Early Rheumatoid Arthritis Study (ERAS)

27/05/04: Dr Adam Young, FRCP, St Albans NHS Trust, UK
 

Rheumatoid Arthritis (RA) is a condition of unknown cause that can vary from a short lasting and self limiting illness to a more chronic course lasting for decades. Treatments vary according to early, intermediate or late stages of disease. The early phase requires control of the main symptoms of joint stiffness, swelling, tenderness and pain using analgesics, anti-inflammatory and disease modifying drugs, or combinations of these. Current drug therapies do not 'cure' Rheumatoid Arthritis, but do suppress the condition. Some patients with mild Rheumatoid Arthritis require little therapy whilst others may need several drugs, singly or in combination. Fortunately this more severe type is less common. Once mechanical damage to joints has occurred, pain and joint deformity often require assistance from remedial therapy, in the form of occupational and physiotherapy, chiropody and podiatry, aids and appliances. Orthopaedic surgery may be required in later stages for large joint replacements (e.g. hip or knee) or stabilisations of small joints (e.g. fusions of wrist or feet).

The solution to all this would be to discover the cause and thus eradicate the trigger of Rheumatoid Arthritis. An alternative would be a drug that targets accurately the aberrant immune process so that any inappropriate inflammation is 'switched off'. At present neither options are available to us, although there have been some advances in the latter recently. What then do patients, their families, professional carers and clinicians need to know to help reduce the problems encountered in this condition?

Most patients want to know what is going to happen to them ('outcome'), the chance of success, and likelihood and type of drug side effects with a particular drug, and whether they will be able to continue their working, family and recreational lives normally, in a limited way or not at all. Clinicians need to answer these questions and to predict, at an early stage, which patients are likely to do badly ('prognosis') so that they can be offered more powerful medication.

The only way of providing some of these answers is to conduct longitudinal studies in ordinary 'true to life' outpatient settings, and thus track the progress of patients from onset, observe what happens to them and what drugs suit or do not suit them. This approach is quite different from clinical trials, which evaluate new drugs in very controlled and slightly artificial settings, and rarely last more than 2-3 years at most. As our drug therapies for Rheumatoid Arthritis continue to improve, and our targets change, follow up studies need to continue over many years. Previous such studies in single rheumatology centres encountered difficulties because not enough patients were available for study over follow up time.

For these reasons, the Early Rheumatoid Arthritis Study (ERAS) was started in 1986 by nine recently appointed rheumatologists in different regions of England. All had a special interest in this area, and had received training in one of only two other such studies, at the Middlesex Hospital London, started in 1966. All ERAS members have remained in post, which has allowed consistent supervision within each of the nine hospitals. All consecutive patients with early Rheumatoid Arthritis prior to treatment have been included and standardised assessments made yearly by trained health professionals. Over 1000 patients are now part of this study, which has allowed analysis of the less common outcomes, according to both different stages of Rheumatoid Arthritis and varying treatment patterns. Because the ERAS centres cover quite different regions of England, including rural, urban and inner city communities, it has been possible to investigate differences in socio-economic effects and resource use on the outcome of Rheumatoid Arthritis.

The conventional treatment of Rheumatoid Arthritis in the late 1980s and early 1990s was non steroidal anti-inflammatory drugs (NSAIDs, e.g. Ibuprofen) followed soon by second line or disease modifying or slow acting anti-arthritic drugs (DMARDs). Most UK rheumatologists used initially either sulphasalazine or methotrexate, and either switched to or combined the first with the other if the first choice was ineffective or if adverse effects were intolerable. Most ERAS patients received these DMARDs early (70% within 3 months of diagnosis), with little variation between centres in timing or type of drug.

Using this type of approach, it has been possible to track and have details on most of the original patients at 5 and 10 years follow up. Information is still missing in around 10% of patients who failed to attend follow up for unknown reasons or could not because they have either moved house, emigrated, or could not attend because of health, social or work related reasons. Even in these latter patients we have at least some idea of their outcome. Previous longitudinal studies have reported larger numbers in this group, enough to interfere with analysis because failure to attend could be due to cure or to very severe Rheumatoid Arthritis, neither of which was known.

