Musculoskeletal ultrasonography: Improving our senses
Musculoskeletal ultrasonography (MUS) has been used for years but traditionally it has been the tool of radiologists, except in mainland Europe where rheumatologists have seen its potential for many years. However, there is a widespread increase in interest and use by rheumatologists (1, 2), which is reflected in the increasing prevalence of rheumatologic-focused literature involving small joint and inflammatory arthritis evaluations (3). Improvements in MUS equipment (especially in digital processing and transducer technology) including the development of subsequent smaller and less expensive machines with better picture resolution, increasing exposure at scientific meetings, and an improving evidence base have all fueled this growth in interest.
MUS is an instrument that potentially extends our clinical examination skills. Before assessing the advantages and disadvantages of MUS for rheumatologists, the reader should know that this editorial is written from the perspective of a rheumatologist who works as both clinician and researcher. The author was medically educated in an ethos that highlighted the importance of careful history taking and skilled examination before considering investigations, and still strongly adheres to these principles. Unlike many readers though, I have had the opportunity to work alongside well-trained rheumatology ultrasonographers for many years in both routine clinical and research settings.
The technology and potential benefits of MUS
Kane and colleagues have recently presented the very interesting history of MUS development, from early 20th century sonar (sound navigation and ranging) to the first description of sonographic rheumatoid synovitis in 1978 (4). In brief, ultrasonography uses the properties of sound waves (at a frequency range above human hearing) such as reflection, absorption, and differing speeds through various tissues to create its images. Higher frequency transducers (the probes that are applied to the patient's skin) provide good resolution but poor penetration, so these have been used in small joint evaluation, whereas lower frequency provides better penetration but at the cost of image resolution. Traditional ultrasonography images are in black and white (called gray-scale ultrasonography); using gray-scale, fluid and bone result in black and white images, respectively (5). Doppler ultrasonography depends on the detection of moving cells, color Doppler ultrasonography has been used for many years to assess carotid stenoses, and power Doppler can detect lower flow-rate in small vessels such as in synovitis (6–8).
Ultrasonography is extremely patient friendly and functions without the radiation exposure of conventional radiographs or the relatively long immobilization required for even patient-friendly peripheral magnetic resonance imaging (MRI) scanners. MUS provides the ability to examine multiple joints during a single visit to the clinic. Patients can often see their particular problem being imaged, aiding their understanding of the structural pathology. The extensive spectrum of pathology that can be imaged with MUS has been reviewed (5, 9) and includes abnormalities of the synovium, tendon, bone, skin, muscle, and nerves.
Realizing the potential
Although potential advantages for MUS have been known for some time, there remains a critical lack of evidence on which to base recommendations for many key areas.
Training and accreditation.
As with the uptake of all new medical procedures, using MUS will require rheumatologists to undertake appropriate training, a topic on which there is scant evidence (10, 11). Training courses have been held under the auspices of the European League Against Rheumatism (EULAR) and the British Society for Rheumatology for a number of years, but to date these courses have given instruction in general principles and have not provided competency certification. In this context, the article by Brown et al in the February issue of Arthritis Care & Research (12) is highly important in establishing an expert-based consensus not only on indications for MUS, but on the appropriate knowledge base and skills required. The authors used a Delphi approach and an international panel of ultrasonographers, including both rheumatologists and radiologists. It is perhaps not surprising but a little disappointing to see the lower response rate of the radiology experts (59%) compared with the rheumatologists (90%). Clearly radiologists often have extensive experience in ultrasonography that is complementary to rheumatologists' musculoskeletal clinical skills, and a partnership would be desirable to further MUS advancement.
The indications for MUS selected by the experts are not surprising and reflect the current belief of the benefits of MUS in inflammatory arthritis diagnosis and monitoring. The presumed benefits of guided joint injections were also reflected by the experts' preferences. The anatomic areas selected as appropriate (predominantly joint-centered) again reflect common rheumatologic practice.
The next steps in training will be to ensure that MUS training courses include the breadth of pathology and anatomic sites recommended in the Brown et al article (12). The composition of training programs might be expected to contain core elements on basic technical understanding and selected modules that allow practitioners to learn specific skills, e.g., joint injection or specific joint examination.
Clearly, highly important questions remain concerning the certification of competency after MUS training and who should be the accrediting organization. Should there be external assessment as part of this process? There are obvious medicolegal implications for both practitioners and rheumatologic societies.
Equipment is another very important issue. Similar to MRI, ultrasound machines have changed remarkably over the last 20 years. Smaller transducers have meant greater accessibility for small joints, and portable machines mean this technology truly can be used very easily in the clinic. However, the ability to detect certain pathologic abnormalities will not only be dependent upon the training of the ultrasonographer but also on the particular equipment that is being used. The equipment itself will need to undergo regular quality control, as is usual of diagnostic instruments in radiology and pathology departments.
Ultrasonography machines have a range of differing capabilities. Perhaps the most important questions relate to the potential advantages of power Doppler over gray-scale MUS. The preliminary evidence to date suggests that power Doppler may be able to distinguish active, more vascular synovitis from inactive, fibrotic tissue (6–8). This ability would be clinically helpful with the common but difficult decision on whether to increase antiinflammatory therapy in the patient with established rheumatoid arthritis (RA) with low-level disease as assessed by clinical examination and systemic biomarkers. However, the use and interpretation of power Doppler signal is fraught with technical difficulties and machine dependence; therefore, additional research is required in this area.
