Method for treating rheumatoid arthritis

Abstract

A method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a first, second and / or third marker panel in a sample from the patient, wherein the agent is selected from an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy.

Description

[0001] METHOD FOR TREATING RHEUMATOID ARTHRITIS FIELD OF THE INVENTION The invention relates to a method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient. The invention further relates to a method for determining whether an RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling (such as etanercept), an agent that downregulates IL-6 mediated signalling (such as tocilizumab) and / or a B cell targeted therapy (such as rituximab). The invention also relates to methods for treating RA patients that are determined to be susceptible or refractory by a method of the invention. BACKGROUND TO THE INVENTION Inflammatory arthritis is a prominent clinical manifestation in diverse autoimmune disorders including rheumatoid arthritis (RA), psoriatic arthritis (PsA), systemic lupus erythematosus (SLE), Sjogren's syndrome and polymyositis. RA is a chronic inflammatory disease that affects approximately 0.5 to 1% of the adult population in northern Europe and North America. It is a systemic inflammatory disease characterized by chronic inflammation in the synovial membrane of affected joints, which ultimately leads to loss of daily function due to chronic pain and fatigue. The majority of patients also experience progressive deterioration of cartilage and bone in the affected joints, which may eventually lead to permanent disability. The long-term prognosis of RA is poor, with approximately 50% of patients experiencing significant functional disability within 10 years from the time of diagnosis. Life expectancy is reduced by an average of 3-10 years. Inflammatory bone diseases, such as RA, are accompanied by bone loss around affected joints due to increased osteoclastic resorption. This process is mediated largely by increased local production of pro-inflammatory cytokines, of which tumor necrosis factor-α (TNF-α) is a major effector. In RA specifically, an immune response is thought to be initiated / perpetuated by one or several antigens presenting in the synovial compartment, producing an influx of acute inflammatory cells and lymphocytes into the joint. Successive waves of inflammation lead to the formation of an invasive and erosive tissue called pannus. This contains proliferating fibroblast-like synoviocytes and macrophages that produce proinflammatory cytokines such as TNF-α and interleukin-1 (IL-I). Local release of proteolytic enzymes, various inflammatory mediators, and osteoclast activation contributes to much of the tissue damage. There is loss of articular cartilage and the formation of bony erosions. Surrounding tendons and bursa may become affected by the inflammatory process. Ultimately, the integrity of the joint structure is compromised, producing disability. B cells are thought to contribute to the immunopathogenesis of RA, predominantly by serving as the precursors of autoantibody-producing cells but also as antigen presenting cells (APC) and pro-inflammatory cytokine producing cells. A number of autoantibody specificities have been identified including antibodies to Type II collagen and proteoglycans, as well as rheumatoid factors and most importantly anti citrullinated protein antibodies (ACPA). The generation of large quantities of antibody leads to immune complex formation and the activation of the complement cascade. This in turn amplifies the immune response and may culminate in local cell lysis. Current standard therapies for RA which are used to modify the disease process and to delay joint destruction are known as disease modifying anti-rheumatic drugs (DMARDs). Methotrexate, leflunomide and sulfasalazine are traditional DMARDs and are often effective as first-line treatment. Biologic agents designed to target specific components of the immune system that play a role in RA are also used as therapeutics. There are various groups of biologic treatments for RA including; TNF-α inhibitors (etanercept, infliximab and adalimumab), human IL-1 receptor antagonist (anakinra) and selective co-stimulation modulators (abatacept). RA patients receive highly-targeted biologic therapies without prior knowledge of target expression levels in the diseased tissue. Approximately 40% of patients do not respond to individual biologic therapies and 5-20% are refractory to all. The mechanisms of non-response are largely unknown and, unlike other medical fields such as cancer where molecular pathology guides the use of targeted therapies, RA targeted therapeutics are prescribed “blindly” and irrespectively of the target expression levels in the diseased tissue. Biologic therapies for RA may be associated with various safety issues, especially infusion- related adverse events and are also very expensive, for example rituximab costs approximately USD 10000 per treatment course. Accordingly, there is a need for methods of predicting whether a given RA patient is likely to respond to a type of treatment. SUMMARY OF THE INVENTION The inventors have applied deep molecular phenotyping and machine learning modelling to synovial RNA-seq from patients included in the biopsy-driven, stratified-medicine STRAP trial. The inventors identified differentially expressed genes between responders and non- responders and examined the relationship between single cell RNA-Seq cell subset modules and response to each drug using deconvolution. Additionally, the inventors defined predictive RNA-seq-based models of response for the three randomized drug types in STRAP, that, when converted into nanoString panels tested in available synovial RNA from STRAP patients, could accurately predict actual observed response in 79-85% of patients. In one aspect, the invention provides a method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a first, second and / or third marker panel in a sample from the patient, wherein the agent is selected from an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, and wherein: (i) the first marker panel comprises one or more markers selected from Table 1.1, wherein the agent is an agent that downregulates TNF mediated signalling; (ii) the second marker panel comprises one or more markers selected from Table 2.1, wherein the agent is an agent that downregulates IL-6 mediated signalling; and (iii) the third marker panel comprises one or more markers selected from Table 3.1, wherein the agent is a B cell targeted therapy. In one aspect, the invention provides a method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a first, second and / or third marker panel in a sample from the patient, wherein the agent is selected from an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, and wherein: (i) the first marker panel comprises one or more markers selected from Table 1.2, wherein the agent is an agent that downregulates TNF mediated signalling; (ii) the second marker panel comprises one or more markers selected from Table 2.2, wherein the agent is an agent that downregulates IL-6 mediated signalling; and (iii) the third marker panel comprises one or more markers selected from Table 3.2, wherein the agent is a B cell targeted therapy. In one aspect, the invention provides a method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy, the method comprising determining the profile of a first, second and / or third marker panel in a sample from the patient, and wherein: (i) the first marker panel comprises one or more markers selected from Table 1.1, wherein the agent is an agent that downregulates TNF mediated signalling; (ii) the second marker panel comprises one or more markers selected from Table 2.1, wherein the agent is an agent that downregulates IL-6 mediated signalling; and (iii) the third marker panel comprises one or more markers selected from Table 3.1, wherein the agent is a B cell targeted therapy. In one aspect, the invention provides a method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy, the method comprising determining the profile of a first, second and / or third marker panel in a sample from the patient, and wherein: (i) the first marker panel comprises one or more markers selected from Table 1.2, wherein the agent is an agent that downregulates TNF mediated signalling; (ii) the second marker panel comprises one or more markers selected from Table 2.2, wherein the agent is an agent that downregulates IL-6 mediated signalling; and (iii) the third marker panel comprises one or more markers selected from Table 3.2, wherein the agent is a B cell targeted therapy. In some embodiments, the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, the method comprising determining the profile of the first marker panel in a sample from the patient. In some embodiments, the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates IL-6 mediated signalling, the method comprising determining the profile of the second marker panel in a sample from the patient. In some embodiments, the method is for determining whether a RA patient is susceptible or refractory to treatment with a B cell targeted therapy, the method comprising determining the profile of the third marker panel in a sample from the patient. In some embodiments, the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling and an agent that downregulates IL-6 mediated signalling, the method comprising determining the profile of the first marker panel and the second marker panel in a sample from the patient. In some embodiments, the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, the method comprising determining the profile of the second marker panel and the third marker panel in a sample from the patient. In some embodiments, the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling and a B cell targeted therapy, the method comprising determining the profile of the first marker panel and the third marker panel in a sample from the patient. In some embodiments, the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, the method comprising determining the profile of the first, second and third marker panels in a sample from the patient. In some embodiments, the first marker panel comprises EIF2B1. In some embodiments, the first marker panel comprises EIF2B1, and CENPBD1. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, and PWWP2B. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, PWWP2B, and AGK. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, PWWP2B, AGK, and RAD52. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, PWWP2B, AGK, RAD52, and ZNF765. In preferred embodiments, the first marker panel comprises EIF2B1, CENPBD1, PWWP2B, AGK, RAD52, ZNF765, and ZNF850. In some embodiments, the first marker panel comprises EIF2B1. In some embodiments, the first marker panel comprises EIF2B1, and CENPBD1. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, and AGK. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, AGK, and RAD52. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, AGK, RAD52, and ZNF765. In preferred embodiments, the first marker panel comprises EIF2B1, CENPBD1, AGK, RAD52, ZNF765, and ZNF850. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, and PWWP2B. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, PWWP2B, and AL161785.1. In preferred embodiments, the first marker panel comprises EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, PWWP2B, AL161785.1, and TRPC6. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, PWWP2B, AL161785.1, TRPC6, and CCDC102B. In preferred embodiments, the first marker panel comprises EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, PWWP2B, AL161785.1, TRPC6, CCDC102B, and IL13RA2. In some embodiments, the first marker panel consists of EIF2B1. In some embodiments, the first marker panel consists of EIF2B1, and CENPBD1. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, and PWWP2B. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, PWWP2B, and AGK. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, PWWP2B, AGK, and RAD52. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, PWWP2B, AGK, RAD52, and ZNF765. In preferred embodiments, the first marker panel consists of EIF2B1, CENPBD1, PWWP2B, AGK, RAD52, ZNF765, and ZNF850. In some embodiments, the first marker panel consists of EIF2B1. In some embodiments, the first marker panel consists of EIF2B1, and CENPBD1. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, and AGK. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, AGK, and RAD52. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, AGK, RAD52, and ZNF765. In preferred embodiments, the first marker panel consists of EIF2B1, CENPBD1, AGK, RAD52, ZNF765, and ZNF850. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, and PWWP2B. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, PWWP2B, and AL161785.1. In preferred embodiments, the first marker panel consists of EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, PWWP2B, AL161785.1, and TRPC6. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, PWWP2B, AL161785.1, TRPC6, and CCDC102B. In preferred embodiments, the first marker panel consists of EIF2B1, CENPBD1, AGK, RAD52, ZNF765, ZNF850, PWWP2B, AL161785.1, TRPC6, CCDC102B, and IL13RA2. In some embodiments, the first marker panel comprises EIF2B1, CENPBD1, PWWP2B, AGK, RAD52, ZNF765, ZNF850, AL161785.1, TRPC6, CCDC102B, SYT1, IL13RA2, HTR2A, IFI44L, and CSF3R. In some embodiments, the first marker panel consists of EIF2B1, CENPBD1, PWWP2B, AGK, RAD52, ZNF765, ZNF850, AL161785.1, TRPC6, CCDC102B, SYT1, IL13RA2, HTR2A, IFI44L, and CSF3R. In some embodiments, the first marker panel comprises all of the markers set forth in Table 1.1. In some embodiments, the first marker panel comprises all of the markers set forth in Table 1.2. In some embodiments, the second marker panel comprises GPNMB. In some embodiments, the second marker panel comprises GPNMB, and MICU1. In some embodiments, the second marker panel comprises GPNMB, MICU1, and NKG7. In some embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, and TOX2. In some embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, and GZMH. In some embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, and CD40LG. In some embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, and SOCS6. In some embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, and CD247. In some embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, and CDKN1B. In some embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, and KLF2. In some embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, KLF2, and TCIM. In preferred embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, KLF2, TCIM, and MERTK. In some embodiments, the second marker panel comprises GPNMB. In some embodiments, the second marker panel comprises GPNMB, and MICU1. In some embodiments, the second marker panel comprises GPNMB, MICU1, and TOX2. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, and CD40LG. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, and SOCS6. In preferred embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, and CD247. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, and NKG7. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, and GZMH. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, and CDKN1B. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, and SP140. In preferred embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, and CD3G. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, and TCIM. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, TCIM, and BLK. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, TCIM, BLK, and FCRL3. In some embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, TCIM, BLK, FCRL3, and KCNA3. In preferred embodiments, the second marker panel comprises GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, TCIM, BLK, FCRL3, KCNA3, and MERTK. In some embodiments, the second marker panel consists of GPNMB. In some embodiments, the second marker panel consists of GPNMB, and MICU1. In some embodiments, the second marker panel consists of GPNMB, MICU1, and NKG7. In some embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, and TOX2. In some embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, TOX2, and GZMH. In some embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, TOX2, GZMH, and CD40LG. In some embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, and SOCS6. In some embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, and CD247. In some embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, and CDKN1B. In some embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, and KLF2. In some embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, KLF2, and TCIM. In preferred embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, KLF2, TCIM, and MERTK. In some embodiments, the second marker panel consists of GPNMB. In some embodiments, the second marker panel consists of GPNMB, and MICU1. In some embodiments, the second marker panel consists of GPNMB, MICU1, and TOX2. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, and CD40LG. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, and SOCS6. In preferred embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, and CD247. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, and NKG7. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, and GZMH. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, and CDKN1B. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, and SP140. In preferred embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, and CD3G. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, and TCIM. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, TCIM, and BLK. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, TCIM, BLK, and FCRL3. In some embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, TCIM, BLK, FCRL3, and KCNA3. In preferred embodiments, the second marker panel consists of GPNMB, MICU1, TOX2, CD40LG, SOCS6, CD247, NKG7, GZMH, CDKN1B, SP140, CD3G, TCIM, BLK, FCRL3, KCNA3, and MERTK. In some embodiments, the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, KLF2, TCIM, MERTK, PTGS2, BLK, SP140, IL2RB, CST7, CD79B, YPEL5, TRPC6, FCRL3, CD3G, IL24, IL23A, KCNA3, CD2, CASP10, ICOS, LTB, LTA, CD6, CD8A, CD3D, and CD3E. In some embodiments, the second marker panel consists of GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, KLF2, TCIM, MERTK, PTGS2, BLK, SP140, IL2RB, CST7, CD79B, YPEL5, TRPC6, FCRL3, CD3G, IL24, IL23A, KCNA3, CD2, CASP10, ICOS, LTB, LTA, CD6, CD8A, CD3D, and CD3E. In some embodiments, the second marker panel comprises all of the markers set forth in Table 2.1. In some embodiments, the second marker panel comprises all of the markers set forth in Table 2.2. In some embodiments, the third biomarker panel comprises CISH. In preferred embodiments, the third biomarker panel comprises CISH, and PPFIA4. In preferred embodiments, the third biomarker panel comprises MICU1. In some embodiments, the third biomarker panel comprises MICU1, and CISH. In some embodiments, the third biomarker panel comprises MICU1, CISH, and PPFIA4. In some embodiments, the third biomarker panel comprises MICU1, CISH, PPFIA4, and MAPK12. In some embodiments, the third biomarker panel comprises MICU1, CISH, PPFIA4, MAPK12, and IL17C. In preferred embodiments, the third biomarker panel comprises MICU1, CISH, PPFIA4, MAPK12, IL17C, and IL5. In some embodiments, the third biomarker panel comprises MICU1, CISH, PPFIA4, MAPK12, IL17C, IL5, and ZNF773. In some embodiments, the third biomarker panel comprises MICU1, CISH, PPFIA4, MAPK12, IL17C, IL5, ZNF773, and MMP9. In some embodiments, the third biomarker panel comprises MICU1, CISH, PPFIA4, MAPK12, IL17C, IL5, ZNF773, MMP9, and IL22. In some embodiments, the third biomarker panel comprises MICU1, CISH, PPFIA4, MAPK12, IL17C, IL5, ZNF773, MMP9, IL22, and CENPH. In some embodiments, the third biomarker panel consists of CISH. In preferred embodiments, the third biomarker panel consists of CISH, and PPFIA4. In preferred embodiments, the third biomarker panel consists of MICU1. In some embodiments, the third biomarker panel consists of MICU1, and CISH. In some embodiments, the third biomarker panel consists of MICU1, CISH, and PPFIA4. In some embodiments, the third biomarker panel consists of MICU1, CISH, PPFIA4, and MAPK12. In some embodiments, the third biomarker panel consists of MICU1, CISH, PPFIA4, MAPK12, and IL17C. In preferred embodiments, the third biomarker panel consists of MICU1, CISH, PPFIA4, MAPK12, IL17C, and IL5. In some embodiments, the third biomarker panel consists of MICU1, CISH, PPFIA4, MAPK12, IL17C, IL5, and ZNF773. In some embodiments, the third biomarker panel consists of MICU1, CISH, PPFIA4, MAPK12, IL17C, IL5, ZNF773, and MMP9. In some embodiments, the third biomarker panel consists of MICU1, CISH, PPFIA4, MAPK12, IL17C, IL5, ZNF773, MMP9, and IL22. In some embodiments, the third biomarker panel consists of MICU1, CISH, PPFIA4, MAPK12, IL17C, IL5, ZNF773, MMP9, IL22, and CENPH. In some embodiments, the third biomarker panel comprises CISH, PPFIA4, ZNF773, ZNF823, MICU1, BCL2, ATF3, IFNL1, IL12B, CREB3, CFLAR, STAT6, MAPK12, P2RY13, ERBB3, IL17C, ZNF331, SHC1, IL22, COL4A6, IFNB1, ZNF75A, CENPH, IL5, and IL20. In some embodiments, the third biomarker panel consists of CISH, PPFIA4, ZNF773, ZNF823, MICU1, BCL2, ATF3, IFNL1, IL12B, CREB3, CFLAR, STAT6, MAPK12, P2RY13, ERBB3, IL17C, ZNF331, SHC1, IL22, COL4A6, IFNB1, ZNF75A, CENPH, IL5, and IL20. In some embodiments, the third biomarker panel comprises all of the biomarkers set forth in Table 3.1. In some embodiments, the third biomarker panel comprises all of the biomarkers set forth in Table 3.2. In some embodiments, determining the profile of a first, second and / or third marker panel in a sample from the patient comprises determining the level of the markers of the first, second and / or third marker panel in the sample from the subject and comparing the level of the markers to the corresponding reference value. In some embodiments, determining the profile of a first marker panel in a sample from the patient comprises determining the level of the markers of the first marker panel in the sample from the subject and comparing the level of the markers to the corresponding reference value. In some embodiments, determining the profile of a second marker panel in a sample from the patient comprises determining the level of the markers of the second marker panel in the sample from the subject and comparing the level of the markers to the corresponding reference value. In some embodiments, determining the profile of a third marker panel in a sample from the patient comprises determining the level of the markers of the third marker panel in the sample from the subject and comparing the level of the markers to the corresponding reference value. In some embodiments: (a) (i) an increased level of one or more markers from Table 1.1.A or 1.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates TNF mediated signalling; (ii) an increased level of one or more markers from Table 1.1.B or 1.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates TNF mediated signalling; (b) (i) an increased level of one or more markers from Table 2.1.A or 2.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates IL-6 mediated signalling; (ii) an increased level of one or more markers from Table 2.1.B or 2.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates IL-6 mediated signalling; and / or (c) (i) an increased level of one or more markers from Table 3.1.A or 3.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with a B cell targeted therapy; (ii) an increased level of one or more markers from Table 3.1.B or 3.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with a B cell targeted therapy. In some embodiments, an increased level of one or more markers from Table 1.1.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates TNF mediated signalling. In some embodiments, an increased level of one or more markers from Table 1.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates TNF mediated signalling. In some embodiments, an increased level of one or more markers from Table 1.1.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates TNF mediated signalling. In some embodiments, an increased level of one or more markers from Table 1.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates TNF mediated signalling. In some embodiments, an increased level of one or more markers from Table 2.1.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates IL-6 mediated signalling. In some embodiments, an increased level of one or more markers from Table 2.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates IL-6 mediated signalling. In some embodiments, an increased level of one or more markers from Table 2.1.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates IL-6 mediated signalling. In some embodiments, an increased level of one or more markers from Table 2.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates IL-6 mediated signalling. In some embodiments, an increased level of one or more markers from Table 3.1.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with a B cell targeted therapy. In some embodiments, an increased level of one or more markers from Table 3.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with a B cell targeted therapy. In some embodiments, an increased level of one or more markers from Table 3.1.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with a B cell targeted therapy. In some embodiments, an increased level of one or more markers from Table 3.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with a B cell targeted therapy. In some embodiments, determining the profile of the first, second and / or third marker panels comprises determining the nucleic acid levels of the markers of the first, second and / or third marker panels. In some embodiments, determining the profile of the first marker panel comprises determining the nucleic acid levels of the markers of the first marker panel. In some embodiments, determining the profile of the second marker panel comprises determining the nucleic acid levels of the markers of the second marker panel. In some embodiments, determining the profile of the third marker panel comprises determining the nucleic acid levels of the markers of the third marker panel. In some embodiments, the nucleic acid level is an mRNA level. In some embodiments, determining the profile of the first, second and / or third marker panels is performed by direct digital counting of nucleic acids, RNA-seq, RT-qPCR, qPCR, multiplex qPCR or RT-qPCR, microarray analysis, or a combination thereof. In some embodiments, determining the profile of the first, second and / or third marker panels is performed by RNA sequencing. In some embodiments, determining the profile of the first, second and / or third marker panels comprises determining the level of gene expression of the markers of the first, second and / or third marker panels. In some embodiments, the determining the level of gene expression is using a multiplexed gene expression assay (e.g. an nCounter assay). In some embodiments, the multiplexed gene expression assay uses the gene panel of Table 5. Suitably, the multiplexed gene expression assay may additionally use some or all (e.g. all) of the housekeeping genes of Table 6. In some embodiments, the one or more sample is a synovial sample. In some embodiments, the agent that downregulates TNF mediated signalling is a TNF inhibitor. In some embodiments, the agent that downregulates TNF mediated signalling is selected from etanercept, infliximab, adalimumab, certolizumab pegol, and golimumab. In preferred embodiments, the agent that downregulates TNF mediated signalling is etanercept. In some embodiments, the agent that downregulates IL-6 mediated signalling is an IL-6 receptor antagonist. In some embodiments, the agent that downregulates IL-6 mediated signalling is selected from the group consisting of tocilizumab, sarilumab, satralizumab and siltuximab. In preferred embodiments, the agent that downregulates IL-6 mediated signalling is tocilizumab. In some embodiments, the B cell targeted therapy is B cell depletion therapy. In some embodiments, the B cell targeted therapy is selected from the group consisting of: rituximab, ocrelizumab, veltuzumab, ofatumumab, epratuzumab, obinutuzumab, ibritumomab and tiuxetan. In preferred embodiments, the B cell targeted therapy is rituximab. In some embodiments, the patient is refractory to DMARD therapy. In some embodiments, the method further comprises administering to the patient an agent that downregulates TNF mediated signalling when the patient is determined to be susceptible to treatment with an agent that downregulates TNF mediated signalling. In some embodiments, the method further comprises administering to the patient an agent that downregulates IL-6 mediated signalling when the patient is determined to be susceptible to treatment with an agent that downregulates IL-6 mediated signalling. In some embodiments, the method further comprises administering to the patient a B cell targeted therapy when the patient is determined to be susceptible to treatment with a B cell targeted therapy. In some embodiments, the method further comprises administering to the patient an alternative therapeutic when the patient is determined to be refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling, and a B cell targeted therapy. In some embodiments, the method comprises using at least one predictive modelling approach to identify the agent or to determine whether the patient is susceptible or refractory to treatment. In some embodiments, the at least one predictive modelling approach is gradient boosted machine (gbm), elastic net regression (glmnet), mixed discriminant analysis (mda), random forest (rf), support vector machine (svm) with polynomial (svmPoly) or radial (svmRadial) kernel, extreme gradient boosting (xgboost) with tree booster (xgbTree) or linear booster (xgbLinear). In some embodiments, populations of patients grouped according to clinical response are used as inputs to the predictive modelling approach. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of an agent that downregulates TNF mediated signalling, wherein the agent is determined to be suitable for treatment by the method of the invention. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of an agent that downregulates TNF mediated signalling, wherein the patient is determined to be susceptible to treatment with an agent that downregulates TNF mediated signalling by the method of the invention. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of an agent that downregulates IL-6 mediated signalling, wherein the agent is determined to be suitable for treatment by the method of the invention. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of an agent that downregulates IL-6 mediated signalling, wherein the patient is determined to be susceptible to treatment with an agent that downregulates IL- 6 mediated signalling by the method of the invention. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of a B cell targeted therapy, wherein the agent is determined to be suitable for treatment by the method of the invention. