Performance of SCORE2, QRISK3 and PREVENT equations in systemic lupus erythematosus
Lévi-Dan Azoulay1,2,3, Thomas Broussaud4, Nadjia Kachenoura1
1Laboratoire d'Imagerie Biomédicale, Sorbonne Université, Inserm, CNRS, Paris, France.
Abstract:
BackgroundAssessment of contemporary cardiovascular risk scores using clinically relevant endpoints is lacking in systemic lupus erythematosus (SLE).AimThis study aimed to assess and compare the performances of SCORE2, QRISK3 and PREVENT equations in SLE.MethodsSLE patients with no prior atherosclerotic cardiovascular disease (ASCVD) who underwent a baseline cardiovascular risk assessment including coronary artery calcium (CAC) scoring at the French national SLE reference center between 2014 and 2024 were retrospectively included. The primary outcome was incident ASCVD events defined as coronary artery disease (CAD), stroke and peripheral artery disease (PAD). The secondary outcome was CAC presence (CAC score >0). Discrimination and calibration were respectively assessed by areas under the curve (AUCs) and observed-to-predicted risk ratios.ResultsA total of 143 patients were included (91% female, median age 51 years [46-60], SLE duration 15 years [8-22]). After a median follow-up of 7 years [3-9], 12 patients (8%) had incident ASCVD events (7 CAD, 4 strokes, 1 PAD). AUCs were 0.81 for SCORE2, 0.76 for QRISK3 and 0.80 for PREVENT and did not significantly differ (all p > .05). Optimal thresholds for clinical events prediction were 3.9% for SCORE2, 9.4% for QRISK3 and 4.3% for PREVENT. Mean observed-to-predicted ratios were 3 for SCORE2, 0.85 for QRISK3 and 2.8 for PREVENT. Similar results were obtained when using CAC as the main outcome.ConclusionRisk scores demonstrated similar and fair discriminative performances but were poorly calibrated except for QRISK3. Application of lower thresholds and use of QRISK3 may improve cardiovascular risk stratification in SLE but requires confirmation from larger studies.


