 41-54, wherein the lymphoma patient is a GCB DLBCL patient.
56. The method of any one of claims 1-14, 21-34, and 41-54, wherein the lymphoma patient is an ABC DLBCL patient.
57. The method of any one of claims 1-14, 21-34, and 41-56, wherein the lymphoma patient is a DHITsig+ DLBCL patient.
58. The method of any one of claims 1-14, 21-27, 33 and 40-56, wherein the lymphoma patient is a DHITsig- DLBCL patient.
PCT/US2020/057860 2019-10-30 2020-10-29 Methods for predicting responsiveness of lymphoma to drug and methods for treating lymphoma WO2021087044A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US201962928305P 2019-10-30 2019-10-30
US62/928,305 2019-10-30

Publications (1)

Publication Number Publication Date
WO2021087044A1 true WO2021087044A1 (en) 2021-05-06

Family

ID=75716333

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2020/057860 WO2021087044A1 (en) 2019-10-30 2020-10-29 Methods for predicting responsiveness of lymphoma to drug and methods for treating lymphoma

Country Status (1)

Country Link
WO (1) WO2021087044A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2022258001A1 (en) * 2021-06-10 2022-12-15 上海岸阔医药科技有限公司 Compound and method for treating chemotherapy-related gastrointestinal side effects
WO2023201348A1 (en) * 2022-04-15 2023-10-19 Celgene Corporation Methods for predicting responsiveness of lymphoma to drug and methods for treating lymphoma

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160032404A1 (en) * 2014-08-01 2016-02-04 Pharmacyclics Llc Biomarkers for predicting response of dlbcl to treatment with a btk inhibitor
US20170073762A1 (en) * 2014-03-03 2017-03-16 Assistance Publique - Hôpitaux De Paris A method for predicting the responsiveness a patient to a treatment with an anti-cd20 antibody

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170073762A1 (en) * 2014-03-03 2017-03-16 Assistance Publique - Hôpitaux De Paris A method for predicting the responsiveness a patient to a treatment with an anti-cd20 antibody
US20160032404A1 (en) * 2014-08-01 2016-02-04 Pharmacyclics Llc Biomarkers for predicting response of dlbcl to treatment with a btk inhibitor

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2022258001A1 (en) * 2021-06-10 2022-12-15 上海岸阔医药科技有限公司 Compound and method for treating chemotherapy-related gastrointestinal side effects
WO2023201348A1 (en) * 2022-04-15 2023-10-19 Celgene Corporation Methods for predicting responsiveness of lymphoma to drug and methods for treating lymphoma

Similar Documents

Publication Publication Date Title
US10870885B2 (en) Dendritic cell response gene expression, compositions of matters and methods of use thereof
AU2016256748A1 (en) Molecular profiling for cancer
US20170038387A1 (en) Methods for treating chronic lymphocytic leukemia and the use of biomarkers as a predictor of clinical sensitivity to immunomodulatory therapies
CA3134825A1 (en) Methods of treating squamous cell carcinomas with farnesyltransferase inhibitors
EP3335041A2 (en) Biomarkers for treatment of alopecia areata
WO2021087044A1 (en) Methods for predicting responsiveness of lymphoma to drug and methods for treating lymphoma
EP2447374B1 (en) In vitro method for the prognosis or prediction of the response in patients with rheumatoid arthritis treated with agents that recognize the cd20 membrane receptor in b lymphocytes
US20200000796A1 (en) Methods of treating cancer with farnesyltransferase inhibitors
AU2019203548C1 (en) Methods of treating cancer with farnesyltransferase inhibitors
US20190119758A1 (en) Methods of selecting cancer patients for treatment with farnesyltransferase inhibitors
US20220242955A1 (en) Targeting galectin-9 as a therapeutic strategy for t-cell exhaustion in t-cell acute lymphoblastic leukemia
US20220305001A1 (en) Methods of treating acute myeloid leukemia with farnesyltransferase inhibitors
US11291663B2 (en) Methods of treating cancer with farnesyltransferase inhibitors
US10806730B2 (en) Methods of treating cancer with farnesyltransferase inhibitors
US20170306402A1 (en) Systems and methods for characterization of multiple sclerosis
WO2023201348A1 (en) Methods for predicting responsiveness of lymphoma to drug and methods for treating lymphoma
AU2018313738B2 (en) Methods of treating cancer with farnesyltransferase inhibitors
US20220168296A1 (en) Methods of treating cancer with farnesyltransferase inhibitors
US20170326133A1 (en) Methods of treating myelodysplastic syndrome with farnesyltransferase inhibitors
EP4208721A1 (en) Methods for predicting the response of a patient to treatment with a pd-1 or pd-l1 immune checkpoint inhibitor
EA043425B1 (en) LEVELS OF IMMUNOGLOBULIN EXPRESSION AS A BIOLOGICAL MARKER FOR RESPONSE TO A PROTEASOMAL INHIBITOR

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20881238

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20881238

Country of ref document: EP

Kind code of ref document: A1