Article Text

FRI0147 Data Driven Lesion-Based Criteria for Defining A Positive Sacroiliac Joint MRI in 2 Cohorts with Axial Spondyloarthritis
  1. U. Weber1,
  2. R.G. Lambert2,
  3. S.J. Pedersen3,
  4. S.M. Chan4,
  5. V. Zubler5,
  6. K. Rufibach6,
  7. M. Ostergaard3,
  8. Z. Zhao7,
  9. W.P. Maksymowych1
  1. 1Rheumatology
  2. 2Radiology, University of Alberta, Edmonton, Canada
  3. 3Rheumatology, University of Copenhagen, Copenhagen, Denmark
  4. 4Ophthalmology, University of Alberta, Edmonton, Canada
  5. 5Radiology, University of Zurich, Zurich
  6. 6Biostatistics, rePROstat, Basel, Switzerland
  7. 7Rheumatology, PLA General Hospital, Beijing, China


Background The ASAS definition of a positive sacroiliac joint (SIJ) MRI for axial spondyloarthritis (SpA) is based exclusively on bone marrow edema (BME). The contextual information of concomitant BME and/or erosion on T1SE and STIR sequences led to the MORPHO definition, which improved sensitivity without worsening specificity compared to the ASAS definition in a non-radiographic axial SpA (nr-axSpA) cohort of short symptom duration.

Objectives To compare MORPHO and ASAS definitions for a positive SIJ MRI with global assessment by expert readers in patients with a broad spectrum of nr-axial SpA. To determine optimal lesion-based definitions for a positive SIJ MRI based on BME and/or erosion.

Methods The sample comprised 2 independent cohorts A/B of 69/88 consecutive back pain patients ≤50 years, median symptom duration 1.3/10.0 years, newly referred to 2 university clinics for suspected SpA (A) or acute anterior uveitis (AAU) and back pain (B), and 20 healthy controls. Patients were classified according to rheumatologist expert opinion based on clinical examination, pelvic radiography and laboratory values as having nr-axSpA (n=51), AS (n=34), or MBP (n=72). SIJ MRI were assessed independently by 4 blinded readers according to standardized modules in which readers first recorded presence/absence of SpA by global evaluation of T1SE and STIR sequences, and subsequently reported whether BME or erosion were present for each SIJ quadrant on all MRI slices. We calculated sensitivity and specificity as mean values over 4 readers for global assessment, ASAS and MORPHO definitions. We derived candidate definitions based on the number of SIJ quadrants with BME and/or erosion to determine which might be optimal in both cohorts.

Results For both cohorts, global assessment and “≥2 SIJ quadrants with erosion” showed the highest specificity. The MORPHO definition had increased sensitivity compared to the ASAS definition, with a slight drop in specificity. “BME in ≥3 SIJ quadrants” improved specificity over ASAS or MORPHO definitions although the combined criterion “≥3 BME and/or ≥2 erosions” showed both a higher sensitivity and specificity.

Table 1.

Sensitivity and specificity of candidate criteria for a positive SIJ MRI in SpA (mean over 4 readers): nr-axSpA versus MBP

Conclusions A stringent lesion-based definition of a positive SIJ MRI, based on both BME and erosion, performed best for classification of axial SpA.

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.3413

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.