BetaEntity Annotation Prototype
← Back to diseases

Annotated abstract

Concordance between presenting features and relapse in granulomatosis with polyangiitis: implications for risk assessment and counselling

rmdopen · 2025-12-31 · canonical JSON source

2 visible annotations · policy: published · automated confidence ≥ 75.00%

Document resource

Objective To investigate the concordance between organ involvement at diagnosis and relapse in granulomatosis with polyangiitis and factors associated with new disease features at relapse.Methods Data from a national database of newly diagnosed patients was analysed. Clinical features were recorded at diagnosis and relapse, grouped by organ system. ORs and HRs were used to assess associations between baseline features and first relapse. Factors independently associated with new organ involvement at relapse were identified using multivariable logistic regression.Results Among 795 patients (median follow-up 3.5 years), 394 (50%) relapsed; organ involvement at relapse was available for 376 patients. Relapses most often affected ear, nose and throat (ENT), lungs and kidneys. Organ involvement at diagnosis was associated with a higher likelihood of relapse in the same organ: eyes (OR 6.69), lungs (OR 3.35), kidneys (OR 3.58), nervous system (OR 2.90), and mucocutaneous (OR 4.53). Major manifestations associated with a higher likelihood of recurrence were scleritis, pachymeningitis, subglottic stenosis and worsening renal function. For 56% of patients, the first relapse affected only the initially involved organs. Of the 165 patients with new organ manifestations, these were rarely isolated (n=34) and usually occurred alongside involvement of at least one previously affected organ (n=131). In multivariable analysis, systemic, ENT and lung manifestations at diagnosis were associated with a lower risk of new organ disease at relapse.Conclusion Although new features can still emerge, organ involvement at diagnosis is associated with a higher likelihood of relapse in the same organ.