Article Text
Abstract
Accurate assessment of disease activity and joint damage in rheumatoid arthritis (RA) is important for monitoring treatment efficiency and for prediction of the outcome of the disease. Therefore, a reliable imaging method needs to be used. Musculoskeletal ultrasound (US) is a sensitive method for the detection of both early inflammatory soft tissue lesions (eg, synovitis, tenosynovitis, and bursitis) and early bone lesions (eg, erosions) in arthritic joint diseases and correlates well with MRI. Several musculoskeletal US scores are used for monitoring RA disease activity. Different qualitative (0/1) and semiquantitative (0–3) systems and quantitative measurements are used. The semiquantitative four-grade system developed by Skudlarek et al, which evaluates joint effusion, synovial thickening, bone erosion and power Doppler activity, is used most often. The new seven-joint ultrasound (US7) score is the first US sum score system which combines soft tissue (synovitis and tenosynovitis/paratenonitis) and destructive lesions (erosions) in a composite scoring system. The US7 score provides a fast overview of disease activity in daily rheumatological practice. This article reviews different US scores and sum scoring systems and current and proposed activity in this field.
- Ultrasonography
- Synovitis
- Tendinitis
- Rheumatoid Arthritis
Statistics from Altmetric.com
Introduction
Medical scoring systems are used to standardise clinical findings, decide on diagnostic procedure, estimate prognosis and reach a therapeutic decision. Imaging scoring systems are developed to standardise and monitor disease activity and determine the therapeutic response. By applying imaging scores in rheumatological cases, immunosuppressive therapy can be adapted. With the numerous available disease-modifying drugs (DMARDs), and especially with the use of biological agents, for the treatment of rheumatoid arthritis (RA), a new era has begun in the field of rheumatology. Recently, an international task force stated that clinical remission was the most important target in the treatment of RA.1 But as the new treatments with biological agents are expensive, reliable scoring methods are needed to determine the effectiveness of these treatments. Standardised monitoring of RA is provided by different clinical scores (ie, 28-joint count Disease Activity Score (DAS28); simplified disease activity index (SDAI); clinical disease activity index (CDAI)), which reflect the clinical disease activity and therapeutic response.
However, clinical scores are limited because, despite clinical remission, subclinical disease activity is observed, which leads to erosive radiographic progression.2 Structural damage is one of the most important consequences of active disease in patients with RA, which can lead to joint deformities and functional limitation.3 Sensitive imaging modalities need to be used to detect the destructive process as early as possible. Musculoskeletal ultrasonography (US) has become an important imaging technique in RA. It can detect both early inflammatory soft tissue lesions (eg, synovitis, tenosynovitis and bursitis) and early erosive bone lesions (eg, erosions) in arthritic joint diseases. Studies show good correlation between US and MRI in the detection of inflammatory soft tissue and erosive bone lesions.4–8 Synovitis plays an important role in the joint-destroying process. It has been shown that no bone destruction occurs in the absence of synovitis.9 ,10 Grey-scale US (GSUS) allows a differentiation between exudative and proliferative synovial changes as it provides a good soft-tissue contrast. The early detection of synovial effusion/proliferation is important in the diagnosis of early arthritis. Additionally, colour and power Doppler ultrasonography (PDUS) can differentiate between an active and inactive joint process,11 and between low disease activity and clinical remission.2 With PDUS, positive colour Doppler signals in patients with clinical DAS28 remission are predictive for a relapse in the next 12 months.12
Definition for sonographic findings according the OMERACT US consensus
Monitoring of disease activity and joint damage in RA is important, and standardisation is therefore essential. During the seventh OMERACT (Outcome Measures in Rheumatology Clinical Trials) conference, the typical RA findings detected by musculoskeletal US, including effusion, synovial hypertrophy/proliferation, tenosynovitis and erosion, were defined as follows13:
Effusion
Abnormal hypoechoic or anechoic intra-articular material that that can be displaced and compressed, but does not exhibit a Doppler signal.
Synovial hypertrophy/proliferation
Abnormal hypoechoic intra-articular tissue that is non-displaceable and poorly compressible and which may exhibit a Doppler signal.
Tenosynovitis
Hypoechoic or anechoic thickened tissue with or without fluid within the tendon sheath with possible signs of Doppler signals, which is seen in two perpendicular planes.
