Article Text

PDF
OP0171 A data driven approach to the definition of a positive MRI of the sacroiliac joints in spondyloarthritis
  1. U. Weber1,
  2. V. Zubler2,
  3. S.J. Pedersen3,
  4. K. Rufibach4,
  5. R.G. Lambert5,
  6. S. Chan6,
  7. M. Ostergaard3,
  8. W.P. Maksymowych7
  1. 1Rheumatology
  2. 2Radiology, Balgrist University Hospital, Zurich, Switzerland
  3. 3Rheumatology, University of Copenhagen, Copenhagen, Denmark
  4. 4Social and Preventive Medicine, University of Zurich, Zurich, Switzerland
  5. 5Radiology
  6. 6Ophthalmology
  7. 7Rheumatology, University of Alberta, Edmonton, Canada

Abstract

Background Inflammation on magnetic resonance imaging (MRI) of the sacroiliac joints (SIJ) in patients with spondyloarthritis (SpA) is a major criterion in the Assessment of SpondyloArthritis (ASAS) classification criteria for axial SpA, which are based on expert clinical opinion as gold standard. The definition of a positive SIJ MRI in the ASAS criteria was generated by consensus among experts. Studies using a data-driven approach to defining a positive SIJ MRI are scarce.

Objectives To generate candidate definitions for a positive MRI based on confidence in the diagnosis of SpA by expert MRI readers.

Methods The study population comprised 220 consecutive patients with back pain ≤50 years newly referred to 2 university clinics, and 79 healthy controls. Patients were classified according to clinical examination and pelvic radiography as having non-radiographic axial SpA (nr-axSpA) (n=74), ankylosing spondylitis (n=60), or mechanical back pain (n=86). SIJ MRI were assessed independently in random order by 4 blinded expert readers according to a standardized module that records SpA features (bone marrow edema (BME), erosion (ER), fat infiltration (FI)) on MRI. Readers recorded their level of confidence in the diagnosis of SpA by global evaluation of the MRI scan on a 0-10 scale (0 = definitely not SpA; 10 = definite SpA). Assignment of a definitive diagnosis of SpA was pre-specified as the majority (at least 3 of 4) of readers recording definite SpA with a confidence level of 8-10. Absence of SpA required all 4 expert readers to record non-SpA (≤4 on 0-10 scale). We estimated the type and extent of involvement according to number of affected SIJ quadrants attaining specificity of 90% and 95% for SpA and corresponding sensitivity. The corresponding area under the curve (AUC) was computed by receiver operating curves using the number of affected SIJ quadrants.

Results 86 subjects met the MRI criterion of definite SpA (≥3 readers scoring ≥8 on 0-10 scale of confidence), and 151 subjects had no SpA (score of ≤4 by all 4 readers).

Table 1. Attained specificity, corresponding sensitivity, cut-off for number of affected SIJ quadrants for a pre-defined specificity of 0.90 and of 0.95, and AUC

Conclusions ER in 1 SIJ quadrant or BME in 3 SIJ quadrants yielded a specificity of over 90% for a positive SIJ MRI and the combination increased sensitivity, compared to ASAS definition, without reducing specificity. This combination represents the best candidate definition for a positive SIJ MRI in SpA for incorporation into future classification criteria.

Disclosure of Interest None Declared

Statistics from Altmetric.com

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.