Longterm Safety Monitoring in Rheumatoid Arthritis. A Proposal from OMERACT

These are exciting times for patients with rheumatoid arthritis (RA).
For years patients have suffered from relative neglect by the pharmaceutical industry. True, nonsteroidal antiinflammatory drugs have proliferated. Little initiative or investment is required to rearrange parts of the molecule within a basic chemical structure in order to change efficacy by a modest amount.

This accepted recipe provides a guarantee of modest profits for the shareholders at little financial outlay for development, but does not necessarily profit the patient. The more interesting collection of drugs with disease modifying action in RA were initially borrowed from other diseases or used in the condition through serendipity.

Some, for example antimalarials and injectable gold, were subjected to little in the way of dose ranging studies to determine optimum dosage. Others, such as azathioprine, relied upon studies performed for different indications1. Penicillamine, by contrast, was the best of several analogs that reduce the titer of rheumatoid factor, imaginatively explored by Jaffe2, its optimum dose subsequently subject to successive reductions as our understanding of the use of the drug grew. The gold standard for disease modification, arguably, remains methotrexate or sulfasalazine.

The former, used in lower doses than in oncology, long championed in the treatment of psoriasis and even psoriatic arthritis, was the subject of
a succession of trials in the USA, each successively showing benefit at ever lower doses3.

In parallel, sulfasalazine was championed in Europe, initially introduced by Nana Svartz for the treatment of rheumatoid disease but later appropriated by gastroenterologists, only to be reclaimed by rheumatologists once its benefit was confirmed by modern trials4.

All this is changing. Considerable effort, initially on the part of many pharmaceutical companies but more recently consolidated through their amalgamations, led to a new generation of disease modifying drugs designed for RA. Some have recently been marketed. Global availability will result, even though demand may initially outstrip supply.

One such compound is leflunomide, novel as a pyrimidine antagonist although synthesized in conventional style at the biochemist's bench5. Another is etanercept, an example of the new generation of biological agents with specific action against the prime cytokine tumor necrosis factor a (TNF-a), essentially the result of immunological research and rather more complex to manufacture6.

Other specific blockers of TNF will follow, the next generation likely to be manufactured through human DNA technology with no antigenicity. Specific collagenase inhibitors, theoretically with a much wider spectrum of use, will soon also emerge.

But all is not necessarily as cosy as it seems in the global village. A trend in the last decade has been for pharmaceutical companies not only
to merge but to become much more efficient in their phase 2 and phase 3 studies. Competitive recruitment between centers is now the order of the day and contract houses spanning many countries (in Europe now stretching from the west into the east) guarantee rapid recruitment to large multicenter trials of adequate power to show very small differences between competing drugs.

The wide inter-center variation in subjective assessments, sometimes performed in many different languages within the same trial, is conveniently ignored, as are differences in the style of recruitment between countries. Regulatory authorities are much faster in processing applications, and the US Food and Drug Administration is now prepared to "fast track" product licence applications for the indication of RA, as they do anticancer drugs and drugs for AIDS.

This feature is not yet prominent in European licensing circles, although the new mechanism for licensing through a small number of Member States in the European Union saves the expense and delay of working separately through up to 15 different licensing bodies in as many countries.

The net effect has been a genuine risk that drugs reach the market with inadequate attention to pharmacokinetics, the optimum dose, the optimum duration of dosing, and whether drugs should be given by themselves or in combination (either in series or in parallel).

Of even greater concern in respect of the newer biological agents is the risk of late, relatively rare, side effects that might be produced by these drugs as they manipulate the immune system, exclusively for the relief of RA. These potent therapies may even alter the natural history of neoplastic disease.

Rheumatologists have long discussed a possible association between RA and lymphoproliferative disorders, the scope for such pursuits
invariably hinting at a "non-proven" verdict7. Oncologists note the occurrence of Sjögren's syndrome with lymphoma in males with Felty's syndrome8. Might disease modification also cause neoplasia in rheumatoid patients?

On the whole, the evidence for azathioprine9 and methotrexate10 has been reassuring, but the new generation of biological agents, much more
potent in this respect, lead us to territory hitherto unexplored.

It is against this background that OMERACT, at its next meeting (OMERACT 5: Toulouse, France, April 3-5, 2000) propose to discuss the development of a large population cohort for longterm safety monitoring in treated RA, described in more detail in this issue11.

Such endeavors are laudable and clearly to be welcomed. They are not necessarily new. The Standing Committee of the European League Against Rheumatism devoted to international clinical research studies including clinical trials has long advocated the benefit of such a register at a pan-European level.

Some governments, often those that spend the greatest proportion of their gross national product on health care, have introduced such registers, often in conjunction with their drug licensing authorities.
These databases are likely to be impartial, concentrating on general associations between drugs, rheumatic disease, and neoplasia, not necessarily restricted to those single specific therapies that, arguably, are most likely to cause the problems.

The pharmaceutical industry certainly maintains such databases, but only within a single company and each specifically devoted to an individual product, their contents usually available upon request for licensing authorities but not necessarily in the public domain.

So the proposals of OMERACT for an independent international effort are timely and worthy of serious consideration and discussion. It is not
entirely clear how they would be funded and it may be unrealistic to expect individual sponsors in the form of pharmaceutical companies to share their findings with other companies worldwide through such a database, unless international law is altered to enforce this principle.

The authors of the proposal (intriguingly drawn from industry as well as academia) concede the existence of other established databases, of which ARAMIS (Arthritis, Rheumatism and Aging Medical Information System) may be the most competitive.

