Evaluation of a novel 7-joint ultrasound score in daily rheumatologic practice: A pilot project

Authors

  • M. Backhaus,

    1. Charité-University Medicine Berlin, Charitéplatz 1, D-10117 Berlin, Germany
    Search for more papers by this author
    • Drs. M. Backhaus and Ohrndorf contributed equally to this work.

    • Dr. M. Backhaus has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from Esaote, Pfizer, Abbott, Wyeth, Essex, Roche, BMS, and Novartis.

  • S. Ohrndorf,

    Corresponding author
    1. Charité-University Medicine Berlin, Charitéplatz 1, D-10117 Berlin, Germany
    • Department of Rheumatology and Clinical Immunology, Charité-Medicine Berlin, Charitéplatz 1, D-10117 Berlin, Germany
    Search for more papers by this author
    • Drs. M. Backhaus and Ohrndorf contributed equally to this work.

  • H. Kellner,

    1. Hospital Neuwittelsbach, Romanstrasse 9, D-80639 Munich, Germany
    Search for more papers by this author
  • J. Strunk,

    1. Hospital Porz am Rhein, Urbacher Weg 19, D-51149 Cologne, Germany
    Search for more papers by this author
    • Dr. Strunk has received consultant fees, speaking fees, and/or honoraria (less than $10,000) from Abbott.

  • T. M. Backhaus,

    1. Charité-University Medicine Berlin, Charitéplatz 1, D-10117 Berlin, Germany
    Search for more papers by this author
  • W. Hartung,

    1. Asklepios Klinikum Bad Abbach, D-93077 Bad Abbach, Germany
    Search for more papers by this author
    • Dr. Hartung has received speaking fees (less than $10,000) from Abbott Immunology.

  • H. Sattler,

    1. Parkklinik Bad Dürkheim, Salinenstrasse 19, D-67098 Bad Dürkheim, Germany
    Search for more papers by this author
  • K. Albrecht,

    1. Benekestrasse 2-8, D-61231 Bad Nauheim, Germany
    Search for more papers by this author
    • Dr. Albrecht has received speaking fees (less than $10,000) from Abbott.

  • J. Kaufmann,

    1. Dachsweg 36-38, D-14974 Ludwigsfelde, Germany
    Search for more papers by this author
    • Dr. Kaufmann has received consultant fees, speaking fees, and/or honoraria (more than $10,000) from Abbott Immunology.

  • K. Becker,

    1. Ulmerstrasse 26, D- 89143 Blaubeuren, Germany
    Search for more papers by this author
  • H. Sörensen,

    1. Argentinische Allee 42, D-14163 Berlin, Germany
    Search for more papers by this author
    • Dr. Sörensen has received speaking fees (less than $10,000) from Abbott Immunology.

  • L. Meier,

    1. Reifenberger Strasse 6, D-65719 Hofheim, Germany
    Search for more papers by this author
  • G. R. Burmester,

    1. Charité-University Medicine Berlin, Charitéplatz 1, D-10117 Berlin, Germany
    Search for more papers by this author
  • W. A. Schmidt

    1. Medical Center for Rheumatology Berlin-Buch, Lindenberger Weg 19, D-13125 Berlin, Germany
    Search for more papers by this author
    • Dr. Schmidt has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from Esaote, General Electrics, Merck (Germany), Merck (US), Pfizer, Abbott, Wyeth, Essex, Roche, BMS, Medac, Ibsen, Actelion, Novartis, and Lilly.


Abstract

Objective

To introduce a new standardized ultrasound score based on 7 joints of the clinically dominant hand and foot (German US7 score) implemented in daily rheumatologic practice.

