Session Information
Session Type: Abstract Submissions (ACR)
Background/Purpose: The definition of a ‘positive’ or ‘negative’ MRI of the sacroiliac joints (MRI-SIJ) according to ASAS is recommended for use in daily practice1. However, in several clinical trials the Spondyloarthritis Research Consortium of Canada (SPARCC)-score is used2. It would be useful to know which SPARCC-score is the best cut-off value as surrogate for the ASAS definition. Therefore, we investigated which SPARCC cut-off value best matches the ASAS definition for a positive MRI-SIJ.
Methods: All MRI-SIJs of two time points of patients included in the SPondyloArthritis Caught Early (SPACE)-cohort in the Leiden University Medical Center (LUMC) were scored independently by 3 readers. The readers, blinded to the time sequence, scored the MRI-SIJs according to the ASAS definition (ASAS-pos) and the SPARCC-score. An MRI-SIJ was marked ASAS-pos if 2/3 readers scored positive. In this analysis, mean SPARCC-scores of the readers that also scored ASAS-pos for that particular case were used. Cross-tabs were used to analyse agreement between several SPARCC-score cut-off values (≥1, ≥2, ≥3 and ≥4) and ASAS-pos, which served as external standard in this comparison.
Results: All available MRI-SIJs were used (n=238 in total; n=148 baseline MRI-SIJs; n=90 follow-up MRI-SIJs). The results of the different tested SPARCC cut-off values are presented in the table. A SPARCC cut-off value ≥1 resulted in 39 (16.4%) false-positive classifications and a cut-off value ≥2 resulted in 11 (4.6%) and both cut-off values had no false-negative classifications compared to ASAS-pos. A SPARCC cut-off value ≥3 resulted in 1 (0.4%) false-positive classification and 4 (1.7%) false-negative classifications, and a SPARCC cut-off value ≥4 resulted in 9 (3.8%) false-negative and 1 (0.4%) false-positive classifications. We found very similar results if baseline MRI-SIJs and follow-up MRI-SIJs were analysed separately.
Conclusion: SPARCC cut-off values of ≥2, ≥3 or ≥4 have all high percentages of correctly classified patients (95.4%, 97.9% and 95.8%, respectively). A SPARCC cut-off value ≥3 shows the most balanced misclassification and the highest agreement with the ASAS definition for a positive MRI-SIJ.
References:
1Rudwaleit M et al. ARD 2009;68:1520-7
2Maksymowych W et al. A&R 2005;53:703-9
|
|
ASAS negative, n (%) |
ASAS positive, n (%) |
|
SPARCC <1, n (%) |
143 (60.1) |
0 (0.0) |
|
SPARCC ≥1, n (%) |
39 (16.4) |
56 (23.5) |
|
Agreement |
|
83.6% |
|
SPARCC <2, n (%) |
171 (71.9) |
0 (0.0) |
|
SPARCC ≥2, n (%) |
11 (4.6) |
56 (23.5) |
|
Agreement |
|
95.4% |
|
SPARCC <3, n (%) |
181 (76.1) |
4 (1.7) |
|
SPARCC ≥3, n (%) |
1 (0.4) |
52 (21.8) |
|
Agreement |
|
97.9% |
|
SPARCC <4, n (%) |
181 (76.1) |
9 (3.8) |
|
SPARCC ≥4, n (%) |
1 (0.4) |
47 (19.7) |
|
Agreement |
|
95.8% |
Disclosure:
R. van den Berg,
None;
M. de Hooge,
None;
V. Navarro-Compán,
None;
M. Reijnierse,
None;
F. van Gaalen,
None;
T. Huizinga,
None;
D. van der Heijde,
None.
« Back to 2012 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/a-spondyloarthritis-research-consortium-of-canada-score-cut-off-%e2%89%a53-as-best-match-for-the-assessment-of-spondyloarthritis-international-society-definition-of-a-positive-mri-of-the-sacroiliac-jo/