What have we learnt and what findings are useful to patients? Is it possible to present this information, obtained from a large number of patients, using averages and an overall picture that clinicians understand, in a clear and relevant form, to the individual patient who has unique experiences and needs?

Although this article attempts to do this, percentages and averages are still needed! 10-15% of patients went into so-called remission ('cure', at least temporarily), with no evidence of persistent disease at follow up 5 years later. About 70-80% had a 'relapsing and remitting' type of Rheumatoid Arthritis, with a mixture of good and bad days, and a variable reduction in normal activities of daily living. A further 10-15% did not fare well by 5 years, in whom hip or knee replacements (7%), or wrist or foot surgery (4%) was required, or appliances and aids (15%) like callipers, home alterations and wheel chair use were needed. Of patients who lost their jobs, Rheumatoid Arthritis was the main cause in 60%. Details of these outcomes are available in our publications, listed at the end.

Most patients had positive experiences with DMARDs, based on our finding that around 60-70% of patients were still on their first DMARD after 3 years from first starting. In the others the most common reason for stopping a drug was lack of or loss of effect, and not because of side effects. Rheumatologists try to keep patients on these drugs using various different products, doses, routes of administration and combinations, and various antidotes to side effects, because previous studies have suggested that patients who stay on these drugs do better in the long term. It was reassuring to find that, in patients who had to stop one or more DMARDs because of side effects (15%), these were short lasting adverse events, and the course of Rheumatoid Arthritis not unduly affected. The safest drug was hydroxychloroquin (Plaquenil), an anti malarial DMARD with mild activity in Rheumatoid Arthritis, but the drugs with the best benefit:risk ratio were sulphasalazine and methotrexate. The number of severe side effects (requiring hospital admission etc) was very small. The NSAIDs used in Rheumatoid Arthritis were more likely to cause temporary side effects, in the form of gastro-intestinal complications, indigestion, nausea, peptic ulcers. As we improve our use of the newer COX-2 anti-inflammatory drugs (e.g. celecoxib, rofecoxib) and DMARDs for Rheumatoid Arthritis, hopefully this will be less of a problem.

ERAS also has information regarding the more permanent effects of the disease and drugs on major organs like heart, lung, kidneys and nerves in the limbs, which can have major effects on patients' lives, especially in addition to joint disease. Most other reports on the frequency of this aspect of Rheumatoid Arthritis have tended to include patients selected usually on the basis of hospitalisation or severity, and not a true picture of Rheumatoid Arthritis in general. ERAS is a more representative example of Rheumatoid Arthritis, and we have found that these complications of the disease ('extra-articular Rheumatoid Arthritis') are fortunately not common within the first 10 years of Rheumatoid Arthritis, around 2-3%. This is still an important group to recognise and treat appropriately from early stages. Do side effects of our drugs cause chronic organ damage? 10year follow up from ERAS suggests that this is uncommon, and if regular supervision by a clinician is carried out, DMARDs are rarely implicated. Long-term use of NSAIDs can uncommonly cause kidney and gastro-intestinal complications, but also need to be carefully monitored.

The NHS has access to a very useful facility, the Office for National Statistics, which allows tracking of patients in such studies who have died. This has allowed ERAS to determine whether the disease itself or its treatment has contributed to cause of death or if the causes are similar to populations without Rheumatoid Arthritis. This has shown us that, firstly no patients died as a direct result of DMARD therapy, and this probably reflects careful drug monitoring schedules agreed within the ERAS centres. Secondly, death rates were only slightly increased in our patients compared to non Rheumatoid Arthritis populations, and this was mainly due to increased cardiovascular deaths (heart attacks). It has been found in other, mainly American studies, that Rheumatoid Arthritis is a risk factor in its own right for cardiovascular disease and death. It is reassuring to find that in England at present, and in the first 10 years of Rheumatoid Arthritis using conventional drug therapy from early stages, this effect is quite small, and in keeping with a similar Dutch study.

