Background/Purpose : To assess the intra- and inter-reader agreement and responsiveness of a novel scoring system for FOI-assessed synovitis.

Methods: FOI of were obtained of both wrists and hands of 46 RA patients inducing or escalating anti-rheumatic therapy and who had ≥1 clinically swollen joint in the hand at baseline, after 3 and 6 months’ follow-up. The hands were placed in the FOI-unit and the patient received a bolus of i.v. ICG-Pulsion (1mg/kg body weight) 10 seconds after starting the examination, which obtained 1 image/second over 6 minutes. The image-sets were anonymized and randomized and were assessed for synovial pathology at the wrist, 1^st-5^th metacarpophalangeal (MCP), 1^st interphalangeal (IP) and 2^nd-4^th proximal interphalangeal (PIP) joint levels in both hands by two readers blinded to patient data but not chronology. The readers performed a calibration session before the exercise. The images of 23 patients were re-anonymized and were assessed as an intra-reader analysis. The scoring system for synovitis was based on the theory that inflamed tissue would demonstrate a more rapid enhancement than surrounding tissues. For each joint, the images were assessed sequentially from start of injection of ICG-Pulsion to peak enhancement. Synovial pathology was defined as a sharply margined enhancement with clear integrity from surrounding tissues and correct anatomical location lasting ≥3 seconds. The thickness of the pathology fulfilling these criteria were measured in the transverse plane of the hand at the 3^rd second of enhancement and were scored as follows: 0: no enhancement, 1: ˂1/3, 2: ≥1/3 but ˂2/3, 3: ≥2/3 of joint thickness. Descriptive statistics and the Wilcoxon signed-rank test were used to assess change in score over time. Intra-/inter-reader for status and change scores were assessed using single measure intra-class correlation coefficients (ICC) and smallest detectable change (SDC, change scores only). Responsiveness was assessed using standardized response mean (SRM).

Results: The median (IQR) change in total synovitis score between baseline and 3/6 months’ follow-up were -5.0 (-10.0;-1.0)/-8.0 (-13.5;-3.0) (p<0.01). Intra- and inter-reader ICC were good to very good for status and change scores at all joint levels and for total scores (Table 1). The SDC were generally low and for the inter-reader SDC, 56%/60% of the patients had a change larger than the SDC between baseline and 3/6 months respectively. The mean SRM for total change scores between baseline and 3/6 months’ follow-up were moderate to good (0.7/0.8).

Conclusion: The novel FOI RA synovitis scoring system showed high reliability and moderate to good responsiveness in the hands. Future studies should focus on comparing the sensitivity and specificity of FOI with ultrasound and magnetic resonance imaging.

Table 1