Presenting features and outcomes of cranial-limited and large-vessel giant cell arteritis: a retrospective cohort study.

To compare the presenting features and outcomes of patients with cranial-limited (C-) and large-vessel (LV-) giant cell arteritis (GCA). Data from our GCA cohort were collected retrospectively. Patients who underwent total-body large-vessel imaging within 10 days after commencing steroid therapy were included. Patients with LV involvement were classified as LV-GCA. Presenting features, treatments, and outcomes of LV-GCA and C-GCA patients were compared. 161 patients were included (LV-GCA, n = 100). At baseline, LV-GCA patients were younger than those with C-GCA (73.2 ± 8.9 vs 76 ± 8.8 years, p = 0.018) and had a longer delay to diagnosis (3.5 ± 4.6 vs 2.3 ± 4.9 months, p = 0.001). C-GCA patients had a higher incidence of headache (p = 0.006) and ischaemic optic neuropathy (p < 0.001), whereas LV-GCA patients had more systemic symptoms (fever, p = 0.002; fatigue, p < 0.001; weight loss, p < 0.001; night sweats, p = 0.015) and dry cough (p = 0.031). Corrected cumulative prednisone dose, relapse-free survival, relapse-rate, and incidence of ascending aortic aneurysms were not significantly different between the two subgroups. A steroid-sparing agent was added in 73% of LV- and 55.7% of C-GCA patients (p = 0.027), but was introduced more frequently at baseline in LV-GCA patients (52% vs 23.5%, p = 0.006). LV-GCA patients initially treated with glucocorticoid monotherapy relapsed sooner (relapse-free survival, HR = 0.56, 95% CI 0.41-0.78, p < 0.001) and had a higher relapse rate (relapses per 10 person-years, 6.73 ± 11.50 vs 3.82 ± 10.83, p = 0.011). LV-GCA patients were younger at diagnosis and suffered a longer diagnostic delay. The outcomes of the two subgroups were similar. An earlier introduction of steroid-sparing agents in LV-GCA patients might have played a positive role.