Is it possible to predict the course and complications of Rheumatoid Arthritis early on? The ability to forecast accurately the likely course of Rheumatoid Arthritis would be very useful to patients, in order to plan their lives, and those of their families, and very helpful to rheumatologists, concerning decisions to offer the more powerful drugs or combination regimens. The ERAS nurses and therapists have been evaluating the use of measures of disease activity and severity (for example, active joint counts, HAQ, ESR, DAS28) since 1986, so that these assessments are now done routinely and efficiently in ordinary outpatients. During the first year, Rheumatoid Arthritis can fluctuate enormously and clinical features in early Rheumatoid Arthritis are unfortunately notoriously unreliable at forecasting what eventually happens. For example, a low grade and insidious onset of Rheumatoid Arthritis can have a poor outcome eventually, and conversely a patient who starts with an acute, explosive and widespread onset can be very well one year later.

A number of studies and observations have previously reported that certain early features are associated with more severe Rheumatoid Arthritis, for example women, older age, impaired ability for normal activities (function), unfavourable socio-economic status, blood tests like ESR and genetic markers. However, these factors are not consistent and the only reliable factors in all these studies were a high rheumatoid factor test and x-ray appearances in the hands and feet. Although these findings appear reproducible in Rheumatoid Arthritis populations as a whole, how good are they for the individual patient? One of the main aims of ERAS was to develop one factor, or combination of factors, which could predict outcome. Our target was that this had to be correct in >90% to be of any use to patients and rheumatologists. This would mean that such a predictive marker would only be wrong in 10% of patients, an acceptable error given that 100% accuracy would be most unlikely to achieve.

Initial studies in the 1970s reported 70% accuracy, but more recently ERAS and some other studies have reported 85-90% accuracy. In ERAS, accurate prediction is not possible on day one, and our best figures are based on predictive markers at 12months, after initial treatment has had a chance to work. The type of predictive marker and its accuracy also depends on which outcome is measured. This is not surprising, and ERAS has shown that loss of paid work (work disability) is more likely with manual types of work and if measures of function (e.g. HAQ) are unfavourable during the first year. On the other hand, the likely need of hip or knee surgery is more likely in patients who have evidence of inflammation in blood tests which persist during the first year (e.g. ESR). Thus it is very difficult to predict the likely outcome when the symptoms first start, but accuracy does improve with time, and even at 12 months, useful information is available. The current aim of ERAS is to continue to try and improve these figures to >90% using recently developed blood tests and genetic markers.

Most DMARDs have been discovered by chance, but greater understanding of the immune system has allowed scientists to develop 'designer' drugs to block certain known steps in the immune pathway which results in inflammation. These TNF alpha blocking drugs (e.g. infliximab and etanercept) have recently been available in the UK and reduce joint swelling and x-ray damage more rapidly than other drugs in about three quarters of patients, but not all. They also help other features of Rheumatoid Arthritis like tiredness, fatigue and general feeling of illness.

Side effects are generally minor and tolerable, and regular blood monitoring is not required to the degree necessary with conventional DMARDs. However, severe adverse events have been reported, especially severe and unusual infections. In view of this and the considerable expense of these drugs (around £10,000/year), the British Society of Rheumatology (BSR) developed criteria for their use by rheumatologists. The National Institute of Clinical Excellence (NICE) did a formal appraisal of the drugs in 2001 in order to determine if there was enough evidence to show that these drugs were cost effective. In other words, whether the extra cost of the drug was balanced by improved disease control, so that less orthopaedic operations were required, less employed patients lost their jobs, and less patients required hospital admissions. Both the BSR and NICE requested long-term outcome information from ERAS in order to develop their respective guidance, and we were pleased to allow both of them access to our anonomised database. This was an unexpected and very important use of the considerable work that has gone into this study, by patients, health professional and rheumatologists. It has made a major contribution towards the evaluation of these new drugs.

A number of our ongoing studies awaiting results include an evaluation of the effect of Rheumatoid Arthritis on household duties, use of complimentary and alternative medicines (CAMs), a novel way of measuring joint deformity, and the manner and speed of x-ray changes with time. The initial aims of ERAS have been achieved. Treatment of Rheumatoid Arthritis is a moving target and all ERAS members wish to continue follow up of their patients to provide 10-15year outcomes. This study has prompted the development of a similar project using the ERAS model, the Early Rheumatoid Arthritis Register (ERAN), which was started in 2002 and has the potential of a UK based register open to all rheumatologists. This will allow continual tracking of progression and outcomes in Rheumatoid Arthritis on a national basis. 20 centres are now recording the basic information described above onto a central Register held at the MRC Unit in London, and a further 20 are on stream to follow. Some of this information could be used to check whether standards for the management of Rheumatoid Arthritis are adhered to, for example those recommended by the Arthritis and Musculoskeletal Alliance (ARMA), the UK umbrella body for service user and professional organisations in collaboration with NRAS and the BSR.

