Objectives To compare clinical characteristics, treatment, long-term follow-up and prognosis of two population-based cohorts of patients with biopsy-proven giant cell arteritis (GCA) from Olmsted County, Minnesota, USA (Olmsted cohort) and the Reggio Emilia area, Northern Italy (Reggio cohort).

Methods All patients residing in Olmsted County and the Reggio Emilia area with a new diagnosis of biopsy-proven GCA in 1986–2007 were retrospectively identified. Patients were followed from GCA diagnosis to death, migration or September 2011. Comparisons were performed using Chi-square and rank sum tests, Kaplan-Meier methods and Cox models.

Results The study included 110 patients in the Olmsted and 144 in the Reggio cohort. Compared with the Olmsted cohort, patients from the Reggio cohort were younger (mean±SD age 74.6±7.4 years vs 77.8±7.6, p=0.002), more likely to have cranial symptoms (93% vs 86%, p=0.048), temporal artery abnormalities at physical examination (68% vs 42%, p<0.001), partial or complete unilateral or bilateral permanent vision loss (21% vs 6%, p=0.001), systemic symptoms (67% vs 46%, p=0.001) and polymyalgia rheumatica at or before GCA diagnosis (47% vs 26%, p<0.001). Scalp tenderness was less common in the Reggio cohort (36% vs 49%, p=0.033). ESR and CRP were higher (mean 88±29 mm/h vs 73±77, p<0.001 and mean 89.0±60.2 mg/L vs 35.2±43.4, p<0.001 respectively) and hemoglobin lower (mean 11.2±1.4 g/dl vs 11.8±1.4, p=0.004) in Reggio than in the Olmsted cohort. Patients from the Olmsted cohort received a higher initial prednisone dose (mean 53.6±15.3 mg/day vs 49.5±12.8, p=0.001). There were no differences in relapse rates, cumulative glucocorticoid (GC) dosages at 1, 2 and 5 years, and time to first GC discontinuation. However, the Reggio cohort reached a prednisone dose <10 mg/day sooner (median 4.9 months vs 7.9, p=0.012) and had a first relapse later (median 13.6 months vs 7.9, p=0.003) than the Olmsted cohort. Patients from the Reggio cohort had a significantly higher mortality compared to those from the Olmsted cohort (HR 1.72, 95% CI 1.12–2.65 adjusted for age and sex).

Conclusions Genetic and/or environmental factors may contribute to the differences in clinical characteristics and disease outcomes observed in this study comparing patients with GCA from North America and Southern Europe.

Disclosure of Interest None declared