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of a B cell targeted therapy, wherein the patient is determined to be susceptible to treatment with a B cell targeted therapy by the method of the invention. In one aspect, the invention provides an agent that downregulates TNF mediated signalling for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the agent is determined to be suitable for treatment by the method of the invention. In one aspect, the invention provides an agent that downregulates TNF mediated signalling for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the patient is determined to be susceptible to treatment with an agent that downregulates TNF mediated signalling by the method of the invention. In one aspect, the invention provides an agent that downregulates IL-6 mediated signalling for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the agent is determined to be suitable for treatment by the method of the invention. In one aspect, the invention provides an agent that downregulates IL-6 mediated signalling for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the patient is determined to be susceptible to treatment with an agent that downregulates IL-6 mediated signalling by the method of the invention. In one aspect, the invention provides a B cell targeted therapy for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the B cell targeted therapy is determined to be suitable for treatment by the method of the invention. In one aspect, the invention provides a B cell targeted therapy for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the patient is determined to be susceptible to treatment with a B cell targeted therapy by the method of the invention. In one aspect, the invention provides a kit for use in the method of the invention. In some embodiments, the kit comprises one or more reagent suitable for detecting the marker(s) of the first, second and / or third marker panels. In some embodiments, the kit comprises reagents for RNA sequencing. In some embodiments, the kit comprises one or more probe or antibody for detecting the markers of the first, second and / or third marker panels. In some embodiments, the kit is in the form of a microchip or microarray. In one aspect, the invention provides a computer-readable medium comprising instructions that when executed cause one or more processors to perform the method of the invention. In one aspect, the invention provides an apparatus comprising: (a) profile determination circuitry to determine the profile of a first, second and / or third panel of markers in a sample from a Rheumatoid Arthritis (RA) patient; and (b) (i) agent identification circuitry to identify, based on the profile determination circuitry, a suitability of an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy for treating RA in the patient; or (ii) patient identification circuitry to identify, based on the profile determination circuitry, a suitability of the patient for treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy. DESCRIPTION OF THE DRAWINGS FIGURE 1. Synovial histological markers associate with baseline genes in response signature a, c and e, Volcano plots of differentially expressed genes from DESeq2 analysis of synovial transcriptomics data at baseline of individuals receiving treatment with a) Etanercept (n=67), c) tocilizumab (n=69), or e) rituximab (n=72). Genes in blue are significant at FDR<0.05 and genes in grey are non-significant. b, d and f, Modular analysis applying QuSAGE with blood microarray-based modulesand WGCNA modules for responders versus non-responders to etanercept (b), tocilizumab (d) and rituximab (f). Log2-fold change of responders (positive values) and non-responders (negative values) are plotted with dots colour coded for q-value (b) or unadjusted p-value (d, f). FIGURE 2. Analysis of common and differential molecular signatures of responsiveness / resistance to tocilizumab, etanercept and rituximab a. Volcano plot showing differentially expressed genes between all ACR20 responder (n=133) and non-responder (n=75) patients to all drugs combined (tocilizumab, etanercept & rituximab) at primary endpoint (16 weeks) to look for shared response genes. b. Shared enriched pathways across responders and non-responders to all drugs combined. Gray dashed lines indicate adjusted p^=^0.05. c.3-way polar plot comparing genes associated with resistance to each individual drug. Red genes (n=103) are significantly upregulated only in non-responder patients treated with etanercept. Genes in green (n=12) are significantly upregulated in rituximab non-responders only. Blue genes (n=9) are significantly upregulated in the tocilizumab non-responder group. d. 3-way polar plot of significantly upregulated genes in responder patients to etanercept (58 genes, red dots), rituximab (24 genes, green dots) and tocilizumab (25 genes, blue dots). e. Forest plots of individual genes showing different log2fold change (responders / non-responders) values for each drug. * P<0.05 , ** P<0.01, *** P<0.001 using FDR adjusted p-values by Wald chi-squared test. FIGURE 3. Single-cell subset patterns of responder and non-responder patients to tocilizumab, etanercept and rituximab a, Forest plot showing fold-changes of single-cell subsets that are differentially present in responders compared to non-responders separately in each medication. b, Forest plot showing fold-changes of single-cell subsets that are differentially present in any responders compared to any non-responders and any responders versus patients difficult to treat. Significant fold-changes are indicated with asterisks (* p < 0.05, ** p < 0.01, *** p < 0.001). c, Boxplots showing modules scores of SC-B1 (IGHD+CD27−naive B-cell), SC-T3 (PD-1+Tph / Tfh), SC-T6 (GZMK+ / GZMB+) T-cell, SC-M1 (IL1B+pro-inflammatory) and SC-M4 (IFN- activated) macrophage subsets for etanercept, rituximab and any treatment (either etanercept, rituximab or tocilizumab) groups. d, Heatmap showing estimated immune cell subset profiles of all individuals at baseline. Individuals (columns) were clustered using Euclidean distance metric and complete linkage clustering method. Upper tracks show ESR, CRP, cell type (B cell rich / poor), pathotype, ACR20 and ACR50 response, randomised medication (treatment) and histological scores for CD3, CD20, CD138, CD68L (lining) and CD68SL (sublining). FIGURE 4. Machine learning predictive models fitted using ten-by-ten-fold cross- validation for response to etanercept, tocilizumab and rituximab a, Multiple types of machine learning (ML) models were fitted to synovial RNA-Seq gene expression data at baseline with the aim of predicting response to each of the 3 trial drugs at the 16 week primary endpoint. Response was defined as by DAS28-ESR < 3.2. Model types: gradient boosted machine (gbm), elastic net regression (glmnet), mixed discriminant analysis (mda), random forest (rf), support vector machine (svm) with polynomial (svmPoly) or radial (svmRadial) kernel, extreme gradient boosting (xgboost) with tree booster (xgbTree) or linear booster (xgbLinear). Unbiased model performance was determined by 10×10-fold nested cross-validation (CV) with 25 repeats, with area under ROC (receiver operating characteristic) curve as performance metric for etanercept and tocilizumab. Coefficient of determination R2was used as performance metric for rituximab models (see Methods) which were fitted to an ordinal (4-level) response outcome as this led to improved final binary response prediction. b, ROC curves for final best models for each drug showing nested CV ROC and ROC calculated from inner CV folds. c, Variable importance plots showing stability of variables selected by final ML model for each drug across nested CV. Error bars show standard error of mean variable importance, size of points shows frequency with which each gene / predictor was selected by models during nested CV. Colour of points shows directionality of association with response: red for genes / predictors upregulated in non-response, blue for genes / predictors upregulated in response. FIGURE 5. Validation of machine learning models using nCounter assay a, Flow diagram outlining the process of converting the RNA-Seq models to a workable nanostring nCounter based assay. Spare baseline synovial biopsy samples from STRAP were subjected to nCounter assay using a custom synovial 506-gene panel. nCounter data was rescaled to RNA-Seq scale (“pseudo-RNA-Seq”) using linear models for each gene. Rescaled nCounter data was passed to machine learning models from Figure 4c and performance of each model was assessed. b, Confusion matrices showing predicted versus actual performance, accuracy and balanced accuracy of nCounter assay applied to baseline synovial biopsies for prediction of response defined as DAS28-ESR < 3.2 after 16 weeks of treatment. c, receiver operating characteristic (ROC) curve plots and area under curve (AUC) measurements for prediction of response to etanercept, tocilizumab and rituximab from nCounter assay applied to baseline synovial biopsies from STRAP. d, proposed algorithm for allocation of a new patient to one of three possible biologic therapies categories (TNF inhibitor, IL6 inhibitor or B-cell depleting agent) dependent on whichever gives the highest probability of response. Individuals with low probability (<0.5) of response to all three classes of biologic are categorised as “biomarker negative” and can be offered an alternative class of therapeutic agent. FIGURE 6. Unsupervised projection of RNA-Seq for data quality control a, Two-dimensional representation of RNA-Seq data displays outlier samples and explains their association with low RNA quality and consequently with low mapping rates. b, PCA of synovial transcriptome showed a clear separation between B-cell rich and poor groups. Similar clustering was also observed for pathotypes where lympho-myeloid and fibroid / pauci-immune groups formed clear clusters, whereas the diffuse-myeloid partially overlapped with these two groups. c, Expression profile of the gene set used to calculate B-cell module scores and to generate B-cell rich / poor categories. Column annotation tracks show B-cell module score and cell type (B^cell rich / poor) quantified using both RNA-Seq and histopathology data, pathotype classification and semi-quantitative histological scores of pathotype markers (CD20 and CD138). FIGURE 7. Machine learning models comparing different response outcomes 25× repeated 10×10-fold nested cross-validation applied to machine learning models for a) etanercept, b) tocilizumab c) rituximab to identify which response outcome resulted in the best predictive models. Model performance for each repeat was measured by area under receiver operating characteristic (ROC) curve (AUC) analysis. Model types: gradient boosted machine (gbm), elastic net regression (glmnet), mixed discriminant analysis (mda), random forest (rf), support vector machine (svm) with polynomial (svmPoly) or radial (svmRadial) kernel, extreme gradient boosting (xgboost) with tree booster (xgbTree) or linear booster (xgbLinear). FIGURE 8. Correlation between nCounter and RNA-Seq applied to synovial biopsies from STRAP a, Frequency histogram for Pearson correlation coefficients for each gene comparing correlation between nCounter and bulk RNA-Seq applied to synovial biopsies from the STRAP cohort. Median Pearson correlation coefficient is shown by dashed red line, showing that the majority of genes show high levels of correlation with RNA-Seq. b, Scatter plot showing that correlation between nCounter and RNA-Seq tends to decrease when mean nCounter expression is low. DETAILED DESCRIPTION OF THE INVENTION The terms “comprising”, “comprises” and “comprised of” as used herein are synonymous with “including” or “includes”; or “containing” or “contains”, and are inclusive or open-ended and do not exclude additional, non-recited members, elements or steps. The terms “comprising”, “comprises” and “comprised of” also include the term “consisting of”. Rheumatoid arthritis (RA) Rheumatoid arthritis (RA) is a chronic, systemic inflammatory disorder that may affect many tissues and organs, but principally attacks synovial joints. It is a disabling and painful condition, which can lead to substantial loss of functioning and mobility if not adequately treated. The disease process involves an inflammatory response of the synovium, secondary to massive immune cell infiltration and proliferation of synovial cells, excess synovial fluid, and the development of fibrous tissue (pannus) in the synovium that attacks the cartilage and sub- chondral bone. This often leads to the destruction of articular cartilage and the formation of bone erosions with secondary ankylosis (fusion) of the joints. RA can also produce diffuse inflammation in the lungs, the pericardium, the pleura, the sclera, and also nodular lesions, most commonly in subcutaneous tissue. RA is considered a systemic autoimmune disease as autoimmunity plays a pivotal role in its chronicity and progression. A number of cell types are involved in the aetiology of RA, including T cells, B cells, monocytes, macrophages, dendritic cells and synovial fibroblasts. Autoantibodies known to be associated with RA include those targeting Rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPA). RA therapy A typical patient with newly diagnosed RA is often treated initially with nonsteroidal anti- inflammatory drugs and disease-modifying anti-rheumatic drugs (DMARDs), such as hydroychloroquine, sulfasalazine, leflunomide or methotrexate (MTX), alone or in combinations. Patients who do not respond to general DMARDs may be termed DMARD- refractory. DMARD-refractory patients are traditionally often progressed to biological therapeutic agents, for example TNF-α antagonists such as Adalimumab, Etanercept, Golimumab and Infliximab. The method of the invention may be performed on a sample from a RA patient who has previously been determined to be refractory to DMARD-therapy. TNF mediated signalling The method of the invention may determine an RA patient as being susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling. In some embodiments, the agent that downregulates TNF mediated signalling is a TNF inhibitor. In some embodiments, the agent that downregulates TNF mediated signalling is selected from etanercept, infliximab, adalimumab, certolizumab pegol, and golimumab. In preferred embodiments, the agent that downregulates TNF mediated signalling is etanercept. IL-6 mediated signalling The method of the invention may determine an RA patient as being susceptible or refractory to treatment with an agent that downregulates interleukin-6 (IL-6) mediated signalling. IL-6 is a cytokine that provokes a broad range of cellular and physiological responses, including inflammation, hematopoiesis and oncogenesis by regulating cell growth, gene activation, proliferation, survival, and differentiation. It is able to directly influence B cell activation state and late stage differentiation towards plasma cells. IL-6 signals through a receptor composed of two different subunits, an alpha subunit that produces ligand specificity and GP (Glycoprotein) 130, a receptor subunit shared in common with other cytokines in the IL-6 family. Binding of IL-6 to its receptor initiates cellular events including activation of JAK (Janus Kinase) kinases and activation of Ras-mediated signalling. Activated JAK kinases phosphorylate and activate STAT transcription factors, particularly STAT3 and SHP2. Phosphorylated STAT3 then forms a dimer and translocates into the nucleus to activate transcription of genes containing STAT3 response elements. STAT3 is essential for GP130-mediated cell survival and G1 to S cell-cycle-transition signals. Both c- Myc and Pim have been identified as target genes of STAT3 and together can compensate for STAT3 in cell survival and cell-cycle transition. SHP2 links cytokine receptor to the Ras / MAP (Mitogen-Activated Protein) kinase pathway and is essential for mitogenic activity. The Ras-mediated pathway, acting through SHC, GRB2 (Growth Factor Receptor Bound protein-2) and SOS1 (Son of Sevenless-1) upstream and activating MAP kinases downstream, activates transcription factors such as Elk1 and NF-IL-6 (C / EBP-β) that can act through their own cognate response elements in the genome. In addition to JAK / STAT and Ras / MAP kinase pathways, IL-6 also activates PI3K (Phosphoinositide-3 Kinase). The PI3K / Akt / NF-KappaB cascade activated by IL-6, functions cooperatively to achieve the maximal anti-apoptotic effect of IL-6 against TGF-β. The anti- apoptotic mechanism of PI3K / Akt is attributed to phosphorylation of the BCL2 family member BAD (BCL2 Associated Death Promoter) by Akt. The phosphorylated BAD is then associated with 14-3-3, which sequesters BAD from BCLXL, thereby promoting cell survival. Regulating the BCL2 family member is also considered as one of the anti-apoptotic mechanisms of STAT3, which may be capable of inducing BCL2 in pro-B cells. The termination of the IL-6- type cytokine signalling is through the action of tyrosine phosphatases, proteasome, and JAK kinase inhibitors SOCS (Suppressor of Cytokine Signaling), PIAS (Protein Inhibitors of Activated STATs), and internalization of the cytokine receptors via GP130. An agent which downregulates IL-6 signalling may interfere with or inhibit any of the above stages involved in IL-6 mediated signalling such that IL-6 signalling and responses are diminished. For example, the agent may be an IL-6 receptor antagonist such as Tocilizumab, which is a humanized monoclonal antibody against the IL-6 receptor. An IL-6 receptor antagonist refers to an agent that reduces the level of IL-6 that is able to bind to the IL-6 receptor. Tocilizumab is a humanized monoclonal IgG1 antibody against the IL-6 receptor that binds to soluble and membrane-bound IL-6 receptor. Tocilizumab inhibits the induction of biological activity due to IL-6 in cells that have expressed both membrane-bound IL-6 receptor and gp130 molecules, and also inhibits the induction of biological activity due to IL-6 / IL-6 receptor complex formation in cells that express gp130 alone. Furthermore, since it has the capacity to dissociate IL-6 / IL-6 receptor complexes that have already formed, it is able to block IL-6 signal transduction. B cell targeted therapy The method of the invention may determine an RA patient as being susceptible or refractory to treatment with a B cell targeted therapy. The term “B cell targeted therapy”, as used herein, may refer to the administration of an agent that interferes with or inhibits the development and / or function of B cells. The B cell targeted therapy may cause B cell depletion or the inhibition of B cell development and maturation. Advantageously, the B cell targeted therapy is directed against B cells in all stages of development other than undifferentiated stem cells and terminally differentiated antibody- producing plasma cells. The agent may be a small molecule drug, such as a Bruton's tyrosine kinase (BTK) inhibitor or other agent which targets B cell signalling pathways. Direct depletion of B cells may be performed through the use of antibodies, such as monoclonal antibodies (mAbs), directed against cell surface markers (e.g. CD20 and CD22). Such antibodies bind to the target antigen and kill the cell by initiating a mixture of apoptosis, complement dependent cytotoxicity (CDC), and antibody-dependent cell-mediated cellular cytotoxicity (ADCC). The B cell targeted therapy used in the invention may be an agent directed against CD20, for example Rituximab, Ocrelizumab, Veltuzumab or Ofatumumab, or an agent directed against CD22 such as Epratuzumab. Rituximab is a chimeric mouse / human immunoglobulin G1 (IgG1) monoclonal antibody to CD20 that stimulates B cell destruction upon binding to CD20. Rituximab depletes CD20 surface-positive naïve and memory B cells from the blood, bone marrow and lymph nodes via mechanisms which include antibody-dependent cellular cytotoxicity (ADCC), complement dependent cytotoxicity (CDC). It does not affect CD20-negative early B cell lineage precursor cells and late B lineage plasma cells in the bone marrow. Ocrelizumab is a humanized anti-CD20 monoclonal antibody that causes CD20+ B cell depletion following binding to CD20 via mechanisms including ADCC and CDC. Veltuzumab is a humanized, second-generation anti-CD20 monoclonal antibody that causes CD20+ B cell depletion following binding to CD20 via mechanisms including ADCC and CDC. Ofatumumab is a human monoclonal IgG1 antibody to CD20 and may inhibit early-stage B lymphocyte activation. Ofatumumab targets a different epitope located closer to the N- terminus of CD20 compared to the epitope targeted by rituximab and includes an extracellular loop, as it binds to both the small and large loops of the CD20 molecule. Ofatumumab stimulates B cell destruction through ADCC and CDC pathways. Epratuzumab is a humanized monoclonal IgG1 antibody to CD22. It contains a murine sequence comprising 5–10% of the molecule, the remainder being human framework sequences. Epratuzumab binds to the CD22 third extracellular domain (epitope B), without blocking the ligand binding site, with measured affinity of Kd = 0.7 nm. In vitro studies showed epratuzumab induces CD22 phosphorylation by binding to its surface. It results in modulation, mostly negative, of BCR activation. Response to therapies in RA patients Methods of assessing a subject’s response to a therapy for rheumatoid arthritis are known in the art and would be familiar to a skilled person. By way of example, well known measures of disease activity in RA include the Disease Activity Score (DAS), a modified version DAS28, and the DAS-based EULAR response criteria. The assessment of response to a therapy for rheumatoid arthritis may use the Clinical Disease Activity Index (CDAI). Susceptibility or refractoriness to treatment of rheumatoid arthritis may, for example, be achievement or not of a CDAI ≥ 50%. Other measures of assessment of response to a therapy for rheumatoid arthritis include CDAI- remission, DAS28(ESR) / (CRP) moderate / good EULAR-response, DAS28(ESR) / (CRP) low- disease-activity, DAS28(ESR) / (CRP) remission and patient reported outcomes, such as fatigue. Markers Susceptibility or refractoriness to an agent that downregulates TNF mediated signalling In one aspect, the invention provides a method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a marker panel in a sample from the patient, wherein the agent is an agent that downregulates TNF mediated signalling, and wherein the marker panel comprises one or more markers selected from Table 1.1. In one aspect, the invention provides a method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, the method comprising determining the profile of a marker panel in a sample from the patient, and wherein the marker panel comprises one or more markers selected from Table 1.1. In some embodiments, the marker panel comprises 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, or all 40 markers from Table 1.1. In some embodiments, the marker panel comprises the first 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 or 39 markers listed in Table 1.1. In some embodiments, the marker panel comprises all 40 markers from Table 1.1. In some embodiments, the marker panel consists of 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, or all 40 markers from Table 1.1. In some embodiments, the marker panel consists of the first 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 or 39 markers listed in Table 1.1. In some embodiments, the marker panel consists of all 40 markers from Table 1.1. Table 1.1. Etanercept model genes ranked by variable importance. Gene Full Gene Name NCBI Gene EnsemblID ID EIF2B1 eukaryotic translation 1967 ENSG00000111361 initiation factor 2B subunit alpha CENPBD1 CENPB DNA-binding domain 92806 ENSG00000177946 containing 1, pseudogene PWWP2B PWWP domain containing 2B 170394 ENSG00000171813 AGK acylglycerol kinase 55750 ENSG00000006530 RAD52 RAD52 homolog, DNA repair 5893 ENSG00000002016 protein ZNF765 zinc finger protein 765 91661 ENSG00000196417 ZNF850 zinc finger protein 850 342892 ENSG00000267041 IL5 interleukin 5 3567 ENSG00000113525 CXorf56 STING1 ER exit protein 1 63932 ENSG00000018610 AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 TRPC6 transient receptor potential 7225 ENSG00000137672 cation channel subfamily C member 6 CCDC102B coiled-coil domain containing 79839 ENSG00000150636 102B SYT1 synaptotagmin 1 6857 ENSG00000067715 IL13RA2 interleukin 13 receptor 3598 ENSG00000123496 subunit alpha 2 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 HTR2A 5-hydroxytryptamine receptor 3356 ENSG00000102468 2A CASP7 caspase 7 840 ENSG00000165806 IFI44L interferon induced protein 44 10964 ENSG00000137959 like CSF3R colony stimulating factor 3 1441 ENSG00000119535 receptor MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 MRPL19 mitochondrial ribosomal 9801 ENSG00000115364 protein L19 TMUB2 transmembrane and ubiquitin 79089 ENSG00000168591 like domain containing 2 APLF aprataxin and PNKP like 200558 ENSG00000169621 factor LTBR lymphotoxin beta receptor 4055 ENSG00000111321 PIK3CA phosphatidylinositol-4,5- 5290 ENSG00000121879 bisphosphate 3-kinase catalytic subunit alpha CD80 CD80 molecule 941 ENSG00000121594 IL17A interleukin 17A 3605 ENSG00000112115 LILRA2 leukocyte immunoglobulin 11027 ENSG00000239998 like receptor A2 MX1 MX dynamin like GTPase 1 4599 ENSG00000157601 SOS1 SOS Ras / Rac guanine 6654 ENSG00000115904 nucleotide exchange factor 1 GPNMB glycoprotein nmb 10457 ENSG00000136235 CREB3L2 cAMP responsive element 64764 ENSG00000182158 binding protein 3 like 2 THY1 Thy-1 cell surface antigen 7070 ENSG00000154096 GUSB glucuronidase beta 2990 ENSG00000169919 RNF19A ring finger protein 19A, RBR 25897 ENSG00000034677 E3 ubiquitin protein ligase IL17RA interleukin 17 receptor A 23765 ENSG00000177663 TNFAIP3 TNF alpha induced protein 3 7128 ENSG00000118503 FGF16 fibroblast growth factor 16 8823 ENSG00000196468 LRIG3 leucine rich repeats and 121227 ENSG00000139263 immunoglobulin like domains 3 HSP90B1 heat shock protein 90 beta 7184 ENSG00000166598 family member 1 Table 1.1.A. Etanercept model genes that increase in responders ranked by variable importance. Gene Full Gene Name NCBI Gene EnsemblID ID EIF2B1 eukaryotic translation 1967 ENSG00000111361 initiation factor 2B subunit alpha AGK acylglycerol kinase 55750 ENSG00000006530 RAD52 RAD52 homolog, DNA repair 5893 ENSG00000002016 protein ZNF765 zinc finger protein 765 91661 ENSG00000196417 ZNF850 zinc finger protein 850 342892 ENSG00000267041 IL5 interleukin 5 3567 ENSG00000113525 CXorf56 STING1 ER exit protein 1 63932 ENSG00000018610 TRPC6 transient receptor potential 7225 ENSG00000137672 cation channel subfamily C member 6 CCDC102B coiled-coil domain containing 79839 ENSG00000150636 102B IL13RA2 interleukin 13 receptor 3598 ENSG00000123496 subunit alpha 2 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 CASP7 caspase 7 840 ENSG00000165806 IFI44L interferon induced protein 44 10964 ENSG00000137959 like MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 MRPL19 mitochondrial ribosomal 9801 ENSG00000115364 protein L19 TMUB2 transmembrane and ubiquitin 79089 ENSG00000168591 like domain containing 2 APLF aprataxin and PNKP like 200558 ENSG00000169621 factor PIK3CA phosphatidylinositol-4,5- 5290 ENSG00000121879 bisphosphate 3-kinase catalytic subunit alpha CD80 CD80 molecule 941 ENSG00000121594 LILRA2 leukocyte immunoglobulin 11027 ENSG00000239998 like receptor A2 MX1 MX dynamin like GTPase 1 4599 ENSG00000157601 SOS1 SOS Ras / Rac guanine 6654 ENSG00000115904 nucleotide exchange factor 1 GPNMB glycoprotein nmb 10457 ENSG00000136235 THY1 Thy-1 cell surface antigen 7070 ENSG00000154096 GUSB glucuronidase beta 2990 ENSG00000169919 RNF19A ring finger protein 19A, RBR 25897 ENSG00000034677 E3 ubiquitin protein ligase IL17RA interleukin 17 receptor A 23765 ENSG00000177663 TNFAIP3 TNF alpha induced protein 3 7128 ENSG00000118503 FGF16 fibroblast growth factor 16 8823 ENSG00000196468 LRIG3 leucine rich repeats and 121227 ENSG00000139263 immunoglobulin like domains 3 Table 1.1.B. Etanercept model genes that increase in non-responders ranked by variable importance. Gene Full Gene Name NCBI Gene EnsemblID ID CENPBD1 CENPB DNA-binding domain 92806 ENSG00000177946 containing 1, pseudogene PWWP2B PWWP domain containing 2B 170394 ENSG00000171813 AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 SYT1 synaptotagmin 1 6857 ENSG00000067715 HTR2A 5-hydroxytryptamine receptor 3356 ENSG00000102468 2A CSF3R colony stimulating factor 3 1441 ENSG00000119535 receptor CREB3L2 cAMP responsive element 64764 ENSG00000182158 binding protein 3 like 2 LTBR lymphotoxin beta receptor 4055 ENSG00000111321 IL17A interleukin 17A 3605 ENSG00000112115 HSP90B1 heat shock protein 90 beta 7184 ENSG00000166598 family member 1 In one aspect, the invention provides a method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a marker panel in a sample from the patient, wherein the agent is an agent that downregulates TNF mediated signalling, and wherein the marker panel comprises one or more markers selected from Table 1.2. In one aspect, the invention provides a method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, the method comprising determining the profile of a marker panel in a sample from the patient, and wherein the marker panel comprises one or more markers selected from Table 1.2. In some embodiments, the marker panel comprises 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, or all 40 markers from Table 1.2. In some embodiments, the marker panel comprises the first 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 or 39 markers listed in Table 1.2. In some embodiments, the marker panel comprises all 40 markers from Table 1.2. In some embodiments, the marker panel consists of 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, or all 40 markers from Table 1.2. In some embodiments, the marker panel consists of the first 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 or 39 markers listed in Table 1.2. In some embodiments, the marker panel consists of all 40 markers from Table 1.2. Table 1.2. Etanercept model genes ranked by variable importance and variable frequency. Gene Full Gene Name NCBI Gene EnsemblID ID EIF2B1 eukaryotic translation 1967 ENSG00000111361 initiation factor 2B subunit alpha CENPBD1 CENPB DNA-binding domain 92806 ENSG00000177946 containing 1, pseudogene AGK acylglycerol kinase 55750 ENSG00000006530 RAD52 RAD52 homolog, DNA repair 5893 ENSG00000002016 protein ZNF765 zinc finger protein 765 91661 ENSG00000196417 ZNF850 zinc finger protein 850 342892 ENSG00000267041 PWWP2B PWWP domain containing 2B 170394 ENSG00000171813 AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 TRPC6 transient receptor potential 7225 ENSG00000137672 cation channel subfamily C member 6 CCDC102B coiled-coil domain containing 79839 ENSG00000150636 102B IL13RA2 interleukin 13 receptor 3598 ENSG00000123496 subunit alpha 2 CXorf56 STING1 ER exit protein 1 63932 ENSG00000018610 SYT1 synaptotagmin 1 6857 ENSG00000067715 IFI44L interferon induced protein 44 10964 ENSG00000137959 like HTR2A 5-hydroxytryptamine receptor 3356 ENSG00000102468 2A CSF3R colony stimulating factor 3 1441 ENSG00000119535 receptor IL5 interleukin 5 3567 ENSG00000113525 LILRA2 leukocyte immunoglobulin 11027 ENSG00000239998 like receptor A2 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 CASP7 caspase 7 840 ENSG00000165806 APLF aprataxin and PNKP like 200558 ENSG00000169621 factor CREB3L2 cAMP responsive element 64764 ENSG00000182158 binding protein 3 like 2 FGF16 fibroblast growth factor 16 8823 ENSG00000196468 MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 MRPL19 mitochondrial ribosomal 9801 ENSG00000115364 protein L19 TMUB2 transmembrane and ubiquitin 79089 ENSG00000168591 like domain containing 2 LTBR lymphotoxin beta receptor 4055 ENSG00000111321 PIK3CA phosphatidylinositol-4,5- 5290 ENSG00000121879 bisphosphate 3-kinase catalytic subunit alpha CD80 CD80 molecule 941 ENSG00000121594 IL17A interleukin 17A 3605 ENSG00000112115 MX1 MX dynamin like GTPase 1 4599 ENSG00000157601 SOS1 SOS Ras / Rac guanine 6654 ENSG00000115904 nucleotide exchange factor 1 GPNMB glycoprotein nmb 10457 ENSG00000136235 THY1 Thy-1 cell surface antigen 7070 ENSG00000154096 GUSB glucuronidase beta 2990 ENSG00000169919 RNF19A ring finger protein 19A, RBR 25897 ENSG00000034677 E3 ubiquitin protein ligase IL17RA interleukin 17 receptor A 23765 ENSG00000177663 TNFAIP3 TNF alpha induced protein 3 7128 ENSG00000118503 LRIG3 leucine rich repeats and 121227 ENSG00000139263 immunoglobulin like domains 3 HSP90B1 heat shock protein 90 beta 7184 ENSG00000166598 family member 1 Table 1.2.A. Etanercept model genes that increase in responders ranked by variable importance and variable frequency. Gene Full Gene Name NCBI Gene EnsemblID ID EIF2B1 eukaryotic translation 1967 ENSG00000111361 initiation factor 2B subunit alpha AGK acylglycerol kinase 55750 ENSG00000006530 RAD52 RAD52 homolog, DNA repair 5893 ENSG00000002016 protein ZNF765 zinc finger protein 765 91661 ENSG00000196417 ZNF850 zinc finger protein 850 342892 ENSG00000267041 TRPC6 transient receptor potential 7225 ENSG00000137672 cation channel subfamily C member 6 CCDC102B coiled-coil domain containing 79839 ENSG00000150636 102B IL13RA2 interleukin 13 receptor 3598 ENSG00000123496 subunit alpha 2 CXorf56 STING1 ER exit protein 1 63932 ENSG00000018610 IFI44L interferon induced protein 44 10964 ENSG00000137959 like IL5 interleukin 5 3567 ENSG00000113525 LILRA2 leukocyte immunoglobulin 11027 ENSG00000239998 like receptor A2 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 CASP7 caspase 7 840 ENSG00000165806 APLF aprataxin and PNKP like 200558 ENSG00000169621 factor FGF16 fibroblast growth factor 16 8823 ENSG00000196468 MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 MRPL19 mitochondrial ribosomal 9801 ENSG00000115364 protein L19 TMUB2 transmembrane and ubiquitin 79089 ENSG00000168591 like domain containing 2 PIK3CA phosphatidylinositol-4,5- 5290 ENSG00000121879 bisphosphate 3-kinase catalytic subunit alpha CD80 CD80 molecule 941 ENSG00000121594 MX1 MX dynamin like GTPase 1 4599 ENSG00000157601 SOS1 SOS Ras / Rac guanine 6654 ENSG00000115904 nucleotide exchange factor 1 GPNMB glycoprotein nmb 10457 ENSG00000136235 THY1 Thy-1 cell surface antigen 7070 ENSG00000154096 GUSB glucuronidase beta 2990 ENSG00000169919 RNF19A ring