Assessing the 10-year risk for atherosclerotic cardiovascular events in cutaneous lupus patients

For patients with lupus erythematosus (LE), there is an urgent need to address their heightened risk for clinical events, chiefly heart attacks and strokes, caused by atherosclerotic cardiovascular disease (ASCVD). We compared different methods of analyzing clinical data from 10 years ago (“Time Zero”), to identify which, if any, can accurately distinguish LE patients in our UPenn Longitudinal Lupus Cohort (ULLC) who developed an ASCVD event in the ensuing decade (Ev) or did not (no-Ev). Time Zero was 10 years before a patient’s most recent documented visit by any provider. If less than 10 years of data was available, but the patient had had a documented ASCVD event after enrollment in the ULLC, Time Zero was enrollment. Using data from Time Zero, we estimated 10-year ASCVD event risk using the ASCVD Risk Estimator Plus, QRisk3, Framingham Risk Score (FRS), and modified FRS (mFRS). Of 131 patients included, age at Time Zero was 48.4±12.7 years (mean±SD), 109 (83.2%) were female, and 44 (33.6%) had an ASCVD event after Time Zero. In distinguishing Ev from no-Ev, areas under the receiver operating characteristic curves (AUC) were only 0.607 for the ASCVD Risk Estimator Plus, 0.582 for the QRisk3, 0.610 for the FRS, and 0.582 for the mFRS. Sensitivities and specificities were 22.7% and 85.1% for the ASCVD Risk Estimator Plus, 38.6 and 66.7% for the QRisk3, 36.4 and 77.0% for the FRS, and 50.0 and 59.8% for the mFRS. Positive and negative predictive values were similar for all risk estimators, ranging from 37.0 to 44.4% and 68.2 to 70.5%, respectively. There were no significant differences in conventional characteristics between Ev and no-Ev groups. Thus, all widely used risk estimators poorly predicted ASCVD events in our cohort of lupus patients. This finding suggests that additional clinical data may be needed to predict which LE patients will have a future ASCVD event.