Power doppler sonography: Improving disease activity assessment in inflammatory musculoskeletal disease

Authors


Musculoskeletal ultrasonography (US) is a powerful tool for evaluating joint and soft tissue pathology and is fast becoming an integral part of routine diagnosis and management in rheumatology practice (1–6). This imaging technique is now being performed by rheumatologists, particularly in Europe, as part of their standard clinical assessment of patients. Increasing evidence supports the use of US in a variety of different locations with demonstrable advantage over standard clinical assessment, enabling more accurate patient diagnosis and facilitating the most appropriate management decisions (7). The trend toward earlier aggressive therapy for inflammatory musculoskeletal disease requires reliable initial diagnosis and optimal disease activity assessment. Interest has therefore been directed toward the use of US as an objective tool for the detection and monitoring of joint and soft tissue inflammation and bone damage (8–15) in early disease.

US has a number of advantages over other imaging techniques. It is safe, noninvasive, and emits no ionizing radiation. The equipment can be situated in the rheumatology outpatient clinic, improving patient access and enabling rapid, “real-time” dynamic examinations of multiple joints in multiple planes at one sitting. In addition, both the capital and running costs of US are significantly lower than those of other imaging modalities, such as magnetic resonance imaging (MRI) and computed tomography (CT).

Traditional gray-scale US has been successfully used for some time for the detection of joint and soft tissue inflammation (1–15). More recently, additional US techniques, including Doppler, have been introduced, offering the potential for improving the accuracy of a US assessment. Doppler US is a technique for making noninvasive measurements of blood flow and was developed from the principles first described by Austrian physicist Christian Doppler in 1842 (16). He was the first to observe the effect of motion on sound when he detected a change in the frequency of a sound wave as a result of movement of either its source or receiver.

There are two main types of Doppler US, color flow Doppler (CFD) and power Doppler (PDS). Both produce a similar color spectral map superimposed onto the gray-scale image (the colors being related to the difference in frequency between the transmitted sound wave and that reflected from the moving interface [the Doppler frequency shift]), but they actually encode different information. CFD represents an estimate of the mean Doppler frequency shift and relates to velocity and direction of red blood cells, whereas PDS denotes the amplitude of the Doppler signal, which is determined by the volume of blood present. In this way, CFD is better suited for evaluating high-velocity flow in large vessels (e.g., carotids), whereas PDS is better suited for assessing low-velocity flow in small vessels (e.g., synovium).

There are a number of particular advantages for using PDS in musculoskeletal assessment. Because PDS provides increased sensitivity to low-volume, low-velocity blood flow at the microvascular level, it is particularly useful for measuring and detecting changes in joints and soft tissue as a consequence of inflammation. PDS also increases the specificity of a US assessment, since it helps in the differentiation among tissue debris, blood clot, fibrin, and a complex effusion which can mimic features of synovial proliferation (17). In addition, compared with CFD, the PDS signal is independent of the transducer angle, there is no aliasing (an image artifact reflecting inadequate signal sampling), and background noise is reduced (18). PDS also provides superior depiction of the continuity and boundaries of blood flow and improved definition of vessel characteristics, such as tortuosity, that are suboptimally imaged with conventional CFD (19).

PDS was pioneered as a cardiology investigation in the 1980s, and its use has gradually extended to other medical specialities. The first description of soft tissue hyperemia using PDS was reported in 1994 by Rubin et al (18), who found pericortical blush reflecting soft tissue blood flow in the kidney and testis. The first reports of the potential value of PDS in musculoskeletal disease appeared in the mid-1990s, when Newman et al (20) demonstrated increased signal at sites of inflammation. Breidahl et al (21) demonstrated increased PDS flow around effusions that were due to inflammatory joint disease compared with effusions that were not of inflammatory origin. Reports of later observational studies have favorably compared PDS with clinical disease activity assessment and traditional gray-scale US (15). More recently, validation has been assessed by comparison with histopathology in the knee in rheumatoid arthritis (RA) and osteoarthritis (22, 23) and by comparison with dynamic MRI in the metacarpophalangeal joints in RA (24). PDS has also been successfully used to assess inflammatory disease activity in RA (15, 25, 26) and to monitor response to treatment (27–29). Other reports suggest the benefit of PDS in the assessment of inflammatory myopathy (30), Achilles tendonopathy (31), and reflex sympathetic dystrophy (32).

In this issue of Arthritis & Rheumatism, D'Agostino et al describe a further use of PDS: the assessment of enthesitis (33). Compared with studies of patients with RA, this area is relatively underinvestigated by US (34–36). D'Agostino et al have previously reported the ability of PDS to demonstrate a response to treatment with infliximab in spondylarthropathy (SpA) patients with inflammatory heel pain (37). Their present cross-sectional study evaluates the assessment of peripheral entheses with either RA or mechanical low back pain (MBP). D'Agostino et al use a detailed 5-point classification of US enthesitis graded according to different combinations of abnormal gray-scale and PDS signal. They studied 4 different areas of the enthesis at 18 anatomic sites in each patient. A subgroup of patients also underwent clinical examination.

