To identify characteristics and factors associated with relapse and glucocorticoid (GC) dependence in a large cohort of consecutive patients with giant-cell arteritis (GCA).

We retrospectively analyzed 326 consecutive patients with GCA diagnosed between 2000 and 2016 and followed them for at least 12 months. Factors associated with relapse were identified in a Cox proportional hazards model. Factors associated with GC dependence were identified in a multivariate logistic regression analysis.

The 326 patients (73% women; median age at diagnosis: 74 [48–92] years) were followed up for 62 [12–262] months. During follow-up, 171 (52%) patients relapsed, including 113 (35%) who developed a GC dependence. Relapsing patients more frequently showed a past history of stroke (P=0.01) and large-vessel involvement (LVI) on imaging (P=0.01) than patients without relapse. Therapeutic strategy did not differ among relapsing and non-relapsing patients. GC-dependent patients were younger (P=0.0001) and more frequently showed a past history of stroke (P=0.004) and LVI on imaging (P=0.005) than patients without GC-dependent disease. In multivariate analyses, LVI was an independent predictive factor of relapse [hazards ratio (HR): 1.49 (95% IC: 1.002–2.12), P=0.04] and GC dependence [odds ratio (OR): 2.19 (95% IC: 1.19–4.05), P=0.01]. Conversely, stroke was a protective factor against relapse [HR: 0.21 (95% CI: 0.03–0.68), P=0.005] and GC dependence disease [OR: 0.10 (95% CI: 0.001–0.31), P=0.0005].

LVI was an independent predictive factor of relapse and GC dependence. Conversely, a previous stroke was protective against both unfavorable treatment outcomes. Further prospective studies are needed to confirm these findings and determine if patients with LVI require a different treatment.