The diagnosis of early rheumatoid arthritis (RA) has inherent difficulties. It requires assessment, not only of the current clinical picture, but of the potential for change. As the pathognomonic feature of RA, is persistence it is not surprising that the American College of Rheumatologists criteria perform better in predicting persistence than severity. An adequate histological/imaging method of diagnosing RA is awaited. In the interim, a pragmatic approach to defining disease has been suggested, which takes a homogeneous group of patients with persistent inflammatory small joint synovitis and secondarily stages them for severity. This proposal is currently being assessed for clinical usefulness.

It is likely that, over the next few years, the rapid technological advancements in ultrafast magnetic resonance imaging and ultrasound resolution improvements, coupled with a drive towards minimally invasive surgery, will further enlarge the scope of imaging guided interventional procedures. Such innovations should increase diagnostic accuracy for optimized treatment regimes as well as probably replace an increasing number of conventional 'open' operative procedures.

Imaging is likely to play an increasingly important role in clinical trials of antirheumatic drugs. The need to evaluate efficacy of a drug quickly often precludes the use of radiological evaluation as an outcome measure, particularly in chronic arthritides. New imaging techniques with the refinement to detect differential early change, often between drugs of comparable potency, would be advantageous. However, the trend to set up large scale multicentre trials that have adequate statistical power for the detection of change mitigates against the use of expensive imaging techniques of restricted access. The majority of drug trials, therefore, still resort to conventional radiology. With the new generation of drugs that have specific actions, sensitivity in comparing changes in synovium, cartilage, bone and adjacent soft tissues would be of value.

The steady improvement in densitometric technique is reviewed in this paper culminating in the newest generation of DXA scanners. The suitability of the lumbar spine, femur and hand as sites for the densitometric measurement of bone loss in rheumatoid disease is assessed. The recent improvement in hand scanning technique using DXA has led to improved clinical utility in the detection of juxta-articular bone loss. The enhanced resolution offered by the newest, solid state detector DXA scanners, and their ability to detect bone loss where no systemic features are evident, makes it possible that DXA scanning of the hand will be a useful modality for the monitoring of early rheumatoid disease.

MRI is a tool of unprecedented capabilities for evaluating arthritis and its progression. Not only can it non-invasively delineate the anatomy of all components of a joint with unparalleled clarity, MRI is also capable of probing important functional and compositional parameters of disease in these tissues. Particularly intriguing is MRI's potential for identifying very early changes of joint disease when clinical symptoms may be minimal or absent. Early detection of patients who are at risk for developing progressive disease may allow appropriate treatment to be initiated earlier, when there may be a greater chance of favourable outcome. MRI can, furthermore, provide objective and quantitative measures of disease progression and treatment response. Certain parameters, such as articular cartilage volume, have been validated cross-sectionally; however, their longitudinal performance has yet to be established. Further work is, therefore, necessary to thoroughly validate and optimize some of these measures so that they can begin to be used in more powerful ways to explore the pathophysiology and potential therapies of arthritic disorders.

The study of the cartilage is relatively easy with modern imaging. If the interpretation of MRI cartilage remains difficult arthro-CT images are easy to obtain. Arthro-CT is much preferable to simple arthrography. These two steps are in fact complementary: the global study of the joint is made with arthrography while the accurate assessment of cartilage is obtained with thin CT native slices acquired in an adapted plane and completed with reformatted images. The detection of small lesions, such as a faint thinning of the cartilage, or a focal fissure is only possible with a long and meticulous examination. The pathological value of the images still depends on the clinical examination, because of the high incidence of asymptomatic chondromalacia in elderly patients. The assessment of the cartilage overlying osteochondritis or osteochondral fractures is of particular clinical interest. The detection of post-traumatic cartilaginous lesions may also be valuable in medicolegal problems.

Patients with a wide variety of rheumatological conditions can be usefully investigated by nuclear medicine techniques and particularly by bone scintigraphy. This aspect of nuclear medicine work is increasing and the trend can be expected to continue. The principal conditions that can be imaged are sports medicine injuries, osteomyelitis, avascular necrosis, reflex sympathetic dystrophy syndrome, enthesopathies and bio-mechanical stress lesions, inflammatory arthropathies, metabolic bone disease and miscellaneous bone conditions such as costo-chondritis. Single photon emission tomography (SPECT) has provided new indications for bone scintigraphy such as the evaluation of spondylolysis and facet syndrome in the spine and of meniscal tears and ligamental lesions.

