Background/Purpose:

Giant Cell Arteritis (GCA) is the most common form of systemic inflammatory vasculitis in elderly people, with prevalence still increasing, and for which there is no consensual diagnostic method. Indeed the relevance of temporal artery wall edema is still a matter of debate, despite two previous meta-analyses by Karassa et al. in 2005 and Arida et al. in 2010.  We re-evaluated the diagnostic value of the ultrasonography-derived signs of the temporal artery for GCA diagnosis.

Methods:

We selected prospective studies concerning patients with suspected GCA published up to February 2015 in the Pubmed, Cochrane library and Embase databases. Studies using ultrasound biomicroscopy were excluded.

Diagnostic performances were determined for the following ultrasonographic signs: halo (unilateral or bilateral), blood flow abnormalities (stenosis and/or occlusions) and unilateral halo associated with BFA. The gold standard used was either the American College of Rheumatology (ACR) 1990 criteria or temporal artery biopsy. Weighed sensitivity and specificity for each sign were assessed, as well as statistical heterogeneity. P-values less than 0.05 were considered as significant.

Results:

We included 16 studies involving 964 patients suspected of GCA. Among these patients, 448 patients were diagnosed by ACR criteria (41% of which were ACR+) and 764 by temporal biopsy (48% of which showed temporal arteritis).

Ultrasound was performed in all patients. The halo sign was found in 29% of cases among which 76% had a positive biopsy. Among patients ACR+, the unilateral halo sign had an overall sensitivity of 68.2% (95% CI 0.58-0.79) and specificity of 93.2% (95 % CI  0.88-0.99). The same sign did not improve when associated to BFA: overall sensitivity of 64.2% (95% CI 0.17-1.17) and specificity of 93.4% (95% CI 0.85-1). The bilateral halo sign among ACR+ patients had an overall sensitivity of 58.2% (95% CI 0.19-0.97). The latter’s specificity was about 100% (95% CI 0.92-1) but was only estimated in one study. Among patients with a positive biopsy,  the unilateral halo sign had an overall sensitivity of 82.4% (95% CI 0.77-0.88) and specificity of 86.8% (95% CI  0.82-0.92) with the same absence of improvement as in the previous group when associated to BFA. The bilateral halo sign achieved an overall sensitivity of 51.7% (95% CI 0.42-0.62) and specificity of 82.4%(95% CI 0.57-1). It is to be noticed that between-study heterogeneity was very significant for all these calculations (I2 > 40%). A positive biopsy was found in 58 patients among the 439 without the halo sign (13% of false-negative). 80.7% of ACR+ patients had a positive biopsy and 19.3% had a normal biopsy.

Conclusion:

Among all ultrasonography-derived signs, the unilateral halo sign may be sufficient in GCA diagnosis. However, this meta-analysis did not confirm the infallibility of these signs, in particular in case of an absence of ultrasonographic signs. This last situation should be explored in further studies.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/diagnostic-value-of-ultrasonography-derived-signs-in-giant-cell-arteritis-literature-review-and-meta-analysis/