Methods

The ultrasound score included the following joints of the clinically dominant hand and foot: wrist, second and third metacarpophalangeal and proximal interphalangeal, and second and fifth metatarsophalangeal joints. Synovitis and synovial/tenosynovial vascularity were scored semiquantitatively (grade 0–3) by gray-scale (GS) and power Doppler (PD) ultrasound. Tenosynovitis and erosions were scored for presence. The scoring range was 0–27 for GS synovitis, 0–39 for PD synovitis, 0–7 for GS tenosynovitis, 0–21 for PD tenosynovitis, and 0–14 for erosions. Patients with arthritis were examined at baseline and after the start or change of disease-modifying antirheumatic drug (DMARD) and/or tumor necrosis factor α (TNFα) inhibitor therapy 3 and 6 months later. C-reactive protein level, erythrocyte sedimentation rate, rheumatoid factor, anti–cyclic citrullinated peptide, Disease Activity Score in 28 joints (DAS28), and radiographs of the hands and feet were performed.

Results

One hundred twenty patients (76% women) with rheumatoid arthritis (91%) and psoriatic arthritis (9%) were enrolled. In 52 cases (43%), erosions were seen in radiography at baseline. Patients received DMARDs (41%), DMARDs plus TNFα inhibitors (41%), or TNFα inhibitor monotherapy (18%). At baseline, the mean DAS28 was 5.0 and the synovitis scores were 8.1 in GS ultrasound and 3.3 in PD ultrasound. After 6 months of therapy, the DAS28 significantly decreased to 3.6 (Δ = 1.4), and the GS and PD ultrasound scores significantly decreased to 5.5 (−32%) and 2.0 (−39%), respectively.

Conclusion

The German US7 score is a viable tool for examining patients with arthritis in daily rheumatologic practice because it significantly reflects therapeutic response.

INTRODUCTION

Tumor necrosis factor α (TNFα) inhibitors have opened a new era in the treatment of rheumatic diseases such as rheumatoid arthritis (RA), psoriatic arthritis, and ankylosing spondylitis. Followup studies investigating their clinical and radiographic effects have shown that TNFα inhibitor therapy, especially in combination with conventional disease-modifying antirheumatic drug (DMARD) therapy, leads to significant improvements in clinical status and significant inhibition of radiographic progression (1–5). A reliable imaging modality enabling visualization of early inflammatory changes is warranted in order to stop destructive processes as early as possible.

Musculoskeletal ultrasound (US) is a valuable imaging tool in the detection of soft tissue lesions and early erosive bone changes. Due to technical improvements with the subsequent development of high-resolution transducers, US has become an established method in the evaluation of changes in superficial musculoskeletal structures, and is being increasingly used in rheumatologic practice and research.

Compared with clinical examination, gray-scale US (GSUS) is a more sensitive method for detecting synovitis and tenosynovitis. US and magnetic resonance imaging (MRI) correlate well in terms of detecting superficial erosive bone processes, and are more sensitive than conventional radiography (6–8). By US, the highest number of erosions is detected in the lateral view of the second metacarpophalangeal (MCP) (9) and fifth metatarsophalangeal (MTP) joints (10).

On developing a novel synovitis scoring system, it was found that synovitis is detected more frequently (in 86% of affected joints) in the palmar proximal area of the MCP and proximal interphalangeal (PIP) joints of RA patients than from the dorsal side, where it was detected in only 14% of affected joints without palmar synovitis. The best results for combined joint counts were achieved using the second through fifth MCP and PIP joints (s4) and the second through fourth MCP and PIP joints (s3) methods (area under the curve [AUC] of 0.90 for both), although similarly good results were seen with the second and third MCP and PIP joints (s2) method (AUC of 0.85). By the area under the receiver operating characteristic curves, the accuracy of the tests that separate groups into those with and without the disease in question is measured. These results allowed for the presence of synovial hypertrophy and synovial fluid to be combined in a single scoring system (11).

Another synovitis scoring system differentiates between synovial thickening and synovial effusion (12). Synovial tissue vascularity can be detected with power Doppler US (PDUS) (13). Increased synovial vascularization represents greater inflammatory activity in RA.

US has become a routinely available bedside method with high patient acceptability. In contrast to MRI, there are no limiting factors such as pacemakers, metal, or claustrophobia, and all of the peripheral joints can be examined as often as necessary. Repeatability of US examination enables accurate clinical evaluation because short-term monitoring is possible. However, it might be time consuming when applied to too many joints. Therefore, a novel 7-joint US score (German US7 score) for use in daily rheumatologic practice was developed. The main focus of this project was to analyze its value in daily rheumatologic practice under real-time conditions. Another objective of this project was to assess the value of this score in the detection of disease activity and therapeutic response.

