Subclavian artery involvement in patients with giant cell arteritis: do we need a modified Halo Score?

Objective: To assess whether adding the subclavian artery examination into the ultrasound (US) Southend Halo Score, as proposed in the modified Halo Score, improves the diagnostic accuracy of giant cell arteritis (GCA) and its relationship with systemic inflammation.

Methods: Retrospective observational study of patients referred to a GCA fast track pathway (FTP) over a 1-year period. Patients underwent US exam of temporal and large vessel (LV) (carotid, subclavian, and axillary) arteries. The extent of inflammation was measured by the halo count, the Southend Halo Score, and the modified Halo Score. The gold standard for GCA diagnosis was clinical confirmation after 6-month follow-up.

Results: Sixty-four patients were evaluated in the FTP, 17 (26.5%) had GCA. Subclavian artery involvement was present only in patients with GCA (29.4% versus 0%, p < 0.001). Overall, the three scores showed excellent diagnostic accuracy for GCA (ROC AUC 0.906, 0.930, and 0.928, respectively) and moderate correlations with acute phase reactants (0.35-0.51, p < 0.01). Only the modified Halo Score correlated with markers of inflammation in patients with LV involvement.

Conclusions: The inclusion of subclavian artery examination in the modified Halo Score does not improve the diagnostic accuracy of GCA. Nevertheless, it correlates better with markers of systemic inflammation in LV-GCA. Key Points • Adding the subclavian artery examination into the Southend Halo Score, as proposed in the modified Halo Score, does not improve the diagnostic accuracy of GCA. • However, the extent of vascular inflammation as quantified by the modified Halo Score correlates better with markers of systemic inflammation in the large vessel (LV) GCA subgroup of patients. • Although the diagnostic value of adding subclavian arteries to the current recommended US examination of GCA is limited, it may have a role in monitoring disease activity as it correlates with the general burden of inflammation in LV GCA. These findings need to be confirmed in additional populations and larger prospective studies.