Page 346 - The Vasculitides, Volume 1: General Considerations and Systemic Vasculitis

320 Vanessa Quick and John Kirwan

Stenosis and occlusion common in elderly patients due to atherosclerosis, neither of
which are specific or sensitive for GCA, may also be noted [103, 104]. Three meta-analyses
[105-107], which demonstrated the usefulness of the halo sign in the diagnosis of GCA with a
sensitivity of 68% and specificity of 91% for a unilateral halo sign, showed 100% specificity
for the bilateral halo sign. However, the findings may be influenced by glucocorticoid use;
moreover there is a presumed limited window of opportunity wherein which to correctly
ascertain the diagnosis [65]. One-half of positive TAUS findings may be missed within seven
days of treatment initiation (Figure 4), implying that in order to avert the low sensitivity of
TAUS in those with suspected GCA, the study should be performed as soon as possible,
preferably before or on the first day of treatment. Some [108] argue that it is never too late to
perform TAUS because even if the sensitivity is diminished, the specificity remains.
Therefore there is a rationale for its use as it has the potential to avert the need for TAB when
a halo sign is noted. Furthermore, it can increase the diagnostic yield of TAB by directing the
incision site to the most affected area of artery [104, 109]. The Temporal Artery Biopsy
Ultrasound Study (TABUL), a prospective multicenter blinded study comparing TAB to
TAUS in the diagnosis of GCA, which plans to recruit 400 patients within six months of
followup, is currently underway.

Other Imaging Studies
Whereas non-invasive vascular imaging may be useful when there is atypical extracranial

large vessel involvement or limb claudication and persistent acute phase reactants or fever of
unknown origin [110], conventional angiography has little if any role unless a surgical
intervention is contemplated [111]. Ultrasound of the thoracic aorta is generally inadequate
but it may provide useful information in proximal upper limb arteries to increases the
diagnostic yield of GCA [112]. Both magnetic resonance imaging (MRI), MR angiography
(MRA), and contrast-enhanced computerized tomographic angiography (CTA) provide useful
images of mural and luminal changes suggestive of large vessel vasculitis in GCA [6, 64,
113, 114] that include circumferential wall swelling, smoothly tapered luminal narrowing of
aortic branches and aortic aneurysm formation [111]. Moreover, MRI and MRA which do not
lead to radiation exposure are favored over CTA by some specialists [38, 101].

Bright mural enhancement of the temporal artery on contrast-enhanced high-resolution
MRI had comparable sensitivity and specificity to TAUS in the diagnosis of GCA in one
retrospective single-center analysis [115], the latter of which decreased in sensitivity over the
first few days of glucocorticoids treatment (85% after 0-1 days, 64% after 2-4 days, 56% after
>4 days) [116].

Whole body 18Fluorodeoxyglucose (FDG) - positron emission tomography (PET)
increased the overall diagnostic accuracy of large vessel involvement in GCA from 54% to
70% [117]. One meta-analysis [118] found that the absence of FDG uptake conferred a
negative predictive value of 88% of GCA, whereas thoracic vascular uptake was highly
suggestive of GCA. However there have not been properly designed trials to assess the
sensitivity and specificity of 18F-FDG PET in GCA, nor does it reliably distinguish between
atherosclerosis and vasculitis [119]. The radiographic features of large-vessel involvement so
noted in CTA and PET imaging decrease rapidly after the initiation of glucocorticoid
treatment; features of large-vessel vasculitis including concentric wall thickening were
significantly more frequent in treatment naive patients compared to patients treated with
corticosteroids glucocorticoids for 1-3 days (77% versus 29%, p = 0.005) [6]. The diagnostic

Complimentary Contributor Copy