Session Information
Date: Sunday, October 21, 2018
Title: 3S082 ACR Abstract: Imaging of Rheumatic Diseases I: MRI & CT (863–868)
Session Type: ACR Concurrent Abstract Session
Session Time: 2:30PM-4:00PM
Background/Purpose:
Bone marrow lesion (BML) in hip osteoarthritis (OA) may be a determinant of prognosis and a target for emerging anti-inflammatory therapies. We developed the OMERACT Hip MRI Scoring System (HIMRISS) to maximize reliability & sensitivity to change. Innovations include electronic overlays (touch or mouse-click a region to identify BML), a web-based online interface, and real-time iterative calibration (RETIC) in which new readers perform a scoring exercise online, learning by observing their ICC vs. expert consensus displayed in real-time. Pre-specified minimum ICC must be achieved before becoming eligible to proceed with formal scoring. We tested feasibility, reliability, and responsiveness of HIMRISS for scoring BML in a large prospective cohort.
Methods:
In HIMRISS, BML is scored as 0/1 (absent/present) in 100 regions on a web-based grid overlay slice-by-slice in coronal hip MRI. In the Edmonton Steroid Injection in Hip OA (STIHO) cohort, we had n=90 patients x 2 hips x 2 MRI: at baseline presenting for hip steroid injection, and at 8 weeks post injection. Coronal T1/STIR sequences were provided, blinded to time point, to 8 readers: 2 radiology fellows, 3 rheumatologists, 3 musculoskeletal radiologists. Each reader prepared by reviewing a comprehensive HIMRISS methodology module, then had to pass an online RETIC training module with proficiency (ICC 0.80 vs. experts) prior to formal scoring. Readers each scored HIMRISS BML in all 360 hips.
Results:
Subjects were 56% male, age 59±13 (mean±SD) years, symptom duration 4.4±4.5 years, K-L OA grade 1/2/3/4 in 20%/28%/34%/18%. Reliability (ICC single measure, absolute agreement, 2 way model) was high for baseline BML status for all readers (0.83) and for 2 experts (0.88-0.92). Change was small, mean (SD) 2.1±10.0/100. Despite this, ICC for change was (0.42-0.76) for all readers, (0.56-0.81) for experts.
Conclusion:
HIMRISS reliably quantifies BML in hip OA at a single time point and can responsively discriminate small changes in BML. Novice readers trained via an online RETIC system achieved reliability close to that of experts. Future studies will assess performance of HIMRISS assessing inflammation in other arthropathies.
|
Reliability of HIMRISS BML Measurements |
Baseline Status |
Change |
||
|
ICC for BML |
All readers |
Experts |
All readers |
Experts |
|
Femoral head |
0.83 (0.65-0.91) |
0.91 (0.87-0.94) |
0.76 (0.65-0.83) |
0.81 (0.72-0.87) |
|
Acetabulum |
0.83 (0.74-0.89) |
0.88 (0.82-0.92) |
0.42 (0.23-0.57) |
0.56 (0.40-0.68) |
|
Total |
0.83 (0.55-0.92) |
0.92 (0.88-0.95) |
0.72 (0.60-0.81) |
0.78 (0.68-0.85) |
|
Change in BML, and HIMRISS Responsiveness |
Scoring range |
Change observed 0-8 wks; mean(SD) |
Smallest detectable change SDC |
Hips with change >SDC |
|
Femoral head |
0-65 |
2.2(9.2) |
3.6 |
45% |
|
Acetabulum |
0-35 |
-0.1(2.3) |
1.9 |
33% |
|
Total |
0-100 |
2.1(10.0) |
4.6 |
36% |
To cite this abstract in AMA style:
Jaremko JL, Lambert RG, Pedersen SJ, Weber U, Lindsay D, Al-Ani Z, Pianta M, Wichuk S, Steer K, Paschke J, Maksymowych WP. International Multi-Reader Validation of the Outcome Measures in Rheumatology Hip MRI Scoring System (OMERACT HIMRISS) for Bone Marrow Lesions in Osteoarthritis [abstract]. Arthritis Rheumatol. 2018; 70 (suppl 9). https://acrabstracts.org/abstract/international-multi-reader-validation-of-the-outcome-measures-in-rheumatology-hip-mri-scoring-system-omeract-himriss-for-bone-marrow-lesions-in-osteoarthritis/. Accessed .« Back to 2018 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/international-multi-reader-validation-of-the-outcome-measures-in-rheumatology-hip-mri-scoring-system-omeract-himriss-for-bone-marrow-lesions-in-osteoarthritis/