Article Text

SAT0204 Do Bone Marrow Edema Lesions in the Sacroiliac Joint Change into Fatty Lesions over A 1-Year Period in Patients with Axial Spondyloarthritis or Possible Spondyloarthritis
  1. P. Bakker,
  2. R. van den Berg,
  3. M. de Hooge,
  4. F. van Gaalen,
  5. M. Reijnierse,
  6. T. Huizinga,
  7. D. van der Heijde
  1. Leiden University Medical Center, Leiden, Netherlands


Background Bone marrow edema (BME) lesions in the sacroiliac joint (SIJ) may change into fatty lesions over time. Fatty lesions are regarded as the earliest sign of chronic changes as a consequence of inflammation but are sometimes also found healthy controls.1,2 At the moment, there is little information on the course over time in patients without treatment of TNF-inhibitors.

Objectives To investigate whether BME lesions in the sacroiliac joint change into fatty lesions over 1 year time in patients with axial SpondyloArthritis (axSpA) or possible SpA and to evaluate the volatility of both lesions general.

Methods Patients in the SPACE cohort (back pain: ≥3 months, ≤2 years, onset <45 years) with (suspicion of) axSpA underwent MRI of the SIJ at baseline and 1-year follow-up (n=76). MRIs were scored independently by 3 well-calibrated readers, blinded for time sequence and patient characteristics (STIR and MRI T1-weighted sequences viewed simultaneously). The presence of BME and fatty lesions was scored. Both lesions were defined present if 1 lesion was seen on ≥2 consecutive slices or if >1 lesion was seen on a single slice and scored in 4 quadrants per SI joint. For the two readers individually, scores of baseline and 1 year were compared on quadrant level (Q). The sum of all 8 Qs was calculated to obtain total scores per patient.

Results 76 patients were completed in the analyses (number of Q=608); of which 39 (51%) were classified axSpA at baseline (ASAS classification criteria), and 37 (41%) as possible SpA. BME or fatty lesions at any time point was found in 27/76 (36%) and 20/76 (26%) patients respectively (reader 1, reader 2). Reader 1 indicated no lesions at baseline or follow-up in 519/608 Qs (85%) and in 45/608 Qs (7%) a change in lesions was visible (new lesions occurred, former lesions disappeared or the type of lesions changed over time). Similar trends were visible amongst the two readers (table). In both readers, fatty lesions newly occurred in 8/76 patients (11%) over 1 year time with agreement on 5 patients. In 3 patients, a transition of BME to fat occurred. On the other hand, fatty lesions disappeared in 2/76 patients (3%, reader 1) and in 5/76 patients (7%, reader 2). In reader 2, concomitant occurrence and disappearance of fatty lesions was observed in 2/76 (3%) of patients. In reader 1, this was not seen.

Conclusions About one third of the patients showed BME/fatty lesions in the SIJ at any time point. There was not much change in site and/or type of lesions visible over 1-year period of time. Fatty lesions occurred more frequently de novo than in quadrants with previous BME and fatty lesions and can also resolve over time. The incremental value of fatty lesions in the SIJ needs to be further re-evaluated. These data show that fatty lesions are not always markers of previous inflammation.


  1. P. Chiowchanwisawakit ARD 2010;69:262.

  2. Song ARD 2011;70:1257.

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.4096

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.