In this analysis, almost all SpA patients (98%) had at least one abnormal enthesis on gray-scale US, and of these, 81% had abnormal vascularization on PDS. Although a smaller number of control patients had evidence of enthesitis on gray-scale US (44% of the MBP patients and 60% of the RA patients), no abnormal PDS flow was found around any of the entheses in these groups. The site of this increased PDS signal was always at the insertion of the enthesis to cortical bone, with only a small number of patients having additional signal in an associated bursa, but no abnormal signal was exhibited elsewhere. In addition, there was a statistically significant difference in the number of abnormal entheses detected by US compared with clinical examination. D'Agostino et al therefore suggest that both gray-scale and PDS should be used to improve the assessment of enthesitis and that the presence of abnormal PDS vascularization at the cortical bone insertion of entheses may be a highly specific feature of SpA.

There are currently a number of potential limiting factors in the use of PDS which need to be addressed in future studies. There is no standardization of the examination technique or technical parameters for performing a PDS assessment, which is reflected in a wide variation in methodology in the literature. Few reports, including the present one, have fully addressed reproducibility, and it is critical that studies be performed to assess both inter- and intraobserver reliability. As with other US techniques, PDS can be affected by variables such as operator experience and training, as well as by the quality of the US machine and image processing. Movement of the transducer or patient (enhancing the Doppler effect) can result in a so-called “flash” artifact, which may compromise interpretation (38). Altering the machine settings, such as increasing the pulse repetition frequency, reducing gain, and altering the persistence, or using a fixed mold for the area to be examined can help to minimize this artifact. Artifacts can also be produced from the bone cortex (39). Excessive pressure from the transducer can result in vessel occlusion, yielding a reduction in blood flow and decreased PDS signal. This can be reduced by using a standoff gel pad. Temperature fluctuations also have the potential to affect the Doppler signal. Another potential disadvantage of PDS is the lack of information that it provides regarding the direction or velocity of blood flow, although this is considered less important in the context of musculoskeletal US (40).

The ability to quantify the PDS signal is clearly desirable, particularly to allow longitudinal assessment and comparison between studies. At present, most researchers have favored a semiquantitative score of the maximum area of enhancement, similar to that proposed by Newman et al (27). A number of investigators have access to software that counts the number of pixels in chosen regions of interest by digital image analysis, producing a quantitative score (15, 22, 25). Interestingly, Walther et al (22) compared the use of a global semiquantitative score with digital image analysis and demonstrated a good correlation between both scoring techniques. There are also issues of finance, since the initial hardware cost of a machine capable of producing good-quality PDS images is high compared with that of a standard gray-scale machine, but the cost is still low compared with that of a CT or MRI scanner.

Although not used in the study by D'Agostino et al or discussed in their current article (33), intravenous microbubble echo-contrast agents have the potential to further increase the sensitivity of the PDS signal by raising the intensity of weak signals to a detectable level (41, 42). These agents have already been used in cardiology and oncology and are currently being evaluated in inflammatory joint diseases. In a study of 40 patients with various arthropathies conducted by Magarelli et al (43), the use of echo contrast resulted in an increase in the Doppler signal intensity in joints with previously low signal, together with an increased number of joints demonstrating PDS flow which previously had no signal. They also demonstrated concordance with contrast-enhanced MRI in all cases. Other investigators have reported a similar increase in the detection rate of Doppler signal flow using this technique (25, 44), but further verification of these findings is required. Future potential applications of contrast agents include the development of transit time curves, bolus arrival times, time to maximum intensity, area under the curve, and wash-in/wash-out characteristics that may further improve the characterization of inflammation. Microbubble-specific imaging modes such as harmonic imaging, as well as 3-dimensional (3-D) and 4-D ultrasound, are other promising future developments.

We believe that PDS and its related techniques offer exciting possibilities for the assessment of inflammatory musculoskeletal disorders. PDS has been shown to provide important additive information to traditional gray-scale US and has the potential to allow improved detection and quantification of inflammation. In early inflammatory disease, gray-scale US has been shown to be superior to clinical examination, and PDS may further increase its sensitivity. In the future, it may be used prognostically to determine which patients will develop more severe disease. In established disease, PDS may assist in differentiating between active inflammation and chronic fibrotic tissue. It may also provide an objective measurement of therapeutic response and disease monitoring. Although the preliminary data are promising, larger, fully blinded, longitudinal controlled studies are required to assess factors such as outcome, prognostic significance, and reproducibility. Standardization of the technique, hardware, software, and operator training are also important issues that need to be addressed. Nevertheless, should the above limitations be overcome, PDS has the potential to become a cost-effective alternative to gadolinium-enhanced MRI.

Acknowledgements

We would like to acknowledge the contributions of other members of the Leeds Imaging Group, in particular those of Drs. Zunaid Karim, Mark Quinn, Philip Conaghan, Dennis McGonagle, and Andrew Grainger.

Ancillary