Details of ERAS centres

Rheumatology centre	Consultant	Clinical Metrologist
District Hospital, Basingstoke, Hants	Peter Prouse	Cathy Boys SRN
Selly Oak, Birmingham, Warwks	A Gough, J.Devlin	Lynn Waterhouse SRN
Broomfield Hospital, Chelmsford, Essex	Paul Davies	Lynne Hill SRN
Medway Hospital, Gillingham, Kent	Peter Williams	Dora White SRN
Diana Princess Wales Hospital, Grimsby, Lincs	David James	Hazel Tait
RJAH Hospital, Oswestry, Shrops	Josh Dixey	Helen Dart SRN
Nether Edge, Sheffield, Yorks	John Winfield
City Hospital, St.Albans, Herts	Adam Young	Annie Seymour SRN
Royal Hants,Winchester, Hants	Nigel Cox	Sue Stafford SRN

ARC academic secretary & coordinator: Cathy Mayes, Rheumatology Dept, City Hospital, Waverly Rd, St.Albans, AL3 5PN, UK. Tel and Fax: 01727 897362

Email: eras@whht.nhs.uk

Early Rheumatoid Arthritis Network (ERAN). Coordinated by Wendy Garwood SRN.
Email: wgarwood@phlexglobal.com

ERAS MAJOR PUBLICATIONS

1. Setting up an early inflammatory arthritis clinic. Gough A, Young A et al. Clinical Rheumatology. 1992; 6:261-284.

2. Short term outcomes in recent onset rheumatoid arthritis. Young A. Br J Rheumatol 1995;34(suppl.2):79-86.

3. Can we predict aggressive disease? Young A, van der Heijde DFM. Clinical Rheumatology. 1997; 11:27-48.

4. Larsen scoring of digitised X-ray images. Solymossy C, et al. Rheumatology 1999;vol 38(11): 1127-9.

5. How does functional disability in early rheumatoid arthritis (RA) affect patients and their lives? Results of 5yrs follow up in 732 patients from the Early Rheumatoid Arthritis Study. Young A, et al. Rheumatology 2000;39:603-611.

6. Socio-economic deprivation and Rheumatoid Arthritis: What lessons for the health service. Young et al. Ann Rheum Dis. 2000;59 (10):794-799.

7. Potential bias in Kaplan-Meier survival analysis applied to rheumatology drug studies. Utley et al. Rheumatology 2000:30:1.

8. Which patients with early Rheumatoid Arthritis stop working? Results from a 5yr inception cohort of 732 patients. Young A, et al. Ann Rheum Dis 2002;61:335-340.

9. Modelling the progression of rheumatoid arthritis: a two country model to estimate costs and consequences of Rheumatoid Arthritis. Kobelt G, Jonsson L, Lindgren P, Young A, Eberhardt K. Arthritis Rheum 2002;46:2310-2319.

10. Modelling the costs and effects of leflunamide in early Rheumatoid Arthritis. Kobelt G, Jonsson L, Lindgren P, Young A. European J Health Economics. 2002;3;130-187.

11. The cost effectiveness of infliximab in treatment of rheumatoid arthritis in Sweden and the United Kingdom based on ATTRACT study. Rheumatology (Oxford) 2002.

12. Prognostic factors for 3year x-ray damage (Larsen scores) in early Rheumatoid Arthritis. Dixey J, et al. J Rheumatology 2003 (in print).

13. Orthopaedic intervention in early Rheumatoid Arthritis (RA). Occurrence and predictive factors in an inception cohort of 1064 patients followed for 5 years. James et al. Rheumatology (Oxford) 2003. In print.

14. A randomised controlled trial of occupational therapy for people with early Rheumatoid Arthritis. Hammond A, Young A, et al. Ann Rheum Dis 2002 (in print).