MRI of the Sacroiliac Joints in Athletes: Do Semi-axial Slices Added to Standard Semi-coronal Scans Facilitate Recognition of Non-specific Bone Marrow Edema?

Ulrich Weber¹, Anne Grethe Jurik ², Anna Zejden ², Ejnar Larsen ³, Steen Hylgaard Jørgensen ⁴, Kaspar Rufibach ⁵, Christian Schioldan ⁶ and Søren Schmidt-Olsen ⁷, ¹Danish Hospital for Rheumatic Diseases, University Hospital of Southern Denmark, Sonderborg, Denmark, ²Department of Radiology, Aarhus University Hospital, Aarhus, Denmark, ³Department of Radiology, North Denmark Regional Hospital, Hjørring, Denmark, ⁴Center for Clinical Research, North Denmark Regional Hospital, Hjørring, Denmark, ⁵Division of Biostatistics, F. Hoffmann-La Roche, Basel, Switzerland, ⁶Department of Physiotherapy, Clinic Benefit, Frederikshavn, Denmark, ⁷Department of Rheumatology and Sports Medicine, North Denmark Regional Hospital, Hjørring, Denmark

Meeting: 2019 ACR/ARP Annual Meeting

Keywords: axial spondyloarthritis, bone marrow lesions and diagnostic imaging, Magnetic resonance imaging (MRI)

Session Information

Date: Sunday, November 10, 2019

Title: 3S083: Spondyloarthritis Including Psoriatic Arthritis – Clinical I: Imaging in Axial Spondyloarthritis & Psoriatic Arthritis (851–856)

Session Type: ACR Abstract Session

Session Time: 2:30PM-4:00PM

Background/Purpose: Low grade bone marrow edema (BME) on magnetic resonance imaging (MRI) of the sacroiliac joints (SIJ) is challenging the discrimination between patients with early axial spondyloarthritis (SpA) and mechanical back pain. In our previous analysis, recreational and elite athletes had on average 3-4 SIJ quadrants with BME, and 30-41% met the Assessment of SpondyloArthritis international Society (ASAS) definition of active sacroiliitis [1]. Potential conditions simulating SIJ BME such as vascular partial volume effect or anatomical SIJ variants could not be explored due to standard semi-coronal MRI scans only. Our goals by assessing combined semi-axial and semi-coronal SIJ MRI scans in 2 cohorts of young athletes were to explore the frequency and topography of non-specific BME by 2 perpendicular MRI planes, its association with 4 constitutional SIJ features, and potential limitation of false positive assignments of ASAS-defined sacroiliitis by standard semi-coronal scans alone.

Methods: Combined semi-axial and semi-coronal SIJ MRI scans of 20 recreational runners before/after running and 22 elite ice-hockey players were evaluated by 3 blinded readers for BME and its association with 4 constitutional SIJ features: vascular partial volume effect, deep iliac ligament insertion, fluid-filled bone cyst, and lumbosacral transitional anomaly. Scans of TNF-treated SpA patients served to mask readers. Pre-test reader agreement for BME on semi-axial scans was calculated by single measure, absolute agreement intra-class correlation coefficient (ICC). We analysed distribution and topography of BME and associated constitutional SIJ features across 8 anatomical SIJ regions (upper/lower ilium/sacrum, subdivided in anterior/posterior slices) descriptively, as concordantly recorded by ≥2/3 readers on both MRI planes. BME confirmed on both scans was compared with previous evaluation of semi-coronal MRI alone which met the ASAS definition for active sacroiliitis.

Results: ICC agreement among 3 readers for semi-axial SIJ BME was 0.86 (0.72-0.95). Perpendicular semi-axial and semi-coronal MRI scans confirmed SIJ BME consistently in 25% and 27% of athletes, preferentially in the anterior upper sacrum. BME associated with 4 constitutional SIJ features was observed in 20-36% of athletes, clustering in the posterior lower ilium. The proportion of ASAS-positive sacroiliitis recorded on semi-coronal plane alone decreased by 30-50% upon amending semi-axial scans (from 30-35% to 20% in runners, from 41% to 18% in ice-hockey players).

Conclusion: Semi-axial combined with standard semi-coronal scans in MRI protocols for sacroiliitis facilitated recognition of non-specific BME, which clustered in the posterior lower ilium in association with constitutional SIJ features, and in the anterior upper sacrum. The proportion of false-positive ASAS assignments of sacroiliitis recorded on semi-coronal plane alone could be substantially reduced by amending semi-axial scans.
Reference. [1] Weber U et al. Arthritis Rheum 2018;70:736.

Figure 1. SIJ MRI with semi-coronal -upper row- and semi-axial -lower row- scans. Scout lines on the left panel visualize the orientation of the 2 perpendicular planes. Arrows indicate subchondral BME in corresponding anatomic location on both planes in a patient with axial SpA.

Table 1

Table 1. Frequency of sacroiliac joint BME on 2 perpendicular MRI planes versus on standard semi-coronal scans fulfilling the ASAS definition of active sacroiliitis, and frequency of BME associated with 4 constitutional sacroiliac joint features. n -%-: number of subjects -percentage- showing a given feature as concordantly reported by ≥2/3 readers.

Disclosure: U. Weber, None; A. Jurik, None; A. Zejden, None; E. Larsen, None; S. Jørgensen, None; K. Rufibach, None; C. Schioldan, None; S. Schmidt-Olsen, None.

To cite this abstract in AMA style:

Weber U, Jurik A, Zejden A, Larsen E, Jørgensen S, Rufibach K, Schioldan C, Schmidt-Olsen S. MRI of the Sacroiliac Joints in Athletes: Do Semi-axial Slices Added to Standard Semi-coronal Scans Facilitate Recognition of Non-specific Bone Marrow Edema? [abstract]. Arthritis Rheumatol. 2019; 71 (suppl 10). https://acrabstracts.org/abstract/mri-of-the-sacroiliac-joints-in-athletes-do-semi-axial-slices-added-to-standard-semi-coronal-scans-facilitate-recognition-of-non-specific-bone-marrow-edema/. Accessed .

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/mri-of-the-sacroiliac-joints-in-athletes-do-semi-axial-slices-added-to-standard-semi-coronal-scans-facilitate-recognition-of-non-specific-bone-marrow-edema/