Objective To compare damage and mortality, from inception up to 10-year follow-up, between SLE patients meeting at baseline the 1997 ACR criteria or only the 2012 SLICC classification criteria. Methods Patients fulfilling the ACR and/or the SLICC classification criteria for SLE were enrolled at inception and followed-up to 10 years at an academic lupus clinic. Damage was defined as SLICC Damage Index (SDI) score ≥1. We assessed with multivariate Cox models the damage and mortality outcomes, according to SLE classification status at inception, adjusting for potential baseline confounders. Results We recruited 192 patients (69.8% fulfilling at inception the ACR criteria and 30.2% only the SLICC criteria). During follow-up, 24.0% of patients accrued organ damage and 4.2% died. Patients meeting ACR criteria compared to those with SLICC criteria alone presented during follow-up with more cases of lupus nephritis (35.1% versus 13.8%, p < 0.01), but less thrombotic antiphospholipid syndrome (4.5% versus 17.2%, p < 0.01). The Cox models showed no significant differences in risk for damage [hazard ratio (HR) (95% CI) 0.991 (0.453-2.167)] or death [hazard ratio (HR) (95% CI) 0.694 (0.107-4.506)] between groups. Conclusion The SLE classification status at inception identified different patterns of clinical phenotype, but did not influence damage accrual or mortality up to 10-year follow-up.
Keywords: Systemic lupus erythematosus; classification criteria; damage; inception cohort; mortality; prognosis.