Ultrasound is an extremely useful and versatile method of assessing soft tissue abnormality in rheumatological conditions. It is best performed as an extension of clinical examination. Ultrasound has the advantage of not only being able to demonstrate abnormalities but also allows transducer compression of those abnormalities to see if it reproduces the patient's characteristic symptoms. It is likely to find even greater use in the clinical setting over future years. In the near future skeletal ultrasound should develop into an essential tool for the extension of physical examination in rheumatology practise. It hopefully will become as vital to a rheumatologist as echocardiography is to a cardiologist. This will however require clinicians to be prepared to undergo sufficient training in order to avoid diagnostic errors. Probably it will only be at that time, when skeletal ultrasound has become a fundamental part of rheumatological diagnosis, that its full potential will be realized.

Percutaneous needle biopsies of synovium are successfully used for diagnosis and investigation of joint disease by an increasing number of groups around the world. This procedure can be done in the office with little morbidity; a large number of samples can minimize the potential limitation of sampling error. Clinical indications for 'imaging the joint' by looking at morphological and other features of the actual tissue include undiagnosed acute or chronic mono- or oligoarthritis, haemarthrosis, suspected deposition diseases, new developments in previous stable disease and less often unexplained polyarthritis. Research into any joint disease can be helped by study of synovium especially using newer immunohistochemical, EM and molecular techniques. This report has reviewed other methods used for obtaining synovium, described the different percutaneous biopsy needles, detailed the methods used for biopsy with the Parker-Pearson needle and described how our group handles tissue so as to obtain maximal impact. The very few side effects of needle biopsy include haemarthrosis and, rarely, needle breakage. Finally, we have provided a brief overview of normal synovium and some aspects of synovium in a variety of joint diseases.

Arthrocentesis has to be considered as a part of the clinical examination. A reasonable amount of aspirated synovial fluid is the best argument in favour of an objective articular disorder. Moreover some very simple evaluations are very helpful to make a diagnosis and to distinguish some particularities of rheumatic diseases. Such evaluations have to include both bacterial and synovial fluid analysis. Moreover, when performing synovial fluid analysis, a search for microcrystals is also performed. Haemarthrosis can easily be distinguished from a traumatic tap if the investigator is observing carefully the synovial fluid entering in the syringe. The diseases responsible for haemarthrosis differ with the age of patients: chondrocalcinosis, together with osteoarthrosis, is the most frequent aetiology in the elderly; disorders of haemostasis and synovial tumours are mostly observed in children and young adults. Paucicellular (< 1000 cells/mm3) synovial fluid is observed in different 'mechanical' disorders. In the case of purulent synovial fluid the primary diagnosis is septic arthritis. However, the most common aetiology is probably crystal-induced acute arthritis. Differential cell count analysis performed in case of 'inflammatory' (> 1000 or 2000 cells/mm3) synovial fluid usually shows a predominance of polymorphonuclear cells. However, high cellularity may sometimes be associated with a predominance of other cells, i.e. lymphocytes, monocytes, eosinophils. In this situation, such a simple evaluation (differential cell count analysis) is very helpful in making a diagnosis, e.g. eosinophilic arthritis, or to distinguish some particularities of rheumatic diseases, e.g. absence of cartilage breakdown in case of lymphocytic arthritis.

Arthroscopy still remains the 'gold standard' for the assessment of articular cartilage and synovium because it provides direct and magnified evaluation of these anatomical structures. Thus, alongside the use of arthroscopy as a diagnostic or therapeutic procedure in knee disorders, a further function of knee arthroscopy, performed under local anaesthesia on an outpatient basis, has been proposed: the monitoring and follow-up of knee chondropathy and synovitis conducted for research purposes on patients suffering from knee osteoarthritis or chronic synovitis. This function is used in order (1) to evaluate the natural history of these diseases, (2) to assess 'de visu' the effect on chondral or synovial lesions of medical treatments or surgical interventions, and (3) to validate non-invasive imaging techniques such as plain radiographs or magnetic resonance imaging. The development of this arthroscopic outcome measurement of chondropathy and synovitis required the establishment and validation of systems for scoring the severity of chondral and synovial lesions. The author reviews both earlier and newer arthroscopic classifications and underlines the need for a quantitative description of cartilage and synovial abnormalities, either global and based on the investigator's overall assessment by using a visual analogue scale of severity, or, more analytically, taking into account the baseline parameters of the lesions, i.e. depth/intensity, extent and location.

Radiographs are characteristically required to define the nature of the disease process in spondylarthritis. They need rarely be repeated, except for complications or unusual manifestations of the underlying disease. To date, new techniques of radio-imaging have provided only minimal advantages, if any. (Berkowitz et al, 1991; Docherty et al, 1992; Gibbon, 1992; Ralston et al, 1992; Deyo, 1994; Jensen et al, 1994). The plain radiograph is still pivotal to our understanding of the disease.