finger protein 19A, RBR 25897 ENSG00000034677 E3 ubiquitin protein ligase IL17RA interleukin 17 receptor A 23765 ENSG00000177663 TNFAIP3 TNF alpha induced protein 3 7128 ENSG00000118503 LRIG3 leucine rich repeats and 121227 ENSG00000139263 immunoglobulin like domains 3 Table 1.2.B. Etanercept model genes that increase in non-responders ranked by variable importance and variable frequency. Gene Full Gene Name NCBI Gene EnsemblID ID CENPBD1 CENPB DNA-binding domain 92806 ENSG00000177946 containing 1, pseudogene PWWP2B PWWP domain containing 2B 170394 ENSG00000171813 AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 SYT1 synaptotagmin 1 6857 ENSG00000067715 HTR2A 5-hydroxytryptamine receptor 3356 ENSG00000102468 2A CSF3R colony stimulating factor 3 1441 ENSG00000119535 receptor CREB3L2 cAMP responsive element 64764 ENSG00000182158 binding protein 3 like 2 LTBR lymphotoxin beta receptor 4055 ENSG00000111321 IL17A interleukin 17A 3605 ENSG00000112115 HSP90B1 heat shock protein 90 beta 7184 ENSG00000166598 family member 1 Susceptibility or refractoriness to an agent that downregulates IL-6 mediated signalling In one aspect, the invention provides a method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a marker panel in a sample from the patient, wherein the agent is an agent that downregulates IL-6 mediated signalling, and wherein the marker panel comprises one or more markers selected from Table 2.1. In one aspect, the invention provides a method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates IL-6 mediated signalling, the method comprising determining the profile of a marker panel in a sample from the patient, and wherein the marker panel comprises one or more markers selected from Table 2.1. In some embodiments, the marker panel comprises 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89 or all 90 markers from Table 2.1. In some embodiments, the marker panel comprises the first 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88 or 89 markers listed in Table 2.1. In some embodiments, the marker panel comprises all 90 markers from Table 2.1. In some embodiments, the marker panel consists of 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89 or all 90 markers from Table 2.1. In some embodiments, the marker panel consists of the first 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88 or 89 markers listed in Table 2.1. In some embodiments, the marker panel consists of all 90 markers from Table 2.1. Table 2.1. Tocilizumab model genes ranked by variable importance. Gene Full Gene Name NCBI Gene EnsemblID ID GPNMB glycoprotein nmb 10457 ENSG00000136235 MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 NKG7 natural killer cell granule 4818 ENSG00000105374 protein 7 TOX2 TOX high mobility group box 84969 ENSG00000124191 family member 2 GZMH granzyme H 2999 ENSG00000100450 CD40LG CD40 ligand 959 ENSG00000102245 SOCS6 suppressor of cytokine 9306 ENSG00000170677 signaling 6 CD247 CD247 molecule 919 ENSG00000198821 CDKN1B cyclin dependent kinase 1027 ENSG00000111276 inhibitor 1B KLF2 KLF transcription factor 2 10365 ENSG00000127528 TCIM transcriptional and immune 56892 ENSG00000176907 response regulator MERTK MER proto-oncogene, 10461 ENSG00000153208 tyrosine kinase JAG1 jagged canonical Notch 182 ENSG00000101384 ligand 1 PTGS2 prostaglandin-endoperoxide 5743 ENSG00000073756 synthase 2 BLK BLK proto-oncogene, Src 640 ENSG00000136573 family tyrosine kinase CSF2RB colony stimulating factor 2 1439 ENSG00000100368 receptor subunit beta SP140 SP140 nuclear body protein 11262 ENSG00000079263 SDC1 syndecan 1 6382 ENSG00000115884 IL2RB interleukin 2 receptor subunit 3560 ENSG00000100385 beta SELE selectin E 6401 ENSG00000007908 CST7 cystatin F 8530 ENSG00000077984 MX1 MX dynamin like GTPase 1 4599 ENSG00000157601 CCDC102B coiled-coil domain containing 79839 ENSG00000150636 102B AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 CD79B CD79b molecule 974 ENSG00000007312 YPEL5 yippee like 5 51646 ENSG00000119801 TRPC6 transient receptor potential 7225 ENSG00000137672 cation channel subfamily C member 6 FCRL3 Fc receptor like 3 115352 ENSG00000160856 P2RX4 purinergic receptor P2X 4 5025 ENSG00000135124 ITGA1 integrin subunit alpha 1 3672 ENSG00000213949 CD3G CD3 gamma subunit of T-cell 917 ENSG00000160654 receptor complex IL24 interleukin 24 11009 ENSG00000162892 IL4 interleukin 4 3565 ENSG00000113520 IL23A interleukin 23 subunit alpha 51561 ENSG00000110944 KCNA3 potassium voltage-gated 3738 ENSG00000177272 channel subfamily A member 3 PIK3R2 phosphoinositide-3-kinase 5296 ENSG00000105647 regulatory subunit 2 TNFRSF10B TNF receptor superfamily 8795 ENSG00000120889 member 10b CD2 CD2 molecule 914 ENSG00000116824 CCL20 C-C motif chemokine ligand 6364 ENSG00000115009 20 CASP10 caspase 10 843 ENSG00000003400 ASNS asparagine synthetase 440 ENSG00000070669 (glutamine-hydrolyzing) EPO erythropoietin 2056 ENSG00000130427 ICOS inducible T cell costimulator 29851 ENSG00000163600 ITGA4 integrin subunit alpha 4 3676 ENSG00000115232 P2RY13 purinergic receptor P2Y13 53829 ENSG00000181631 CX3CL1 C-X3-C motif chemokine 6376 ENSG00000006210 ligand 1 PPFIA4 PTPRF interacting protein 8497 ENSG00000143847 alpha 4 RALGPS2 Ral GEF with PH domain and 55103 ENSG00000116191 SH3 binding motif 2 LTB lymphotoxin beta 4050 ENSG00000227507 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 LTA lymphotoxin alpha 4049 ENSG00000226979 CEBPB CCAAT enhancer binding 1051 ENSG00000172216 protein beta CXCL14 C-X-C motif chemokine 9547 ENSG00000145824 ligand 14 GZMK granzyme K 3003 ENSG00000113088 IL2RG interleukin 2 receptor subunit 3561 ENSG00000147168 gamma GNLY granulysin 10578 ENSG00000115523 CD6 CD6 molecule 923 ENSG00000013725 PABPC1 poly(A) binding protein 26986 ENSG00000070756 cytoplasmic 1 CD8A CD8 subunit alpha 925 ENSG00000153563 CD3D CD3 delta subunit of T-cell 915 ENSG00000167286 receptor complex PIK3R3 phosphoinositide-3-kinase 8503 ENSG00000117461 regulatory subunit 3 TRAF2 TNF receptor associated 7186 ENSG00000127191 factor 2 PIAS2 protein inhibitor of activated 9063 ENSG00000078043 STAT 2 DAB2IP DAB2 interacting protein 153090 ENSG00000136848 ZBP1 insulin like growth factor 2 10642 ENSG00000159217 mRNA binding protein 1 GZMB granzyme B 3002 ENSG00000100453 HTR2A 5-hydroxytryptamine receptor 3356 ENSG00000102468 2A CCND3 cyclin D3 896 ENSG00000112576 IRF4 interferon regulatory factor 4 3662 ENSG00000137265 CD3E CD3 epsilon subunit of T-cell 916 ENSG00000198851 receptor complex TRDC T cell receptor delta constant 28526 NA GFAP glial fibrillary acidic protein 2670 ENSG00000131095 PLCG1 phospholipase C gamma 1 5335 ENSG00000124181 RIPK1 receptor interacting 8737 ENSG00000137275 serine / threonine kinase 1 ISG15 ISG15 ubiquitin like modifier 9636 ENSG00000187608 STAM signal transducing adaptor 8027 ENSG00000136738 molecule LY9 lymphocyte antigen 9 4063 ENSG00000122224 APLF aprataxin and PNKP like 200558 ENSG00000169621 factor BTLA B and T lymphocyte 151888 ENSG00000186265 associated IL7R interleukin 7 receptor 3575 ENSG00000168685 FOSL1 FOS like 1, AP-1 transcription 8061 ENSG00000175592 factor subunit PLCG2 phospholipase C gamma 2 5336 ENSG00000197943 SLAMF6 SLAM family member 6 114836 ENSG00000162739 IRF9 interferon regulatory factor 9 10379 ENSG00000213928 ATF3 activating transcription factor 467 ENSG00000162772 3 PNOC prepronociceptin 5368 ENSG00000168081 TNFRSF13C TNF receptor superfamily 115650 ENSG00000159958 member 13C MS4A1 membrane spanning 4- 931 ENSG00000156738 domains A1 GZMA granzyme A 3001 ENSG00000145649 ITGB4 integrin subunit beta 4 3691 ENSG00000132470 Table 2.1.A. Tocilizumab model genes that increase in responders ranked by variable importance. Gene Full Gene Name NCBI Gene EnsemblID ID GPNMB glycoprotein nmb 10457 ENSG00000136235 MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 SOCS6 suppressor of cytokine 9306 ENSG00000170677 signaling 6 MERTK MER proto-oncogene, 10461 ENSG00000153208 tyrosine kinase AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 P2RX4 purinergic receptor P2X 4 5025 ENSG00000135124 PIK3R2 phosphoinositide-3-kinase 5296 ENSG00000105647 regulatory subunit 2 CCL20 C-C motif chemokine ligand 6364 ENSG00000115009 20 EPO erythropoietin 2056 ENSG00000130427 PPFIA4 PTPRF interacting protein 8497 ENSG00000143847 alpha 4 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 CEBPB CCAAT enhancer binding 1051 ENSG00000172216 protein beta PIAS2 protein inhibitor of activated 9063 ENSG00000078043 STAT 2 HTR2A 5-hydroxytryptamine receptor 3356 ENSG00000102468 2A STAM signal transducing adaptor 8027 ENSG00000136738 molecule APLF aprataxin and PNKP like 200558 ENSG00000169621 factor FOSL1 FOS like 1, AP-1 transcription 8061 ENSG00000175592 factor subunit Table 2.1.B. Tocilizumab model genes that increase in non-responders ranked by variable importance. Gene Full Gene Name NCBI Gene EnsemblID ID NKG7 natural killer cell granule 4818 ENSG00000105374 protein 7 TOX2 TOX high mobility group box 84969 ENSG00000124191 family member 2 GZMH granzyme H 2999 ENSG00000100450 CD40LG CD40 ligand 959 ENSG00000102245 CD247 CD247 molecule 919 ENSG00000198821 CDKN1B cyclin dependent kinase 1027 ENSG00000111276 inhibitor 1B KLF2 KLF transcription factor 2 10365 ENSG00000127528 TCIM transcriptional and immune 56892 ENSG00000176907 response regulator JAG1 jagged canonical Notch 182 ENSG00000101384 ligand 1 PTGS2 prostaglandin-endoperoxide 5743 ENSG00000073756 synthase 2 BLK BLK proto-oncogene, Src 640 ENSG00000136573 family tyrosine kinase CSF2RB colony stimulating factor 2 1439 ENSG00000100368 receptor subunit beta SP140 SP140 nuclear body protein 11262 ENSG00000079263 SDC1 syndecan 1 6382 ENSG00000115884 IL2RB interleukin 2 receptor subunit 3560 ENSG00000100385 beta SELE selectin E 6401 ENSG00000007908 CST7 cystatin F 8530 ENSG00000077984 MX1 MX dynamin like GTPase 1 4599 ENSG00000157601 CCDC102B coiled-coil domain containing 79839 ENSG00000150636 102B CD79B CD79b molecule 974 ENSG00000007312 YPEL5 yippee like 5 51646 ENSG00000119801 TRPC6 transient receptor potential 7225 ENSG00000137672 cation channel subfamily C member 6 FCRL3 Fc receptor like 3 115352 ENSG00000160856 ITGA1 integrin subunit alpha 1 3672 ENSG00000213949 CD3G CD3 gamma subunit of T-cell 917 ENSG00000160654 receptor complex IL24 interleukin 24 11009 ENSG00000162892 IL4 interleukin 4 3565 ENSG00000113520 IL23A interleukin 23 subunit alpha 51561 ENSG00000110944 KCNA3 potassium voltage-gated 3738 ENSG00000177272 channel subfamily A member 3 TNFRSF10B TNF receptor superfamily 8795 ENSG00000120889 member 10b CD2 CD2 molecule 914 ENSG00000116824 CASP10 caspase 10 843 ENSG00000003400 ASNS asparagine synthetase 440 ENSG00000070669 (glutamine-hydrolyzing) ICOS inducible T cell costimulator 29851 ENSG00000163600 ITGA4 integrin subunit alpha 4 3676 ENSG00000115232 P2RY13 purinergic receptor P2Y13 53829 ENSG00000181631 CX3CL1 C-X3-C motif chemokine 6376 ENSG00000006210 ligand 1 RALGPS2 Ral GEF with PH domain and 55103 ENSG00000116191 SH3 binding motif 2 LTB lymphotoxin beta 4050 ENSG00000227507 LTA lymphotoxin alpha 4049 ENSG00000226979 CXCL14 C-X-C motif chemokine 9547 ENSG00000145824 ligand 14 GZMK granzyme K 3003 ENSG00000113088 IL2RG interleukin 2 receptor subunit 3561 ENSG00000147168 gamma GNLY granulysin 10578 ENSG00000115523 CD6 CD6 molecule 923 ENSG00000013725 PABPC1 poly(A) binding protein 26986 ENSG00000070756 cytoplasmic 1 CD8A CD8 subunit alpha 925 ENSG00000153563 CD3D CD3 delta subunit of T-cell 915 ENSG00000167286 receptor complex PIK3R3 phosphoinositide-3-kinase 8503 ENSG00000117461 regulatory subunit 3 TRAF2 TNF receptor associated 7186 ENSG00000127191 factor 2 DAB2IP DAB2 interacting protein 153090 ENSG00000136848 ZBP1 insulin like growth factor 2 10642 ENSG00000159217 mRNA binding protein 1 GZMB granzyme B 3002 ENSG00000100453 CCND3 cyclin D3 896 ENSG00000112576 IRF4 interferon regulatory factor 4 3662 ENSG00000137265 CD3E CD3 epsilon subunit of T-cell 916 ENSG00000198851 receptor complex TRDC T cell receptor delta constant 28526 NA GFAP glial fibrillary acidic protein 2670 ENSG00000131095 PLCG1 phospholipase C gamma 1 5335 ENSG00000124181 RIPK1 receptor interacting 8737 ENSG00000137275 serine / threonine kinase 1 ISG15 ISG15 ubiquitin like modifier 9636 ENSG00000187608 LY9 lymphocyte antigen 9 4063 ENSG00000122224 BTLA B and T lymphocyte 151888 ENSG00000186265 associated IL7R interleukin 7 receptor 3575 ENSG00000168685 PLCG2 phospholipase C gamma 2 5336 ENSG00000197943 SLAMF6 SLAM family member 6 114836 ENSG00000162739 IRF9 interferon regulatory factor 9 10379 ENSG00000213928 ATF3 activating transcription factor 467 ENSG00000162772 3 PNOC prepronociceptin 5368 ENSG00000168081 TNFRSF13C TNF receptor superfamily 115650 ENSG00000159958 member 13C MS4A1 membrane spanning 4- 931 ENSG00000156738 domains A1 GZMA granzyme A 3001 ENSG00000145649 ITGB4 integrin subunit beta 4 3691 ENSG00000132470 In one aspect, the invention provides a method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a marker panel in a sample from the patient, wherein the agent is an agent that downregulates IL-6 mediated signalling, and wherein the marker panel comprises one or more markers selected from Table 2.2. In one aspect, the invention provides a method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates IL-6 mediated signalling, the method comprising determining the profile of a marker panel in a sample from the patient, and wherein the marker panel comprises one or more markers selected from Table 2.2. In some embodiments, the marker panel comprises 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89 or all 90 markers from Table 2.2. In some embodiments, the marker panel comprises the first 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88 or 89 markers listed in Table 2.2. In some embodiments, the marker panel comprises all 90 markers from Table 2.2. In some embodiments, the marker panel consists of 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89 or all 90 markers from Table 2.2. In some embodiments, the marker panel consists of the first 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88 or 89 markers listed in Table 2.2. In some embodiments, the marker panel consists of all 90 markers from Table 2.2. Table 2.2. Tocilizumab model genes ranked by variable importance and variable frequency. Gene Full Gene Name NCBI Gene EnsemblID ID GPNMB glycoprotein nmb 10457 ENSG00000136235 MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 TOX2 TOX high mobility group box 84969 ENSG00000124191 family member 2 CD40LG CD40 ligand 959 ENSG00000102245 SOCS6 suppressor of cytokine 9306 ENSG00000170677 signaling 6 CD247 CD247 molecule 919 ENSG00000198821 NKG7 natural killer cell granule 4818 ENSG00000105374 protein 7 GZMH granzyme H 2999 ENSG00000100450 CDKN1B cyclin dependent kinase 1027 ENSG00000111276 inhibitor 1B SP140 SP140 nuclear body protein 11262 ENSG00000079263 CD3G CD3 gamma subunit of T-cell 917 ENSG00000160654 receptor complex TCIM transcriptional and immune 56892 ENSG00000176907 response regulator BLK BLK proto-oncogene, Src 640 ENSG00000136573 family tyrosine kinase FCRL3 Fc receptor like 3 115352 ENSG00000160856 KCNA3 potassium voltage-gated 3738 ENSG00000177272 channel subfamily A member 3 MERTK MER proto-oncogene, 10461 ENSG00000153208 tyrosine kinase CST7 cystatin F 8530 ENSG00000077984 CD6 CD6 molecule 923 ENSG00000013725 CD3E CD3 epsilon subunit of T-cell 916 ENSG00000198851 receptor complex KLF2 KLF transcription factor 2 10365 ENSG00000127528 PTGS2 prostaglandin-endoperoxide 5743 ENSG00000073756 synthase 2 IL2RB interleukin 2 receptor subunit 3560 ENSG00000100385 beta IL23A interleukin 23 subunit alpha 51561 ENSG00000110944 LTB lymphotoxin beta 4050 ENSG00000227507 CD79B CD79b molecule 974 ENSG00000007312 TRPC6 transient receptor potential 7225 ENSG00000137672 cation channel subfamily C member 6 CASP10 caspase 10 843 ENSG00000003400 ICOS inducible T cell costimulator 29851 ENSG00000163600 RALGPS2 Ral GEF with PH domain and 55103 ENSG00000116191 SH3 binding motif 2 JAG1 jagged canonical Notch 182 ENSG00000101384 ligand 1 IL24 interleukin 24 11009 ENSG00000162892 CD2 CD2 molecule 914 ENSG00000116824 GZMK granzyme K 3003 ENSG00000113088 GNLY granulysin 10578 ENSG00000115523 CD3D CD3 delta subunit of T-cell 915 ENSG00000167286 receptor complex LY9 lymphocyte antigen 9 4063 ENSG00000122224 SDC1 syndecan 1 6382 ENSG00000115884 YPEL5 yippee like 5 51646 ENSG00000119801 ASNS asparagine synthetase 440 ENSG00000070669 (glutamine-hydrolyzing) LTA lymphotoxin alpha 4049 ENSG00000226979 ZBP1 insulin like growth factor 2 10642 ENSG00000159217 mRNA binding protein 1 GZMB granzyme B 3002 ENSG00000100453 GZMA granzyme A 3001 ENSG00000145649 CSF2RB colony stimulating factor 2 1439 ENSG00000100368 receptor subunit beta SELE selectin E 6401 ENSG00000007908 MX1 MX dynamin like GTPase 1 4599 ENSG00000157601 AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 ITGA1 integrin subunit alpha 1 3672 ENSG00000213949 PIK3R2 phosphoinositide-3-kinase 5296 ENSG00000105647 regulatory subunit 2 TNFRSF10B TNF receptor superfamily 8795 ENSG00000120889 member 10b CX3CL1 C-X3-C motif chemokine 6376 ENSG00000006210 ligand 1 CD8A CD8 subunit alpha 925 ENSG00000153563 PIAS2 protein inhibitor of activated 9063 ENSG00000078043 STAT 2 CCND3 cyclin D3 896 ENSG00000112576 IRF4 interferon regulatory factor 4 3662 ENSG00000137265 CCDC102B coiled-coil domain containing 79839 ENSG00000150636 102B P2RX4 purinergic receptor P2X 4 5025 ENSG00000135124 IL4 interleukin 4 3565 ENSG00000113520 CCL20 C-C motif chemokine ligand 6364 ENSG00000115009 20 EPO erythropoietin 2056 ENSG00000130427 ITGA4 integrin subunit alpha 4 3676 ENSG00000115232 P2RY13 purinergic receptor P2Y13 53829 ENSG00000181631 PPFIA4 PTPRF interacting protein 8497 ENSG00000143847 alpha 4 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 CEBPB CCAAT enhancer binding 1051 ENSG00000172216 protein beta CXCL14 C-X-C motif chemokine 9547 ENSG00000145824 ligand 14 IL2RG interleukin 2 receptor subunit 3561 ENSG00000147168 gamma PABPC1 poly(A) binding protein 26986 ENSG00000070756 cytoplasmic 1 PIK3R3 phosphoinositide-3-kinase 8503 ENSG00000117461 regulatory subunit 3 TRAF2 TNF receptor associated 7186 ENSG00000127191 factor 2 DAB2IP DAB2 interacting protein 153090 ENSG00000136848 HTR2A 5-hydroxytryptamine receptor 3356 ENSG00000102468 2A TRDC T cell receptor delta constant 28526 NA GFAP glial fibrillary acidic protein 2670 ENSG00000131095 PLCG1 phospholipase C gamma 1 5335 ENSG00000124181 RIPK1 receptor interacting 8737 ENSG00000137275 serine / threonine kinase 1 ISG15 ISG15 ubiquitin like modifier 9636 ENSG00000187608 STAM signal transducing adaptor 8027 ENSG00000136738 molecule APLF aprataxin and PNKP like 200558 ENSG00000169621 factor BTLA B and T lymphocyte 151888 ENSG00000186265 associated IL7R interleukin 7 receptor 3575 ENSG00000168685 FOSL1 FOS like 1, AP-1 transcription 8061 ENSG00000175592 factor subunit PLCG2 phospholipase C gamma 2 5336 ENSG00000197943 SLAMF6 SLAM family member 6 114836 ENSG00000162739 IRF9 interferon regulatory factor 9 10379 ENSG00000213928 ATF3 activating transcription factor 467 ENSG00000162772 3 PNOC prepronociceptin 5368 ENSG00000168081 TNFRSF13C TNF receptor superfamily 115650 ENSG00000159958 member 13C MS4A1 membrane spanning 4- 931 ENSG00000156738 domains A1 ITGB4 integrin subunit beta 4 3691 ENSG00000132470 Table 2.2.A. Tocilizumab model genes that increase in responders ranked by variable importance and variable frequency. Gene Full Gene Name NCBI Gene EnsemblID ID GPNMB glycoprotein nmb 10457 ENSG00000136235 MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 SOCS6 suppressor of cytokine 9306 ENSG00000170677 signaling 6 MERTK MER proto-oncogene, 10461 ENSG00000153208 tyrosine kinase AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 PIK3R2 phosphoinositide-3-kinase 5296 ENSG00000105647 regulatory subunit 2 PIAS2 protein inhibitor of activated 9063 ENSG00000078043 STAT 2 P2RX4 purinergic receptor P2X 4 5025 ENSG00000135124 CCL20 C-C motif chemokine ligand 6364 ENSG00000115009 20 EPO erythropoietin 2056 ENSG00000130427 PPFIA4 PTPRF interacting protein 8497 ENSG00000143847 alpha 4 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 CEBPB CCAAT enhancer binding 1051 ENSG00000172216 protein beta HTR2A 5-hydroxytryptamine receptor 3356 ENSG00000102468 2A STAM signal transducing adaptor 8027 ENSG00000136738 molecule APLF aprataxin and PNKP like 200558 ENSG00000169621 factor FOSL1 FOS like 1, AP-1 transcription 8061 ENSG00000175592 factor subunit Table 2.2.B. Tocilizumab model genes that increase in non-responders ranked by variable importance and variable frequency. Gene Full Gene Name NCBI Gene EnsemblID ID TOX2 TOX high mobility group box 84969 ENSG00000124191 family member 2 CD40LG CD40 ligand 959 ENSG00000102245 CD247 CD247 molecule 919 ENSG00000198821 NKG7 natural killer cell granule 4818 ENSG00000105374 protein 7 GZMH granzyme H 2999 ENSG00000100450 CDKN1B cyclin dependent kinase 1027 ENSG00000111276 inhibitor 1B SP140 SP140 nuclear body protein 11262 ENSG00000079263 CD3G CD3 gamma subunit of T-cell 917 ENSG00000160654 receptor complex TCIM transcriptional and immune 56892 ENSG00000176907 response regulator BLK BLK proto-oncogene, Src 640 ENSG00000136573 family tyrosine kinase FCRL3 Fc receptor like 3 115352 ENSG00000160856 KCNA3 potassium voltage-gated 3738 ENSG00000177272 channel subfamily A member 3 CST7 cystatin F 8530 ENSG00000077984 CD6 CD6 molecule 923 ENSG00000013725 CD3E CD3 epsilon subunit of T-cell 916 ENSG00000198851 receptor complex KLF2 KLF transcription factor 2 10365 ENSG00000127528 PTGS2 prostaglandin-endoperoxide 5743 ENSG00000073756 synthase 2 IL2RB interleukin 2 receptor subunit 3560 ENSG00000100385 beta IL23A interleukin 23 subunit alpha 51561 ENSG00000110944 LTB lymphotoxin beta 4050 ENSG00000227507 CD79B CD79b molecule 974 ENSG00000007312 TRPC6 transient receptor potential 7225 ENSG00000137672 cation channel subfamily C member 6 CASP10 caspase 10 843 ENSG00000003400 ICOS inducible T cell costimulator 29851 ENSG00000163600 RALGPS2 Ral GEF with PH domain and 55103 ENSG00000116191 SH3 binding motif 2 JAG1 jagged canonical Notch 182 ENSG00000101384 ligand 1 IL24 interleukin 24 11009 ENSG00000162892 CD2 CD2 molecule 914 ENSG00000116824 GZMK granzyme K 3003 ENSG00000113088 GNLY granulysin 10578 ENSG00000115523 CD3D CD3 delta subunit of T-cell 915 ENSG00000167286 receptor complex LY9 lymphocyte antigen 9 4063 ENSG00000122224 SDC1 syndecan 1 6382 ENSG00000115884 YPEL5 yippee like 5 51646 ENSG00000119801 ASNS asparagine synthetase 440 ENSG00000070669 (glutamine-hydrolyzing) LTA lymphotoxin alpha 4049 ENSG00000226979 ZBP1 insulin like growth factor 2 10642 ENSG00000159217 mRNA binding protein 1 GZMB granzyme B 3002 ENSG00000100453 GZMA granzyme A 3001 ENSG00000145649 CSF2RB colony stimulating factor 2 1439 ENSG00000100368 receptor subunit beta SELE selectin E 6401 ENSG00000007908 MX1 MX dynamin like GTPase 1 4599 ENSG00000157601 ITGA1 integrin subunit alpha 1 3672 ENSG00000213949 TNFRSF10B TNF receptor superfamily 8795 ENSG00000120889 member 10b CX3CL1 C-X3-C motif chemokine 6376 ENSG00000006210 ligand 1 CD8A CD8 subunit alpha 925 ENSG00000153563 CCND3 cyclin D3 896 ENSG00000112576 IRF4 interferon regulatory factor 4 3662 ENSG00000137265 CCDC102B coiled-coil domain containing 79839 ENSG00000150636 102B IL4 interleukin 4 3565 ENSG00000113520 ITGA4 integrin subunit alpha 4 3676 ENSG00000115232 P2RY13 purinergic receptor P2Y13 53829 ENSG00000181631 CXCL14 C-X-C motif chemokine 9547 ENSG00000145824 ligand 14 IL2RG interleukin 2 receptor subunit 3561 ENSG00000147168 gamma PABPC1 poly(A) binding protein 26986 ENSG00000070756 cytoplasmic 1 PIK3R3 phosphoinositide-3-kinase 8503 ENSG00000117461 regulatory subunit 3 TRAF2 TNF receptor associated 7186 ENSG00000127191 factor 2 DAB2IP DAB2 interacting protein 153090 ENSG00000136848 TRDC T cell receptor delta constant 28526 NA GFAP glial fibrillary acidic protein 2670 ENSG00000131095 PLCG1 phospholipase C gamma 1 5335 ENSG00000124181 RIPK1 receptor interacting 8737 ENSG00000137275 serine / threonine kinase 1 ISG15 ISG15 ubiquitin like modifier 9636 ENSG00000187608 BTLA B and T lymphocyte 151888 ENSG00000186265 associated IL7R interleukin 7 receptor 3575 ENSG00000168685 PLCG2 phospholipase C gamma 2 5336 ENSG00000197943 SLAMF6 SLAM family member 6 114836 ENSG00000162739 IRF9 interferon regulatory factor 9 10379 ENSG00000213928 ATF3 activating transcription factor 467 ENSG00000162772 3 PNOC prepronociceptin 5368 ENSG00000168081 TNFRSF13C TNF receptor superfamily 115650 ENSG00000159958 member 13C MS4A1 membrane spanning 4- 931 ENSG00000156738 domains A1 ITGB4 integrin subunit beta 4 3691 ENSG00000132470 Susceptibility or refractoriness to a B cell targeted therapy In one aspect, the invention provides a method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a marker panel in a sample from the patient, wherein the agent is a B cell targeted therapy, and wherein the marker panel comprises one or more markers selected from Table 3.1. In one aspect, the invention provides a method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with a B cell targeted therapy, the method comprising determining the profile of a marker panel in a sample from the patient, and wherein the marker panel comprises one or more markers selected from Table 3.1. In some embodiments, the marker panel comprises 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80 or all 81 markers from Table 3.1. In some embodiments, the marker panel comprises the first 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79 or 80 markers listed in Table 3.1. In some embodiments, the marker panel comprises all 81 markers from Table 3.1. In some embodiments, the marker panel consists of 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80 or all 81 markers from Table 3.1. In some embodiments, the marker panel consists of the first 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79 or 80 markers listed in Table 3.1. In some embodiments, the marker panel consists of all 81 markers from Table 3.1. Table 3.1. Rituximab model genes ranked by variable importance. Gene Full Gene Name NCBI Gene EnsemblID ID CISH cytokine inducible SH2 1154 ENSG00000114737 containing protein PPFIA4 PTPRF interacting protein 8497 ENSG00000143847 alpha 4 ZNF773 zinc finger protein 773 374928 ENSG00000152439 CDKN1B cyclin dependent kinase 1027 ENSG00000111276 inhibitor 1B ZNF823 zinc finger protein 823 55552 ENSG00000197933 MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 BCL2 BCL2 apoptosis regulator 596 ENSG00000171791 ATF3 activating transcription factor 467 ENSG00000162772 3 IFNL1 interferon lambda 1 282618 ENSG00000182393 IL12B interleukin 12B 3593 ENSG00000113302 CREB3 cAMP responsive element 10488 ENSG00000107175 binding protein 3 CFLAR CASP8 and FADD like 8837 ENSG00000003402 apoptosis regulator STAT6 signal transducer and 6778 ENSG00000166888 activator of transcription 6 MAPK12 mitogen-activated protein 6300 ENSG00000188130 kinase 12 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 FGFRL1 fibroblast growth factor 53834 ENSG00000127418 receptor like 1 MMP9 matrix metallopeptidase 9 4318 ENSG00000100985 P2RY13 purinergic receptor P2Y13 53829 ENSG00000181631 ERBB3 erb-b2 receptor tyrosine 2065 ENSG00000065361 kinase 3 ISY1 ISY1 splicing factor homolog 57461 ENSG00000240682 IL17C interleukin 17C 27189 ENSG00000124391 TRDC T cell receptor delta constant 28526 NA IL23A interleukin 23 subunit alpha 51561 ENSG00000110944 ASNS asparagine synthetase 440 ENSG00000070669 (glutamine-hydrolyzing) MSH6 mutS homolog 6 2956 ENSG00000116062 STAM signal transducing adaptor 8027 ENSG00000136738 molecule PLCG1 phospholipase C gamma 1 5335 ENSG00000124181 SOX9 SRY-box transcription factor 6662 ENSG00000125398 9 ZNF331 zinc finger protein 331 55422 ENSG00000130844 FADD Fas associated via death 8772 ENSG00000168040 domain SHC1 SHC adaptor protein 1 6464 ENSG00000160691 QPRT quinolinate 23475 ENSG00000103485 phosphoribosyltransferase TNF tumor necrosis factor 7124 ENSG00000232810 IL22 interleukin 22 50616 ENSG00000127318 EPO erythropoietin 2056 ENSG00000130427 PCBP3 poly(rC) binding protein 3 54039 ENSG00000183570 HIVEP3 HIVEP zinc finger 3 59269 ENSG00000127124 CEACAM3 CEA cell adhesion molecule 1084 ENSG00000170956 3 FGF7 fibroblast growth factor 7 2252 ENSG00000140285 PIK3CB phosphatidylinositol-4,5- 5291 ENSG00000051382 bisphosphate 3-kinase catalytic subunit beta COL4A6 collagen type IV alpha 6 1288 ENSG00000197565 chain LAMA3 laminin subunit alpha 3 3909 ENSG00000053747 IFNB1 interferon beta 1 3456 ENSG00000171855 IKBKE inhibitor of nuclear factor 9641 ENSG00000263528 kappa B kinase subunit epsilon ZNF75A zinc finger protein 75a 7627 ENSG00000162086 CENPH centromere protein H 64946 ENSG00000153044 CASP3 caspase 3 836 ENSG00000164305 IL12A interleukin 12A 3592 ENSG00000168811 CNTF ciliary neurotrophic factor 1270 ENSG00000242689 IL5 interleukin 5 3567 ENSG00000113525 SOCS7 suppressor of cytokine 30837 ENSG00000274211 signaling 7 SOCS5 suppressor of cytokine 9655 ENSG00000171150 signaling 5 CENPBD1 CENPB DNA-binding domain 92806 ENSG00000177946 containing 1, pseudogene CD180 CD180 molecule 4064 ENSG00000134061 SLC2A3 solute carrier family 2 6515 ENSG00000059804 member 3 IL4 interleukin 4 3565 ENSG00000113520 APLF aprataxin and PNKP like 200558 ENSG00000169621 factor ATF6B activating transcription factor 1388 ENSG00000213676 6 beta CCND2 cyclin D2 894 ENSG00000118971 IQCG IQ motif containing G 84223 ENSG00000114473 PSMC3 proteasome 26S subunit, 5702 ENSG00000165916 ATPase 3 AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 IL5RA interleukin 5 receptor subunit 3568 ENSG00000091181 alpha HLA-B major histocompatibility 3106 ENSG00000234745 complex, class I, B IL19 interleukin 19 29949 ENSG00000142224 TRIT1 tRNA isopentenyltransferase 54802 ENSG00000043514 1 MSH3 mutS homolog 3 4437 ENSG00000113318 PDGFRL platelet derived growth factor 5157 ENSG00000104213 receptor like IL22RA2 interleukin 22 receptor 116379 ENSG00000164485 subunit alpha 2 LRIG3 leucine rich repeats and 121227 ENSG00000139263 immunoglobulin like domains 3 IL20 interleukin 20 50604 ENSG00000162891 IL17A interleukin 17A 3605 ENSG00000112115 PANK2 pantothenate kinase 2 80025 ENSG00000125779 C19orf12 chromosome 19 open 83636 ENSG00000131943 reading frame 12 GTF2H2 general transcription factor 2966 ENSG00000145736 IIH subunit 2 MCL1 MCL1 apoptosis regulator, 4170 ENSG00000143384 BCL2 family member PIAS1 protein inhibitor of activated 8554 ENSG00000033800 STAT 1 IL17F interleukin 17F 112744 ENSG00000112116 SOS2 SOS Ras / Rho guanine 6655 ENSG00000100485 nucleotide exchange factor 2 IFNA1 interferon alpha 1 3439 ENSG00000197919 IL13RA2 interleukin 13 receptor 3598 ENSG00000123496 subunit alpha 2 Table 3.1.A. Rituximab model genes that increase in responders ranked by variable importance. Gene Full Gene Name NCBI Gene EnsemblID ID CISH cytokine inducible SH2 1154 ENSG00000114737 containing protein ZNF773 zinc finger protein 773 374928 ENSG00000152439 CDKN1B cyclin dependent kinase 1027 ENSG00000111276 inhibitor 1B BCL2 BCL2 apoptosis regulator 596 ENSG00000171791 ATF3 activating transcription factor 467 ENSG00000162772 3 MAPK12 mitogen-activated protein 6300 ENSG00000188130 kinase 12 TRDC T cell receptor delta constant 28526 NA MSH6 mutS homolog 6 2956 ENSG00000116062 ZNF331 zinc finger protein 331 55422 ENSG00000130844 QPRT quinolinate 23475 ENSG00000103485 phosphoribosyltransferase TNF tumor necrosis factor 7124 ENSG00000232810 IL22 interleukin 22 50616 ENSG00000127318 HIVEP3 HIVEP zinc finger 3 59269 ENSG00000127124 PIK3CB phosphatidylinositol-4,5- 5291 ENSG00000051382 bisphosphate 3-kinase catalytic subunit beta LAMA3 laminin subunit alpha 3 3909 ENSG00000053747 IFNB1 interferon beta 1 3456 ENSG00000171855 CENPH centromere protein H 64946 ENSG00000153044 CASP3 caspase 3 836 ENSG00000164305 CD180 CD180 molecule 4064 ENSG00000134061 APLF aprataxin and PNKP like 200558 ENSG00000169621 factor ATF6B activating transcription factor 1388 ENSG00000213676 6 beta CCND2 cyclin D2 894 ENSG00000118971 IQCG IQ motif containing G 84223 ENSG00000114473 AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 IL5RA interleukin 5 receptor subunit 3568 ENSG00000091181 alpha HLA-B major histocompatibility 3106 ENSG00000234745 complex, class I, B TRIT1 tRNA isopentenyltransferase 54802 ENSG00000043514 1 MSH3 mutS homolog 3 4437 ENSG00000113318 PDGFRL platelet derived growth factor 5157 ENSG00000104213 receptor like IL17A interleukin 17A 3605 ENSG00000112115 PANK2 pantothenate kinase 2 80025 ENSG00000125779 C19orf12 chromosome 19 open 83636 ENSG00000131943 reading frame 12 IL17F interleukin 17F 112744 ENSG00000112116 SOS2 SOS Ras / Rho guanine 6655 ENSG00000100485 nucleotide exchange factor 2 IFNA1 interferon alpha 1 3439 ENSG00000197919 Table 3.1.B. Rituximab model genes that increase in non-responders ranked by variable importance. Gene Full Gene Name NCBI Gene EnsemblID ID PPFIA4 PTPRF interacting protein 8497 ENSG00000143847 alpha 4 ZNF823 zinc finger protein 823 55552 ENSG00000197933 MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 IFNL1 interferon lambda 1 282618 ENSG00000182393 IL12B interleukin 12B 3593 ENSG00000113302 CREB3 cAMP responsive element 10488 ENSG00000107175 binding protein 3 CFLAR CASP8 and FADD like 8837 ENSG00000003402 apoptosis regulator STAT6 signal transducer and 6778 ENSG00000166888 activator of transcription 6 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 FGFRL1 fibroblast growth factor 53834 ENSG00000127418 receptor like 1 MMP9 matrix metallopeptidase 9 4318 ENSG00000100985 P2RY13 purinergic receptor P2Y13 53829 ENSG00000181631 ERBB3 erb-b2 receptor tyrosine 2065 ENSG00000065361 kinase 3 ISY1 ISY1 splicing factor homolog 57461 ENSG00000240682 IL17C interleukin 17C 27189 ENSG00000124391 IL23A interleukin 23 subunit alpha 51561 ENSG00000110944 ASNS asparagine synthetase 440 ENSG00000070669 (glutamine-hydrolyzing) STAM signal transducing adaptor 8027 ENSG00000136738 molecule PLCG1 phospholipase C gamma 1 5335 ENSG00000124181 SOX9 SRY-box transcription factor 6662 ENSG00000125398 9 FADD Fas associated via death 8772 ENSG00000168040 domain SHC1 SHC adaptor protein 1 6464 ENSG00000160691 EPO erythropoietin 2056 ENSG00000130427 PCBP3 poly(rC) binding protein 3 54039 ENSG00000183570 CEACAM3 CEA cell adhesion molecule 1084 ENSG00000170956 3 FGF7 fibroblast growth factor 7 2252 ENSG00000140285 COL4A6 collagen type IV alpha 6 1288 ENSG00000197565 chain IKBKE inhibitor of nuclear factor 9641 ENSG00000263528 kappa B kinase subunit epsilon ZNF75A zinc finger protein 75a 7627 ENSG00000162086 IL12A interleukin 12A 3592 ENSG00000168811 CNTF ciliary neurotrophic factor 1270 ENSG00000242689 IL5 interleukin 5 3567 ENSG00000113525 SOCS7 suppressor of cytokine 30837 ENSG00000274211 signaling 7 SOCS5 suppressor of cytokine 9655 ENSG00000171150 signaling 5 CENPBD1 CENPB DNA-binding domain 92806 ENSG00000177946 containing 1, pseudogene SLC2A3 solute carrier family 2 6515 ENSG00000059804 member 3 IL4 interleukin 4 3565 ENSG00000113520 PSMC3 proteasome 26S subunit, 5702 ENSG00000165916 ATPase 3 IL19 interleukin 19 29949 ENSG00000142224 IL22RA2 interleukin 22 receptor 116379 ENSG00000164485 subunit alpha 2 LRIG3 leucine rich repeats and 121227 ENSG00000139263 immunoglobulin like domains 3 IL20 interleukin 20 50604 ENSG00000162891 GTF2H2 general transcription factor 2966 ENSG00000145736 IIH subunit 2 MCL1 MCL1 apoptosis regulator, 4170 ENSG00000143384 BCL2 family member PIAS1 protein inhibitor of activated 8554 ENSG00000033800 STAT 1 IL13RA2 interleukin 13 receptor 3598 ENSG00000123496 subunit alpha 2 In one aspect, the invention provides a method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a marker panel in a sample from the patient, wherein the agent is a B cell targeted therapy, and wherein the marker panel comprises one or more markers selected from Table 3.2. In one aspect, the invention provides a method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with a B cell targeted therapy, the method comprising determining the profile of a marker panel in a sample from the patient, and wherein the marker panel comprises one or more markers selected from Table 3.2. In some embodiments, the marker panel comprises 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80 or all 81 markers from Table 3.2. In some embodiments, the marker panel comprises the first 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79 or 80 markers listed in Table 3.2. In some embodiments, the marker panel comprises all 81 markers from Table 3.2. In some embodiments, the marker panel consists of 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80 or all 81 markers from Table 3.2. In some embodiments, the marker panel consists of the first 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79 or 80 markers listed in Table 3.2. In some embodiments, the marker panel consists of all 81 markers from Table 3.2. Table 3.2. Rituximab model genes ranked by variable importance and variable frequency. Gene Full Gene Name NCBI Gene EnsemblID ID MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 CISH cytokine inducible SH2 1154 ENSG00000114737 containing protein PPFIA4 PTPRF interacting protein 8497 ENSG00000143847 alpha 4 MAPK12 mitogen-activated protein 6300 ENSG00000188130 kinase 12 IL17C interleukin 17C 27189 ENSG00000124391 IL5 interleukin 5 3567 ENSG00000113525 ZNF773 zinc finger protein 773 374928 ENSG00000152439 MMP9 matrix metallopeptidase 9 4318 ENSG00000100985 IL22 interleukin 22 50616 ENSG00000127318 CENPH centromere protein H 64946 ENSG00000153044 ZNF823 zinc finger protein 823 55552 ENSG00000197933 CREB3 cAMP responsive element 10488 ENSG00000107175 binding protein 3 ZNF331 zinc finger protein 331 55422 ENSG00000130844 SHC1 SHC adaptor protein 1 6464 ENSG00000160691 ZNF75A zinc finger protein 75a 7627 ENSG00000162086 IL20 interleukin 20 50604 ENSG00000162891 ATF3 activating transcription factor 467 ENSG00000162772 3 CFLAR CASP8 and FADD like 8837 ENSG00000003402 apoptosis regulator STAT6 signal transducer and 6778 ENSG00000166888 activator of transcription 6 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 P2RY13 purinergic receptor P2Y13 53829 ENSG00000181631 BCL2 BCL2 apoptosis regulator 596 ENSG00000171791 IFNL1 interferon lambda 1 282618 ENSG00000182393 ERBB3 erb-b2 receptor tyrosine 2065 ENSG00000065361 kinase 3 PCBP3 poly(rC) binding protein 3 54039 ENSG00000183570 COL4A6 collagen type IV alpha 6 1288 ENSG00000197565 chain CDKN1B cyclin dependent kinase 1027 ENSG00000111276 inhibitor 1B IL12B interleukin 12B 3593 ENSG00000113302 TRDC T cell receptor delta constant 28526 NA ASNS asparagine synthetase 440 ENSG00000070669 (glutamine-hydrolyzing) MSH6 mutS homolog 6 2956 ENSG00000116062 IFNB1 interferon beta 1 3456 ENSG00000171855 IL17F interleukin 17F 112744 ENSG00000112116 FGFRL1 fibroblast growth factor 53834 ENSG00000127418 receptor like 1 ISY1 ISY1 splicing factor homolog 57461 ENSG00000240682 IL23A interleukin 23 subunit alpha 51561 ENSG00000110944 SOX9 SRY-box transcription factor 6662 ENSG00000125398 9 FADD Fas associated via death 8772 ENSG00000168040 domain TNF tumor necrosis factor 7124 ENSG00000232810 EPO erythropoietin 2056 ENSG00000130427 IL12A interleukin 12A 3592 ENSG00000168811 CNTF ciliary neurotrophic factor 1270 ENSG00000242689 CENPBD1 CENPB DNA-binding domain 92806 ENSG00000177946 containing 1, pseudogene SLC2A3 solute carrier family 2 6515 ENSG00000059804 member 3 PSMC3 proteasome 26S subunit, 5702 ENSG00000165916 ATPase 3 HLA-B major histocompatibility 3106 ENSG00000234745 complex, class I, B STAM signal transducing adaptor 8027 ENSG00000136738 molecule PLCG1 phospholipase C gamma 1 5335 ENSG00000124181 QPRT quinolinate 23475 ENSG00000103485 phosphoribosyltransferase HIVEP3 HIVEP zinc finger 3 59269 ENSG00000127124 CEACAM3 CEA cell adhesion molecule 1084 ENSG00000170956 3 FGF7 fibroblast growth factor 7 2252 ENSG00000140285 PIK3CB phosphatidylinositol-4,5- 5291 ENSG00000051382 bisphosphate 3-kinase catalytic subunit beta LAMA3 laminin subunit alpha 3 3909 ENSG00000053747 IKBKE inhibitor of nuclear factor 9641 ENSG00000263528 kappa B kinase subunit epsilon CASP3 caspase 3 836 ENSG00000164305 SOCS7 suppressor of cytokine 30837 ENSG00000274211 signaling 7 SOCS5 suppressor of cytokine 9655 ENSG00000171150 signaling 5 CD180 CD180 molecule 4064 ENSG00000134061 IL4 interleukin 4 3565 ENSG00000113520 APLF aprataxin and PNKP like 200558 ENSG00000169621 factor ATF6B activating transcription factor 1388 ENSG00000213676 6 beta CCND2 cyclin D2 894 ENSG00000118971 IQCG IQ motif containing G 84223 ENSG00000114473 AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 IL5RA interleukin 5 receptor subunit 3568 ENSG00000091181 alpha IL19 interleukin 19 29949 ENSG00000142224 TRIT1 tRNA isopentenyltransferase 54802 ENSG00000043514 1 MSH3 mutS homolog 3 4437 ENSG00000113318 PDGFRL platelet derived growth factor 5157 ENSG00000104213 receptor like IL22RA2 interleukin 22 receptor 116379 ENSG00000164485 subunit alpha 2 LRIG3 leucine rich repeats and 121227 ENSG00000139263 immunoglobulin like domains 3 IL17A interleukin 17A 3605 ENSG00000112115 PANK2 pantothenate kinase 2 80025 ENSG00000125779 C19orf12 chromosome 19 open 83636 ENSG00000131943 reading frame 12 GTF2H2 general transcription factor 2966 ENSG00000145736 IIH subunit 2 MCL1 MCL1 apoptosis regulator, 4170 ENSG00000143384 BCL2 family member PIAS1 protein inhibitor of activated 8554 ENSG00000033800 STAT 1 SOS2 SOS Ras / Rho guanine 6655 ENSG00000100485 nucleotide exchange factor 2 IFNA1 interferon alpha 1 3439 ENSG00000197919 IL13RA2 interleukin 13 receptor 3598 ENSG00000123496 subunit alpha 2 Table 3.2.A. Rituximab model genes that increase in responders ranked by variable importance and variable frequency. Gene Full Gene Name NCBI Gene EnsemblID ID CISH cytokine inducible SH2 1154 ENSG00000114737 containing protein MAPK12 mitogen-activated protein 6300 ENSG00000188130 kinase 12 ZNF773 zinc finger protein 773 374928 ENSG00000152439 IL22 interleukin 22 50616 ENSG00000127318 CENPH centromere protein H 64946 ENSG00000153044 ZNF331 zinc finger protein 331 55422 ENSG00000130844 ATF3 activating transcription factor 467 ENSG00000162772 3 BCL2 BCL2 apoptosis regulator 596 ENSG00000171791 CDKN1B cyclin dependent kinase 1027 ENSG00000111276 inhibitor 1B TRDC T cell receptor delta constant 28526 NA MSH6 mutS homolog 6 2956 ENSG00000116062 IFNB1 interferon beta 1 3456 ENSG00000171855 IL17F interleukin 17F 112744 ENSG00000112116 TNF tumor necrosis factor 7124 ENSG00000232810 HLA-B major histocompatibility 3106 ENSG00000234745 complex, class I, B QPRT quinolinate 23475 ENSG00000103485 phosphoribosyltransferase HIVEP3 HIVEP zinc finger 3 59269 ENSG00000127124 PIK3CB phosphatidylinositol-4,5- 5291 ENSG00000051382 bisphosphate 3-kinase catalytic subunit beta LAMA3 laminin subunit alpha 3 3909 ENSG00000053747 CASP3 caspase 3 836 ENSG00000164305 CD180 CD180 molecule 4064 ENSG00000134061 APLF aprataxin and PNKP like 200558 ENSG00000169621 factor ATF6B activating transcription factor 1388 ENSG00000213676 6 beta CCND2 cyclin D2 894 ENSG00000118971 IQCG IQ motif containing G 84223 ENSG00000114473 AL161785.1 Long non-coding RNA NR_121587.1 ENSG00000224307 LINC02975 IL5RA interleukin 5 receptor subunit 3568 ENSG00000091181 alpha TRIT1 tRNA isopentenyltransferase 54802 ENSG00000043514 1 MSH3 mutS homolog 3 4437 ENSG00000113318 PDGFRL platelet derived growth factor 5157 ENSG00000104213 receptor like IL17A interleukin 17A 3605 ENSG00000112115 PANK2 pantothenate kinase 2 80025 ENSG00000125779 C19orf12 chromosome 19 open 83636 ENSG00000131943 reading frame 12 SOS2 SOS Ras / Rho guanine 6655 ENSG00000100485 nucleotide exchange factor 2 IFNA1 interferon alpha 1 3439 ENSG00000197919 Table 3.2.B. Rituximab model genes that increase in non-responders ranked by variable importance and variable frequency. Gene Full Gene Name NCBI Gene EnsemblID ID MICU1 mitochondrial calcium uptake 10367 ENSG00000107745 1 PPFIA4 PTPRF interacting protein 8497 ENSG00000143847 alpha 4 IL17C interleukin 17C 27189 ENSG00000124391 IL5 interleukin 5 3567 ENSG00000113525 MMP9 matrix metallopeptidase 9 4318 ENSG00000100985 ZNF823 zinc finger protein 823 55552 ENSG00000197933 CREB3 cAMP responsive element 10488 ENSG00000107175 binding protein 3 SHC1 SHC adaptor protein 1 6464 ENSG00000160691 ZNF75A zinc finger protein 75a 7627 ENSG00000162086 IL20 interleukin 20 50604 ENSG00000162891 CFLAR CASP8 and FADD like 8837 ENSG00000003402 apoptosis regulator STAT6 signal transducer and 6778 ENSG00000166888 activator of transcription 6 BIRC2 baculoviral IAP repeat 329 ENSG00000110330 containing 2 P2RY13 purinergic receptor P2Y13 53829 ENSG00000181631 IFNL1 interferon lambda 1 282618 ENSG00000182393 ERBB3 erb-b2 receptor tyrosine 2065 ENSG00000065361 kinase 3 PCBP3 poly(rC) binding protein 3 54039 ENSG00000183570 COL4A6 collagen type IV alpha 6 1288 ENSG00000197565 chain IL12B interleukin 12B 3593 ENSG00000113302 ASNS asparagine synthetase 440 ENSG00000070669 (glutamine-hydrolyzing) FGFRL1 fibroblast growth factor 53834 ENSG00000127418 receptor like 1 ISY1 ISY1 splicing factor homolog 57461 ENSG00000240682 IL23A interleukin 23 subunit alpha 51561 ENSG00000110944 SOX9 SRY-box transcription factor 6662 ENSG00000125398 9 FADD Fas associated via death 8772 ENSG00000168040 domain EPO erythropoietin 2056 ENSG00000130427 IL12A interleukin 12A 3592 ENSG00000168811 CNTF ciliary neurotrophic factor 1270 ENSG00000242689 CENPBD1 CENPB DNA-binding domain 92806 ENSG00000177946 containing 1, pseudogene SLC2A3 solute carrier family 2 6515 ENSG00000059804 member 3 PSMC3 proteasome 26S subunit, 5702 ENSG00000165916 ATPase 3 STAM signal transducing adaptor 8027 ENSG00000136738 molecule PLCG1 phospholipase C gamma 1 5335 ENSG00000124181 CEACAM3 CEA cell adhesion molecule 1084 ENSG00000170956 3 FGF7 fibroblast growth factor 7 2252 ENSG00000140285 IKBKE inhibitor of nuclear factor 9641 ENSG00000263528 kappa B kinase subunit epsilon SOCS7 suppressor of cytokine 30837 ENSG00000274211 signaling 7 SOCS5 suppressor of cytokine 9655 ENSG00000171150 signaling 5 IL4 interleukin 4 3565 ENSG00000113520 IL19 interleukin 19 29949 ENSG00000142224 IL22RA2 interleukin 22 receptor 116379 ENSG00000164485 subunit alpha 2 LRIG3 leucine rich repeats and 121227 ENSG00000139263 immunoglobulin like domains 3 GTF2H2 general transcription factor 2966 ENSG00000145736 IIH subunit 2 MCL1 MCL1 apoptosis regulator, 4170 ENSG00000143384 BCL2 family member PIAS1 protein inhibitor of activated 8554 ENSG00000033800 STAT 1 IL13RA2 interleukin 13 receptor 3598 ENSG00000123496 subunit alpha 2 In some embodiments, the marker panel may further comprise one or more clinical parameters. In some embodiments, the marker panel may further comprise a tender joint count parameter, for example a square root (tender joint count). In some embodiments, the marker panel may further comprise an erythrocyte sedimentation rate parameter, for example a square root (erythrocyte sedimentation rate). In some embodiments, the marker panel may further comprise a visual analogue score (VAS). Tender joint count is the 28-joint tender joint count, which is part of the DAS28 clinical assessment. It is an integer score that runs from 0 to 28. Erythrocyte sedimentation rate is a standard blood test which is routinely measured. Its units are in mm / hour. It may be carried out, for example, by adding a sample to a tube which may then be left to stand for 1 hour and the amount of sedimented blood measured in mm. Automated systems may also be used that test viscosity of blood and convert this to an ESR reading. Visual analogue score (VAS) is a psychometric response scale with which patients are asked to rate their overall arthritis activity from 0 to 100. These three clinical parameters are routinely determined for RA patients and the skilled person would be readily familiar with methods for their determination and analysis. The methods of the invention may apply statistical methods as would be understood by the skilled person. For example, the methods of the invention may apply a model such as an elastic net regression model or a gradient boosting machine model, such as a model disclosed herein in the Examples (e.g. using one or more coefficient or variable importance disclosed therein, e.g. at Tables 7-10). In some embodiments, the method comprises inputting the profile of the first, second and / or third marker panel into a trained machine learning algorithm. Suitably, the trained machine learning algorithm is arranged to: (i) compare the level of the one or more marker to the corresponding reference value; and (ii) output whether the agent is suitable for treating the Rheumatoid Arthritis (RA) patient. Suitably, the trained machine learning algorithm is arranged to: (i) compare the level of the one or more marker to the corresponding reference value; and (ii) output whether the Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy. Suitably, the trained machine learning algorithm is a gradient boosted machine (gbm), elastic net regression (glmnet), mixed discriminant analysis (mda), random forest (rf), support vector machine (svm) with polynomial (svmPoly) or radial (svmRadial) kernel, extreme gradient boosting (xgboost) with tree booster (xgbTree) or linear booster (xgbLinear) model. The trained machine learning algorithm may be trained based on training data. Suitably, the training data may comprise: first data including marker profiles for a plurality of patients before and after administration of the agent or therapy; and second data identifying whether for each of the plurality of patients the patient was susceptible or refractory to treatment with the agent or therapy. In some embodiments, the method for determining if an agent that downregulates TNF mediated signalling is suitable for treating a Rheumatoid Arthritis (RA) patient applies an elastic net regression model (e.g. glmnet) (e.g. as disclosed herein in the Examples, for example using one or more coefficient or variable importance as disclosed therein, e.g. at Tables 7-10). In some embodiments, the method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling applies an elastic net regression model (e.g. glmnet) (e.g. as disclosed herein in the Examples, for example using one or more coefficient or variable importance as disclosed therein, e.g. at Tables 7-10). In some embodiments, the method for determining if an agent that downregulates IL-6 mediated signalling is suitable for treating a Rheumatoid Arthritis (RA) patient applies a gradient boosted machine model (e.g. gbm) (e.g. as disclosed herein in the Examples, for example using one or more coefficient or variable importance as disclosed therein, e.g. at Tables 7-10). In some embodiments, the method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates IL-6 mediated signalling applies a gradient boosted machine model (e.g. gbm) (e.g. as disclosed herein in the Examples, for example using one or more coefficient or variable importance as disclosed therein, e.g. at Tables 7-10). In some embodiments, the method for determining if a B cell targeted therapy is suitable for treating a Rheumatoid Arthritis (RA) patient applies an extreme gradient boosting model (e.g. xgbLinear) (e.g. as disclosed herein in the Examples, for example using one or more coefficient or variable importance as disclosed therein, e.g. at Tables 7-10). In some embodiments, the method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with a B cell targeted therapy applies an extreme gradient boosting model (e.g. xgbLinear) (e.g. as disclosed herein in the Examples, for example using one or more coefficient or variable importance as disclosed therein, e.g. at Tables 7-10). In some embodiments, the method comprises converting levels of gene expression of the markers of the first, second and / or third marker panels (e.g. determined using a gene expression assay, preferably a multiplexed gene expression assay (e.g. a non-sequencing based assay), such as an nCounter assay) to equivalent expected values for another method of determining the profile of the first, second and / or third marker panels (preferably RNA-Seq). In some embodiments, the converting applies a linear model for each gene. The linear models may use predetermined linear regression components (e.g. intercept and slope coefficient). Each model may be determined based on comparison of first data including a marker profile as determined by a first method (e.g. RNA-Seq) with second data including a marker profile as determined by a second method (e.g. nCounter). In one aspect, the invention provides a method of diagnosis or determination of suitability of an agent for treating a patient, wherein the method comprises a step of converting levels of gene expression of markers of one or more marker panel (e.g. determined using a gene expression assay, preferably a multiplexed gene expression assay (e.g. a non-sequencing based assay), such as an nCounter assay) to equivalent expected values for another method of determining the profile of the one or more marker panel (preferably RNA-Seq). In some embodiments, the converting applies a linear model for each gene. The linear models may use predetermined linear regression components (e.g. intercept and slope coefficient). Each model may be determined based on comparison of first data including a marker profile as determined by a first method (e.g. RNA-Seq) with second data including a marker profile as determined by a second method (e.g. nCounter). Determining marker profile In some embodiments, determining the profile of a first, second and / or third marker panel in a sample from the patient comprises determining the level of the markers of the first, second and / or third marker panel in the sample from the subject. In some embodiments, determining the profile of a first marker panel in a sample from the patient comprises determining the level of the markers of the first marker panel in the sample from the subject. In some embodiments, determining the profile of a second marker panel in a sample from the patient comprises determining the level of the markers of the second marker panel in the sample from the subject. In some embodiments, determining the profile of a third marker panel in a sample from the patient comprises determining the level of the markers of the third marker panel in the sample from the subject. Optionally, the level of the marker(s) is compared to the corresponding reference value. Methods for determining marker levels are well known in the art and would be familiar to the skilled person. For example, the level of a marker may be determined by measuring gene expression for the marker gene (for example, using RTPCR) or by detecting the protein product of the marker gene (for example, using an immunoassay). In some embodiments, determining the level of the marker(s) comprises determining the level of gene expression of the marker(s). In some embodiments, the level is a nucleic acid level. In some embodiments, the nucleic acid level is an mRNA level. In some embodiments, the marker level is determined by direct digital counting of nucleic acids (e.g. by Nanostring), RNA-seq, RT-qPCR, qPCR, multiplex qPCR or RT-qPCR, microarray analysis, or a combination thereof. In preferred embodiments, the marker level is determined by RNA sequencing In some embodiments, the level is a protein level. In some embodiments, the marker level is determined by an immunoassay, liquid chromatography-mass spectrometry (LC-MS), nephelometry, aptamer technology, or a combination thereof. In some embodiments, the marker level is an average of the marker level. In some embodiments, the average of the marker level is an average of a normalised marker level. In some embodiments, the marker level is a median of the marker level. In some embodiments, the median of the marker level is a median of a normalised marker level. In some embodiments, the marker level is the marker level normalised to a reference gene. Reference values The method of the invention may comprise the step of comparing the marker level to a corresponding reference value. As used herein, the term “reference value” may refer to a level against which another level (e.g. the level of one or more marker disclosed herein) is compared (e.g. to make a diagnostic (e.g. predictive and / or prognostic) and / or therapeutic determination). For example, the reference value may be derived from level(s) in a reference population (preferably the median level in a reference population), for example the population of patients disclosed in the Examples herein; a reference sample; and / or a pre-assigned value (e.g. a cut-off value which was previously determined to significantly separate a first subset of individuals who are susceptible to treatment with an agent and a second subset of individuals who are resistant to the treatment; or a cut-off value which was previously determined to significantly separate a first subset of individuals who are refractory to treatment with an agent and a second subset of individuals who are susceptible to the treatment). In some embodiments, the cut-off value may be the median or mean (preferably median) level in the reference population. In some embodiments, the reference level may be the top 40%, the top 30%, the top 20%, the top 10%, the top 5% or the top 1% of the expression level in the reference population. The reference value may, for example, be based on a mean or median level of the biomarker in a control population of subjects, e.g.5, 10, 100, 1000 or more subjects (who may be age- and / or gender-matched, or unmatched to the test subject). In certain embodiments the reference value may have been previously determined, or may be calculated or extrapolated without having to perform a corresponding determination on a control sample with respect to each test sample obtained. The increase in the level of the one or more marker compared to the corresponding reference value (when the level of the one or more marker is greater than the corresponding reference value) may, for example, be an increase in the level of at least about 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 96%, 97%, 98% or 99% or greater relative to the reference value. The increase in the level of the one or more marker compared to the corresponding reference value (when the level of the one or more marker is greater than the corresponding reference value) may, for example, be an increase in the level of at least about 1.1x, 1.2x, 1.3x, 1.4x, 1.5x, 1.6x, 1.7x, 1.8x, 1.9x, 2x, 2.1x, 2.2x, 2.3x, 2.4x, 2.5x, 2.6x, 2.7x, 2.8x, 2.9x, 3x, 3.5x, 4x, 4.5x, 5x, 6x, 7x, 8x, 9x, 10x, 15x, 20x, 30x, 40x, 50x, 100x, 500x or 1000x relative to the reference value. A decrease in a level of one or more marker compared to the corresponding reference values (when the level of the one or more marker is less than the corresponding reference value) may, for example, be a decrease in the level of at least about 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 96%, 97%, 98% or 99% or greater relative to the reference value. Sample The method of the invention may be carried out on one or more sample obtained from a subject, for example a patient with RA. Samples may be obtained from a joint of a subject, for example from a biopsy. Samples may be obtained from a synovial tissue sample from a subject. In preferred embodiments, the one or more sample is a synovial sample. In some embodiments, the synovial sample is a synovial tissue sample or a synovial fluid sample. As used herein, the term “synovial sample” refers to a sample derived from a synovial joint. Typically, the synovial sample will be derived from a synovial joint of a RA patient. A synovial sample may be a synovial tissue biopsy and the synovial joint may display active inflammation at the time the sample is taken. Samples may be biological samples taken from a patient. In some embodiments, a sample is blood. In some embodiments, a sample is serum (e.g. the fluid and solute component of blood without the clotting factors). In some embodiments, a sample is plasma (e.g. the liquid portion of blood). Methods for obtaining samples, such as synovial tissue samples are well known in the art and are familiar to the skilled person. For example, techniques such as ultrasound (US)-guided biopsies may be used to obtain tissue samples. In some embodiments, the sample is obtained by synovial biopsy, preferably ultrasound- guided synovial biopsy. Patient In preferred embodiments, the patient is a human. In preferred embodiments, the patient is an adult human. In some embodiments, the patient may be a child or an infant. In preferred embodiments, the RA patient is refractory to DMARD therapy Antibodies The term “antibody” is used herein to relate to an antibody or a functional fragment thereof. By functional fragment, it is meant any portion of an antibody which retains the ability to bind to the same antigen target as the parental antibody. As used herein, “antibody” means a polypeptide having an antigen binding site which comprises at least one complementarity determining region (CDR). The antibody may comprise 3 CDRs and have an antigen binding site which is equivalent to that of a domain antibody (dAb). The antibody may comprise 6 CDRs and have an antigen binding site which is equivalent to that of a classical antibody molecule. The remainder of the polypeptide may be any sequence which provides a suitable scaffold for the antigen binding site and displays it in an appropriate manner for it to bind the antigen. The antibody may be a whole immunoglobulin molecule or a part thereof such as a Fab, F(ab)’2, Fv, single chain Fv (ScFv) fragment or Nanobody. The antibody may be a conjugate of the antibody and another agent or antibody, for example the antibody may be conjugated to a polymer (e.g. PEG), toxin or label. The antibody may be a bifunctional antibody. The antibody may be non-human, chimeric, humanised or fully human. Methods of treatment In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of an agent that downregulates TNF mediated signalling, wherein the agent is determined to be suitable for treatment by the method of the invention. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of an agent that downregulates TNF mediated signalling, wherein the patient is determined to be susceptible to treatment with an agent that downregulates TNF mediated signalling by the method of the invention. In some embodiments, the agent that downregulates TNF mediated signalling is a TNF inhibitor. In some embodiments, the agent that downregulates TNF mediated signalling is selected from etanercept, infliximab, adalimumab, certolizumab pegol, and golimumab. In preferred embodiments, the agent that downregulates TNF mediated signalling is etanercept. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of an agent that downregulates IL-6 mediated signalling, wherein the agent is determined to be suitable for treatment by the method of the invention. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of an agent that downregulates IL-6 mediated signalling, wherein the patient is determined to be susceptible to treatment with an agent that downregulates IL- 6 mediated signalling by the method of the invention. In some embodiments, the agent that downregulates IL-6 mediated signalling is an IL-6 receptor antagonist. In some embodiments, the agent that downregulates IL-6 mediated signalling is selected from the group consisting of tocilizumab, sarilumab, satralizumab and siltuximab. In preferred embodiments, the agent that downregulates IL-6 mediated signalling is tocilizumab. