Early prediction of treatment response in rheumatoid arthritis by quantitative macrophage PET

OBJECTIVE: To determine whether macrophage positron emission tomography (PET)/computed tomography (CT) imaging using (R)-[(11)C]PK11195 at 0 and 2 weeks is associated with clinical response at 13 weeks in patients with early rheumatoid arthritis (RA). METHODS: Whole-body (R)-[(11)C]PK11195 PET/CT scans were performed at baseline and after 2 weeks of COBRA-light (combination therapy of methotrexate and prednisone) treatment in 35 patients with clinically active early RA. Clinical assessment (Disease Activity Score of 44 joints (DAS44)) was performed at 0, 2 and 13 weeks of treatment. PET/CT scans were assessed visually by two blinded, experienced readers, and by calculating standardised uptake values (SUVs) for shoulders, elbows, hips, knees, and hand and feet joints. Clinical and PET variables were compared using (multivariate) linear regression. RESULTS: 18 males and 17 females were included (baseline DAS44=3.2 ± 1.0). 171 out of 1470 joints were visually PET positive at baseline, decreasing to 100 joints after 2 weeks. In general, small feet joints showed the highest uptake at baseline, and the largest decrease after 2 weeks (Δ(0-2)). Neither baseline nor Δ(0-2) PET measures correlated with DAS44 at 13 weeks. However, at 2 weeks, average SUV of the feet significantly correlated with DAS44 at 13 weeks (R(2)=0.14, p=0.04). In a multivariable model, DAS44 and average SUV of the feet at 2 weeks showed substantial combined predictive value (combined R(2)=0.297, p<0.01). CONCLUSION: Quantitative macrophage PET assessment of feet joints, together with DAS44, after 2 weeks of COBRA light treatment in patients with early RA correlates with clinical response after 3 months of treatment.