RA bone erosion
An intra-articular discontinuity of the bone surface that is visible in two perpendicular planes.
Standardised multiplanar scans have been developed for each joint region according to the guidelines of the European League Against Rheumatism (EULAR).14 Therefore, complete joint scanning is recommended. Furthermore, a dynamic examination is necessary in order to detect small collections of fluid. The additional use of PDUS helps in differentiating active and inactive synovial/tenosynovial processes, especially in small joints.
US findings can be scored using quantitative (mm) or semiquantitative (0–3) scoring systems to estimate the degree of synovial/tenosynovial and erosive processes. Additionally, US findings can be described on a binary (1/0) basis. International/European US expert groups are working towards the development of a standardised US scoring system which will reflect a patient's global disease activity. A scoring system for synovitis in RA which combines GSUS and PDUS findings in a semiquantitative (0–3) grading system is in the process of development, but consensus about the joint regions and the optimal (minimal) number of joints has not yet been reached. Therefore, an internationally available and acceptable musculoskeletal US scoring system is still needed.
Ultrasound (US) scoring systems
Different imaging scoring systems have been developed to standardise clinical findings. Several musculoskeletal US scores exist to monitor RA disease activity and the therapeutic response to DMARDs. Recently, Mandl et al systematically analysed US joint count and scoring systems to assess synovitis according to the OMERACT filter.15 In this paper different US scoring systems are described in chronological order, which assess synovitis, tenosynovitis/paratenonitis and erosions.
Wakefield et al7 described the first semiquantitative scoring system for the measurement of erosions as follows: normal: <2 mm; small erosion: 2 mm; moderate erosion: >2–<4 mm; large erosion: ≥4 mm. Their study examined 100 patients with RA, assessing metacarpophalangeal (MCP) joints II–V of the clinically most affected hand from ulnar, radial, palmar and dorsal aspects. Most erosion (73%) was found either from radial or ulnar aspects, and was more extensive than found by conventional radiography. Interobserver values were examined in this study and κ values between two observers were at least 0.76 for present/absent erosions.
Szkudlarek et al16 developed a four-step semiquantitative US grading system separately for joint effusion, synovial thickening, bone erosion and PD activity of five preselected small joints of 30 patients with RA (unilateral MCP II, III, proximal interphalangeal (PIP) II and metatarsophalangeal (MTP) I, II joints examined from a dorsal aspect). Joint effusion was defined as a compressible anechoic intracapsular area and the amount of fluid semiquantitatively scored as follows: grade 0: no effusion; grade 1: minimal amount; grade 2: moderate (without distension of the joint capsule); grade 3: extensive (with distension of the joint capsule). Synovial thickening was defined as a non-compressible hypoechoic intracapsular area scored as follows: grade 0: none; grade 1: minimal synovial thickening; grade 2: synovial thickening bulging over the line linking tops of the periarticular bones without extension along the bone diaphysis; grade 3: synovial thickening bulging over the line linking tops of the periarticular bones with extension to at least one of the bone diaphyses. Bone erosions were defined as follows: grade 0: normal bone surface; grade 1: bone surface irregularity without the defect being seen in two planes; grade 2: defect of the surface in two planes; grade 3: bone defect creating extensive bone destruction. Semiquantitative grading of the PD evaluation was as follows: grade 0: no flow; grade 1: single-vessel signals; grade 2: less than half of the area of the synovium filled with vessels; grade 3: more than half of the area of the synovium filled with vessels.
This group proved the reproducibility of this scoring system by examination of the interobserver reliability of two investigators with different backgrounds (rheumatologist vs radiologist). They found fair to good interobserver agreement rates (κ values from 0.48 to 0.68) for the identification of synovitis and bone erosions using this newly introduced semiquantitative scoring system. They came to the conclusion that US was a reproducible method for the examination of finger and toe joints of patients with RA.15
The first US synovitis sum scoring system for the evaluation of RA finger joints was introduced by Scheel et al.17 They examined semiquantitatively (0–3) and quantitatively (mm) the clinically most affected MCP II–V and PIP II–V joints of 46 patients with RA from dorsal and palmar aspects. Synovitis was defined as both synovial hypertrophy and synovial fluid. They found that synovitis was more frequently detected in the palmar proximal area of the MCP and PIP joints than from the dorsal view. There were no significant differences between semiquantitative (0–3) scores and quantitative measurements. The best results for the synovitis sum scores were achieved using the ‘s4’ (sum of MCP II–V and PIP II–V) and the ‘s3’ (sum of MCP II–IV, PIP II–IV) scores (area under the curve for each 0.9). Good results were also detected for the ‘s2’ (sum of MCP II–III, PIP II–III; area under the curve 0.85) score—the minimum number of examined joints count. Scheel et al concluded that examination of a reduced numbers of joints is preferable, in view of the shorter examination time required.
