Rheumatoid patients with intractable knee effusions may benefit from medical or radio-isotopic synoviorthesis. These offer more convenient, less costly alternatives to surgery with similar long-term outcome. Temporary symptomatic relief may be obtained, but disease progression is unaffected. Potential adverse effects include development of osteoarthrosis with osmic acid and teratogenicity and mutagenicity with alkylating agents and radioisotopes.

Choice of diet is one way in which an individual can influence his/her own health, and it is to be expected that patients will seek their physician's expert opinion regarding dietary matters. Respect for the legitimacy of these enquiries and balanced informed discussion, which includes general advice for a prudent diet, as well as disease-specific recommendations when indicated, can be the key to a productive relationship between patients and physician. The issue of dietary advice has an impact on the management of most forms of arthritis including osteoarthritis (obesity/energy balance), gout (dietary purines, energy balance, alcohol, fluid intake) and rheumatoid arthritis (n-3 fatty acids). Food hypersensitivity appears to be a rare cause of polyarthritis, and elimination diets and fasting have little or no place in routine practice. Strategies under investigation include oral tolerization, the utility of which remains to be established.

New interest in the use of antibiotics in the treatment of arthritis was stimulated by two factors: (1) observations that, in some forms of chronic arthritis, microbial antigens persist in the synovial membrane, and (2) the increasing knowledge of the anti-inflammatory and immunosuppressive effects of antibiotics. Recently, several published controlled studies reported a beneficial effect of tetracyclines on RA and reactive arthritis. Whether the anti-arthritic activity of the tetracyclines investigated is mediated by the antimicrobial, anti-inflammatory or immunomodulatory properties remains to be determined. It may be concluded from these studies that tetracyclines have a beneficial effect on RA, especially when laboratory parameters are considered. The effect on the clinical parameters is not unequivocal. The adverse effects seem to be mild but the long-term efficacy and safety of tetracyclines as disease-modifying antirheumatic drugs remain to be demonstrated.

A need remains for the development of more effective therapies for the treatment of rheumatoid arthritis (both NSAIDs and DMARDs). The NSAIDs remain the cornerstone of symptomatic therapy, but concern remains about their safety, potential for the delay in commencing definitive therapy and theoretical pro-inflammatory effects. Each of the NSAIDs reviewed here do provide an advantage over therapies previously available and should prove to be useful additions to the rheumatologists' therapeutic armament.

Cyclosporin A (CSA) Cyclosporin inhibits IL-2 release and T-cell activation and, secondarily, affects B-cell function. It also inhibits bone resorption, at least in vitro. This drug's bio-availability averages 25-35% but is highly variable. Food and grapefruit juice enhance bio-availability and newer formulations may make its absorption more reliable. It is highly concentrated in fatty tissues and red blood cells but does not cross the blood-brain barrier. CSA is metabolized to numerous metabolites by the liver and its elimination half-life is 6-12 hours in the absence of severe liver disease. Biliary excretion accounts for 94% of CSAs elimination. Because it is highly metabolized, its metabolism can be inhibited by other drugs (e.g. ketoconazole and erythromycin) or its metabolism can be induced (e.g. anticonvulsants). Cyclosporin is more effective than placebo for the treatment of rheumatoid arthritis and as effective as other antirheumatics. There is potential for the use of CSAs in DMARD combinations. The principal toxicities of cyclosporin are gastro-intestinal and renal, with the latter being of more concern. Leflunomide (LF). Leflunomide may be a pyrimidine synthesis inhibitor, although tyrosine kinase inhibitor may also be part of its mechanism of action. Its active metabolite is excreted renally to a large degree, with a prolonged elimination half-life of about 11 days. Since LF is activated by liver metabolism, renal failure may have less effect on kinetics than severe liver disease. Early data on efficacy indicate efficacy at 10-25 mg/day, although more well-controlled data is necessary. Toxicity relates to the skin, liver and GI tract, although some degree of weight loss was also found. Nitrogen mustard (NM). Nitrogen mustard is an alkylating agent whose pharmacokinetics are poorly understood. Small, open studies in RA indicate that NM has a potential for relatively rapid response (1-2 weeks) but, clearly, much work remains to be done. As an alkylating agent, GI and hematological toxicities are of greatest concern.

