TY - JOUR
T1 - Reliable semiquantitative whole-joint MRI score for the shoulder joint
T2 - The shoulder osteoarthritis severity (SOAS) score
AU - Jungmann, Pia M.
AU - Gersing, Alexandra S.
AU - Woertler, Klaus
AU - Dietrich, Tobias J.
AU - Baum, Thomas
AU - Baumann, Frederic
AU - Bensler, Susanne
N1 - Publisher Copyright:
© 2018 International Society for Magnetic Resonance in Medicine
PY - 2019/6
Y1 - 2019/6
N2 - Background: Shoulder osteoarthritis causes severe pain and functional disability. Preventive surgical procedures aiming to halt the progression of degenerative changes are increasingly applied. However, no MRI-based score exists that may be applied for scoring of osteoarthritic changes and their progression. Purpose: To establish a semiquantitative MRI-based shoulder osteoarthritis severity (SOAS) evaluation system and to test its reliability. Study Type: Retrospective. Subjects: A total of N = 60 patients (73.2 ± 7.3 years; 30/60 female) was included; n = 15 subjects for each of the four radiographic grades of osteoarthritis (Samilson score 0 to 3). Assessment: Based on the MRIs of the shoulder, the SOAS scoring system was created. All MRIs were assessed by six readers. The severity of degeneration was evaluated for: rotator cuff, labral-bicipital-complex, cartilage, osseous findings, joint capsule, and acromion. The total SOAS score ranged between 0 (absence of osteoarthritis) and 100 (most severe osteoarthritis). SOAS scores were correlated with radiographic Samilson, Hamada and Kellgren–Lawrence (KL) gradings. Statistics: Pearson correlations, t-tests, receiver operating characteristics (ROC) and interclass correlation coefficients (ICC). Results: Intra- (ICC = 0.99) and interreader agreement (ICC 0.96–0.98) for the total SOAS score was excellent. The range of SOAS scores was from 1 to 88. SOAS correlated significantly with radiographic Samilson and KL scores (R = 0.82, P < 0.001), but not with Hamada scores (R = –0.07, P = 0.60). The highest correlations with Samilson scores were found for cartilage (R = 0.82, P < 0.001) and osseous findings (R = 0.86, P < 0.001). SOAS scores were significantly different between different Samilson grades (Samilson 0, 13.4 ± 7.6; Samilson 1, 26.0 ± 9.1; Samilson 2, 38.2 ± 19.2; Samilson 3, 65.5 ± 13.0; P < 0.05). The ability of the SOAS score to predict incident radiographic shoulder OA (KL grade ≥2) was excellent (AUC = 0.91; P < 0.001). Data Conclusion: The newly developed semiquantitative MRI-based SOAS score represents the severity of global shoulder OA and structure-specific shoulder degeneration with excellent reliability in a standardized manner and may therefore be helpful in MRI research studies of the shoulder. Level of Evidence: 3. Technical Efficacy: Stage 2. J. Magn. Reson. Imaging 2018.
AB - Background: Shoulder osteoarthritis causes severe pain and functional disability. Preventive surgical procedures aiming to halt the progression of degenerative changes are increasingly applied. However, no MRI-based score exists that may be applied for scoring of osteoarthritic changes and their progression. Purpose: To establish a semiquantitative MRI-based shoulder osteoarthritis severity (SOAS) evaluation system and to test its reliability. Study Type: Retrospective. Subjects: A total of N = 60 patients (73.2 ± 7.3 years; 30/60 female) was included; n = 15 subjects for each of the four radiographic grades of osteoarthritis (Samilson score 0 to 3). Assessment: Based on the MRIs of the shoulder, the SOAS scoring system was created. All MRIs were assessed by six readers. The severity of degeneration was evaluated for: rotator cuff, labral-bicipital-complex, cartilage, osseous findings, joint capsule, and acromion. The total SOAS score ranged between 0 (absence of osteoarthritis) and 100 (most severe osteoarthritis). SOAS scores were correlated with radiographic Samilson, Hamada and Kellgren–Lawrence (KL) gradings. Statistics: Pearson correlations, t-tests, receiver operating characteristics (ROC) and interclass correlation coefficients (ICC). Results: Intra- (ICC = 0.99) and interreader agreement (ICC 0.96–0.98) for the total SOAS score was excellent. The range of SOAS scores was from 1 to 88. SOAS correlated significantly with radiographic Samilson and KL scores (R = 0.82, P < 0.001), but not with Hamada scores (R = –0.07, P = 0.60). The highest correlations with Samilson scores were found for cartilage (R = 0.82, P < 0.001) and osseous findings (R = 0.86, P < 0.001). SOAS scores were significantly different between different Samilson grades (Samilson 0, 13.4 ± 7.6; Samilson 1, 26.0 ± 9.1; Samilson 2, 38.2 ± 19.2; Samilson 3, 65.5 ± 13.0; P < 0.05). The ability of the SOAS score to predict incident radiographic shoulder OA (KL grade ≥2) was excellent (AUC = 0.91; P < 0.001). Data Conclusion: The newly developed semiquantitative MRI-based SOAS score represents the severity of global shoulder OA and structure-specific shoulder degeneration with excellent reliability in a standardized manner and may therefore be helpful in MRI research studies of the shoulder. Level of Evidence: 3. Technical Efficacy: Stage 2. J. Magn. Reson. Imaging 2018.
KW - magnetic resonance imaging
KW - osteoarthritis
KW - shoulder
UR - http://www.scopus.com/inward/record.url?scp=85052641268&partnerID=8YFLogxK
U2 - 10.1002/jmri.26251
DO - 10.1002/jmri.26251
M3 - Article
C2 - 30079543
AN - SCOPUS:85052641268
SN - 1053-1807
VL - 49
SP - e152-e163
JO - Journal of Magnetic Resonance Imaging
JF - Journal of Magnetic Resonance Imaging
IS - 7
ER -