Objective
Clinical joint count assessment is important for detecting synovitis but its reliability is a subject of controversy. This study was undertaken to assess the correlation of positron emission tomography (PET)–derived parameters in 68 joints with disease activity and to compare the reliability of joint counts between PET with computed tomography (CT) and clinical assessment in patients with rheumatoid arthritis (RA).
Methods
We enrolled 91 patients with active RA (69 in a development group and 22 in a validation group) who underwent concurrent 18F‐fluorodeoxyglucose (18F‐FDG)–PET‐CT and clinical disease activity evaluation. PET‐derived parameters were compared with disease activity assessed using clinical joint count parameters. A Disease Activity Score (DAS) using counts of PET‐positive joints was developed, and then validation studies were performed in an independent group.
Results
The number of PET‐positive joints (of 28 and 68 joints) was significantly correlated with the swollen joint count (SJC) and tender joint count (TJC) and the DAS in 28 joints using the erythrocyte sedimentation rate (DAS28‐ESR). Intraobserver and interobserver reliability of PET for the affected joint counts were excellent. Interobserver reliability between nuclear medicine physicians and rheumatologists was good for the SJC and TJC in both 28 joints and 68 joints. After multivariate analyses, including ESR and patient's global assessment of disease activity (PtGA) in addition to PET‐derived parameters, the PET/DAS was derived as (0.063 × number of PET‐positive joints in 28 joints) + (0.011 × ESR) + (0.030 × PtGA). A significant correlation between the PET/DAS and the DAS28‐ESR was confirmed in the validation group (P < 0.001).
Conclusion
PET‐CT could serve as a sensitive and reliable method in the evaluation of disease activity in RA patients, and may be applicable as a research tool, particularly in clinical trials.