Comparison of methods of quantifying global synovial metabolic activity with FDG-PET/CT in rheumatoid arthritis

2-deoxy-2-[¹⁸ F]fluoro-d-glucose positron emission tomography/computed tomography (FDG-PET/CT) can portray increased glycolysis due to inflammation in rheumatoid arthritis (RA). The aim of this study was to evaluate and compare the reliability and construct validity of two methods of quantifying RA disease activity using FDG-PET/CT.

Nineteen RA patients and 19 healthy controls matched to sex and age underwent prospective FDG-PET/CT imaging. Metabolic and volumetric metrics were calculated using fixed and adaptive thresholding techniques and partial volume correction. Fixed thresholds segmented regions above average maximum physiological uptake in controls. Differences of means between RA and controls were assessed using t tests, and discrimination was assessed using receiver operating characteristics. Spearman correlation analysis was used to assess associations between FDG-PET/CT measures and clinical assessments of disease activity.

All FDG-PET/CT measures were substantially different and nearly non-overlapping between RA and controls (all P < .001). Area under the curve (AUC) for adaptive threshold parameters ranged from 0.986 to 0.997, and AUC for fixed threshold parameters ranged from 0.898 to 0.945. PET parameters were found to correlate positively with various clinical features, namely C-reactive protein, erythrocyte sedimentation rate, interleukin (IL)-6, IL-1, and swollen joint count.

All FDG-PET/CT parameters reflecting global RA disease activity differentiated between RA and controls, indicating high clinical utility to diagnose and assess RA. Adaptive thresholds can be used in a wider setting without control data, but methods utilizing fixed thresholds were more reproducible and more closely associated with indications of inflammation.