Naredo et al18 investigated the validity of reduced joint counts including large and small joints on both sides. A 12-joint score including bilateral wrist, MCPs II and III, PIPs II and III and knees was used to determine effusion, synovitis and PD activity, and correlated highly with a corresponding 60-joint score. The 12-joint score effectively reflected overall joint inflammation in patients with RA (at a patient level) and is therefore useful for monitoring treatment.
The new seven-joint ultrasound (US7) score was proposed by Backhaus et al.19 This is the first US composite scoring system, combining soft tissue lesions (synovitis and tenosynovitis/paratenonitis) and destructive processes (erosions) in a single scoring system. The US7 score includes US examination of the following joints of the clinically more affected side: wrist, MCP II and III, PIP II and III, MTP II and V. The joints are examined by GSUS and PDUS for synovitis and tenosynovitis/paratenonitis from a dorsal and palmar/plantar aspect, and for erosions from a dorsal, palmar/plantar and radial/lateral (only MCP II and MTP V) aspect. Synovitis in GSUS is analysed semiquantitatively (0–3) according to Scheel et al17: 0: absence; 1: small hypoechoic/anechoic line beneath the joint capsule (mild synovitis); 2: elevation of the joint capsule parallel to the joint area (moderate synovitis); 3: strong distension of the joint capsule (severe synovitis) (figures 1 and 3). The PDUS evaluation for synovitis (figures 2 and 4) and tenosynovitis/paratenonitis is scored according to the scoring system of Szkudlarek et al.16 Tenosynovitis/paratenonitis and erosions in GSUS are registered on a binary basis (0/1) (figures 5 and 6). The seven preselected joints are assessed in a standardised manner according to the EULAR guidelines.14
In the first multicentre evaluation, 120 patients were examined at three visits (baseline and after 3 and 6 months) before (baseline) and after onset of treatment or change of treatment using the US7 score. The clinical DAS28 score and laboratory parameters (C-reactive protein; erythrocyte sedimentation rate) were evaluated at the same visits. All parameters (US7 score, clinical and laboratory data) were reduced after 3 (except PDUS synovitis and erosion score) and 6 months (except erosion score). Clinical, laboratory and US parameters improved when different treatments were used—DMARDs and/or tumour necrosis factor (TNFα) inhibitors vs DMARDs alone. Furthermore, the US7 synovitis score correlated with DAS28 at 3 and 6 months. Backhaus et al concluded that the use of the US7 score would provide a fast overview of disease activity in daily rheumatological practice18 and would be helpful in monitoring treatment.
Ellegaard et al20 found that a standardised Doppler US examination of the wrist joint was useful in addition to clinical examination in RA. They examined 109 patients with RA and found significant correlations between colour Doppler activity and C-reactive protein, erythrocyte sedimentation rate, swollen joint count and DAS28, but not with the Health Assessment Questionnaire or tender joint count, concluding that one single affected joint can be used as a measure of disease activity.
Dougados et al21 assessed several US synovitis scores in comparison with clinical examination. In a multicentre study, 28 joints (DAS28), 20 joints (both MCP I–V and MTP I–V) and 38 joints (28 joints and MTP I–V) of patients with RA were ultrasonographically (GSUS and PDUS either binary (0/1) or a 0–3 grade, only dorsal view) examined during 4 months’ treatment with anti-TNF agents. They found that US evaluation as an outcome measure was no better than clinical examination.
Recently, Hammer et al22 published a study, in which a comprehensive US score including 78 joints was compared with reduced joint counts (7-, 12-, 28- and 44-joint scores) at different time points (baseline, 1, 3, 6, and 12 months) during adalimumab treatment. They found high correlation between the reduced joint scores and the 78-joint score at all examination time points (range 0.79–0.99 for GSUS, and 0.77–0.99 for PDUS; each p<0.001). The seven-joint count was the smallest joint count with a good correlation with the 78-joint count.
