Maximising Therapeutic Utility for Rheumatoid Arthritis using genetic and genomic tissue responses to stratify medicines.

Rheumatoid arthritis (RA) is a disease in which inflammation of joints can cause pain, stiffness, lasting damage and disability. It affects an estimated 500,000 people in the UK and has important impacts on their lives. For example, within 5 years of diseases onset, 33% people are unable to remain in full-time work.
There have been dramatic advances in treatment of RA in recent years. First, patients are treated with drugs as soon as they are diagnosed with RA and the dose is increased quickly to try and suppress all joint inflammation. The most common drug used is called methotrexate (MTX). Second, those people who fail to respond to MTX can be given a biologic drug, so-called because they inhibit a biological pathway involved in inflammation. There are a number of biologic drugs that can be used; most commonly, anti-tumour necrosis factor (TNF) therapies are used as the first choice biologic. Patients who fail to respond to those are then switched to try a biologic drug that acts by removing B cells from the circulation (rituximab, RTX). More recently, a biologic drug that acts by blocking another inflammatory pathway, the IL6 pathway, has been recommended for use (tocilizumab, TOC). Whilst each of these drugs has been proven to control inflammation and slow or prevent lasting joint damage, they don't work for every patient. Thus, 45% patients fail to respond to MTX whilst 30-40% fail to respond to each of the biologic drugs. At the moment, we have no means of predicting which patients will respond best to which drug and so the drugs are prescribed on a 'trial and error' basis. However, we know that the longer it takes to find an effective therapy, the more joint damage accumulates and the worse the long-term outlook is for patients. Therefore, the aim of this study is to identify predictors of response to MTX, anti-TNF, RTX and TOC so that patients with RA can be given the drug that they are most likely to respond to, as soon as they are diagnosed.
The response predictors may be changes in DNA, proteins or other molecules and these changes may be detectable in blood samples or in the joint tissues that are inflamed. In this program of work, we plan a comprehensive series of experiments to investigate this. We will study both tissue samples taken from the joints of patients with RA as well as blood samples collected from large numbers of RA patients treated with each of the 4 drugs to be investigated. We will study genetic changes, protein levels and other molecules. We will combine the information obtained to test whether combinations of markers predict response better than studying one marker at a time. We already have evidence to suggest that the number of B cells in the inflamed joint tissue can predict whether patients are likely to respond well to the B cell depleting therapy, RTX. Therefore, we also plan to test how well this works in practice by performing a randomized clinical trial whereby patients will be randomly assigned to receive RTX or another biologic therapy based on the level of B cell infiltration seen in their joint tissue.
The outcome of this study will be the identification of blood or tissue markers that help to target the right treatments to the right patients with RA. This is particularly important because the biologic drugs are very expensive (£8-10,000 per patient per year). Crucially, as the annual cost of biologics to the NHS is ~£160 million, the identification of the 30-40% of non-responders prior to "blind" therapy would potentially save £13-18 million annually because approximately 4-5,000 patients start anti-TNF therapy each year in the UK.

Research objectives: Two parallel workstrands (WS) will investigate synovial tissue and blood correlates (WS.1) and large scale blood based screening (WS.2) to identify biomarkers of response to four drugs which from the main steps in treatment escalation in rheumatoid arthritis (RA): methotrexate (MTX), anti-tumour necrosis factor (anti-TNF) and rituximab (RTX) and tocilizumab (TOC).
Methodology and experimental design: WS.1: We shall search for tissue-driven biomarkers and blood correlates in the largest synovial tissue biobank of this type (2 time points 0 & 6 months biopsy) in the world & clinical datasets (over 200 patients immediately available). In addition, in a prospective randomised clinical trial (adaptive design) we will test the hypothesis that discrete cellular and molecular signatures in the synovial tissue ("pathotypes") will enrich for response to existing biologic therapies. WS.2: we will take advantage of the large observational cohorts of patient samples either already collected or which could be collected for minimal cost to undertake a comprehensive analysis to identify that will reliably identify responders/non-responders to each of the drugs. These collections include biological samples from over 3,000 anti-TNF-, 1,500 MTX- and 1,200 RTX-treated subjects.
Proposed techniques and approaches: The two WSs are fully integrated through "multi-omic" approaches that constitute cross-cutting themes (CTs) incorporating genetic, genomic, transcriptomic, and proteomic studies. Both WSs converge in a large analytical and modelling package driven by experts in Bioinformatics and Statistics to identify individual or, more likely, a combination of factors that predict treatment response to each of the four drugs. A preliminary health economic assessment of identified biomarkers will also be undertaken.
Exploitation of results: In collaboration with our Industry partners we have drawn up plans to commercialize potential diagnostics for patient benefit

Commercial beneficiaries:
Immediate commercial beneficiaries are the commercial parties of this consortium including large pharma e.g. Amgen, Pfizer, UCB Pharma, Pfizer, diagnostics companies e.g Qiagen and small biotech e.g. Activiomics.

We have developed a strong industry engagement network that has led to a major Industry response with firm intellectual, in-kind and financial contributions from 10 commercial partners (see MICA forms). Industrial partners have been selected to provide expertise in their areas of research, development and commercialisation, as global leaders in the field of therapeutic and (companion) diagnostic development.

Testing of the biomarker will have to be standardized before widespread introduction into clinical practice. Thus commercial diagnostics companies will benefit from development of a point-of-care companion diagnostic or delivery of a clinical test in a CLIA lab. Our industry collaborators will provide a means to facilitate direct engagement with large pharmaceutical companies and SMEs.

Policy makers: Given the expense of biologic drugs and their significant failure rate, policy makers will benefit from this research. Thus our data will contribute to assessments by the National Institute of Health and Clinical Excellence (NICE) to optimize and rationalize treatment for RA.

Patient and public beneficiaries: Ultimately, patients should benefit from this research because it will enable them to be treated earlier in their care pathway with a drug that is most likely to induce remission of symptoms and least likely to induce serious adverse events. This will have downstream consequences for employment and care providers thus having a wider social and economic impact.