Background/Purpose: PET/CT enables assessment of active vascular inflammation in Takayasu arteritis (TAK), which is often difficult to evaluate with conventional imaging. Although the distribution of affected arteries has been classified into patterns by both conventional imaging and PET/CT, its relationship with clinical outcomes remains unclear. This study aimed to evaluate the significance of PET/CT-based arterial involvement in treatment-naïve TAK and its association with relapse and cardiovascular events.

Methods: We conducted a retrospective study of patients with TAK who underwent PET/CT before treatment between 2010 and 2024. Hierarchical clustering was performed based on the distribution of affected arteries, with 18F-fluorodeoxyglucose uptake assessed by a four-grade visual scale; the score ≥1 was considered positive. Clinical characteristics and outcomes were compared across clusters. Relapse and large vessel events were defined according to the 2018 EULAR recommendations and analyzed using survival analysis. Longitudinal changes in PET/CT and clinical activity were evaluated in patients with post-treatment PET/CT. Continuous variables were expressed as medians with interquartile ranges.

Results: Thirty patients were included, with a follow-up of 4.8 years (1.4–9.9), age at diagnosis of 48 years (28–58), and 90% female. All met Japanese diagnostic criteria; 80% met the 2022 ACR/EULAR criteria. Numano’s classification based on CT was as follows; I, 2; II, 10; III, 3; V, 15. Relapse occurred in 12 patients (40%) with one major relapse; and large vessel events in nine patients (30%); aortic regurgitation, 2; aortic aneurysm, 5; aortic stenosis. 4; cardiovascular death, 1. Hierarchical clustering revealed three distinct patterns: (A) thoracic type (mainly thoracic aorta with less frequent branch involvement; 43%), (B) diffuse type (extensive involvement of the aorta and its branches; 30%), and (C) localized type (mainly branch lesions; 27%) (Figure 1). The proportion of type V was 78% in A, 38% in B, and 38% in C (p=0.35), indicating inconsistency between PET/CT and angiographic classifications. PET vascular activity scores were significantly different among clusters: 9 (7–13) in A, 15 (12–21) in B, and 3 (1–5) in C (p< 0.01). CRP levels were as follows; 3.1 (1.7–3.8), 3.4 (1.4–5.6), and 0.7 (0.1–7.1) mg/dL (p=0.28). Relapse rates were the highest in B (23%, 78%, and 25%, p=0.04), whereas large vessel events tended to occur more in A (46%, 22%, and 13%; p=0.09) (Figure 2). Numano's classification was not associated with relapse or large vessel events. Among 17 patients with repeated scans, PET/CT activity was observed despite clinical inactivity, indicating a risk of relapse in such patients. Meanwhile, a relapse was also observed in a patient whose PET/CT activity had normalized (Figure 3).

Conclusion: PET/CT-based arterial pattern in TAK was classified into three clusters with distinct prognostic profiles. Diffuse type was associated with higher relapse, whereas thoracic type tended to present vascular complications. Conventional angiographic classification failed to predict these outcomes, and thus PET/CT would provide superior risk stratification in TAK.

Figure 1. Hierarchical clustering based on the distribution of affected arteries revealed three distinct patterns.

Figure 2. Survival curve analysis showed higher relapse rates in the diffuse type and a trend toward more large vessel events in the thoracic type.

Figure 3. Sankey diagram illustrating temporal changes in clinical and PET/CT disease activity.

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/pet-ct-based-distribution-of-arterial-involvement-and-its-association-with-clinical-outcomes-in-takayasu-arteriti/