It is accepted that combination DMARD therapy is a useful tool in current rheumatological practice. However, well-designed, large, long-term, controlled clinical trials are needed to determine which combinations, dosage schedules, and sequences of administration are most beneficial and least toxic. Until we develop treatment regimens that reliably induce and sustain acceptable control of disease manifestations in all patients for the rest of their natural lifespan, daily oral prednisone will continue to be a troublesome component of 'bridge' therapy, as it becomes the sole surviving constant in complex regimens whose other components are eventually discontinued because of toxicity, lack of efficacy, or non-compliance. We have often seen patients in whom the replacement of a well-tolerated but presumable ineffective DMARD with another DMARD has led to worsening of disease, when the modest benefits of the discontinued DMARD were lost before the hoped for onset of benefit from its replacement became evident. Since the toxicity of combinations of DMARDs has not appeared to be excessive, one can reasonably add the second DMARD to the first, while carefully monitoring for adverse effects and planning ton continue the combination until increased benefit occurs. Subsequently, if the second DMARD is not tolerated, the partial benefit from the first has not been given up, and a longer duration of treatment with the initial DMARD is sometimes associated with satisfactory improvement. If better control of rheumatoid arthritis is evident after 3-6 months of treatment with the combination of DMARDs, one must still decide whether to stop the first DMARD, stop the second, or continue with the combination. In the absence of major toxicity, we are most likely to choose to continue the combination if the patient has had a good response, thus inadvertently embarking on prolonged combined DMARD therapy (Paulus, 1990). Of course, other drugs besides those discussed above are available to control different aspects of joint damage; they should be considered in any combination therapy. Drugs which potentially protect cartilage from damage, such as orgotein, glycosaminoglycan polysulphate (Arteparon), and Rumalon, may prove useful in rheumatoid arthritis; they have been studied in osteoarthritis, but there is evidence that they protect cartilage from breakdown by inflammation in some animal models. As one of the many goals of treatment in rheumatoid arthritis is to protect cartilage, these chondroprotective agents might also be considered as part of the combinations to be studied. The combination of modest clinical efficacy with minimal toxicity reported with minocycline treatment of rheumatoid arthritis make this another potentially interesting addition to combination therapy regimens (Tilley et al, 1995). It is also important to continue the development of so-called 'biological agents', such as interleukin-2 receptor antibodies, anti-CD4 antibodies, anti-TNF-alpha agents and anti-thymocyte globulin. Combinations which include such agents have not yet been evaluated, although is seems logical considering that these agents offer the possibility of precise intervention directed at specific steps of the immuno-inflammatory process; their combination may thus be more effective than the use of single agents alone. While we await results of well-designed studies of these newer agents in RA therapy, we should continue to consider creative ways of using drugs that are already available.

There is considerable evidence to implicate T cells in the pathogenesis of rheumatoid arthritis (RA). They initiate and sustain inflammation and therefore are attractive targets for immunotherapy. Several strategies targeting T cells have been tried in RA. The use of monoclonal antibodies to deplete T cells have been used extensively but with little success. Studies have shown that T cell depleting antibodies produce profound peripheral blood lymphopenia but they are less effective in depleting lymphocytes in the joint. Since clinical efficacy is likely to depend on depleting almost all synovial lymphocytes, high doses of monoclonal antibodies would have to be given. However, the invariably severe peripheral blood lymphopenia induced by such a regimen is likely to result in profound immunosuppression. Therefore, this strategy has been abandoned and recent attempts have been made to induce tolerance in RA. In animal models of RA, treatment with high dose non-depleting anti-CD4 monoclonal antibody protects them from arthritis induced by injection of streptococcal cell wall. In addition, it leads to a state of anergy which protects the animals from arthritis induction without further treatment with anti-CD4 monoclonal antibody. This is currently being used in clinical trials of RA. Other tolerance inducing treatment strategies include T cell or T cell receptor vaccination and oral tolerance. The former is particularly difficult since the rheumatoid arthritogenic antigen and the pathogenic T cell remain unknown. The latter has shown promise in placebo controlled trials although the ideal dosage remains unknown. The mechanism of action of oral tolerance involves either immunosuppressive T cell cytokines, T cell anergy or depletion.

The past few years have seen an explosion of knowledge concerning the role of cytokines and their naturally occurring inhibitors in the promotion and modulation of inflammatory disease. In RA, this knowledge has been translated into the clinic, with ongoing evaluation of specific cytokine inhibitors, including those targeting tumour necrosis factor-alpha (TNF-alpha), interleukin-1 (IL-1) and interleukin-6 (IL-6). In this review, we outline the scientific data supporting anti-cytokine therapies in RA and describe the results of published and unpublished clinical trials with biological agents. At least for anti-TNF therapy, short-term clinical efficacy and good tolerability have been confirmed in randomized, placebo controlled trials. The results of IL-1 blockade in vivo also appear encouraging, although detailed descriptions of trail outcomes are awaited. Problems associated with long-term administration of biological agents are discussed, including the development of antiglobulin responses to injected monoclonal antibodies and poor pharmacokinetics of low-molecular-weight inhibitors. Ways of facilitating the long-term use of current biological agents and alternative means for inhibiting cytokine function in future studies in RA are presented.

