EULAR 2025 classification criteria for haemochromatosis arthropathy.

OBJECTIVES

This study aims to develop classification criteria for haemochromatosis arthropathy (HA).

METHODS

A task force was convened per European Alliance of Associations for Rheumatology standardised operating procedures, including physicians experienced in genetic haemochromatosis and its arthropathy and patient research partners. Candidate classification criteria items were selected via systematic literature review, consensus, and Delphi process. Derivation cohorts of patients with C282Y homozygous HFE gene mutation, joint pain, and iron loading, along with disease mimics (primary generalised osteoarthritis or calcium pyrophosphate deposition disease), were recruited from routine care and assessed against candidate items. Group comparison and diagnostic performance analyses identified variables with the greatest discriminatory capacity. A subset of these variables, agreed upon by consensus (considering both statistical results and the HA disease construct), was modelled using multivariable logistic regression and receiver operating characteristic curve analysis to develop classification criteria models. The final model was agreed upon through discussion, voting, and consensus, taking both statistical performance and disease Gestalt (face validity) into account.

RESULTS

A derivation cohort of 154 patients with HA and 120 patients with disease mimics was recruited. A point-based classification model was developed using 8 variables covering age of symptom onset, clinical, and radiographic features at the metacarpophalangeal, distal interphalangeal, and ankle joints, and surgery of hips or ankles. A score ≥5 out of 11 provides 93.3% specificity and 71.4% sensitivity for classifying HA.

CONCLUSIONS

An international multicentre task force has developed the first classification criteria for HA from a unique derivation cohort using rigorous methodology. Criteria should be used to include patients in research studies and await external validation in separate cohorts.