Molecular Prediction of Osteoarthritis to enable its Prevention: Post-traumatic Osteoarthritis as an exemplar

Osteoarthritis (OA) is the commonest form of arthritis, affecting 8.5 million people in the UK. It is the main reason for hip and knee joint replacement surgery, with an associated healthcare budget estimated at £2 billion, with much greater societal cost. We currently have no drug treatments that prevent, slow or cure OA. Knee joint injury, such as anterior cruciate ligament rupture, is the biggest risk factor for future knee OA. About half of all people with knee injuries will develop OA; surgery to treat the injury does not reduce this risk. This type of OA is known as 'post traumatic osteoarthritis' (PTOA). Individuals are often younger when they develop PTOA, but we don't know if this form of OA is otherwise different from 'usual' OA. There is an inflammation response in the knee to the injury which varies considerably between people. Our laboratory studies link this initial 'joint injury response' to later OA. We can measure different aspects of the response: protein 'markers' in knee joint fluid, message levels of genes in blood, an individual's genetic makeup and clinical factors like age, sex or type of injury. Studying the joint injury response gives us an opportunity to understand the processes which cause OA, and to aim to prevent PTOA by picking out those who are at high risk and treating them at the time of their injury.

My aim is to identify and test predictors of progression to PTOA after knee injury which can be used in the clinic. I will:
-Identify clinical factors and measurable markers at the time of the knee injury which predict PTOA (in the joint and in blood)
-Assess the genetic risk of PTOA
-Develop a test or risk score which can rate a person's individual risk of future PTOA
-Improve our ability to design clinical trials with the aim of preventing PTOA

We will achieve this by using 3 different approaches: (1) work in specific knee injury 'cohorts' (groups) of individuals who have been followed over time with the collection of clinical samples, questionnaires and scans or X-rays; (2) join together a number of other international knee injury cohorts to ask genetic questions in PTOA, but also to check our findings and (3) access much larger existing research studies of individuals with associated healthcare data: UK Biobank (~32,000 individuals have knee OA and associated genetic information), and a large group of general OA studies (Genetics of Osteoarthritis) has an additional 61,000 cases of knee OA. These big numbers are needed to ask genetic questions. We will ask whether known genetic risks for 'usual' OA are the same for PTOA, and whether we can identify any new inherited risk factors. We will also look at 5000 markers in a single joint fluid sample at once, and look at how genes and markers relate to each other.

If we identify markers that predict OA, we can develop a test or risk score which gives an individual their personal risk of OA. This would have a number of benefits: helping those with injury, with work and life planning and choices around exercise and potential treatments. It would enable clinical trials in this area, because for the first time we would be able to pick out those at the highest risk. This would make clinical trials more acceptable to participants, and also increase our chances of measuring a true effect of a treatment, allowing smaller numbers and greater certainty about our findings. It may also give us faster, more accurate answers in trials. Lastly, the test could be used in the clinic, channelling scarce healthcare resources to those who are at greatest risk and not giving unnecessary treatments. It is possible that these types of markers may also be relevant to those with early OA of other causes and we will test this. Given the high and growing frequency of OA, finding a way of making any difference in this area is likely to reduce our future health and social care costs.

The most important stakeholders to benefit will be patients. Currently, there is no way of predicting osteoarthritis (OA) after joint injury, which often affects young people. In a previous discussion group run by our Centre for those with a recent significant knee injury, some of them described their current uncertainty. P2"Where am I going to be in 20 years?";"I want to know what the exact implication of my injury is for me." P3"I am keen to know the facts. I would like to know my personal risk of arthritis. I would also want to know what could be done. It would help me plan my fitness and lifestyle: at the moment I feel like I'm 'flying blind'." A predictive test, alone, or incorporated into a risk prediction tool which is fit for the clinic would address their questions directly. This is the primary goal of this work and could be achievable within 5-7 years. Until we can identify those at the highest risk, trials of prevention of post-traumatic OA (PTOA) remain unlikely. Although there are other challenges in delivering such trials, a need for risk stratification is agreed to be essential. Patients understand that re-stabilising knee surgery does not reduce their risk of PTOA and they want new treatments to prevent disease. P3"I know it is difficult, but if certain things could be done to prevent [OA], that would be good." P1"Prevention is better than cure: Can we stop people getting arthritis? Understanding the very first triggers for disease is key."

The second group benefitting would be Pharma and regulators. Industrial investment in OA has wavered in recent years. One of the biggest challenges has not been the identification of novel targets, but in trial design. Most OA does not progress rapidly or predictably, making trials long. Combined with insensitive outcome measures, this leads to large expensive trials which are unattractive to Pharma. A linked issue is one of safety. Exposing all those to an agent, including individuals with good prognosis is not attractive, either to regulators, patients or indeed healthcare providers. This issue is augmented in prevention trials. The FDA has no current label for prevention of OA after injury. In a recent workshop, major needs identified by Pharma to allow progress were the ability to stratify at the time of joint injury and development of surrogate endpoints. This work would directly address this need, with output within 4-7 years. We may also identify novel therapeutic targets. Although this is less predictable, our work in year 4 would assess potential targets for their feasibility to take forward, ideally in collaboration with Pharma.

The scientific community working on PTOA currently work in silos. Other than an increase in knowledge, this fellowship would benefit the community by bringing groups together to work on the shared question of genetic risk. No individual group can do this alone, and this fellowship will broker a new way of working, which may yield future benefits. This output is within 4 years as a primary objective.

UK plc could benefit from IP and manufacturing of a predictive test, or from identification of new therapies. Both would have global healthcare implications.

Lastly healthcare providers and budgets would benefit. The cost of OA after joint injury is not known, but will be disproportionately represented in the overall spend on OA of £2 billion per annum in the UK, given that individuals are younger (affected longer, more work disability), with high healthcare costs (often needing early joint replacement and subsequent revisions). Any progress in this area, in the assessment of risk (to focus existing secondary prevention strategies such as weight loss and exercise), to developing new treatments, would likely have a positive impact on healthcare provision and associated cost. This would be magnified if our findings were transferable to those with early non-traumatic OA, with impact in this area possible within 10 years.