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of a B cell targeted therapy, wherein the agent is determined to be suitable for treatment by the method of the invention. In one aspect, the invention provides a method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising administering to the patient an effective amount of a B cell targeted therapy, wherein the patient is determined to be susceptible to treatment with a B cell targeted therapy by the method of the invention. In some embodiments, the B cell targeted therapy is B cell depletion therapy. In some embodiments, the B cell targeted therapy is selected from the group consisting of: rituximab, ocrelizumab, veltuzumab, ofatumumab, epratuzumab, obinutuzumab, ibritumomab and tiuxetan. In some embodiments, the B cell targeted therapy is selected from the group consisting of: rituximab, ocrelizumab, veltuzumab, ofatumumab and epratuzumab. In preferred embodiments, the B cell targeted therapy is rituximab. Kits The present invention also provides a kit suitable for performing the method as disclosed herein. In particular, the kit may comprise reagents suitable for detecting the markers disclosed herein, or a marker combination as disclosed herein. The kit may also comprise instructions for use. The kit may also comprise an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling, or a B cell targeted therapy. The skilled person will understand that they can combine all features of the invention disclosed herein without departing from the scope of the invention as disclosed. Preferred features and embodiments of the invention will now be described by way of non- limiting examples. The practice of the present invention will employ, unless otherwise indicated, conventional techniques of chemistry, biochemistry, molecular biology, microbiology and immunology, which are within the capabilities of a person of ordinary skill in the art. Such techniques are explained in the literature. See, for example, Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd Edition, Cold Spring Harbor Laboratory Press; Ausubel, F.M. et al. (1995 and periodic supplements) Current Protocols in Molecular Biology, Ch.9, 13 and 16, John Wiley & Sons; Roe, B., Crabtree, J. and Kahn, A. (1996) DNA Isolation and Sequencing: Essential Techniques, John Wiley & Sons; Polak, J.M. and McGee, J.O’D. (1990) In Situ Hybridization: Principles and Practice, Oxford University Press; Gait, M.J. (1984) Oligonucleotide Synthesis: A Practical Approach, IRL Press; and Lilley, D.M. and Dahlberg, J.E. (1992) Methods in Enzymology: DNA Structures Part A: Synthesis and Physical Analysis of DNA, Academic Press. Each of these general texts is herein incorporated by reference. EXAMPLES EXAMPLE 1 RESULTS Differential gene expression analysis identifies signatures of drug responsiveness in synovial tissue biopsies 223 patients were included in the primary analysis of the STRAP trial, of which 208 had complete RNA-Seq data from synovial biopsies. Of these, 72 had been randomised to rituximab, 69 to tocilizumab and 67 to etanercept. Baseline characteristics, disease activity, and synovial B cell group are reported in Table 4. Patients were assessed for clinical response 16 weeks after starting treatment using the ACR20 criteria with 38 responding to etanercept, 18 responding to tocilizumab and 44 responding to rituximab. Differentially Expressed Gene (DEG) analysis using DESeq2 identified 171 genes differentially expressed between responders and non-responders to etanercept, 41 genes in tocilizumab and 46 genes for rituximab response (FDR<0.05) in baseline synovial biopsies (Figure.1a,c,e). Etanercept and rituximab response was associated with immunoglobulin genes (e.g. IGLV4-60) and TNF receptor superfamily member 13C (TNFRSF13C), while tocilizumab and rituximab responders had genes in common such FLG (filaggrin) and keratin genes (KRT2). However, other keratin genes (KRT1, KRT14, KRT15) together with MYH7, FLG, NRAP and EEF1A2 were associated with non-response to etanercept, likewise LTF, TCL1A associated with tocilizumab non-response and SFN, MYH1 with rituximab. Notably, protein phosphatase catalytic subunits were observed to be upregulated in the non-response group across all three drugs (PPP1R14A in rituximab, PPP1R9A in tocilizumab and PPP2R3A in etanercept). QuSAGE modular analysis of differentially expressed genes showed an increase in multiple B cell modules in responders to etanercept (Figure.1b), while dendritic cells and interferon alpha modules were upregulated in responders to tocilizumab (Figure.1d). For rituximab, several modules related to B cells, T peripheral helper cells (Tph), CD8+T and NK cells were increased in responders, while fibroblast associated modules were associated with non- response (Figure.1f). Defining shared and differential gene signatures of response / resistance across drugs To investigate common molecular patterns of response / resistance to treatment we conducted a drug-independent analysis comparing responders (n=133) and non-responders (n=75) in the whole cohort. DEG analysis (Figure.2a) showed 190 up-regulated genes (FDR<0.05) in the responder group that were associated with inflammation, immunoregulatory interactions and B-cell proliferation such as MS4A1, PAX5, CXCL13, as confirmed by pathway enrichment analysis (Figure.2b). Genes upregulated in the non-responder group were more abundant and heterogeneous (266 genes at FDR<0.05) (Figure.2a). Among these, many myosin-related genes (MYH7, MYL2, MYLK3) together with ATP genes (ATP1A2, ATP2A1, ATP2B2) consistent with the presence of myofibroblasts in synovial tissue, with associated pathways by enrichment analysis (Figure.2b). Multiple protein phosphatase catalytic subunits (PPP) were also found to be upregulated in this group (PPP1R3A, PPP1R3B, PPP1R3C) along with pro-fibrotic genes such as SCN7A and FGF13. The heterogeneity of resistance mechanisms found in the DEG analysis led to a separate analysis aimed at identifying and comparing the different molecular signatures driving resistance to each drug. To analyse specific differences between each drug, genes associated with lack of response to each drug were visualised with a 3-way polar plot (Figure.2c). This revealed 103 genes specifically upregulated in non-responders to etanercept, 12 in rituximab and 9 in tocilizumab respectively. Drug-specific responder genes are shown in Figure.2d revealing greater abundance of genes along the etanercept responder axis (58 genes), and fewer genes for the rituximab-specific and tocilizumab-specific groups (24 and 25 genes respectively). These comparisons highlighted the differential impact of keratin and filaggrin genes on treatment response: KRT1, KRT5, KRT14 and KRT15 were found to be significantly upregulated in etanercept non-responders (Figure.2c,e) but were upregulated in tocilizumab responders (Figure.2d,e). Keratins are among a group of cartilage proteins which are under investigation as possible serum biomarkers of cartilage turnover and joint destruction in RA. Filaggrin genes FLG and FLG2, which represent some of the most important epitopes for pathogenic autoantibody generation in RA, showed a similar pattern of behaviour; high in etanercept non-responders while also being high in tocilizumab responders. High levels of B- cell genes (MS4A1, PAX5) were significantly associated with response to etanercept with an expected trend to rituximab (Figure.2e), while myosin gene MYH7 was significantly upregulated in responders to rituximab while being upregulated in non-responders to etanercept. Cellular composition determined by clustering analysis and single-cell RNA-Seq subset deconvolution is associated with different patterns of drug response To explore the association of immune cells with response to treatment (ACR20), we used clustering analysis and a modular enrichment approach to estimate the relative abundance of cells in the synovial tissue of RA, including fibroblasts, macrophages, B-cells and T-cells based on single-cell RNA-Seq subsets identified in RA synovium by Zhang (Zhang et al. (2019) Nat Immunol 20: 928-42). Clustering of patients using their immune cell subset profiles at baseline was performed to examine the RNA-Seq data for underlying structure and the presence of biological subgroups. This revealed four clusters which partially mapped to the three pathotypes as determined by histology (Figure.3a). The pauci-immune fibroid patients were split across two clusters, namely cluster-1 which was almost entirely composed of fibroid patients and cluster-2 which contained the remainder of fibroid patients. Cluster-3 contained a mixture of diffuse myeloid patients and lympho-myeloid patients, while Cluster-4 was almost entirely composed of lympho-myeloid patients with the highest levels of B / T-cells by CD20 / CD3 histology score and plasma cells by CD138 score. Abundance of DKK3+fibroblasts (SC-F3) and CD34+sublining fibroblasts (SC-F1) delineated cluster-1, while CD55+lining fibroblasts and NUPR1+macrophages defined cluster-2, the other fibroid associated cluster, with intermediate levels of both cells in cluster-3. This suggests that specific synovial single cell subtypes can define additional subgroups into which the original three pathotypes can be further subdivided. Differential abundance of synovial single cell subtypes between responders and non- responders to any of the drugs combined (Figure.3b) was compared with differential abundance to each drug specifically (Figure.3c). This showed that HLA-DRAhighsublining fibroblasts (SC-F2), a proinflammatory subset associated with leukocyte-rich synovial infiltration in RA, was significantly higher (PF2_any=0.008) in responders in the any treatment comparison, which is in parallel to previous findings reported earlier in the R4RA cohort. In R4RA, DKK3+sublining fibroblasts (SC-F3) were increased in patients refractory to treatment. Here, DKK3+sublining fibroblasts (SC-F3) showed a trend to a non-significant increase in etanercept and rituximab treated patients, while CD34+sublining fibroblasts (SC-F1) were significantly increased in non-responders to etanercept. Patients treated with rituximab did not show significant changes in the populations of monocyte / macrophage subsets; however, those who responded to etanercept had a significantly higher proportion of IL1B+pro-inflammatory macrophages (SC-M1) compared to non-responders, and tocilizumab responders had a higher proportion of NUPR1+and IFN- activated macrophages (SC-M2 and SC-M4). The signatures for IGHD+CD27–naive (SC-B1) and IGHG3+CD27+memory (SC-B2) B-cell subsets were significantly higher in responders who received etanercept and any treatment at baseline (PB1_eta=0.036, PB1_any=0.018, PB2_eta=0.032, PB2_any=0.004). The Tph subset (SC-T3) showed one of the greatest fold-change among all immune cell subsets and was significantly upregulated in three responder groups: etanercept, rituximab and any treatment (Figure.3c,d). A significant increase in FOXP3+regulatory T-cells (SC-T2) was seen in the any treatment comparison and in rituximab responders specifically. Similarly, SC-T4 (GZMK+CD8+T-cells) and SC-T6 (GZMK+ / GZMB+T- cells) were also increased in responders in the any treatment comparison. Machine learning models predict drug response to etanercept, rituximab and tocilizumab. Machine learning (ML) predictive models were developed to establish the ability of baseline synovial tissue gene expression and clinical parameters to predict treatment response with the aim of converting RNA-Seq to an nCounter based assay. We capitalised on a synovial specific 506-gene nCounter panel, custom-made with nanoString to cover genes linked to synovial biology, pathotypes, RA pathogenesis and response gene signatures identified in previous RNA-Seq studies. Models were built from RNA-Seq data restricted to these 506 genes and performance tested in an unbiased manner maintaining separation of samples for training and testing using 10^10-fold nested cross-validation with 25 repeats. Eight model types were tested. Predictive performance was measured by AUC (area under ROC curve) for response to each drug, comparing the primary endpoint ACR20 response against five other response endpoints (Figure.7). This showed that for etanercept and tocilizumab the best prediction of response was seen for target DAS28-ESR (<3.2), which was thus selected for model optimisation. Due to difficulties with training a binary response model for rituximab, ordinal regression on DAS28-ESR response categorised into 4 levels were used for model training, followed by converting the fitted model to a binary response for final performance estimation by AUC. For etanercept, the best model system was an elastic net regression model (glmnet) (Figure.4a), with a final model of 15 genes with nested CV AUC of 0.765 (Figure.4b). For tocilizumab the final model was a gradient boosted machine model (gbm) using 35 genes which had a nested CV AUC of 0.736. The final rituximab extreme gradient boosting model (xgbLinear) had 3 clinical parameters and 25 genes leading to a nested CV AUC of 0.754. Variables were ranked according to their importance in models which was averaged across the outer CV folds and the final model to give a reliable estimate of variable importance and its variance (Figure.4c). Conversion of machine learning models into nCounter panels demonstrates high predictive value for patient stratification by response to specific biologics RNA-Seq is a highly effective research tool, but technical / time barriers prevent its use in routine clinical practice, while nCounter is a rapid multiplexed gene expression assay, which has achieved FDA certification for clinical adoption in the field of breast cancer. Thus we converted the RNA-Seq machine learning models into an nCounter based system, which we validated using spare RNA from synovial biopsies (n=118) from the STRAP trial utilising the custom 506-gene panel described above. Comparison of nCounter normalised counts against RNA-Seq on the same samples showed strong correlation between nCounter and RNA-Seq for the vast majority of genes (Figure.8). To be able to input nCounter gene counts into the machine learning models constructed on the RNA-Seq data (Figure.4b,c), we developed an algorithm to convert nCounter count data for each gene to RNA-Seq scale (pseudo-RNA-Seq), using linear models fitted for each gene. Figure.5a outlines the pipeline used to test the validity of the nCounter biomarker panel for predicting response to each of the three biologic drugs randomised in STRAP. nCounter data for each cohort (etanercept n=39, tocilizumab n=34, rituximab n=45, total n=118) was converted to pseudo-RNA-Seq and inputted into the relevant ML model (described in Figure.4b,c) for each cohort. We tested predicted response against actual response at 16 weeks (reference) to the drug received by each patient in the STRAP trial. A balanced accuracy of 0.79 was achieved for the etanercept and tocilizumab models and 0.81 for the rituximab model (Figure.5b), while an AUC of 0.87 was observed for etanercept response, 0.82 for tocilizumab response and 0.87 for rituximab response (Figure.5c). Thus, all three RNA-Seq derived machine learning models were successfully converted and validated using an alternative, non-sequencing-based assay. This led to the development of an algorithm for future clinical use where the nCounter biomarker panel can assign patients to either i) a TNF-inhibitor, ii) an IL6-inhibitor, iii) a B-cell depleting agent, or iv) a biomarker-negative group if patients have low probability (<50%) of responding to all three classes of drugs (Figure.5d). DISCUSSION We performed a comprehensive analysis of RNA-Seq data from synovial biopsies from the STRAP trial that led to the identification of gene signatures associated with response to etanercept, tocilizumb and rituximab (Figure.1a,c,e). B-cell gene modules had a strong theme predicting response to both rituximab and etanercept (Figure.1b,f), while chemokine and cytokine gene modules were associated with response to tocilizumab (Figure.1d). Comparative analysis identified both shared gene response and non-response signatures (Figure.2a,b) with Fc receptor mediated inflammatory signalling as a strong underlying theme. Drug-specific response signatures were also dissected, showing that specific genes could be associated with response to one drug and non-response to another, for example selected fibroblast genes such as KRT1 being associated with response to tocilizumab and non- response to etanercept, while B cell associated genes MS4A1 which encodes CD20 and the B cell specific transcription factor PAX5 were upregulated in etanercept responders. Deconvolution of the bulk RNA-Seq data using single cell subsets from the Accelerated Medicines Partnership (AMP) consortium showed that specific cellular subsets were associated with response, for example SC-M2 and SC-M4 macrophage subsets were associated with tocilizumab response while SC-M1 macrophage, and SC-B1 and SC-B2 B cell subsets were associated with etanercept response (Figure.3b,c). These results are consistent with other recent single-cell work examining the role of specific synovial macrophages subsets in maintaining or limiting remission in RA patients. Single-cell subset analysis of the bulk data segregated with pathotypes (Figure.3d) but suggested greater complexity than the pathotypes we originally described, with for example the pauci-immune fibroid pathotype being split into two clusters based on specific cell types including DKK3+(SC-F3) and CD34+(SC-F1) fibroblasts in cluster-1 and CD55+fibroblasts (SC-F4) in cluster-2. The machine learning analyses identified three predictive models of response to each drug at 16 weeks based on baseline synovial gene expression (Figure.4b) that, for practical clinical use, were converted into a nanoString panel validated using a 506-synovial-gene panel applied to spare synovial RNA (n=118) from the STRAP trial. This analysis showed that each model was successfully validated with accuracy ranging from 79-85% and AUC ranging from 0.82-0.87 (Figure.5b,c). This has led to the development of a novel algorithm which can assign patients to one of three classes of biologic therapy (TNF inhibitor, IL6 inhibitor, B-cell depleting agent) based on their probability of response predicted by machine learning models applied to nCounter assay on a patient’s synovial biopsy (Figure.5d). This algorithm can also predict whether patients have a low probability (<0.5) of responding to all three drugs, in which case they are labelled as “biomarker negative” and can be offered an alternative therapeutic class, thus reducing costs and unnecessary drug exposure to agents unlikely to be effective. In summary, the in-depth analysis presented here has shown that machine learning identifies synovial biopsy biomarkers that predict response to three specific biologic classes, namely TNF-inhibitors, IL6-inhibitors and B-cell depleting agents, targeting three major disease pathways. The machine learning biomarker models were successfully translated into an nCounter assay, which was applied to spare STRAP synovial biopsy RNA, with its clinical utility validated against actual observed outcomes in the trial. Thus, this study provides the underpinning evidence towards changing the prescribing paradigm away from ‘trial-and-error’ and may transform future clinical practice through successful bedside tests to help stratify patients to the most effective drugs from disease outset. METHODS Patients and intervention A total of 226 patients aged 18 years or over, fulfilling 2010 ACR / EULAR classification criteria for RA who were eligible for treatment with anti-TNF therapy according to UK National Institute for Health and Care Excellence (NICE) guidelines, i.e. failing or intolerant to conventional synthetic disease-modifying anti-rheumatic drug (csDMARD) therapy were recruited when fulfilling the trial inclusion / exclusion criteria. Due to issues with drug supply, a separate trial ‘STRAP-EU’ was opened which replicated the STRAP trial in the UK, and recruitment was expanded to 4 other countries in Europe in 2018. For the analyses, the data from both trials were combined. Briefly, patients underwent a synovial biopsy of a clinically active joint at entry to the trial performed according to the expertise of local centre as either ultrasound-guided or arthroscopic procedure. Following synovial biopsy, patients were randomised to receive either rituximab as two 1000-mg infusions at an interval of 2 weeks administered at baseline, tocilizumab 162mg administered as a weekly subcutaneous injection, or etanercept 50mg administered as a weekly subcutaneous injection. Patients were followed up every 4 weeks (+ / - 1 week) throughout the 48-week trial treatment period where RA disease activity measurements and safety data were collected. Follow-up of STRAP patients recruited in the UK after 1st January 2019 ended at 24 weeks from baseline, whereas the follow-up of all STRAP-EU patients continued to end at 48 weeks from baseline. Optionally, a repeated synovial biopsy of the same joint sampled at baseline was performed at 16 weeks. The study was conducted in compliance with the Declaration of Helsinki, International Conference on Harmonisation Guidelines for Good Clinical Practice, and local country regulations. The final protocol, amendments and documentation of consent were approved by the institutional review board of each study centre and relevant independent ethics committees. RNA-Seq data processing A total of 283^paired-end RNA-Seq samples from 50^million reads of 150-bp length were mapped to reference human transcriptome (Gencode v29, GRCh38.p12) and transcripts were then quantified using Salmon version 0.13.1. Tximport version 1.13.10 was used to aggregate transcript-level expression data to genes, then counts were subjected to variance-stabilizing transformation (VST) using the DESeq2 version 1.25.9 package. Following RNA-Seq quality control, with Principal Component Analysis (PCA), 5 baseline and 3 follow-up, in total 8^samples were excluded due to poor mapping rate that was originated from low RNA quality (Figure 6a). Thus RNA-Seq data from 208^patients were available for subsequent analysis at baseline (65 samples at later time points are not analysed here). Baseline characteristics of patients with available RNA-Seq are shown in Table 4. Differential expression and modular analysis of RNA-Seq data at baseline Patients classed as responders and non-responders based on their 16-week assessment using ACR20 criteria were compared for each individual treatment group. The groups showed no significant differences for baseline characteristics including histological and molecular B- cell status, gender or disease duration (Table 4). Low expressed genes (expressed in fewer than 19 samples with at least a normalized count of 9) were excluded from analysis. Remaining genes were subjected to Differentially Expressed Gene (DEG) analysis based on general linear regression models with negative binomial distribution using RNA-Seq count data using DESeq2 (version 1.34.0) which uses a Wald test to compare differences between treatment response groups in synovium RNA-Seq samples. P-values were false discovery rate (FDR) adjusted using Storey’s q-value. A cut-off of q < 0.05 was used to identify significant DEGs. DEGs were illustrated by volcano plots and DESeq2 outputs were used for modular analysis using Bioconductor package Quantitative Set Analysis for Gene Expression (QuSAGE, v2.30.0). Weighted gene correlation network analysis (WGCNA) gene modules from Li et al. (Li et al. (2014) Nat Immunol 15: 195-204) were selected for gene set enrichment and relevant modules were summarized in plots. 2x3 way analysis In order to visualise specificity of genes for predicting responsiveness to each gene, DEG identified by the responders vs non-responders contrast in each drug cohort were displayed in three-way polar plots using the volcano3D R package (version 2.0.6) as a 2x3-way analysis. DEGs upregulated in responders were categorised E+, R+, T+ based on the Wald Chi- squared test p-value significance for responder vs non-responder analysis for etanercept, rituximab and tocilizumab respectively. Genes whose FDR result was significant for two drugs were labelled as mixed categories (E+R+, E+T+ or R+T+). Genes significantly upregulated in responders with all three drugs were excluded from this analysis. The graphical position of the genes along the three axes of the polar plot was calculated using estimated log fold change for each gene. The same procedure was applied with directionality reversed to plot genes significantly upregulated in the non-responder group of each drug. Enrichment analysis Multiple GO / pathway analyses were performed to enrich genes of relevance coming from DEG analyses. enrichR (version 3.1) was used to extract GO / pathways from the following repositories: Reactome_2016 and GO_Biological_Process_2021. Gene expression integration into cell specific modules Modular approaches for gene set enrichment analysis and relative quantification of cell subsets are widely used in the recent molecular studies. Here, we integrated gene expression into cell specific modules to characterize the association of synovial immune cells in RA with multidrug resistance. For the enrichment of 18 single-cell subsets identified in scRNA-Seq of RA synovial tissue that are composed of four fibroblast subtypes (SC-F1: CD34+ sublining, SC-F2: HLA+ sublining, SC-F3: DKK3+ sublining and SC-F4: CD55+ lining), four macrophages subtypes (SC-M1: IL1B+ pro-inflammatory, SC-M2: NUPR1+, SC-M3: C1QA+, SC-M4: IFN-activated), six T-cell subtypes (SC-T1: CCR7+ CD4+, SC-T2: FOXP3+ Tregs, SC-T3: PD-1+ Tph / Tfh, SC-T4: GZMK+ CD8+, SC-T5: GNLY+ GZMB+, SC-T6: GZMK+ / GZMB+), and four B-cell subtypes (SC-B1: IGHD+ CD27- naive, SC-B2: IGHG3+ CD27+ memory, SC-B3: Autoimmune associated, SC-B4: Plasmablasts), we scored each subtype using a modular approach that integrates previously published gene signatures. The top 5 exclusively differentially expressed genes for (based on AUC scores) were utilised as cell subtype-specific gene sets. Module scores for each subtype were calculated using the AddModuleScore function from R package Seurat. Linear model method as implemented in R package Limma, was used to detect the differentially abundant cell subpopulations when comparing responders and non-responders. The main aim of STRAP clinical trial is to test the utility of analysing synovial B-cell infiltrates as a potential biomarker to guide therapeutic decisions in patients failing DMARD therapy. To stratify patients according to their synovial B-cell infiltrates into B cell poor / rich pathotypes, we used the B-cell module gene set that contains 71 genes derived from FANTOM5 (see Figure 6c) (FANTOM Consortium et al. (2014) Nature 507: 462-70). First, we assigned the mean of normalized and scaled gene expression values to this module as previously described and validated. Then, patients were categorized as B-cell poor or B-cell rich using a predefined cut- off -0.0413, the median B-cell module values of RA patients recruited in R4RA clinical trial. Building classifier models for prediction of response to rituximab, tocilizumab and etanercept Machine learning models were built to predict target DAS28ESR, target DAS28CRP, ACR20 response, CDAI 50% response, EULAR DAS28-ESR good v moderate / non-response and EULAR DAS28-CRP good vs moderate / non-response to either rituximab, tocilizumab, or etanercept treatment at the primary endpoint (16 weeks). The model feature space was created using RNA-Seq data restricted to 506 genes relevant to synovial biology based on nCounter custom panel (see description of the panel below). For the rituximab models, baseline clinical parameters were included to improve response prediction. Clinical features included: baseline tender joint count (TJC), swollen joint count (SJC), the square root of TJC and SJC, CDAI, Arthritis Activity, Physicians Global Assessment, HAQ Score, rheumatoid factor (RF), erythrocyte sedimentation rate (ESR), C- reactive protein (CRP), and disease activity score based on ESR and CRP (DAS28ESRand DAS28CRP). Genes were filtered to those with mean expression ≥ 6 on VST scale to remove low expressed genes which were less reliably detected by nCounter. Following processing, data was split into 10 × 10 nested CV folds using the nestedcv R package (version 0.7.3). Feature selection was performed within outer CV folds using a t-test filter, with the top n genes by two-tailed t-test p-value being retained for fitting of models. An alternative feature selection method which was used for rituximab response prediction models, was based on fitting a glmnet elastic net regression model to the whole data and selecting genes which were retained in the glmnet model. The number of features selected was chosen to limit model size to between 25 to 40 predictors to design practical models more likely to be feasible in real life clinical situation. Model hyperparameters were tuned by inner 10-fold cross- validation based on log loss. Overall model performance was determined by 10-fold outer cross-validation with 25 repeats to give averaged unbiased estimates of model accuracy. Elastic net penalised regression using the glmnet package was compared against seven machine learning models from the caret package (version 6.0): random forest (RF), least- squares support vector machine (SVM) with radial basis function kernel (svmRadial), least- squares SVM with polynomial kernel (svmPoly), gradient boosted machine (GBM), mixture discriminant analysis (MDA), extreme gradient boosting (xgboost) using trees (xgbTree) or linear regression (xgbLinear). None of these models, with the exception of glmnet, are sparse, which means that they incorporate all predictors during modelling. This leads to models with large numbers of predictors which have a tendency to fail validation in real clinical situations. Hence feature selection was used to limit model size to between 25 to 40 predictors to design practical models more likely to be feasible in a real-life clinical situation. Previous studies have shown that with gene expression data filtering genes with a simple t-test often performs better than more complex feature selection methods. Thus for models predicting etanercept or tocilizumab response we used a t-test filter. However it was more difficult to obtain a good quality prediction model for rituximab, so a glmnet-based filter was used. In this two-step process a LASSO regression model is fitted to the training folds to select optimal features, which are then passed to other models for fitting. To evaluate overall model performance from each repeat of 10 × 10-fold nested CV, predictions from left-out outer CV test folds were pooled and performance determined compared to ground truth. Multiple metrics were used including area under curve (AUC) from receiver operating characteristic (ROC) curves computed using R package pROC, accuracy and balanced accuracy. Tuning parameters for the final model were determined by a final round of CV on the whole data with the final model fitted to the whole dataset. Performance of each model for predicting response to each drug was measured by AUC (area under ROC curve) with 25 repeats of 10×10 nested CV to compare the primary endpoint ACR20 response against five other response endpoints to identify the optimal endpoint (Figure 7): DAS28-ESR / CRP < 3.2 (target DAS28-ESR / CRP); CDAI 50% response, EULAR good vs moderate / non-response for DAS28-ESR / CRP. This showed that for etanercept and tocilizumab the best prediction of response was seen for target DAS28-ESR (<3.2). So this response outcome measure was selected for model optimisation. For etanercept and tocilizumab, models were trained with a binary outcome with response defined as DAS28- ESR < 3.2 at 16 weeks. For the rituximab model, due to difficulties with obtaining a reliable binary classification model, an ordinal outcome was used, namely DAS28-ESR status at 16 weeks which has 4 levels: high (DAS28-ESR > 5.1), moderate (3.2 < DAS28 < 5.1) or low disease activity (2.6 < DAS28 < 3.2) and remission (DAS28 < 2.6). This four-level outcome fits alongside the original binary outcome as low disease activity / remission corresponds directly to response, and moderate / high disease activity at 16 weeks corresponds to non- response. Rituximab prediction models were fitted to this ordinal outcome as a regression, then converted to a binary outcome after the final model was fitted and performance calculated for the binary outcome. Feature importance was measured across outer CV folds as well as the final model to rank predictors in terms of importance and estimate stability of variables in models by the frequency with which variables were selected across outer CV folds and estimate the variance across variable importance across the outer CV folds and final model. nCounter analysis A custom synovium nanoString nCounter panel was developed containing 506 genes relevant to synovial pathobiology based on previous studies of synovial gene expression (see Tables 5 and 6). The panel also contains 17 housekeeping genes selected for detectable and stable expression in synovial tissue and technical probes (6 positive, 8 negative). RNA samples were assayed using the nanoString nCounter Sprint Profiler from 100ng of synovial tissue RNA, following the manufacturer’s instructions. Raw nCounter counts were extracted from RCC files and normalized using nanoString’s official R package (nanoStringNCTools version 1.6.0). Housekeeping genes and panel standard probes (synthetic oligos) normalization methods were applied to the raw data to scale and standardize the data. nCounter pseudo-RNA-Seq conversion and prediction using machine learning models Linear models for each gene were fitted between variance stabilised transformed bulk RNA- Seq gene counts and log transformed normalised nCounter gene counts. nCounter assay data was converted to RNA-Seq scale (referred to as pseudo-RNA-Seq) based on stored regression models for each gene. Using predefined linear regression components (intercept and slope coefficient), nCounter data was converted to pseudo-RNA-Seq and then passed as fresh data input into the finalised machine learning fitted nestedcv package model objects (glmnet model for etanercept; gbm model for tocilizumab; xgbLinear model for rituximab) to determine a predicted probability of response for each sample. Predicted response probability was compared with actual outcome in the STRAP trial to determine confusion matrices of predicted binary response vs actual response, accuracy, balanced accuracy and ROC AUC for the nCounter assay. Data availability The datasets generated during and / or analyzed during the current study are available on an interactive web interface that allows direct data exploration (https: / / strap.hpc.qmul.ac.uk / ). A searchable interface is available to examine relationships between individual synovial gene transcript levels and histological and clinical parameters, and clinical response at 16 weeks. The website was constructed using R shiny server 1.5.16, with interactive plots generated with R plotly 4.9.3. RNA-Seq data is available at ArrayExpress accession ID 18203. The datasets can be downloaded from https: / / www.ebi.ac.uk / arrayexpress / experiments / E-MTAB-18203.