The different US scores are presented in chronological order in the table 1.
Current and future perspectives in the development of US scores
Musculoskeletal US scoring systems are used to show clinically detected RA disease activity and the therapeutic response to DMARDs. Semiquantitative (0–3) scores can be used to estimate the distension of the synovial/tenosynovial inflammation, and the degree of erosion for each examined joint (at a joint level), whereas US sum/composite scores of a reduced joint count reflect the overall RA disease activity in a short examination time (at a patient level).
In the development of US sum scores, several studies have analysed different numbers of joint scores, and all studies reached the conclusion that reduced sum scores provide a good reflection of overall inflammatory activity in RA.16–18 ,20 ,21 All the scoring systems presented assessed frequently affected joints in RA, but in most cases only inflammatory signs such as synovitis, effusion and PD activity were included (see table 1).
Up to now, the US7 score is the first and only score that combines examination of all typical RA pathologies (synovitis, tenosynovitis/paratenonitis and erosion) in a composite scoring system; thereby each feature (synovitis, tenosynovitis/paratenonitis, erosion) is described separately. The US7 synovitis score (in GSUS and PDUS) showed significant correlation with the clinical score DAS28 during treatment with DMARDs and/or TNFα inhibitors over a period of 6 months; the US7 score is therefore sensitive to change.18 In our opinion, therefore, the US7 score should be used in addition to DAS28, as differentiation between clinical and subclinical disease activity is possible only by US which enables future damage to be predicted.23 ,24
A recently published US7 reliability study investigated the interobserver and intraobserver reliability of sonographers, determining the US7 score under conditions similar to those of daily rheumatological practice. The median overall κ value for detecting synovitis was κ=0.51, for tenosynovitis/paratenonitis κ=0.57 and for erosions κ=0.45. This study also showed that the US7 score has the potential also in multicentre analysis to assess structural changes with moderate to good reliability.25 In a recently published study, 432 patients with RA were examined by the US7 score and different treatment groups were compared. Patients who switched from one biological agent to another exhibited a significant decline in the US7 erosion score after 12 months, leading to the conclusion that this score might differentiate between different therapeutic regimens.26
In summary, the US7 score is a feasible, reliable and responsive (to different therapeutic options) composite scoring system in daily rheumatological practice. Future evaluation should assess if additional joint regions should be included—for example, the dorsal part of the included finger joints MCP II, III and PIP II, III for the assessment of synovitis by GSUS, or even whether some joint regions should be excluded. US examination for synovitis is mostly carried out from the dorsal aspect of the finger joint, although Scheel et al17 showed that synovitis was most often detected in the palmar proximal area (86%) of the affected finger joints.17 Recently, Vlad et al27 came to the same conclusion. They showed that palmar US findings correlated more strongly with clinical scores (SDAI; CDAI) than dorsal US findings. Therefore, palmar US examinations should accompany the dorsal examination both in clinical practice and in clinical trials.27 Those results might explain the observations of Dougados et al,21 who found that US evaluation of synovitis (only assessed from the dorsal aspect) as an outcome measurement was no better than clinical examination.
For further analysis of the US7 score, its use in cohorts of patients with early RA needs to be evaluated to prove its efficacy in predicting therapeutic response to different treatment regimens (DMARDs/biological agents). The outcome of the disease might also be predicted by US, so that it could play a role as a ‘biomarker’. This point has been discussed for other sensitive imaging modalities.28
Current activities of the EULAR/OMERACT US task force include the development of a Global OMERACT Sonography Scoring (GLOSS) system in RA. Its feasibility, sensitivity to change and value over standard clinical care are being tested (data not yet published). GLOSS examines a number of small joints for synovitis, and the results evaluated by GSUS and PDUS are combined in this scoring system.29
In summary, the implementation of musculoskeletal US as a patient-friendly, reliable, bedside method in daily rheumatological practice is helpful and is essential for the objective examination of joints assumed to be affected. The use of a representative US sum/composite score, in which mostly active joint regions are included, reduces examination time and, at the same time, reflects a patient's overall disease activity. Currently, the use of US is being studied mostly for RA, but its use in other diseases (eg, psoriatic arthritis) will follow.
References
Footnotes
-
Competing interests None.
-
Provenance and peer review Commissioned; externally peer reviewed.