PATIENTS AND METHODS

This prospective national project started in November 2006. One hundred twenty patients underwent US examinations at 3 visits (baseline and after 3 and 6 months) at 23 sites. Seven joints of the clinically dominant hand and foot were sonographically evaluated after the onset of therapy or change of actual therapy (DMARDs and/or TNFα inhibitors). For each patient, the hand and foot that was clinically more affected by tenderness and/or swelling had been chosen for US examination. Several parameters were assessed.

Clinical assessment

At baseline and after 3 and 6 months, 28 joints (bilateral PIPs, MCPs, wrists, elbows, shoulders, and knees) were clinically assessed for swelling and tenderness. The visual analog scale for disease activity (range 0–100 mm) was also reported at each patient's visit. In addition, the following data were recorded on report sheets: year of birth, sex, height, weight, onset of typical symptoms, current rheumatologic therapy including DMARDs, TNFα inhibitors, and nonsteroidal antiinflammatory drugs, and glucocorticoid dose at each visit.

Disease activity assessment

The Disease Activity Score in 28 joints (DAS28) was used to assess disease activity at each visit. Disease activity was graded by the following classification criteria: DAS28 ≤2.6 = clinical remission, ≤3.2 = mild disease activity, ≤5.2 = moderate disease activity, and >5.2 = severe disease activity (14).

Laboratory evaluation

C-reactive protein (CRP) level (normal level <5 mg/liter) and erythrocyte sedimentation rate (ESR; normal level <20 mm/hour) were obtained at each visit. IgM rheumatoid factor (RF; normal level <24 IU) and anti–cyclic citrullinated peptide (anti-CCP) antibodies (normal level <20 units/ml) were assessed at baseline.

US examination

The wrist, MCP2 and MCP3, and PIP2 and PIP3 joints, as well as the MTP2 and MTP5 joints of the clinically dominant side, were sonographically examined in a standardized manner according to German (15) and European League Against Rheumatism (EULAR) (16) guidelines. All joint regions were assessed by GSUS and PDUS (where available). PDUS was used in 80% of patients.

GSUS

GSUS was performed as follows: the wrist was examined for synovitis and tenosynovitis from the dorsal, palmar, and ulnar aspects (Figure 1). In the dorsal aspect, the probe was parallel to the extensor digitorum tendons (dorso-median). For the palmar wrist examination, the probe was placed parallel to the median nerve (palmo-median), and for the ulnar aspect, the probe was set parallel to the extensor carpi ulnaris tendon.

Figure 1.

Gray-scale ultrasound (US) and power Doppler (PD) US performed for synovitis, tenosynovitis/paratenonitis, and erosions from the dorsal, palmar, and ulnar aspects of the wrist, metacarpophalangeal (MCP), proximal interphalangeal (PIP), and metatarsophalangeal (MTP) joints.

MCP2 and MCP3 joints were evaluated for synovitis and tenosynovitis from the palmar view, and for paratenonitis from the dorsal aspect. Erosions were detected from the dorsal, palmar, and radial (MCP2 joint) aspects, or from the dorsal and palmar aspects (MCP3 joint). PIP2 and PIP3 joints were assessed for synovitis from the palmar aspect, and for erosions from the dorsal and palmar aspects (Figure 1).

MTP2 and MTP5 joints were examined for synovitis from the dorsal aspect, and erosions were detected from the dorsal and plantar aspects (MTP2 joint) and from the dorsal, plantar, and lateral (MTP5 joint) aspects (Figure 1).

Synovitis by GSUS was analyzed semiquantitatively (0 = absence, 1 = mild, 2 = moderate, and 3 = severe synovitis). Grade 1 describes a small hypoechoic/anechoic line beneath the joint capsule. In grade 2, the joint capsule is elevated parallel to the joint area. Grade 3 characterizes a strong distension of the joint capsule (Figure 2).

Figure 2.