[0002] Table 4. Baseline characteristics of patients with available RNA-seq. Data are n (%), median [IQR], mean (SD). CDAI=Clinical disease activity index. DAS28=28 joint count disease activity score. CRP=C-reactive protein. ESR=erythrocyte sedimentation rate. P values were obtained using a two-sided Chi-Square test for categorical and binary variables, whereas a two-sided Student’s t test or Mann-Whitney U test (based on normality, tested through Shapiro-Wilk test) was used for continuous variables. No adjustment for multiplicity was applied.

[0003] Table 5. Gene Symbol Type Gene Symbol TypeFGF16 Endogenous PWWP2B EndogenousIGIP Endogenous CAT EndogenousICAM1 Endogenous CSF1R EndogenousIL20RA Endogenous CCL20 EndogenousLIFR Endogenous CFLAR EndogenousOLR1 Endogenous GRB2 EndogenousNFAM1 Endogenous TNFRSF10B EndogenousIL27A Endogenous ZBP1 EndogenousIFI44 Endogenous CD19 EndogenousGK Endogenous FAM8A1 EndogenousGPR84 Endogenous COL4A6 EndogenousFBLN1 Endogenous SYK EndogenousPLA2G2D Endogenous CXorf56 EndogenousCD86 Endogenous CASP3 EndogenousC1QA Endogenous CASP8 EndogenousC19orf12 Endogenous FKBP11 EndogenousMX1 Endogenous IL7 EndogenousIL7R Endogenous IL18R1 EndogenousDNM1L Endogenous EIF2B1 EndogenousIL17RA Endogenous CREB3L1 EndogenousVEGFC Endogenous CD8A EndogenousITGAM Endogenous LILRB2 Endogenous IL6ST Endogenous SLC25A35 Endogenous TLR8 Endogenous CCL2 Endogenous BAK1 Endogenous OSMR EndogenousIGF1 Endogenous FOXP3 EndogenousSTAT2 Endogenous CSF1 EndogenousCCND3 Endogenous APLF EndogenousBLK Endogenous IL27RA EndogenousPIK3CA Endogenous FGFBP1 EndogenousITCH Endogenous EPSTI1 EndogenousRIPK1 Endogenous IFNGR2 EndogenousLIF Endogenous FCGR1A EndogenousIL21R Endogenous MX2 EndogenousCD80 Endogenous SOCS2 EndogenousPARP9 Endogenous ZNF773 EndogenousAKT1 Endogenous IL17F EndogenousCDKN1B Endogenous FRY EndogenousTHY1 Endogenous CLEC17A EndogenousMZB1 Endogenous RAD52 EndogenousSTAT1 Endogenous GSK3B EndogenousGAB1 Endogenous TOX2 EndogenousBIRC2 Endogenous PAX5 EndogenousIL11 Endogenous IFNA2 EndogenousABHD2 Endogenous SOX9 EndogenousMT2A Endogenous MMP14 EndogenousLTA Endogenous MSH6 EndogenousCYBB Endogenous PARVG EndogenousXBP1 Endogenous SOCS4 EndogenousIL22 Endogenous STAM EndogenousRALGPS2 Endogenous CEACAM6 EndogenousPNOC Endogenous FBXL3 EndogenousIFNG Endogenous IL10RA EndogenousLEPR Endogenous MS4A1 EndogenousIL12A Endogenous GTF2H2 EndogenousINPP5D Endogenous TNFRSF17 EndogenousSHC3 Endogenous LEP EndogenousPIAS2 Endogenous SPRY2 EndogenousAKT3 Endogenous ANGPT1 EndogenousLCP2 Endogenous IL17D EndogenousZNF823 Endogenous ITGAX EndogenousIL22RA1 Endogenous CD79A EndogenousEFHD2 Endogenous ASPN EndogenousTNFAIP3 Endogenous SLC2A13 EndogenousCU633906.1 / CU633904.1 Endogenous AL161785.1 EndogenousYPEL5 Endogenous CD84 EndogenousGZMK Endogenous IFNA7 / 17 EndogenousIL6 Endogenous COL9A3 EndogenousSPP1 Endogenous ZNF765 EndogenousFCER1G Endogenous PIK3CB EndogenousCYBRD1 Endogenous CFL1 EndogenousCSF2RB Endogenous ITGA1 EndogenousMAP3K5 Endogenous GZMA EndogenousSIRPB2 Endogenous CD40LG Endogenous IFNA8 Endogenous IL6R EndogenousISG15 Endogenous WASF3 EndogenousIL2RB Endogenous MAP3K14 EndogenousIL17RC Endogenous P2RY13 EndogenousIL13RA2 Endogenous HMGB1 EndogenousGLMN Endogenous S100B EndogenousDAB2IP Endogenous CD53 EndogenousSLC2A3 Endogenous PIK3R2 EndogenousBCL2 Endogenous BIRC3 EndogenousMICU1 Endogenous FCRL5 EndogenousCEP85 Endogenous CSF2RA EndogenousIL5RA Endogenous PSMC3 EndogenousCASP6 Endogenous SLAMF7 EndogenousPTGS2 Endogenous TRAF3 EndogenousSLAMF6 Endogenous CD3G EndogenousJAK3 Endogenous MAVS EndogenousHLA-DOB Endogenous LGALS3BP EndogenousMAPK13 Endogenous STK40 EndogenousAOX1 Endogenous IL17RE EndogenousATF6B Endogenous ITGB8 EndogenousCREB5 Endogenous IQCG EndogenousSH2B1 Endogenous GNLY EndogenousFGF1 Endogenous CNTF EndogenousTRAF1 Endogenous MRC1 EndogenousFGFBP2 Endogenous CD79B EndogenousIL17RB Endogenous SDC1 EndogenousOSM Endogenous LAMC2 EndogenousCCDC102B Endogenous TLR10 EndogenousSTX1A Endogenous LY6E EndogenousATF4 Endogenous IL1B EndogenousFGFRL1 Endogenous HLA-DRB1 EndogenousMED13 Endogenous TRADD EndogenousSYT1 Endogenous HTR2A EndogenousCXCL13 Endogenous IL13RA1 EndogenousPRELP Endogenous EPO EndogenousCREB3L2 Endogenous CD247 EndogenousPDHA1 Endogenous CD83 EndogenousTRAF2 Endogenous BID EndogenousSOCS1 Endogenous JAG1 EndogenousCTSL Endogenous TLR4 EndogenousIL10RB Endogenous SHC1 EndogenousMMP9 Endogenous IL2RA EndogenousCD69 Endogenous STAT3 EndogenousBCL2L1 Endogenous PTPN11 EndogenousPLCG1 Endogenous TNFRSF1B EndogenousVPREB3 Endogenous IRS1 EndogenousCLEC5A Endogenous CAV2 EndogenousPIAS1 Endogenous IFNL2 EndogenousC1QB Endogenous HSPA1A EndogenousVAV1 Endogenous LYN EndogenousIL12RB1 Endogenous CD180 EndogenousCREB3 Endogenous PPP1R1A Endogenous CX3CL1 Endogenous MTOR EndogenousIL19 Endogenous NR4A2 EndogenousCSF3R Endogenous NLRP3 EndogenousCOL4A5 Endogenous GFAP EndogenousSOCS7 Endogenous CENPH EndogenousIFITM3 Endogenous MMP3 EndogenousIL22RA2 Endogenous C5AR1 EndogenousGZMB Endogenous GPNMB EndogenousIFNGR1 Endogenous TNFRSF1A EndogenousSOS1 Endogenous HLA-C EndogenousHLA-B Endogenous CEACAM3 EndogenousSTAP1 Endogenous PIK3R3 EndogenousKDM6B Endogenous FGFR1 EndogenousCD2 Endogenous POU2AF1 EndogenousKLF2 Endogenous JUND EndogenousCD52 Endogenous DEF6 EndogenousTRAF5 Endogenous LILRA2 EndogenousEDN1 Endogenous CEBPB EndogenousAGK Endogenous HIVEP3 EndogenousRNF19A Endogenous PABPC1 EndogenousHCK Endogenous SEP10 EndogenousEIF4E3 Endogenous CD27 EndogenousIL3RA Endogenous IL3 EndogenousHERC5 Endogenous CCND2 EndogenousICOS Endogenous CXCL14 EndogenousMAPK12 Endogenous SOCS6 EndogenousCCND1 Endogenous NCAM1 EndogenousEFNA5 Endogenous CSF3 EndogenousFGF2 Endogenous SLC37A3 EndogenousTNF Endogenous FPR3 EndogenousATF3 Endogenous PRKCB EndogenousITGB2 Endogenous FOSL1 EndogenousINPPL1 Endogenous ZEB2 EndogenousLINC00494 Endogenous CBL EndogenousLMNB2 Endogenous TMEM51 EndogenousSIRPA Endogenous TGFB1 EndogenousSELE Endogenous DIRAS3 EndogenousCREB3L4 Endogenous IL20 EndogenousIL21 Endogenous CSF2 EndogenousP2RX5 Endogenous CD6 EndogenousIL12B Endogenous FADD EndogenousCASP7 Endogenous IL5 EndogenousAPOOL Endogenous FAM30A EndogenousTRDC Endogenous CXCR4 EndogenousMAMDC2 Endogenous CLEC7A EndogenousFOSB Endogenous PLS3 EndogenousIL15RA Endogenous TLR1 EndogenousGZMH Endogenous HLA-DPB1 EndogenousCD22 Endogenous FCRLA EndogenousSP140 Endogenous PANK2 EndogenousFAS Endogenous SEL1L3 EndogenousSPIB Endogenous DENND2D Endogenous SSR4 Endogenous SERPINA1 EndogenousTLR2 Endogenous SOCS5 EndogenousZNF331 Endogenous PLEC EndogenousQPRT Endogenous PDCD1 EndogenousRAF1 Endogenous ASNS EndogenousCISH Endogenous PIM1 EndogenousPPARG Endogenous IL20RB EndogenousSTAT6 Endogenous CD38 EndogenousHRAS Endogenous IFNL1 EndogenousTCIM Endogenous IL17A EndogenousPLSCR4 Endogenous PTPRC EndogenousIL4 Endogenous ITGB4 EndogenousLAMA3 Endogenous HLA-DMA EndogenousRPS6KA4 Endogenous PGAM5 EndogenousPIAS4 Endogenous VSIG4 EndogenousFOXO3 Endogenous IL10 EndogenousMERTK Endogenous CD74 EndogenousPTPN2 Endogenous IKBKE EndogenousIL9 Endogenous HLA-DMB EndogenousPDGFRL Endogenous TRPC6 EndogenousFCRL3 Endogenous MLKL EndogenousPILRA Endogenous TMSB10 EndogenousIFI44L Endogenous IFNB1 EndogenousTNFRSF13B Endogenous P2RX4 EndogenousIL17C Endogenous IFNAR1 EndogenousJUNB Endogenous PIM2 EndogenousTSLP Endogenous PPFIA4 EndogenousRSAD2 Endogenous CTF1 EndogenousPCBP3 Endogenous IFNA1 EndogenousCTLA4 Endogenous CENPBD1 EndogenousCASP10 Endogenous BTK EndogenousJCHAIN Endogenous ERBB3 EndogenousATXN1L Endogenous FHL1 EndogenousBCL3 Endogenous CD3E EndogenousLY9 Endogenous IL11RA EndogenousPIAS3 Endogenous TMEM156 EndogenousAKT2 Endogenous IKZF2 EndogenousSFRP1 Endogenous STAT5B EndogenousIFNLR1 Endogenous TRIT1 EndogenousMSH3 Endogenous HLA-DPA1 EndogenousFCRL2 Endogenous PTPN1 EndogenousRALGPS1 Endogenous ZNF75A EndogenousEPOR Endogenous VCAM1 EndogenousNKG7 Endogenous SASH3 EndogenousPRKCQ Endogenous STAM2 EndogenousCD163 Endogenous CNR2 EndogenousLTB Endogenous IGLL5 EndogenousIL15 Endogenous PURA EndogenousBTLA Endogenous TIGIT EndogenousIGHV3.53 Endogenous CD14 EndogenousMMP1 Endogenous IRF4 EndogenousFGF7 Endogenous DKK3 Endogenous FAP Endogenous IL23A EndogenousCCL19 Endogenous IL23R EndogenousTNFRSF13C Endogenous FGF9 EndogenousCTSS Endogenous OLFML2B EndogenousMCL1 Endogenous ISY1 EndogenousRIPK3 Endogenous STK11IP EndogenousATAD3C Endogenous TFRC EndogenousCST7 EndogenousIFNAR2 EndogenousMYD88 EndogenousFCN1 EndogenousHSP90B1 EndogenousSOS2 EndogenousPIK3CD EndogenousIL24 EndogenousKCNA3 EndogenousSOCS3 EndogenousIRF9 EndogenousSLC11A1 EndogenousCD68 EndogenousCD3D EndogenousIL2RG EndogenousPLCG2 EndogenousITGA4 EndogenousIL12RB2 EndogenousIL27R EndogenousFGFR2 EndogenousCNTFR EndogenousZNF343 EndogenousPRLR EndogenousPTPN6 EndogenousLRIG3 EndogenousBAG4 EndogenousZNF850 EndogenousPTK7 EndogenousAFAP1L2 EndogenousCD40 EndogenousPIK3R1 EndogenousLTBR EndogenousTable 6. Gene Symbol Type TRIM56 Housekeeping SF3A1 Housekeeping TBC1D10B Housekeeping WDR55 Housekeeping MRPL19 Housekeeping GUSB Housekeeping TLK2 Housekeeping ERCC3 Housekeeping DNAJC14 Housekeeping PUM1 Housekeeping ABCF1 Housekeeping G6PD Housekeeping PSMC4 Housekeeping UBXN4 Housekeeping SDHA Housekeeping TBP Housekeeping TMUB2 Housekeeping NEG-1 Negative NEG-2 Negative NEG-3 Negative NEG-4 Negative NEG-5 Negative NEG-6 Negative NEG-7 Negative NEG-8 Negative POS-1 Positive POS-2 Positive POS-3 Positive POS-4 Positive POS-5 Positive POS-6 Positive Table 7. Coefficients for Etanercept glmnet model. Coef (Intercept) -41.37013389 EIF2B1 3.464604902 CENPBD1 -1.945547771 PWWP2B -1.567093775 RAD52 1.286691473 AGK 1.248986069 ZNF765 0.945241068 ZNF850 0.790651987 HTR2A -0.503999904 IFI44L 0.500062906 SYT1 -0.395486675 CCDC102B 0.359013787 CSF3R -0.283218854 IL13RA2 0.254948826 TRPC6 0.2172973 AL161785.1 -0.127164512

