Background/Purpose: Fat metaplasia has been shown to follow resolution of inflammation in axial SpA and this may be evident within one year (1). Fat metaplasia at vertebral corners has also been shown to predict the development of new syndesmophytes (2). Consequently, the scoring of fat lesions in the spine may constitute both an important measure of treatment efficacy as well as a responsive surrogate for new bone formation. Because of this, we developed and validated a new scoring system for fat lesions in the spine, the CanDen Fat SpA Spine Score (FASSS), which addresses the localization and phenotypic diversity of fat lesions in SpA.

Methods: In 2007, the Canada-Denmark MRI working group developed anatomy-based definitions of fat lesions on T1 weighted sagittal MRI scans of the spine (3). In 2011, further definitions, a reference image module, and an online spinal unit schematic for data entry were developed, which formed the basis for a scoring method. The method comprises six different types of fat lesions defined according to their anatomical location, which are recorded dichotomously (present/absent) at each vertebral endplate from C2 lower to S1 upper (scoring range per disco-vertebral unit in C spine: 0-8, and in T and L spine: 0-18). Two rheumatologists assessed spine MRI scans obtained at two time points (mean 1.5 years (SD 0.5)) in chronological order of 66 patients with axial SpA (exercise 1). Scoring methodology and reference image set were revised based on discrepant cases in exercise 1. Then the two readers re-read 30 randomly selected pairs of MRI scans from the first exercise together with 40 new pairs of MRI scans (exercise 2) (mean 1.7 years (SD: 0.8)) (exercise 2).  Inter-observer reliability were assessed by intra-class correlation coefficients (ICCs), and responsiveness as mean change per year by standardized response mean (SRM).

Results: In exercise 2, the change in fat scores ranged from -34 to 52 for reader A and from -38 to 61 for reader B. Inter-observer ICC scores were high to very high for the FASSS baseline scores, and improved substantially in change scores from exercise 1 to 2 (Table). This was particularly notable for the 30 patients evaluated in both exercises. Inter-observer ICCs for baseline scores were high for all spinal segments, and change scores improved from small to moderate (C spine), and from moderate to high (L spine) and very high (T spine). For total FASSS score the mean change/year and SRM were 2.4 and 0.34 for reader A and 3.7 and 0.26 for reader B.

Conclusion: The FASSS meets essential validation criteria for further assessment in axial SpA, and may thus be useful for follow-up of SpA in clinical trials and practice.

References: 1. Chiowchanwisawakit et al. ARD 2010; 2. Chiowchanwisawakit et al. AR 2011; 3. Østergaard et al. J Rheum 2009

Table

Interobserver ICCs for spinal segments scored according to the CanDen Fat SpA Spine Score (FASSS)

*The 30 patients, that were scored two times, were a subgroup of patients in exercise 1 and 2.