Synovitis by gray-scale ultrasound. a, grade 0, b, grade 1, c, grade 2, d, grade 3.ft = flexor tendon; pp = phalanx proximalis; * = synovitis; bar = joint capsule elevation.

Tenosynovitis/paratenonitis and erosions were registered as being absent (0) or present (1). Erosion was defined as an interruption of the bone surface in 2 perpendicular planes. Tenosynovitis is a hypoechoic or anechoic thickened tissue with or without fluid within the tendon sheath (17). Paratenonitis is identified in US as an echo-poor halo around the tendon and often shows increased vascularity on Doppler imaging (18).

PDUS

PDUS was performed for synovitis and tenosynovitis/paratenonitis from the palmar and dorsal aspects in each joint region evaluated, except for MTP joints from the plantar (Figure 1). The semiquantitative findings of PDUS activity for synovitis and tenosynovitis/paratenonitis were scored as follows: grade 0 = no intraarticular color signal, grade 1 = up to 3 color signals or 2 single and 1 confluent signal in the intraarticular area, grade 2 = greater than grade 1 to <50% of the intraarticular area filled with color signals, and grade 3 = ≥50% of the intraarticular area filled with color signals (19) (Figure 3).

Figure 3.

Synovitis by power Doppler ultrasound. a, grade 0, b, grade 1, c, grade 2, d, grade 3. t = tendon; r = radius; s = synovitis; l = lunate bone.

Based on these results, a 7-joint US score (US7) was performed, including the sum of synovitis scores in the GSUS (0–27) and PDUS (0–39) modes, tenosynovitis/paratenonitis in the GSUS (0–7) and PDUS (0–21) modes, and erosions (0–14) in the GSUS mode. US examination of each patient took 10–20 minutes, including documentation.

Radiographic evaluation

According to standard rheumatologic criteria, posteroanterior scans of the patients' hands and forefeet were recorded at baseline by conventional radiography. The presence of erosions was evaluated by the Steinbrocker score (20).

US inter- and intraobserver reliability

Thirty readers took part in the US reliability study. The reading of 2 specialists in musculoskeletal US (MB and WAS) was used as the US imaging gold standard. Thirty-three stored US images were scored for synovitis in GSUS, synovitis in PDUS, tenosynovitis, and erosions under blinded conditions. Analysis was performed on a yes/no basis and also semiquantitatively (grade 0–3) for synovitis in GSUS and PDUS. For intrareader testing, 4 US images were shown twice.

Statistical analysis

Statistical analysis was performed with SPSS statistical software, version 15.0 (SPSS, Chicago, IL). For quantitative parameters (e.g., number of patients, age of examined patients, and their disease activity), we used the mean ± SD and range. Significant changes were calculated by the 2-sided exact Wilcoxon's test. P values less than 0.05 were considered statistically significant. Correlations between changes of the different examination modalities (DAS28 and US) throughout followup were evaluated by 2-sided exact Spearman's correlation coefficients.

Inter- and intrareader agreement was calculated using kappa coefficients between the readers. The kappa coefficients were divided as follows: <0.0 = poor, 0–0.20 = slight, 0.21–0.40 = fair, 0.41–0.60 = moderate, 0.61–0.80 = substantial, and 0.81–1.0 = almost perfect agreement (21). Percentage agreements were also calculated.

RESULTS

Characteristics of patients

One hundred twenty patients (76% women) with a mean ± SD age of 55 ± 14 years (range 21–80) and a mean ± SD disease duration of 8.3 ± 9.3 years (range 0.2–47.6) were examined at 3 visits (baseline and 3 and 6 months later). Ninety-one percent had RA and 9% had psoriatic arthritis. In 52 cases (43%), radiographic erosions were seen in the hands and/or forefeet at baseline. On US evaluation of the dominant hand and forefoot, bone erosions were detected in 43% of the 120 patients. At inclusion, 72% of this group was RF positive and 64% were anti-CCP positive.