[0004] Table 8. Variable importance for Etanercept model. mean sd sem frequency sign final EIF2B1 21.49049819 4.420390825 1.332797981 11 1 yes CENPBD1 10.09703098 3.667353867 1.105748796 11 -1 yes PWWP2B 9.688896586 4.096939375 1.2352737 10 -1 yes AGK 9.306753772 5.382396695 1.622853665 11 1 yes RAD52 6.804152033 1.975206611 0.595547201 11 1 yes ZNF765 6.555559757 5.39640061 1.627076004 11 1 yes ZNF850 5.18286493 1.941619622 0.585420343 11 1 yes IL5 3.235061393 4.511710442 1.360331882 4 1 no CXorf56 2.716795367 3.29740129 0.994203897 6 1 no AL161785.1 2.209440454 1.503288269 0.453258467 10 -1 yes TRPC6 1.963767244 1.267495008 0.382164124 10 1 yes CCDC102B 1.959968489 2.242269608 0.676069725 8 1 yes SYT1 1.828230415 1.784539843 0.538059008 6 -1 yes IL13RA2 1.536470076 1.585038287 0.477907025 8 1 yes BIRC2 1.531458612 3.449514569 1.040067776 2 1 no HTR2A 1.531286696 2.053604039 0.619184915 5 -1 yes CASP7 1.487579609 3.326737454 1.003049083 2 1 no IFI44L 1.164951992 1.320523285 0.398152751 6 1 yes CSF3R 1.134063148 1.36253978 0.410821201 5 -1 yes MICU1 1.110209265 3.682147572 1.110209265 1 1 no MRPL19 0.96633071 3.20495639 0.96633071 1 1 no TMUB2 0.867107147 2.875869061 0.867107147 1 1 no APLF 0.849085251 1.940728452 0.585151645 2 1 no LTBR 0.771758674 2.559633951 0.771758674 1 -1 no PIK3CA 0.692696431 2.297414156 0.692696431 1 1 no CD80 0.540230145 1.791740693 0.540230145 1 1 no IL17A 0.486503175 1.613548492 0.486503175 1 -1 no LILRA2 0.43604496 0.955608606 0.288126836 3 1 no MX1 0.378661684 1.255878729 0.378661684 1 1 no SOS1 0.342276377 1.135202318 0.342276377 1 1 no GPNMB 0.29618934 0.982348907 0.29618934 1 1 no CREB3L2 0.22453413 0.503175206 0.151713033 2 -1 no THY1 0.138816867 0.460403461 0.138816867 1 1 no GUSB 0.121332049 0.402412881 0.121332049 1 1 no RNF19A 0.105705221 0.350584555 0.105705221 1 1 no IL17RA 0.08726771 0.289434251 0.08726771 1 1 no TNFAIP3 0.07775861 0.257896134 0.07775861 1 1 no FGF16 0.03638788 0.113716397 0.034286784 2 1 no LRIG3 0.025626371 0.084993059 0.025626371 1 1 no HSP90B1 0.020648259 0.068482528 0.020648259 1 -1 no Table 9. Variable importance for Tocilizumab model. mean sd sem frequency sign final GPNMB 63.53177963 29.1262733 8.781901826 11 1 yes MICU1 63.3538808 28.8335764 8.69365039 11 1 yes NKG7 49.44105535 41.73119253 12.58242797 10 -1 yes TOX2 46.84732569 24.77391334 7.46961592 11 -1 yes GZMH 42.99926466 39.27216251 11.84100252 10 -1 yes CD40LG 42.80790412 27.53624835 8.302491264 11 -1 yes SOCS6 40.13211682 24.09617042 7.265268743 11 1 yes CD247 37.49282988 20.97256541 6.323466397 11 -1 yes CDKN1B 36.89018516 18.82496396 5.675940195 10 -1 yes KLF2 28.75815392 34.29979686 10.34177787 6 -1 yes TCIM 24.38487777 22.69637453 6.843214401 9 -1 yes MERTK 23.9979812 30.37587734 9.158671619 8 1 yes JAG1 21.16074371 30.10041857 9.075617677 4 -1 no PTGS2 12.18802702 17.75746815 5.354078099 6 -1 yes BLK 10.21614975 10.16216035 3.064006632 9 -1 yes CSF2RB 9.975456407 22.20389761 6.694727023 2 -1 no SP140 9.522577732 11.59379353 3.495660276 10 -1 yes SDC1 8.66692978 15.89860534 4.793609873 3 -1 no IL2RB 8.207629378 12.43353384 3.748851507 6 -1 yes SELE 7.605432702 21.57426532 6.504885745 2 -1 no CST7 7.429498917 6.43110884 1.939052273 7 -1 yes MX1 7.261997625 16.98559976 5.121351023 2 -1 no CCDC102B 6.773657653 22.46568089 6.773657653 1 -1 no AL161785.1 6.57802253 14.65461631 4.418533068 2 1 no CD79B 6.368177473 8.521272146 2.569260222 5 -1 yes YPEL5 6.292642648 11.55031639 3.482551424 3 -1 yes TRPC6 5.769555435 7.951658072 2.397515117 5 -1 yes FCRL3 5.380348889 6.572557196 1.981700557 9 -1 yes P2RX4 5.023734308 16.66184175 5.023734308 1 1 no ITGA1 4.908597429 11.50481968 3.46883365 2 -1 no CD3G 4.799836399 4.831950237 1.456887813 10 -1 yes IL24 4.54880445 11.85067191 3.573112023 4 -1 yes IL4 4.3672383 14.48449081 4.3672383 1 -1 no IL23A 4.220574029 5.665791378 1.708300377 6 -1 yes KCNA3 3.951175297 3.217626815 0.970150987 9 -1 yes PIK3R2 3.865031242 8.616534908 2.597983026 2 1 no TNFRSF10B 3.860590581 10.56892545 3.186650922 2 -1 no CD2 3.739423698 6.430874863 1.938981727 4 -1 yes CCL20 3.550921412 11.77707398 3.550921412 1 1 no CASP10 3.473802395 4.946665598 1.491475796 5 -1 yes ASNS 3.295596831 5.923480393 1.785996538 3 -1 no EPO 3.204643491 10.62860005 3.204643491 1 1 no ICOS 3.048443111 5.601163578 1.688814362 5 -1 yes ITGA4 3.01842148 10.01097151 3.01842148 1 -1 no P2RY13 2.961992849 9.823818913 2.961992849 1 -1 no CX3CL1 2.891563157 6.468615571 1.950360978 2 -1 no PPFIA4 2.8452012 9.436464832 2.8452012 1 1 no RALGPS2 2.840189757 6.099202834 1.838978847 5 -1 no LTB 2.553666646 3.940653993 1.188151884 6 -1 yes BIRC2 2.449184875 8.123027272 2.449184875 1 1 no LTA 2.351915475 4.330347064 1.305648766 3 -1 yes CEBPB 2.03125657 6.736915895 2.03125657 1 1 no CXCL14 2.021870775 6.705786734 2.021870775 1 -1 no GZMK 2.016406859 3.501661411 1.05579064 4 -1 no IL2RG 1.964490399 6.515477559 1.964490399 1 -1 no GNLY 1.942521075 3.304871073 0.996456121 4 -1 no CD6 1.938367702 2.548733226 0.768471982 7 -1 yes PABPC1 1.937904971 6.427303669 1.937904971 1 -1 no CD8A 1.713094355 3.811620147 1.149246715 2 -1 yes CD3D 1.698611776 3.24751989 0.979164088 4 -1 yes PIK3R3 1.610622764 5.341831388 1.610622764 1 -1 no TRAF2 1.547612646 5.132850469 1.547612646 1 -1 no PIAS2 1.533202824 4.057540229 1.22339441 2 1 no DAB2IP 1.513599995 5.020043267 1.513599995 1 -1 no ZBP1 1.387469579 2.594387056 0.782237129 3 -1 no GZMB 1.21116 2.608101561 0.786372209 3 -1 no HTR2A 1.19640118 3.968013811 1.19640118 1 1 no CCND3 1.091017365 2.497539496 0.753036492 2 -1 no IRF4 1.051073915 2.430988638 0.732970653 2 -1 no CD3E 1.042905803 1.499995713 0.452265724 7 -1 yes TRDC 0.972756749 3.226269148 0.972756749 1 -1 no GFAP 0.960597072 3.185940063 0.960597072 1 -1 no PLCG1 0.925261857 3.068746413 0.925261857 1 -1 no RIPK1 0.876590095 2.907320439 0.876590095 1 -1 no ISG15 0.856545317 2.840839433 0.856545317 1 -1 no STAM 0.817290094 2.710644588 0.817290094 1 1 no LY9 0.755701007 1.572518734 0.474132238 4 -1 yes APLF 0.697280576 2.312618043 0.697280576 1 1 no BTLA 0.679247931 2.252810527 0.679247931 1 -1 no IL7R 0.644921995 2.138964277 0.644921995 1 -1 no FOSL1 0.556760696 1.846566326 0.556760696 1 1 no PLCG2 0.45933625 1.523445993 0.45933625 1 -1 no SLAMF6 0.430732454 1.428577935 0.430732454 1 -1 no IRF9 0.384760855 1.276107391 0.384760855 1 -1 no ATF3 0.289995473 0.961806174 0.289995473 1 -1 no PNOC 0.241923768 0.802370368 0.241923768 1 -1 no TNFRSF13C 0.238088002 0.789648571 0.238088002 1 -1 no MS4A1 0.189142377 0.627314298 0.189142377 1 -1 no GZMA 0.180187042 0.330204677 0.099560456 3 -1 no ITGB4 0.020410282 0.067693246 0.020410282 1 -1 no Table 10. Variable importance for Rituximab model. mean sd sem frequency sign final sqTJ 100 0 0 11 -1 yes sqESR 62.69807881 14.06333273 4.240254362 11 -1 yes MICU1 7.417789574 10.26487953 3.094977628 11 -1 yes Arthritis.Activity 2.585246839 0.950492308 0.286584214 11 -1 yes CISH 18.85770173 17.26439538 5.205411064 10 1 yes PPFIA4 16.59061249 11.09746953 3.346012959 10 -1 yes MAPK12 3.599183421 4.967317133 1.497702468 10 1 yes IL17C 2.000493027 3.720584867 1.121798546 10 -1 yes IL5 0.370164351 0.572909506 0.172738716 10 -1 yes ZNF773 12.00793712 11.86712433 3.578072613 9 1 yes MMP9 2.960098191 4.733879311 1.427318316 8 -1 no IL22 0.93185227 2.779109738 0.837933114 8 1 yes CENPH 0.473097069 0.729993904 0.220101443 8 1 yes ZNF823 8.325911986 11.2348039 3.387420831 7 -1 yes CREB3 4.699418027 7.096338683 2.139626618 7 -1 yes ZNF331 1.201629774 1.520153972 0.458343668 7 1 yes SHC1 1.052845443 1.492020036 0.449860967 6 -1 yes ZNF75A 0.523667321 0.948056006 0.285849641 6 -1 yes IL20 0.017101792 0.032316251 0.009743716 6 -1 yes ATF3 5.813823513 10.18483722 3.070843966 5 1 yes CFLAR 4.316961043 8.012318555 2.415804941 5 -1 yes STAT6 3.728558816 6.852789489 2.066193773 5 -1 yes BIRC2 3.146647305 5.391016934 1.625452764 5 -1 no P2RY13 2.917097682 4.157206036 1.253444781 5 -1 yes BCL2 6.12218385 11.4970667 3.466496038 4 1 yes IFNL1 5.6326392 10.44862752 3.150379733 4 -1 yes ERBB3 2.755553869 5.090552367 1.534859289 4 -1 yes PCBP3 0.833071015 1.713369844 0.516600445 4 -1 no COL4A6 0.587890828 1.671013585 0.503829553 4 -1 yes CDKN1B 8.899977138 15.71651741 4.738708296 3 1 no IL12B 5.470503976 12.18085871 3.672667088 3 -1 yes TRDC 1.693782122 4.694567111 1.415465242 3 1 no ASNS 1.553353203 4.828132062 1.45573659 3 -1 no MSH6 1.470625829 4.212343574 1.270069375 3 1 no IFNB1 0.535768065 1.209385844 0.364643552 3 1 yes IL17F 0.001047727 0.001885294 0.000568437 3 1 no FGFRL1 3.029312646 10.03164936 3.024656088 2 -1 no ISY1 2.255544104 5.142767307 1.550602686 2 -1 no IL23A 1.684244275 4.72810254 1.425576554 2 -1 no SOX9 1.203961737 3.049132097 0.919347918 2 -1 no FADD 1.078510177 3.203560657 0.965909881 2 -1 no TNF 1.029533772 3.12902743 0.943437268 2 1 no EPO 0.838420751 2.751104394 0.829489185 2 -1 no IL12A 0.453081114 1.441521358 0.434635043 2 -1 no CNTF 0.44488034 1.474405172 0.444549886 2 -1 no CENPBD1 0.315828565 0.868730335 0.261932052 2 -1 no SLC2A3 0.280972302 0.928397653 0.279922425 2 -1 no PSMC3 0.146471765 0.450012905 0.135683996 2 -1 no HLA-B 0.065532893 0.205594992 0.061989222 2 1 no STAM 1.418508589 4.70466075 1.418508589 1 -1 no PLCG1 1.391553887 4.615262119 1.391553887 1 -1 no QPRT 1.048577488 3.477738092 1.048577488 1 1 no HIVEP3 0.78062402 2.589036976 0.78062402 1 1 no CEACAM3 0.672066484 2.228992361 0.672066484 1 -1 no FGF7 0.646390649 2.143835251 0.646390649 1 -1 no PIK3CB 0.59955886 1.98851178 0.59955886 1 1 no LAMA3 0.581728477 1.929375087 0.581728477 1 1 no IKBKE 0.523919432 1.737644176 0.523919432 1 -1 no CASP3 0.472853573 1.568277881 0.472853573 1 1 no SOCS7 0.347947228 1.154010401 0.347947228 1 -1 no SOCS5 0.344748662 1.14340196 0.344748662 1 -1 no CD180 0.281043073 0.932114422 0.281043073 1 1 no IL4 0.20076048 0.665847184 0.20076048 1 -1 no APLF 0.160571922 0.532556817 0.160571922 1 1 no ATF6B 0.156328609 0.518483341 0.156328609 1 1 no CCND2 0.151162004 0.501347649 0.151162004 1 1 no IQCG 0.147110857 0.487911515 0.147110857 1 1 no AL161785.1 0.115015793 0.381464232 0.115015793 1 1 no IL5RA 0.065681527 0.217840981 0.065681527 1 1 no IL19 0.045952815 0.152408244 0.045952815 1 -1 no TRIT1 0.027836201 0.092322233 0.027836201 1 1 no MSH3 0.026568441 0.088117549 0.026568441 1 1 no PDGFRL 0.022951807 0.076122532 0.022951807 1 1 no IL22RA2 0.018893001 0.062660995 0.018893001 1 -1 no LRIG3 0.017347581 0.057535417 0.017347581 1 -1 no IL17A 0.00998865 0.033128604 0.00998865 1 1 no PANK2 0.009489971 0.031474673 0.009489971 1 1 no C19orf12 0.006968404 0.02311158 0.006968404 1 1 no GTF2H2 0.006787139 0.022510394 0.006787139 1 -1 no MCL1 0.004064503 0.013480432 0.004064503 1 -1 no PIAS1 0.001415655 0.004695195 0.001415655 1 -1 no SOS2 0.000923092 0.00306155 0.000923092 1 1 no IFNA1 0.000804329 0.002667658 0.000804329 1 1 no 4.32241689456903e- 4.32241689456903e- IL13RA2 05 0.000143358 05 1 -1 no Various features and embodiments of the present invention will now be described with reference to the following numbered paragraphs (paras): 1. A method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a first, second and / or third marker panel in a sample from the patient, wherein the agent is selected from an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, and wherein: (i) the first marker panel comprises one or more markers selected from Table 1.1 or 1.2, wherein the agent is an agent that downregulates TNF mediated signalling; (ii) the second marker panel comprises one or more markers selected from Table 2.1 or 2.2, wherein the agent is an agent that downregulates IL-6 mediated signalling; and (iii) the third marker panel comprises one or more markers selected from Table 3.1 or 3.2, wherein the agent is a B cell targeted therapy. 2. A method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy, the method comprising determining the profile of a first, second and / or third marker panel in a sample from the patient, and wherein: (i) the first marker panel comprises one or more markers selected from Table 1.1 or 1.2, wherein the agent is an agent that downregulates TNF mediated signalling; (ii) the second marker panel comprises one or more markers selected from Table 2.1 or 2.2, wherein the agent is an agent that downregulates IL-6 mediated signalling; and (iii) the third marker panel comprises one or more markers selected from Table 3.1 or 3.2, wherein the agent is a B cell targeted therapy. 3. The method of para 2, wherein the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, the method comprising determining the profile of the first marker panel in a sample from the patient. 4. The method of para 2, wherein the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates IL-6 mediated signalling, the method comprising determining the profile of the second marker panel in a sample from the patient. 5. The method of para 2, wherein the method is for determining whether a RA patient is susceptible or refractory to treatment with a B cell targeted therapy, the method comprising determining the profile of the third marker panel in a sample from the patient. 6. The method of para 2, wherein the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling and an agent that downregulates IL-6 mediated signalling, the method comprising determining the profile of the first marker panel and the second marker panel in a sample from the patient. 7. The method of para 2, wherein the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, the method comprising determining the profile of the second marker panel and the third marker panel in a sample from the patient. 8. The method of para 2, wherein the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling and a B cell targeted therapy, the method comprising determining the profile of the first marker panel and the third marker panel in a sample from the patient. 9. The method of para 2, wherein the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, the method comprising determining the profile of the first, second and third marker panels in a sample from the patient. 10. The method of any one of paras 1-3, 6, 8 and 9, wherein the first marker panel comprises EIF2B1. 11. The method of any one of paras 1-3, 6, and 8-10, wherein the first marker panel comprises EIF2B1, CENPBD1, PWWP2B, and AGK. 12. The method of any one of paras 1-3, 6, and 8-11, wherein the first marker panel comprises EIF2B1, CENPBD1, PWWP2B, AGK, RAD52, ZNF765, and ZNF850. 13. The method of any one of paras 1-3, 6, and 8-12, wherein the first marker panel comprises all of the markers set forth in Table 1.1 or 1.2. 14. The method of any one of paras 1, 2, 4, 6, 7, and 9-13, wherein the second marker panel comprises GPNMB and MICU1. 15. The method of any one of paras 1, 2, 4, 6, 7, and 9-14, wherein the second marker panel comprises GPNMB, MICU1, NKG7, and TOX2. 16. The method of any one of paras 1, 2, 4, 6, 7, and 9-15, wherein the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, and SOCS6. 17. The method of any one of paras 1, 2, 4, 6, 7, and 9-16, wherein the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, and CDKN1B. 18. The method of any one of paras 1, 2, 4, 6, 7, and 9-17, wherein the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, KLF2, TCIM, and MERTK. 19. The method of any one of paras 1, 2, 4, 6, 7, and 9-18, wherein the second marker panel comprises all of the markers set forth in Table 2.1 or 2.2. 20. The method of any one of paras 1, 2, 5, and 7-19, wherein the third biomarker panel comprises CISH, and PPFIA4. 21. The method of any one of paras 1, 2, 5, and 7-20, wherein the third biomarker panel comprises CISH, PPFIA4, ZNF773, ATF3, MAPK12, and BCL2. 22. The method of any one of paras 1, 2, 5, and 7-21, wherein the third biomarker panel comprises all of the biomarkers set forth in Table 3.1 or 3.2. 23. The method of any one of the preceding paras, wherein determining the profile of a first, second and / or third marker panel in a sample from the patient comprises determining the level of the markers of the first, second and / or third marker panel in the sample from the subject and comparing the level of the markers to the corresponding reference value. 24. The method of any one of the preceding paras, wherein: (a) (i) an increased level of one or more markers from Table 1.1.A or 1.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates TNF mediated signalling; (ii) an increased level of one or more markers from Table 1.1.B or 1.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates TNF mediated signalling; (b) (i) an increased level of one or more markers from Table 2.1.A or 2.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates IL-6 mediated signalling; (ii) an increased level of one or more markers from Table 2.1.B or 2.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates IL-6 mediated signalling; and / or (c) (i) an increased level of one or more markers from Table 3.1.A or 3.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with a B cell targeted therapy; (ii) an increased level of one or more markers from Table 3.1.B or 3.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with a B cell targeted therapy. 25. The method of any one of the preceding paras, wherein determining the profile of the first, second and / or third marker panels comprises determining the nucleic acid levels of the markers of the first, second and / or third marker panels. 26. The method of para 25, wherein the nucleic acid level is an mRNA level. 27. The method of any preceding para, wherein determining the profile of the first, second and / or third marker panels is performed by direct digital counting of nucleic acids, RNA- seq, RT-qPCR, qPCR, multiplex qPCR or RT-qPCR, microarray analysis, or a combination thereof. 28. The method of any preceding para, wherein the step of determining the profile of the first, second and / or third marker panels is performed by RNA sequencing. 29. The method of any preceding para, wherein determining the profile of the first, second and / or third marker panels comprises determining the level of gene expression of the markers of the first, second and / or third marker panels. 30. The method of any preceding para, wherein the one or more sample is a synovial sample. 31. The method of any preceding para, wherein the agent that downregulates TNF mediated signalling is a TNF inhibitor. 32. The method of any preceding para, wherein the agent that downregulates TNF mediated signalling is selected from etanercept, infliximab, adalimumab, certolizumab pegol, and golimumab. 33. The method of any preceding para, wherein the agent that downregulates TNF mediated signalling is etanercept. 34. The method of any preceding para, wherein the agent that downregulates IL-6 mediated signalling is an IL-6 receptor antagonist. 35. The method of any preceding para, wherein the agent that downregulates IL-6 mediated signalling is selected from the group consisting of tocilizumab, sarilumab, satralizumab and siltuximab. 36. The method of any preceding para, wherein the B cell targeted therapy is B cell depletion therapy. 37. The method of any preceding para, wherein the B cell targeted therapy is selected from the group consisting of: rituximab, ocrelizumab, veltuzumab, ofatumumab, epratuzumab, obinutuzumab, ibritumomab and tiuxetan. 38. The method of any preceding para, wherein the B cell targeted therapy is rituximab. 39. The method of any preceding para, wherein the patient is refractory to DMARD therapy. 40. The method of any preceding para, wherein the method further comprises: (a) administering to the patient an agent that downregulates TNF mediated signalling when the patient is determined to be susceptible to treatment with an agent that downregulates TNF mediated signalling; (b) administering to the patient an agent that downregulates IL-6 mediated signalling when the patient is determined to be susceptible to treatment with an agent that downregulates IL-6 mediated signalling; (c) administering to the patient a B cell targeted therapy when the patient is determined to be susceptible to treatment with a B cell targeted therapy; or (d) administering to the patient an alternative therapeutic when the patient is determined to be refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling, and a B cell targeted therapy. The method of any one of the preceding paras, wherein the method comprises using at least one predictive modelling approach to identify the agent or to determine whether the patient is susceptible or refractory to treatment. The method of para 41, wherein the at least one predictive modelling approach is gradient boosted machine (gbm), elastic net regression (glmnet), mixed discriminant analysis (mda), random forest (rf), support vector machine (svm) with polynomial (svmPoly) or radial (svmRadial) kernel, extreme gradient boosting (xgboost) with tree booster (xgbTree) or linear booster (xgbLinear). The method of para 41 or para 42, wherein populations of patients grouped according to clinical response are used as inputs to the predictive modelling approach. A method for treating Rheumatoid Arthritis (RA) in a patient, the method comprising: (a) administering to the patient an effective amount of an agent that downregulates TNF mediated signalling, wherein the agent is determined to be suitable for treatment or the patient is determined to be susceptible to treatment with an agent that downregulates TNF mediated signalling by the method of any one of paras 1-43; (b) administering to the patient an effective amount of an agent that downregulates IL-6 mediated signalling, wherein the agent is determined to be suitable for treatment or the patient is determined to be susceptible to treatment with an agent that downregulates IL-6 mediated signalling by the method of any one of paras 1-43; or (c) administering to the patient an effective amount of a B cell targeted therapy, wherein the agent is determined to be suitable for treatment or the patient is determined to be susceptible to treatment with a B cell targeted therapy by the method of any one of paras 1-43. 45. An agent that downregulates TNF mediated signalling for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the agent is determined to be suitable for treatment or the patient is determined to be susceptible to treatment with an agent that downregulates TNF mediated signalling by the method of any one of paras 1-43. 46. An agent that downregulates IL-6 mediated signalling for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the agent is determined to be suitable for treatment or the patient is determined to be susceptible to treatment with an agent that downregulates IL-6 mediated signalling by the method of any one of paras 1-43. 47. A B cell targeted therapy for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the B cell targeted therapy is determined to be suitable for treatment or the patient is determined to be susceptible to treatment with a B cell targeted therapy by the method of any one of paras 1-43. 48. A kit for use in the method of any preceding para. 49. The kit for use according to para 48, wherein the kit comprises one or more reagent suitable for detecting the markers of the first, second and / or third marker panels. 50. The kit for use according to para 48 or para 49, wherein the kit comprises reagents for RNA sequencing. 51. The kit of any one of paras 48-50, wherein the kit comprises one or more probe or antibody for detecting the markers of the first, second and / or third marker panels. 52. The kit of any one of paras 48-51 which is in the form of a microchip or microarray. 53. A computer-readable medium comprising instructions that when executed cause one or more processors to perform the method of any one of paras 1-43. 54. An apparatus comprising: (a) profile determination circuitry to determine the profile of a first, second and / or third panel of markers in a sample from a Rheumatoid Arthritis (RA) patient; and (b) (i) agent identification circuitry to identify, based on the profile determination circuitry, a suitability of an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy for treating RA in the patient; or (ii) patient identification circuitry to identify, based on the profile determination circuitry, a suitability of the patient for treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy. All publications mentioned in the above specification are herein incorporated by reference. Various modifications and variations of the disclosed methods, agents, uses, kits, computer- readable media, or apparatus of the invention will be apparent to the skilled person without departing from the scope and spirit of the invention. Although the invention has been disclosed in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Indeed, various modifications of the disclosed modes for carrying out the invention, which are obvious to the skilled person are intended to be within the scope of the following claims.

