Phenotypic subgroups and classification criteria performance in paediatric Behçet’s disease: insights on systemic involvement

Abstract Objectives Behçet’s disease (BD) is a multisystem vasculitis with variable paediatric presentations. Early diagnosis may be difficult due to heterogeneous manifestations and limited applicability of existing criteria. This study compared clinical characteristics by age at onset and examined factors contributing to diagnostic delay. Methods A retrospective cohort of 38 paediatric BD patients from a tertiary centre was analysed. Patients were grouped as early-onset (6 years) or late-onset (≥6 years). Demographic features, clinical manifestations, laboratory findings, treatments and fulfilment of ISG, ICBD and PEDBD criteria were evaluated. Statistical analyses included Mann–Whitney U, χ2, Fisher’s exact tests and Spearman correlation. Results Early-onset patients had younger symptom onset but a diagnostic delay similar to late-onset cases. Clinical manifestations, laboratory findings and treatment approaches were comparable between the groups. ISG, ICBD and PEDBD scores showed no differences and demonstrated strong inter-correlations. Mucocutaneous features—including genital ulcers, oral aphthae and skin lesions—were significantly less common in ICBD-negative patients, while PEDBD positivity corresponded to mucocutaneous-dominant phenotypes. CNS and vascular involvement frequently co-occurred and were associated with fewer mucocutaneous findings. None met ISG, and only one-third met ICBD, indicating limited sensitivity of current criteria for systemic-dominant presentations. No significant associations were found for ocular, joint or febrile symptoms, and multivariate analysis was not feasible due to small subgroup sizes. Conclusion Early-onset BD shows similar clinical profiles to later-onset disease despite a tendency towards longer diagnostic delay. Current criteria insufficiently detect neurologic and vascular phenotypes, underscoring the need for paediatric-specific revisions and validation in larger cohorts.