Distribution Patterns of 18F-fluorodeoxyglucose in Large Vessels of Takayasu's and Giant Cell Arteritis Using Positron Emission Tomography

Clin Exp Rheumatol. Mar-Apr 2018;36 Suppl 111(2):99-106. Epub 2018 May 18.


Objectives: To compare patterns of vascular involvement using 18F-fluorodeoxyglucose-positron emission tomography computed tomography (FDG PET/CT) in patients with giant cell arteritis (GCA) and Takayasu's arteritis (TAK).

Methods: A total of 130 consecutive 18F-FDG PET/CT scans performed during the disease course for evaluating disease activity in 15 GCA and 13 TAK patients were retrospectively examined by two nuclear physicians blinded to clinical data. Standardised uptake values (SUVmax) in 14 vascular districts including all the aortic segments and the main tributaries were measured. The average SUVmax value for each vascular district was also calculated. Principal component analysis (PCA) and agglomerative hierarchical cluster analysis (CA) were used to explore distribution patterns of vascular FDG uptake.

Results: The aortic segments showed the highest SUV max values among the different districts in both GCA and TAK. SUV max values measured in the different districts were significantly higher in GCA compared to TAK, except for the axillary arteries. Regarding thoracic and abdominal aorta, ascending aorta and aortic arch had the highest correlation in both vasculitis (p<0.0001). CA confirmed that carotid, axillary, subclavian, iliac and femoral arteries clustered with their contralateral counterpart in both vasculitis. The 3 components of thoracic aorta clustered with abdominal aorta in TAK, while aortic arch clustered only with ascending aorta, and descending and abdominal aorta grouped together with iliac and femoral arteries in GCA. PCA analysis identified 3 different components for TAK and GCA explaining 72% and 71% of the total variance respectively in these two vasculitis. Confirming CA, a component including the entire aortic district was identified in TAK, but not in GCA. Similar results in PCA using averaged data were observed.

Conclusions: Strong similarities, but also a subtle skewing in terms of distribution patterns of arterial involvement assessed by SUVmax values were observed between GCA and TAK.

MeSH terms

  • Adult
  • Aged
  • Aorta, Abdominal / diagnostic imaging
  • Aorta, Thoracic / diagnostic imaging
  • Aortitis / diagnostic imaging*
  • Aortitis / etiology
  • Axillary Artery / diagnostic imaging
  • Carotid Arteries / diagnostic imaging
  • Carotid Artery Diseases / diagnostic imaging*
  • Carotid Artery Diseases / etiology
  • Cluster Analysis
  • Female
  • Femoral Artery / diagnostic imaging
  • Fluorodeoxyglucose F18
  • Giant Cell Arteritis / diagnostic imaging*
  • Humans
  • Iliac Artery / diagnostic imaging
  • Male
  • Middle Aged
  • Positron Emission Tomography Computed Tomography
  • Principal Component Analysis
  • Radiopharmaceuticals
  • Retrospective Studies
  • Subclavian Artery / diagnostic imaging
  • Takayasu Arteritis / diagnostic imaging*


  • Radiopharmaceuticals
  • Fluorodeoxyglucose F18