Medication

A total of 79% of the patients received a prednisolone equivalent, with a mean daily dosage of 6.1 mg. They received either DMARDs (n = 49 [41%]), a combination of DMARDs plus TNFα inhibitors (n = 49 [41%]), or TNFα inhibitor monotherapy (n = 22 [18%]). The following DMARDs were administered: methotrexate, leflunomide, hydroxychloroquine, sulfasalazine, cyclosporine, and aurothiomalate (only 1 patient). Concerning the group receiving TNFα inhibitors, 84% of these patients received adalimumab, 13% received etanercept, and 3% received infliximab.

US, clinical, and laboratory parameters

US, clinical, and laboratory data are shown in Table 1 for the entire patient group. The mean synovitis score in GSUS was 8.1 at baseline, and it had significantly decreased to 5.5 (−32.1%; P < 0.05) after 6 months of followup. The mean synovitis score in PDUS was 3.3 before changing actual therapy, and after 6 months it had significantly decreased to 2.0 (−39.4%; P < 0.05). The mean qualitatively evaluated number of erosions remained the same over 6 months (2.6 at both visits; P = 0.94). A significant reduction in mean DAS28 was observed over 6 months (from 5.0 to 3.6; P < 0.05). The mean ESR was reduced from 30.4 mm/hour at baseline to 21.4 mm/hour after 6 months (P < 0.05). The mean CRP level also decreased significantly (from 22.8 to 8.0 mg/liter; P < 0.05).

Table 1. US, clinical, and laboratory data (n = 120)*
US scores/disease activityBaselineAfter 3 monthsAfter 6 months
  • *

    Values are the mean ± SD (range). US = ultrasound; GSUS = gray-scale US; max = maximum; PDUS = power Doppler US; DAS28 = Disease Activity Score in 28 joints; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein.

  • P < 0.05 (2-sided exact significance by Wilcoxon's test).

Synovitis score by GSUS (max 27)8.1 ± 5.8 (0–25)6.1 ± 5.5 (0–22)5.5 ± 5.5 (0–20)
Synovitis score by PDUS (max 39)3.3 ± 4.0 (0–16)2.4 ± 3.6 (0–20)2.0 ± 3.5 (0–18)
Tenosynovitis/paratenonitis score by GSUS (max 7)1.3 ± 1.8 (0–7)0.6 ± 1.2 (0–7)0.6 ± 1.3 (0–7)
Tenosynovitis/paratenonitis score by PDUS (max 21)0.8 ± 1.7 (0–7)0.4 ± 0.9 (0–4)0.2 ± 0.8 (0–4)
Erosions score (max 14)2.6 ± 3.6 (0–14)2.6 ± 3.9 (0–14)2.6 ± 3.6 (0–14)
DAS285.0 ± 1.3 (2–8)3.6 ± 1.4 (1–8)3.6 ± 1.3 (1–7)
ESR, mm/hour30.4 ± 20.3 (2–88)21.9 ± 17.2 (2–96)21.4 ± 17.9 (2–90)
CRP level, mg/liter22.8 ± 31.6 (0–153)9.7 ± 16.1 (0–94)8.0 ± 12.6 (0–62)

US, clinical, and laboratory data for the subgroups in relation to their medical treatment (only DMARD therapy in comparison with TNF inhibitor therapy alone or in combination with DMARDs) are shown in Tables 2 and 3. Comparing the effects of the different therapies, both clinical and ultrasonographic parameters were significantly improved in both groups.

Table 2. US, clinical, and laboratory data of 49 patients (41%) treated with disease-modifying antirheumatic drugs*
US scores/disease activityBaselineAfter 3 monthsAfter 6 months
  • *

    Values are the mean ± SD (range). See Table 1 for definitions.

  • P < 0.05 (2-sided exact significance by Wilcoxon's test).

