Background/Purpose: Previous studies in systemic inflammatory joint diseases have suggested that Fluorescence Optical Imaging (FOI) is a reliable and sensitive tool for the detection of altered microcirculation as a proxy for inflammation. Inflammation is commonly present in hand osteoarthritis (OA), and plays a role for symptoms and disease progression. Feasible, sensitive and reliable tools to evaluate inflammation are important in hand OA trials. Our aim was to explore the inter-reader reliability of FOI in hand OA, which has not been assessed in any previous trials.

Methods: Hand OA patients with no systemic inflammatory joint disease or psoriasis were included in the Nor-Hand study. A fluorescence dye was administered intravenously (ICG-pulsion, 0.1 mg/kg body weight), and one image was obtained every second for 6 minutes using the Xiralite scanner. Two readers independently scored the FOI scans from 21 patients (19 women, mean (standard deviation) age 63.9 (5.6) years). Fluorescence enhancement in 32 joints including the bilateral distal (DIP), proximal interphalangeal (PIP), metacarpophalangeal (MCP), first carpometacarpal (CMC1) and wrist joints was scored on 0-3 scales. The readers evaluated three different phases (phase 1-3) based on the wash-in and wash-out of the dye and the prima vista mode (PVM; a composite of the 240 first images). We assessed the agreement between the two readers by calculating weighted kappa values at joint level as well as intraclass correlation coefficients (ICC) and smallest detectable difference (SDD) values for sum scores when using the two readers« average scores (=1.96*standard deviation of the difference between readers/Ã2).

Results: Fluorescence enhancement was most commonly seen in phase 2 of the examination. Enhancement in phase 3 and especially phase 1, which is thought to represent active inflammation, was uncommon (Table). In phase 2, the readers agreed upon enhancement in 152/210 (72.3%) PIP and 83/168 (49.4%) DIP joints, but only 3/41 (7.3%) wrist joints. No enhancement was found in the CMC1 joints in any of the phases. The inter-reader ICC values were very good for the PVM and phase 2-3. In these phases, the inter-reader kappa values were good/very good for PIP joints and good for DIP joints. The reliability was lower for MCP and wrist joints. Fluorescence enhancement in phase 1 could not be reliably assessed.

Conclusion: In hand OA, signal alterations by FOI as a proxy for inflammation were most commonly found in DIP and PIP joints. FOI seemed insensitive for the detection of inflammation in CMC1 joints. The inter-reader reliability was good for all phases, except phase 1.