The length of time that patients remain on anti-rheumatic therapy is an important measure of the effectiveness of that therapy since length of time on therapy is a composite measure that accounts for sustained, positive therapeutic benefit as well as negative therapeutic benefit (e.g. adverse reactions, unacceptable costs and loss of efficacy), and accounts for noise (non-compliance, psychological factors, misunderstanding, etc.). Effectiveness is a measure of how well a drug does work, while efficacy, the measure used in randomized controlled trials, means that a drug can work; however, efficacy may or may not translate to usefulness in the clinic. To understand drug effectiveness we reviewed studies of 5809 patients receiving various SMARDs. The average median time on drug ranged from 1.10 to 2.27 years, excluding methotrexate, with shortest survival times falling to sulfasalazine (1.10) and auranofin (1.16), intermediate times to hydroxychloroquine (1.59), penicillamine (1.42), IM gold (1.40), and the longest time to azathioprine (2.27). Overall, excluding methotrexate, the average median survival time was 1.41 for 3998 patients. Median time on drug was 3.3 times greater for all other drugs combined, averaging 4.61 years. Expressed in terms of '5-year survival,' an average of 55.7% of patients remained on methotrexate 5 years after it was started. Better results noted here for methotrexate stand in contradistinction to short-term randomized controlled trials which find most SMARDs to be equal in efficacy. Other factors that may influence drug survival time include age, age, education level, psychological status, presence of fibromyalgia, rank order of SMARD administration, disease severity or corticosteroid administration. Studies can provide more information if they also measure clinical variables as well as time on drug, providing area-under-the-curve measurements.

The article details the history, concept, definition and assessment of the still enigmatic condition of 'fibrositis' or, as it has more recently been called, 'fibromyalgia'. The concept and diagnosis became popular, especially in North America, in the 1970s, after the seminal publications of Hugh Smythe (1972) and Smythe and Moldofsky (1977). It is noticeable that there does not appear to be an early case report as there is for instance for gout, rheumatoid arthritis or certain vasculitides. This may be one reason why we still lack a commonly shared clinical image of the 'typical' case. After Smythe and coworkers, operational definitions and classification criteria were given by Yunus et al, Lautenschläger et al (both in 1989) and Wolfe et al in 1990. The latter received the endorsement of the American College of Rheumatology and are now the most widely used. They identify fibromyalgia as a musculoskeletal disorder with spontaneous widespread pain and exaggerated tenderness as prominent and distinctive features. The other two criteria sets refer to a different concept of fibromyalgia as a 'functional' or 'dysfunctional' disorder. These and other nosological differences pose problems for clinical as well as epidemiological research. They may be of minor importance if it is accepted that any present definition is arbitrary and that a wide range of possible elements are more relevant to research than a uniform concept of a disease called fibromyalgia.

Classification of back pain is a difficult task. Traditional schemes have focused on the small percentage of cases which have specific causes. Structural anomalies observed on X-ray examination explain only a small proportion of back pain cases, and the emphasis placed on these in the traditional schemes is, as Anderson put it, the tail wagging the dog (Anderson, 1977). Many syndrome classifications are based on arbitrary notions of cause, with little empirical justification and no evidence that they can reliably and usefully be applied in practice. More pragmatic approaches start with the separation of the serious from the less serious, and the distinction between spinal pain and pain arising from outside the spine. The classification of the large majority of back pain cases which are 'non-specific' is best approached by grading the severity of the clinical and psychological features of back pain and their disabling consequences. Such grading schemes also provide the most appropriate outcome measures for clinical and epidemiological back pain research.

Two complementary notions need to be considered in the definition of osteoporosis: a reduction in bone mineral density and the occurrence of fracture. Bone mineral density is an established determinant of the risk of future fracture, but current interventions are most usefully targeted before bone density has fallen to the levels at which fractures usually occur. Bone density may now be measured accurately and precisely, and the sensitivity and specificity of this technology for fracture risk prediction are high. It therefore seems reasonable to classify osteoporosis in terms of both a reduction in bone density and the occurrence of fracture. On this basis, a four point scale has recently been proposed by the World Health Organization: 1) normal, 2) low bone mass, 3) osteoporosis and 4) established osteoporosis. While it is likely that this diagnostic classification will change as experience of osteoporosis increases, it is important to have widely adopted guidelines of this type which act as a framework for further research.

