Sacroiliitis-like and degenerative sacroiliac joint MRI findings in the large population-based cohort SHIP.
TL;DR
Small, focal BME-like and degenerative SIJ findings are common on this non-dedicated whole-body protocol, and protocol-specific imaging background frequencies do not estimate axSpA prevalence or diagnostic accuracy; whole-body and dedicated SIJ MRI are not interchangeable.
Abstract
Objectives
To determine the protocol-specific frequency, extent, and anatomic distribution of sacroiliitis-like and degenerative sacroiliac joint MRI findings in a large population-based cohort.
Materials And Methods
This retrospective cross-sectional analysis used prospectively acquired SHIP-Trend-0 whole-body MRI (2008-2012). Of 4,420 participants, 2,187 underwent MRI; 124 exclusions for missing PD-FS or insufficient image quality left 2,063 evaluable (age, 48.8 ± 14.6 years; 999 men). Non-dedicated axial PD-FS was the primary sequence, with coronal TIRM supported lesion assessment. Two readers recorded bone marrow edema (BME)-like signal, extent, quadrants, structural/degenerative findings, and joint-space width.
Results
BME-like signal was present in 225/2,063 (10.9%; 95% CI 9.6-12.3%): women, 12.1%; men, 9.6%; unadjusted odds ratio (OR), 1.30 (95% CI 0.98-1.72; p = 0.067). Maximum depth was 9.84 ± 3.53 mm; 85/225 (37.8%; 95% CI 31.4-44.5%) had depths ≥ 10 mm. Mean slice extent was 2.91 ± 0.82. Among 222 location-coded participants, 156 (70.3%; 95% CI 63.8-76.2%) had one quadrant, predominantly superior sacral. Bilateral signal occurred in 52/2,060 (2.5%; 95% CI 1.9-3.3%). Any structural or degenerative finding occurred in 349/2,063 (16.9%; 95% CI 15.3-18.6%).
Conclusion
Small, focal BME-like and degenerative SIJ findings are common on this non-dedicated whole-body protocol. These protocol-specific imaging background frequencies do not estimate axSpA prevalence or diagnostic accuracy; whole-body and dedicated SIJ MRI are not interchangeable. Isolated signal requires morphologic and clinical context.