National databases also exist, for example, the Norfolk Arthritis Register in the UK, supported by the Arthritis Research Campaign (formerly the Arthritis and Rheumatism Council)12,13, but these are epidemiological ventures, concentrating on incidence and prevalence of disease within a small circumscribed area, not necessarily adaptable to small increases in the frequency of neoplasms that are in need of detection.

Nevertheless, experience acquired with such a database does draw attention to some of the pitfalls in their collection, presumably still to be considered by OMERACT. One concern is continued contact with the patients enrolled for a sufficiently long time period for late changes to be detected.

 OMERACT tantalizingly offer a survey by postcard, telephone, or Internet on an annual basis, but it is not clear how patients can be persuaded to notify the investigator of their change of address, always assuming they maintain their interest and willingness to cooperate over such a long period.

While the case control within the proposed database appears impeccable and is easily described on paper at this early stage, enrolment to the survey may present more problems. Should this be done by the patients, their pharmacist (not only fastidious by training but perhaps encouraged by a legal obligation), their primary care physician, their consultant rheumatologist, a representative of the company manufacturing the drug, or a trained research nurse? Enrolment may need to be adapted between different countries to suit local customs.

The new drugs will not come cheap. Perhaps the database will be skewed worldwide, restricted either to affluent countries or, of more concern, to affluent sectors of society within a single country. Control subjects will need to be recruited with discretion to ensure they provide adequate control for the group of patients who have such access to treatment. The authors concede the need for the control pool to accommodate up to three simultaneous predispositions (of which disease duration, number of prior therapies, and types of prior therapies are those cited), but the last two of these must surely vary from country to country.

Simple factors of age and sex as well as those as complex as industrial carcinogens and other environmental influences will all need to be considered if an association between drugs and neoplasia is demonstrated beyond doubt. Endpoints will also need to be considered carefully.

Even the most basic, that of death, may be subject to variation in reporting through death certificates between countries. In Third World countries, often with a different genetic pool but where an increasing number of new compounds are being tested, death may even go undetected.

As polypharmacy becomes more accepted14, the task of unravelling which side effect has resulted from which drug will become even more complex. Even the simplest ploy of adding steroid or methotrexate to a biological anti-TNF compound (which may conveniently serve to reduce its use and therefore cost) may seriously affect immunological activity, further complicating an easy unscrambling of whether the drug alone or the drug in combination was the culprit.

Perhaps those subjects who, on genetic grounds, were most likely to develop the severest disease (and therefore need the most potent drugs) are also those most genetically predisposed to neoplasia.

These many reservations apart, the authors of the OMERACT proposal11 are still to be congratulated on their determination and should be encouraged to develop their proposal at their meeting in April 2000,
after which their further deliberations will no doubt be presented to the rheumatological community.

HOWARD A. BIRD, MA, MD, FRCP,
Professor of Pharmacological Rheumatology,
University of Leeds,
Clinical Pharmacology Unit,
Chapel Allerton Hospital,
Chapeltown Road,
Leeds LS7 4SA UK
Address reprint requests to Dr. Bird.

REFERENCES
1.Bird HA. Long term agents for rheumatoid arthritis (anti-rheumatoid
drugs). In: Bird HA, Wright V, editors. Applied drug therapy of the
rheumatic diseases. Bristol: Wright PSG; 1982:103-39.
2.Huskisson EC, Jaffe IA, Scott J, et al. 5-thiopyridoxine in RA:
clinical and experimental studies. Arthritis Rheum 1980;23:106-10.
3.Scully CJ, Anderson CJ, Cannon GW. Long term methotrexate therapy for
rheumatoid arthritis. Semin Arthritis Rheum 1991;20:317-31.
4.Neumann VC, Grindulis KA, Hubball S, et al. Comparison between
penicillamine and sulphasalazine in rheumatoid arthritis:
Leeds-Birmingham trial. BMJ 1983;287:1099-102.
5.Plosker GL, Wagstaff AJ. Leflunomide. Clin Immunother 1996;6:300-6.
6.Jarvis B, Faulds D. Etanercept: a review of its use in rheumatoid
arthritis. Drugs 1999;57:945-66.
7.Gridley G, McLaughlin JK, Ekbom A, et al. Incidence of cancer among
men with rheumatoid arthritis. J Natl Cancer Inst 1993;85:307-11.
8.Gridley G, Klippel J, Hoover R, et al. Incidence of cancer among men
with the Felty syndrome. Ann Intern Med 1994;120:35-9.
9.Lewis P, Hazleman BL, Hanka R, et al. Cause of death in patients with
RA with particular reference to azathioprine. Ann Rheum Dis
1980;39:457-61.
10.Kamel OW. Lymphomas during long-term methotrexate therapy
[editorial]. Arch Dermatol 1997;133:903-4.
11.Lipani JA, Strand CV, Woodworth TG, et al. A proposal for developing
a large patient population cohort for longterm safety monitoring in
rheumatoid arthritis. J Rheumatol 2000:27:827-30.
12.Wiles N, Symmons DPM, Harrison B, et al. Estimating the incidence of
rheumatoid arthritis. Arthritis Rheum 1999;42:1339-46.
13.Wiles N, Barrett J, Barrett E, et al. Disability in patients with early inflammatory polyarthritis cannot be "tracked" from year to year; an examination of the hypothesis underlying percentile reference charts.
J Rheumatol 1999;26:800-4.
14.Kremer JM. Combination therapy with biologic agents in rheumatoid arthritis: perils and promise [editorial]. Arthritis Rheum 1998;41:1548-51.

© 2000. The Journal of Rheumatology Publishing Company Limited.
All rights reserved.