Synovitis score by GSUS (max 27)6.7 ± 4.8 (0–18)4.7 ± 5.1 (0–20)4.1 ± 4.7 (0–16)
Synovitis score by PDUS (max 39)2.1 ± 2.9 (0–10)1.8 ± 2.7 (0–12)1.0 ± 2.0 (0–10)
Tenosynovitis/paratenonitis score by GSUS (max 7)1.5 ± 1.9 (0–7)0.8 ± 1.2 (0–4)0.5 ± 1.2 (0–4)
Tenosynovitis/paratenonitis score by PDUS (max 21)0.7 ± 1.8 (0–7)0.3 ± 0.8 (0–3)0.1 ± 0.7 (0–4)
Erosions score (max 14)1.3 ± 2.5 (0–9)1.2 ± 2.5 (0–10)1.3 ± 2.4 (0–9)
DAS284.8 ± 1.5 (2–7)3.3 ± 1.3 (1–8)3.4 ± 1.2 (2–6)
ESR, mm/hour31.7 ± 21.5 (4–79)21.2 ± 14.8 (3–60)20.9 ± 15.4 (2–60)
CRP level, mg/liter28.2 ± 38.1 (0–152)8.3 ± 8.8 (0–32)9.0 ± 13.8 (0–60)
Table 3. US, clinical, and laboratory data of 71 patients (59%) treated with tumor necrosis factor α inhibitors with or without disease-modifying antirheumatic drugs*
US scores/disease activityBaselineAfter 3 monthsAfter 6 months
  • *

    Values are the mean ± SD (range). See Table 1 for definitions.

  • P < 0.05 (2-sided exact significance by Wilcoxon's test).

Synovitis score by GSUS (max 27)8.8 ± 6.2 (0–25)6.8 ± 5.5 (0–22)6.0 ± 5.8 (0–20)
Synovitis score by PDUS (max 39)3.8 ± 4.1 (0–16)2.8 ± 4.0 (0–20)2.3 ± 4.1 (0–18)
Tenosynovitis/paratenonitis score by GSUS (max 7)1.1 ± 1.6 (0–7)0.6 ± 1.1 (0–7)0.5 ± 1.2 (0–7)
Tenosynovitis/paratenonitis score by PDUS (max 21)0.9 ± 1.7 (0–7)0.4 ± 1.0 (0–4)0.3 ± 0.84 (0–3)
Erosions score (max 14)3.1 ± 3.8 (0–13)3.3 ± 4.2 (0–14)3.1 ± 4.0 (0–14)
DAS285.0 ± 1.3 (2–8)3.7 ± 1.5 (1–8)3.6 ± 1.4 (1–7)
ESR, mm/hour30.1 ± 20.0 (2–88)21.9 ± 18.8 (2–96)22.6 ± 20.5 (2–90)
CRP level, mg/liter19.1 ± 27.1 (0–153)10.8 ± 19.8 (0–94)7.9 ± 12.5 (0–62)

Longitudinal correlation between US and clinical parameters

There were significant correlations between changes in the US parameters for synovitis and the DAS28 changes through 3 months (GSUS/DAS28 and PDUS/DAS28: each r = 0.44, P < 0.05) and 6 months of followup (GSUS/DAS28: r = 0.38, P < 0.05; PDUS/DAS28: r = 0.31, P < 0.05). There was also a significant correlation for tenosynovitis/paratenonitis (GSUS and PDUS) and DAS28 changes through 6 months of followup (r = 0.26 and r = 0.24, respectively; P < 0.05). There were no significant correlations between US parameters for tenosynovitis/paratenonitis (GSUS and PDUS) and DAS28 changes through 3 months of followup (r = 0.11). There were no significant correlations between the change in erosions and the DAS28 changes through 3 and 6 months of followup (r = 0.11 and r = 0.16, respectively).

Inter- and intraobserver reliability of the US assessment

The mean kappa value of 0.65 (mean interreader agreement 86%) was reached for the qualitative (yes/no basis) scoring of all stored images by 30 US readers. For the semiquantitative (0–3) scoring of the stored images for synovitis in GSUS and PDUS, the mean kappa value was 0.60 (mean interreader agreement 71%).

With regard to different pathologies, the US readers obtained a kappa of 0.62 (interreader agreement 90.2%) for the qualitative reading of synovitis in GSUS, and a kappa of 0.84 (interreader agreement 93.1%) for the qualitative PDUS analysis. For the semiquantitative results of synovitis in GSUS, a kappa of 0.55 (interreader agreement 66%) was achieved, and in PDUS, a kappa of 0.67 was achieved (interreader agreement 75.7%). An interreader agreement of 64.3% for tenosynovitis in GSUS was obtained. In the detection of erosions, this group attained a kappa of 0.56 (interreader agreement 76.7%).

