Does the halo count on temporal and axillary ultrasound predict time to relapse in giant cell arteritis?

OBJECTIVES: To determine whether the halo count (HC) on temporal and axillary artery US (TAUS) predicts time to relapse in giant cell arteritis (GCA). METHODS: We conducted a single-centre retrospective study of GCA patients. HC, the number of vessels with non-compressible halo on the TAUS at diagnosis, was determined by retrospective review of the US report and images. Relapse was defined as increase in GCA disease activity requiring treatment escalation. Cox proportional hazard regression was used to identify predictors of time to relapse. RESULTS: A total of 72 patients with confirmed GCA were followed up for a median of 20.9 months. Thirty-seven of 72 (51.4%) relapsed during follow-up, at a median prednisolone dose of 9 mg (range 0–40 mg). Large-vessel (axillary artery) involvement did not predict relapse. On univariable analysis, a higher HC was associated with shorter time to relapse (per-halo hazard ratio 1.15, 95% CI 1.02, 1.30; P = 0.028). However, statistical significance was lost when the 10 GCA patients with an HC of zero were excluded from analysis. CONCLUSION: In this real-world setting, relapse occurred at a wide range of glucocorticoid doses and was not predicted by axillary artery involvement. GCA patients with a higher HC at diagnosis were significantly more likely to relapse, but significance was lost on excluding those with HC of zero. HC is feasible in routine care and may be worth incorporating into future prognostic scores. Further research is required to determine whether confirmed GCA patients with negative TAUS represent a qualitatively different subphenotype within the GCA disease spectrum.