Giant cell arteritis: a systematic review of the qualitative and semiquantitative methods to assess vasculitis with 18F-fluorodeoxyglucose positron emission tomography

Abstract

Giant cell arteritis (GCA) is the most common vasculitis affecting medium and large vessels. It shows a close clinical association with polymyalgia rheumatica (PMR), a musculoskeletal inflammatory disorder, which is clinically characterized by girdles pain and stiffness. 18F-Fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) is an effective tool for the diagnosis, grading, and follow-up of patients affected by GCA involving the aorta and its proximal branches, but the lack of a standardized method for the assessment of vascular inflammation remains a critical issue, potentially leading to misclassification. In our systematic review, including 19 original articles for a total of 442 GCA patients (with or without PMR symptoms) and 535 healthy controls, we described the different qualitative, semiquantitative and combined methods that have been proposed throughout the literature for assessing the presence and grading the severity of GCA-related vascular inflammation on 18F-FDG PET scans, focusing on the diagnostic performance and examining their respective advantages and limitations. The majority of the included studies adopted qualitative methods of PET image analysis, which are less sensitive but more specific than semiquantitative ones. Among the semiquantitative approaches, the aortic-to-blood pool uptake ratio of the aortic arch seems to be the most accurate method.

Figure 1

Flowchart of the review process.

Figure 2

65-year-old female patient with 18F FDG PET-CT findings indicating the clinical association of polymyalgia rheumatica and giant cell arteritis. Coronal PET (a) and PET-CT (b) images demonstrate a significant tracer uptake of the walls of the ascending aorta, aortic arch (void arrows in (a) and (b)), and subclavian arteries (arrowheads in (a) and (b)). The second pair of coronal PET (c) and PET-CT (d) images demonstrate the inflammatory involvement of the abdominal aorta (arrowheads in (c) and (d)). A bilateral uptake of the tracer of the glenohumeral joints is also seen (solid arrows).

Figure 3

77-year-old female patient with clinical and PET-CT findings of giant cell arteritis. Coronal PET (a) and PET-CT (b) scans demonstrate the inflammatory involvement of aortic arch and abdominal aorta (void arrows), which is clearly appreciable by means of an immediate qualitative assessment of the images.

Figure 4

80-year-old male patient with clinical and PET-CT findings of giant cell arteritis. In this patient, coronal PET (a) and PET-CT (b) images demonstrate 18F FDG uptake of the walls of the ascending thoracic aorta (void arrows). The tracer uptake is similar to that of the liver parenchyma (asterisk), corresponding to grade 2 (significant vascular inflammation) according to the visual grading score proposed by Meller et al.

Figure 5

65-year-old female patient with polymyalgia rheumatica and suspected giant cell arteritis. An immediate qualitative visual assessment of the coronal PET-CT scan ((a) and (b)) led to the diagnosis of inflammatory involvement of the ascending thoracic aorta (white arrows). In this patient, the semiquantitative method of analysis proposed by Hautzel et al. was further applied (aorta-to-liver SUVmax ratio). Placing a ROI on the ascending thoracic aorta in the coronal PET image (c), a SUVmax of 1.6 was obtained. The SUVmax obtained by drawing the same ROI comprehensive on the liver (c) was 2.2, and the resulting aorta-to-liver SUVmax ratio was 0.7, below the cut-off value of 1 for diagnosing significant vascular inflammation. This is an example of discrepancy between qualitative and semiquantitative methods of image analysis.