It is important to realize that MUS technology is changing rapidly. For example, 3-dimensional evaluation that has been commonplace in MRI is now possible with ultrasonography (13); therefore, the continual evolution of technology will have an impact on all the issues discussed in this editorial.
The performance properties of MUS.
Many questions remain unanswered about the validity, reliability, and sensitivity to change of MUS measurements. The pertinent literature in this area (related to RA on which most work has been done) has recently been reviewed in detail (14). Although this review highlighted particular gaps in the field in assessment of interscanner variability, responsiveness, and predictive validity, work in these areas has been hampered in part by the lack of consensus definitions of MUS pathology, and a group of interested ultrasonographers (working within Outcome Measures in Rheumatology Clinical Trials and EULAR frameworks) have only recently established preliminary definitions for further testing (15). Such definitions will be strengthened by reports on MUS findings in healthy individuals (16, 17). Of course all of these issues again relate to the technology used (e.g., gray-scale versus power Doppler) and operator issues. The operator-dependent nature, which has been a major concern in the uptake of MUS, can only be addressed with appropriate training.
Impact on clinical practice.
There is a small but growing collection of literature in the vitally important area of MUS clinical utility (9). In terms of diagnosis, tests add the most value to the clinical scenario where there is diagnostic uncertainty, most notably when the test can provide information not obtained on history and examination. Ultrasound is therefore extremely useful in several common scenarios. First, when the clinical diagnosis at the individual joint level is unclear. This is the case in swollen joints where synovitis, effusion, and tenosynovitis may be confused, or where there are complex joint structures, i.e., at the wrist and ankle joint. In this situation, MUS can make a major difference in the local (or joint-specific) diagnosis (18), which may also impact the overall diagnosis. Second, MUS is more sensitive than clinical examination in detecting synovitis (19). This may be extremely important in the patient with oligoarthritis clinically, who with the application of MUS is found to have a rheumatoid-like pattern of polyarthritis. Preliminary data suggest that this subclinical synovitis is clinically relevant (20, 21), but further studies are required in this area. It is important, however, to realize that even with increased detection sensitivity, the presence of synovitis does not automatically imply an inflammatory arthritis; synovitis (when defined as synovial hypertrophy or effusion) is a common finding in osteoarthritis (OA) of the knee examined by MUS (22) and is presumably also common in small joint OA.
MUS enables guided aspiration of traditionally difficult-to-enter joints (23). Also, a distinction between synovial hypertrophy and effusion can aid clinical decisions on whether or not to attempt aspiration. These factors should improve diagnosis in the acutely-inflamed joint, especially with respect to gout. This may also require a greater willingness on the part of rheumatologists to perform synovial fluid analyses.
MUS provides the examiner with the ability to detect features that our fingers never could: bony erosions. These may be a rheumatoid-type or bony irregularity associated with entheseal disorders, both of which may also be useful in diagnosis. Ultrasound is more sensitive than radiographs in detecting RA erosions, especially in early RA (24); this sensitivity will depend on the particular joint scanned (related to the access of the ultrasonography transducer) and the number of joints scanned (when compared, for example, with the large number of joints included in a set of hand and feet radiographs).
MUS-guided injections could potentially have an impact on therapy. It is often quoted that many intraarticular injections are not appropriately placed, even in large joints (succinctly reviewed in a recent editorial by Hall and Buchbinder ). MUS has been demonstrated to improve placement of intraarticular injections, which should at least improve short-term pain outcomes (26, 27). The effects of such guidance are seen in patients with RA, where intraarticular injections have obvious long-term potential benefits (28); however, it is certainly unlikely that many biomechanically driven joint disorders will be cured because corticosteroid injections are more accurately placed (29), and more studies will be required to determine the role of MUS-guided injections in the clinic. It is too early to determine whether the advances in clinical examination due to MUS will actually improve our prognostic ability in inflammatory and mechanical diseases.
The introduction of new procedures generally implies increased service costs. Although it could be argued that the benefits of MUS-guided injections, for example, may actually reduce number of patient visits, it is likely that an increased number of examinations will occur if the rheumatologist-practitioner has the technology available in the consulting room (as opposed to referring the patient to a colleague at another site). Longer examination times, training in MUS, and the length of the procedure itself will all have implications for reimbursement. It is therefore attendant upon us as responsible physicians to ensure evidence-based guidelines for the use of MUS. This will require a more robust analysis in the scientific literature than currently exists. A considerable expense is still required when purchasing an MUS machine, and the rheumatology-ultrasonographer would be well advised to have carefully considered the many issues raised here before rushing to add this equipment.
Is peripheral MRI an alternative?
The sparse (but increasing) literature on peripheral, office-based MRI does not yet support the inclusion of this question, but is mentioned to highlight the rapidly-changing technology in both the MRI and ultrasonography fields and our need to be rigorous in critically evaluating the clinical use of technology.
The future is now!
The use of MUS by rheumatologists offers an incremental and measurable improvement in our diagnostic and therapeutic skills. As with all procedures, it should never replace the need for careful history and examination. Substantial research is required to determine crucial issues on validity, reliability, and clinical usefulness. However, MUS is now well entrenched in rheumatology, and I believe that on balance, given the potential advantages but limited clinical impact data, consideration should be given to incorporating MUS in rheumatology training programs, at least as an optional component. The training required is substantial and the work of Brown and colleagues (12) is an essential step to ensure that such training is relevant for rheumatologists. Clearly these MUS training programs, practitioner competency accreditation, and revalidation over time now need the attention of all national rheumatologic societies.