Radiographs are a suitable outcome measure in patients with rheumatoid arthritis. They reflect the history of the joint pathology and provide a permanent record necessary for serial evaluation of the disease. Great care should be taken to overcome technical problems with radiographs to ensure that good quality films are available to score. Many scoring methods have been described ranging from a global score for the whole patient to the more sophisticated methods of scoring erosions and joint space narrowing in a selected number of joints. These latter abnormalities give additive information and are the most important features in scoring radiographs in rheumatoid arthritis. An overview of the most important methods is given with an emphasis on four selected methods: the Larsen method, a modification of this by Rau and Herborn, the Sharp method and a modification of this by van der Heijde. All four methods produce sufficient intra- and inter-observer reliability. Although data are scarce, the Sharp method and its modifications seem the most sensitive methods for detecting changes over time. However, these are more time-consuming than the Larsen method or its modifications. Depending on the type of study a choice can be made between the two types of methods. For clinical trials where small differences are important, the (modified) Sharp method seems the most appropriate. In working with large data sets, time might be a more crucial factor and, therefore, the (modified) Larsen method could be chosen.

Radiographic joint space narrowing is the hallmark of progression in osteoarthritis (OA). An accurate measure of progression for OA requires a precise quantitative method. New techniques have been designed to quantitatively assess knee and hip joint space using computer analysis of digitally stored radiographs. The interobserver coefficient of variation of the technique ranges from 1% to 3.3% and is clearly better than that obtained by the use of dividers and rule. A precise standardization of the radiological procedure as well as use of films of good quality are required for a good reproducibility of the method. The computerized measurement of joint space size from standard radiographs is applicable to monitor progression of hip and knee OA, and appears to be of value in the evaluation of disease-modifying therapies for OA.

This article has reviewed the use of semiquantitative individual feature scales based on standard photographs in published atlases, and quantitative techniques, including chondrometry, high-definition microfocal radiography, and computerized digital image analysis, and commented on their use in longitudinal studies of osteoarthritis where data are obtained from the reading of serial radiographs. For investigators planning future studies involving the reading of already available radiographs, the use of computerized digital image analysis, if available, is recommended for measurement of joint space because of its superior reliability; if this is not available, then chondrometry should be used. Other individual features should be scored using a standard atlas. For investigators planning new studies, especially therapeutic trials, recent guidelines recommended by a World Health Organization Satellite Workshop on 'The standardization of methods for the assessment of articular cartilage changes in osteoarthritis of the knee and hip joint' should be followed (Dieppe, 1995).

Quantitative macroradiography of patients with hand and knee OA, employing standardized radiographic and mensural procedures, permit: accurate and reproducible measurements of all radiographic features. significant changes in JSW and osteophyte number and size to be detected within as short a period as 18 months. therapeutic effects upon articular cartilage to be determined from changes in JSW. a better understanding of the disease process from an evaluation of the relative changes in the different radiographic features of OA.

Radiological assessment of joint space width on plain radiographs is the most frequently used primary outcome criterion in clinical trials of disease modifying drugs. Changes in measurement over time are slow. Therefore, in order to improve the capacity of measurement to detect changes, the sources of variability (i.e. radiographic procedure, joint positioning, process of measurements, etc.) must be limited. Precise guidelines concerning radiographic procedure, joint positioning and methods of reading are required to achieve this goal.

In the light of more modern techniques such as sonography and magnetic resonance imaging, the reader may well ask if plain radiography has still a role in the diagnostic work-up of rheumatoid arthritis. However, in daily routine, the value of diagnostic radiography in support of the clinical and laboratory diagnosis of rheumatoid arthritis is unrivaled. It allows differentiation from other joint diseases, such as osteoarthritis or crystal arthropathies, when the ARA criteria are not conclusive for the diagnosis of rheumatoid arthritis. Further, plain radiography is part of the basic documentation of the disease in measuring disease progression. Therapeutic decisions, such as systemic versus local therapy, and selection of drugs, as well as the form of local therapy, are heavily dependent on radiographs. However, the limitations of radiography in evaluating disease progression have to be recognized. Ultrasonography, as a 'bedside method', and MRI are indispensable adjuncts to radiography, because they are superior in detecting synovitis, early forms of cartilage damage as well as bone reaction such as erosions and cysts. A superior assessment of the degree of synovial changes is also possible with MRI.

Modulation of the immune network through the hormonal system is an exciting emerging concept. It results from a decade of experimental research obtained on rodents. Unfortunately, the therapeutic response on NZB/NZW F1 mice with murine lupus or in adjuvant induced arthritis are not transferable to man. This new field opens new therapeutic perspectives with well known and relatively safe drugs (hormone agonists, antagonists or inhibitors of hormonal secretion). Today we lack controlled, randomized clinical studies, with long-term follow-up. We also need to identify the subgroup of patients who could benefit from hormonal immunomodulation (sex, basal hormonal status, menopausal status for women, disease activity).