The systemic vasculitides are a group of rare inflammatory conditions resulting in inflammation and necrosis of blood vessel walls. They are somewhat commoner than previously believed with an annual incidence approaching 40 per million. Furthermore the annual incidence of rheumatoid vasculitis is 12.5 per million and Wegener's granulomatosis is 8.5 per million. The first useful classification system for systemic vasculitis was published in 1952, since then a number of different schemes have been published. The major changes have been the recognition of the importance of dominant blood vessel size, the distinction between primary and secondary vasculitis and the incorporation of pathogenetic markers such as ANCA (see Table 6). Until relatively recently there were no widely agreed diagnostic or classification criteria. In 1990 the ACR published criteria for the diagnosis of polyarteritis nodosa, Churg-Strauss syndrome, Wegener's granulomatosis, hypersensitivity vasculitis, Henoch-Schönlein purpura, giant cell arteritis and Takayasu's arteritis. The criteria were provided in both traditional and tree format. Sensitivity and specificity rates varied considerably: 71.0-95.3% for sensitivity and 78.7-99.7% for specificity. The criteria were not tested against the general population or against patients with other connective tissue diseases or rheumatic conditions. In 1993/94 the Chapel Hill Consensus Conference developed and published definitions for the nomenclature of systemic vasculitis based on clinical features. These have not met with universal acceptance. However, they are a useful addition, since their use should result in different centres studying more homogeneous populations of patients and facilitate comparison of data between different centres. Assessment of vasculitis comprises an activity score (BVAS), damage index and quality of life/health status (SF-36). These are recent developments which are still undergoing validation.

The original descriptions of polymyalgia rheumatica (PMR) and giant cell arteritis (GCA) in the medical literature date back to 1888 and 1890, respectively. Classification criteria for PMR and GCA are not standardized since most authors used subjective criteria based on their personal experience. Only one study has evaluated criteria for PMR and has found seven variables with high discriminant value. Criteria for GCA are less varied because a positive biopsy of the temporal artery is diagnostic. However, combinations of different clinical and laboratory features have been used for diagnosis when biopsy is negative or missing. Assessment of PMR/GCA is based on the serial determination of markers of acute phase such as ESR, CRP, or plasma viscosity. However, their value in predicting recurrence of the diseases is poor. New immunological factors including soluble interleukin-2 receptors, interleukin-6, serum soluble CD8, and serum soluble intercellular adhesion molecule-1 are presently under investigation.

New information regarding myositis specific autoantibodies, histopathologic analysis of muscle biopsy specimens, and immunogenetic features of the different serologic subsets of disease has greatly increased our understanding of the pathogenesis of the inflammatory myopathies. The clinical descriptions of inclusion body myositis and 'amyopathic dermatomyositis' (Euwer and Sontheimer, 1993) are examples of our expanded descriptive capabilities in the evaluation of patients with myopathy. Finally, newer techniques such as cytokine analysis and magnetic resonance imaging may help in the ongoing assessment of disease activity in patients with myositis. The combination of these recently described clinical and laboratory parameters are enough to force a reconsideration of the previously described classification and diagnostic criteria in the inflammatory myopathies.

The review describes the historical development and nomenclature of Sjögren's syndrome--primary and secondary. The strengths and weaknesses as well as similarities and dissimilarities for establishing the diagnosis of keratoconjunctivitis sicca and xerostomia used among the five most commonly classification criteria sets are focused upon in text and table. Validation of the different sets of criteria are performed. Primary and secondary disease activity markers and outcome measures as well as future research plans for improving diagnosis are dealt with.

In practice, the classification of scleroderma is less problematic than that of most other connective tissue diseases given the distinctive pattern of skin involvement. Classification of early and limited disease is more problematic and is difficult to separate from severe forms of Raynaud's phenomenon. Division of scleroderma into two groups depending on the presence or absence, early in the disease, of truncal skin involvement makes biological and clinical sense. Measurement using a clinical skin score remains the best approach to monitoring disease progress.

Assessing patients with SLE is difficult because of the heterogeneity of the disease. The Revised 1982 ARA Classification Criteria set has been widely accepted for classifying SLE patients for inclusion in clinical studies, but it is not appropriate for making the diagnosis of SLE in an individual and is not helpful for classifying patients with early or mild disease in population based epidemiological studies. Further refinement of this criteria set to meet these objectives and to facilitate subdivision of patients with SLE into those with similar clinical, serological or genetic features poses a future challenge for the clinical epidemiologist. A number of valid and reliable indices are available for measuring clinical disease activity in SLE (Table 3). Despite their different approaches they have been shown to correlate well with each other and hence would appear to be measuring the same thing. The exact choice of instrument will be dictated by the purpose for which it is required. Although none is perfect they are useful for monitoring groups of patients in outcome studies in a research setting. Practically, disease activity indices are unlikely to be appropriate for dictating treatment decisions in individual cases: an instrument comprehensive and flexible enough for this purpose would undoubtedly be far too cumbersome for widespread use. Ultimately such fine tuning will continue to rely upon the skill and intuition of experienced physicians.