Claims

CLAIMS 1. A method for determining if an agent is suitable for treating a Rheumatoid Arthritis (RA) patient, the method comprising determining the profile of a first, second and / or third marker panel in a sample from the patient, wherein the agent is selected from an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, and wherein: (i) the first marker panel comprises one or more markers selected from Table 1.1 or 1.2, wherein the agent is an agent that downregulates TNF mediated signalling; (ii) the second marker panel comprises one or more markers selected from Table 2.1 or 2.2, wherein the agent is an agent that downregulates IL-6 mediated signalling; and (iii) the third marker panel comprises one or more markers selected from Table 3.1 or 3.2, wherein the agent is a B cell targeted therapy.

2. A method for determining whether a Rheumatoid Arthritis (RA) patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy, the method comprising determining the profile of a first, second and / or third marker panel in a sample from the patient, and wherein: (i) the first marker panel comprises one or more markers selected from Table 1.1 or 1.2, wherein the agent is an agent that downregulates TNF mediated signalling; (ii) the second marker panel comprises one or more markers selected from Table 2.1 or 2.2, wherein the agent is an agent that downregulates IL-6 mediated signalling; and (iii) the third marker panel comprises one or more markers selected from Table 3.1 or 3.2, wherein the agent is a B cell targeted therapy.

3. The method of claim 2, wherein: (a) the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, themethod comprising determining the profile of the first marker panel in a sample from the patient; (b) the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates IL-6 mediated signalling, the method comprising determining the profile of the second marker panel in a sample from the patient; (c) the method is for determining whether a RA patient is susceptible or refractory to treatment with a B cell targeted therapy, the method comprising determining the profile of the third marker panel in a sample from the patient; (d) the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling and an agent that downregulates IL-6 mediated signalling, the method comprising determining the profile of the first marker panel and the second marker panel in a sample from the patient; (e) the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, the method comprising determining the profile of the second marker panel and the third marker panel in a sample from the patient; (f) the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling and a B cell targeted therapy, the method comprising determining the profile of the first marker panel and the third marker panel in a sample from the patient; or (g) the method is for determining whether a RA patient is susceptible or refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and a B cell targeted therapy, the method comprising determining the profile of the first, second and third marker panels in a sample from the patient.

4. The method of any one of claims 1-3, wherein the first marker panel comprises EIF2B1, CENPBD1, PWWP2B, AGK, RAD52, ZNF765, and ZNF850.

5. The method of any one of claims 1-4, wherein the first marker panel comprises all of the markers set forth in Table 1.1 or 1.2.

6. The method of any one of claims 1-5, wherein the second marker panel comprises GPNMB, MICU1, NKG7, TOX2, GZMH, CD40LG, SOCS6, CD247, CDKN1B, KLF2, TCIM, and MERTK.

7. The method of any one of claims 1-6, wherein the second marker panel comprises all of the markers set forth in Table 2.1 or 2.

2.

8. The method of any one of claims 1-7, wherein the third biomarker panel comprises CISH, PPFIA4, ZNF773, ATF3, MAPK12, and BCL2.

9. The method of any one of claims 1-8, wherein the third biomarker panel comprises all of the biomarkers set forth in Table 3.1 or 3.

2.

10. The method of any one of the preceding claims, wherein determining the profile of a first, second and / or third marker panel in a sample from the patient comprises determining the level of the markers of the first, second and / or third marker panel in the sample from the subject and comparing the level of the markers to the corresponding reference value.

11. The method of any one of the preceding claims, wherein: (a) (i) an increased level of one or more markers from Table 1.1.A or 1.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates TNF mediated signalling; (ii) an increased level of one or more markers from Table 1.1.B or 1.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates TNF mediated signalling; (b) (i) an increased level of one or more markers from Table 2.1.A or 2.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with an agent that downregulates IL-6 mediated signalling; (ii) an increased level of one or more markers from Table 2.1.B or 2.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with an agent that downregulates IL-6 mediated signalling; and / or(c) (i) an increased level of one or more markers from Table 3.1.A or 3.2.A in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of susceptibility to treatment with a B cell targeted therapy; (ii) an increased level of one or more markers from Table 3.1.B or 3.2.B in one or more sample obtained from the patient as compared to the corresponding reference value is indicative of refractoriness to treatment with a B cell targeted therapy.

12. The method of any one of the preceding claims, wherein determining the profile of the first, second and / or third marker panels comprises determining the nucleic acid levels of the markers of the first, second and / or third marker panels, optionally wherein the nucleic acid level is an mRNA level.

13. The method of any preceding claim, wherein the one or more sample is a synovial sample.

14. The method of any preceding claim, wherein: (a) the agent that downregulates TNF mediated signalling is a TNF inhibitor, optionally wherein the agent that downregulates TNF mediated signalling is selected from etanercept, infliximab, adalimumab, certolizumab pegol, and golimumab, preferably wherein the agent that downregulates TNF mediated signalling is etanercept; (b) the agent that downregulates IL-6 mediated signalling is an IL-6 receptor antagonist, optionally wherein the agent that downregulates IL-6 mediated signalling is selected from the group consisting of tocilizumab, sarilumab, satralizumab and siltuximab; and / or (c) the B cell targeted therapy is B cell depletion therapy, optionally wherein the B cell targeted therapy is selected from the group consisting of: rituximab, ocrelizumab, veltuzumab, ofatumumab, epratuzumab, obinutuzumab, ibritumomab and tiuxetan, preferably wherein the B cell targeted therapy is rituximab.

15. The method of any preceding claim, wherein the patient is refractory to DMARD therapy.

16. The method of any preceding claim, wherein the method further comprises: (a) administering to the patient an agent that downregulates TNF mediated signalling when the patient is determined to be susceptible to treatment with an agent that downregulates TNF mediated signalling; (b) administering to the patient an agent that downregulates IL-6 mediated signalling when the patient is determined to be susceptible to treatment with an agent that downregulates IL-6 mediated signalling; (c) administering to the patient a B cell targeted therapy when the patient is determined to be susceptible to treatment with a B cell targeted therapy; or (d) administering to the patient an alternative therapeutic when the patient is determined to be refractory to treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling, and a B cell targeted therapy.

17. The method of any one of the preceding claims, wherein the method comprises using at least one predictive modelling approach to identify the agent or to determine whether the patient is susceptible or refractory to treatment, optionally wherein the at least one predictive modelling approach is gradient boosted machine (gbm), elastic net regression (glmnet), mixed discriminant analysis (mda), random forest (rf), support vector machine (svm) with polynomial (svmPoly) or radial (svmRadial) kernel, extreme gradient boosting (xgboost) with tree booster (xgbTree) or linear booster (xgbLinear).

18. The method of claim 17, wherein populations of patients grouped according to clinical response are used as inputs to the predictive modelling approach.

19. An agent that downregulates TNF mediated signalling for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the agent is determined to be suitable for treatment or the patient is determined to be susceptible to treatment with an agent that downregulates TNF mediated signalling by the method of any one of claims 1-18.

20. An agent that downregulates IL-6 mediated signalling for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the agent is determined to be suitable for treatment or the patient is determined to be susceptible to treatment with an agent that downregulates IL-6 mediated signalling by the method of any one of claims 1-18.

21. A B cell targeted therapy for use in treating Rheumatoid Arthritis (RA) in a patient, wherein the B cell targeted therapy is determined to be suitable for treatment or the patient is determined to be susceptible to treatment with a B cell targeted therapy by the method of any one of claims 1-18.

22. A kit for use in the method of any preceding claim.

23. The kit for use according to claim 22, wherein the kit comprises one or more reagent suitable for detecting the markers of the first, second and / or third marker panels.

24. A computer-readable medium comprising instructions that when executed cause one or more processors to perform the method of any one of claims 1-18.

25. An apparatus comprising: (a) profile determination circuitry to determine the profile of a first, second and / or third panel of markers in a sample from a Rheumatoid Arthritis (RA) patient; and (b) (i) agent identification circuitry to identify, based on the profile determination circuitry, a suitability of an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy for treating RA in the patient; or (ii) patient identification circuitry to identify, based on the profile determination circuitry, a suitability of the patient for treatment with an agent that downregulates TNF mediated signalling, an agent that downregulates IL-6 mediated signalling and / or a B cell targeted therapy.

Images