The following results were achieved for intrareader reliability: a mean kappa value of 0.83 (intrareader agreement 93.7%) was obtained for qualitative intrareader testing, and a mean kappa value of 0.64 (intrareader agreement 84.5%) was obtained for semiquantitative intrareader testing of 4 images with signs of synovitis in GSUS, PDUS, and erosions.

DISCUSSION

Musculoskeletal US has become an imaging method that is routinely used in the daily practice of many rheumatologists for both disease activity and therapy monitoring, despite the fact that US examination can be a very time-consuming imaging modality, especially if multiple joints need to be scanned repeatedly. That is why we initiated a novel 7-joint US score (German US7 score). This score examines 7 joints (wrist, MCP2, MCP3, PIP2, PIP3, MTP2, and MTP5 joints) of the clinically dominant hand and foot of patients with inflammatory rheumatologic disease, especially RA, because this score predominantly includes the joints that are most affected by RA.

To the best of our knowledge, the German US7 score is the first score that combines soft tissue changes such as synovitis and tenosynovitis as well as erosive bone lesions in a single US scoring system. PDUS also helps to differentiate active from inactive synovial proliferation. Evaluating synovitis semiquantitatively using GSUS and PDUS also allows for the degree of disease activity to be estimated.

Naredo et al examined US scores including effusion, synovitis, and PD activity with different numbers of joints. They found out that a 12-joint (both MCP2 and MCP3 joints, PIP2 and PIP2 joints, wrists, and knees) score correlated well with a 60-joint score and with clinical and laboratory parameters. They concluded that a reduced US score of 12 joints effectively reflects overall joint inflammatory activity in RA (22).

Loeuille et al developed a new semiquantitative US synovitis score (ScUSI) for both GSUS and PDUS evaluating 7 joint regions (wrist, MCP2, MCP3, MCP5, MTP2, MTP3, and MTP5 joints) and compared ScUSI results with the radiographic Sharp score and the disease activity parameter DAS28. They found US synovitis to be a better predictive factor in the detection of destructive joint processes than the DAS28 (23). The ScUSI does not include US examination of PIP joints. Scheel et al found that extensive synovitis (grade 3) was detected in PIP joints more often than in MCP joints (11), and that this difference was significant (P < 0.002).

Hensch et al showed that GSUS and PDUS are precise tools in detecting and monitoring synovial processes in patients with active RA during treatment with TNFα inhibitors. This group used a semiquantitative synovitis score including the examination of 8 joints of 4 fingers called s4 (second through fifth MCP and PIP joints [11]) that reflected changes in inflammatory processes earlier than DAS28 and that also correlated well with the DAS28 and CRP level (24).

In the development of US scores for erosions, Chary-Valckenaere et al evaluated a semiquantitative US score (grade 0–3) including MCP2, MCP3, MCP5, MTP2, MTP3, and MTP5 joints (ScUSST) that correlated well with the radiographic Sharp score. In this study, they also found that the MTP5 joint was the most frequently involved joint, followed by the MCP2 joint (10). These joints are also included in the German US7 score.

This project was initiated in order to investigate the value of the German US7 score in daily rheumatologic practice by assessing its ability to reflect patients' disease activity and therapeutic response. It was possible to show a significant decrease in the German US7 score including parameters such as synovitis (GSUS and PDUS) and tenosynovitis in comparison with DAS28, ESR, and CRP level under immunosuppressant therapy (DMARD and/or TNF inhibitors).

This study also shows a significant correlation between changes in synovitis seen in GSUS and PDUS and between DAS28 changes through 3 and 6 months. Filippucci et al describe a slight positive correlation between PDUS and the DAS28 in a study that examines PDUS in comparison with clinical (and laboratory) parameters in RA patients treated with adalimumab (25).

