Background The objective detection and quantification of inflammatory disease activity is critical for achieving optimal therapy results. Fluorescence optical imaging (FOI) is a novel modality designed for imaging the hands and wrists, and the automated quantification of the ensuing scans using DACT (Disease ACTivity)-FOI is a novel algorithm for analyzing these images.
Objectives To determine the utility of DACT-FOI in the assessment of hand and wrist inflammation.
Methods Bilateral finger and wrist joints (n=1360) of 40 patients with inflammatory arthritis were studied. Synovitis was defined as tender and swollen joints on clinical examination, presence of synovial thickening/effusion and intra-articular Doppler signals on ultrasound (MSUS), and abnormal focal optical signal intensities on FOI, respectively. The DACT score used an automatically generated algorithm of the composite images (of 240 frames per second) to achieve a quantified score for each patient. Using dedicated image parameters and size correction, the enhanced pixels were extracted automatically from the image background, and the high signal intensities calculated. The DACT-FOI formula was based on fluorescence intensity curve thresholds that were used to discriminate intensity variations, and then divided by the 95th centile of intensities in normal individuals as the reference value. DACT-FOI ≤1 was referred to as normal digital activity signals. Subclinical synovitis was defined as being clinically non-inflamed but inflamed on MSUS.
Results Out of the 1360 joints evaluated, 215 (16%) were inflamed clinically, 329 (24%) by MSUS, and 347 (26%) by FOI. For overall hand and wrist disease activity (n=40), the number (mean ± SD) of active joints detected by clinical, MSUS and semi-quantitative FOI was 5.4±7.0; 8.2±7.8; and 8.7±7.8, respectively. The automated digital activity (±SD) calculation by DACT-FOI was 3.8 (±2.1). Correlations of high statistical significance was denoted as ** when p<0.01. A strong positive correlation (r =0.458**; p=0.003) between clinical synovitis and DACT-FOI was demonstrated. The mean DACT values also correlated significantly with MSUS (r =0.442**; p=0.004) and semi-quantitative FOI (r =0.439**; p=0.005). There was a highly significant correlation of synovitis detection between clinical examination and MSUS (r =0.730**; p=0.000) and between clinical examination and semi-quantitative FOI (r =0.577**; p=0.000). Agreement between MSUS and FOI in synovitis detection was good, and revealed strong correlations (0.816**; p=0.000). Out of the non-inflamed joints by clinical examination, 142/1145 (12%) were inflamed by MSUS, of which 102/142 (72%) were also inflamed by FOI. Thus, for detecting subclinical synovitis, the sensitivity, specificity, and positive and negative predictive values of FOI were 72% (102/142), 93% (934/1003), 77% and 91%, respectively.
Conclusions FOI and the automated analysis DACT-FOI were technically feasible with high reproducibility and strong agreement with clinical scoring. Therefore, this objective digitally quantified measurement of inflammatory disease activity in the hands & wrists may be useful both in diagnosis and in monitoring the effects of clinical therapy.
Disclosure of Interest None declared