The results obtained with the German US7 score appear to indicate that this method is more sensitive than the clinical instrument DAS28 in the characterization of inflammatory joint processes. Although the DAS28 was similar in both treatment groups at baseline (4.8 versus 5.0, according to moderate disease activity), the synovitis scores in GSUS and PDUS were higher in the TNF inhibitor (plus DMARD) therapy group at baseline.

In a recent large study including 367 patients receiving TNFα inhibitor therapy, Naredo et al showed that PDUS parameters significantly improved parallel to the disease activity parameter DAS28 and laboratory parameters. This group also established that PDUS might be of predictive value in relation to the changes detected by radiographs (26).

Thus far, we do not know how bone lesions detected by the US7 score develop in comparison with bone lesions seen in conventional radiography. The number of erosive cases at baseline was similar in both GSUS and conventional radiography. Therefore, it might be interesting to observe how erosive bone lesions detected by the 2 imaging methods will develop after one year, especially with different treatment (DMARDs versus TNFα inhibitors with or without DMARDs). Throughout the followup periods of 3 and 6 months, we did not find any changes in erosions detected by GSUS independent from therapy (DMARDs alone versus TNF inhibitors alone or in combination with DMARDs).

The fact that patients were examined by different US observers using different US machines may present a limitation to this study. US is still considered to be operator dependent. However, Scheel et al were able to show moderate to good interreader agreement in the first interobserver reliability study performed by 14 experts of the EULAR working group (27). Confirming these findings in a larger study, Naredo et al also found moderate to good interobserver reliability between 23 European musculoskeletal US experts (28). D'Agostino et al developed a learning curve for ultrasonographers with little or no experience in the assessment of synovitis in MCP, PIP, and MTP joints that shows that at least 70 examinations were necessary to develop US competence in assessing synovitis in small joints in patients with RA (29).

Musculoskeletal US is an essential part of rheumatologic specialist training in Germany, which requires at least 300 examinations before being admitted to the final examination. Therefore, most of the US participants are well experienced in musculoskeletal US. Furthermore, compact atlases have been compiled for this project in order to explain the German US7 score by means of US images showing typical pathologic findings included in this score. Moreover, participants were trained in specialized courses before starting to include patients in this project.

In order to verify inter- and intraobserver agreement of the arthrosonographic examiners taking part in this project, we performed a reliability study. The results of this study showed moderate to substantial kappa values. Hence, US findings evaluated by different examiners taking part in this project might be comparable.

Concerning intermachine reliability, D'Agostino et al recently showed that there is good reliability between experts using different types of US machines. Repeated measurements of the same joint by 2 different sonographers or by the same sonographers on 2 different machines would increase the reliability coefficient (analog to kappa) to greater than 0.65, i.e., patients relate to the greatest part of variability (30).

In conclusion, the German US7 score is a valuable tool for the US examination of inflamed joint activity in rheumatologic disease, especially in RA. The German US7 score is suitable for monitoring therapy in daily practice. Concentration on a small number of active joint regions reduces examination time (approximately 10 to 20 minutes), thereby making it possible to integrate the US7 score in daily rheumatologic practice. It is an appropriate supplementary instrument to DAS28 for gaining information about the activity of joint processes.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Ohrndorf had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. M. Backhaus, Ohrndorf, Strunk, Sattler, Albrecht, Burmester, Schmidt.

Acquisition of data. M. Backhaus, Ohrndorf, Kellner, T. M. Backhaus, Hartung, Sattler, Albrecht, Kaufmann, Becker, Sörensen, Meier, Schmidt.

Analysis and interpretation of data. M. Backhaus, Ohrndorf, Sattler, Burmester, Schmidt.

ROLE OF THE STUDY SPONSOR

Abbott GmbH & Co. KG, Germany, supported the logistic procedure of this ultrasound project. Abbott agreed to submit the manuscript for publication and approved the manuscript, but did not have any influence on the study design, data analysis, or writing of the manuscript.

Acknowledgements

We thank Peter Martus, PhD, Department of Medical Statistics, Charité University Hospital, Berlin, Germany, and the US and clinical research investigators involved in this study.

Ancillary