POS0551 ERS-RA as a Tool for Cardiovascular Risk Prediction in Established Rheumatoid Arthritis: An External Validation

Le résumé fourni par la source

Background: Patients with rheumatoid arthritis (RA) are at increased risk for cardiovascular disease (CVD), but risk scores based on traditional CVD risk factors have underperformed in RA populations. The Expanded Risk Model for Rheumatoid Arthritis (ERS-RA) incorporates traditional CV risk factors and RA-related measures of disease activity and was observed to have a c-index of 0.76 and net reclassification of 40% compared to a traditional risk factor model [1]. Additional external validations are needed to better understand the utility of ERS-RA in different patient groups with RA. Objectives: To externally validate a previously published cardiovascular risk score, the ERS-RA, in a cohort of participants with established RA, assessing different definitions of CV events. Methods: We used data from the Brigham and Women's Hospital Rheumatoid Arthritis Sequential Study (BRASS), a longitudinal cohort of participants with confirmed RA enrolled from 2003-2023. We obtained cardiovascular endpoints from three data sources: 1) self-reported CV events in BRASS, assessed annually, 2) linkage to Medicare/Medicaid claims, and 3) linkage to electronic health records. We used validated claims-based algorithms to define CV events in Medicare/Medicaid administrative claims. For CV outcomes identified in electronic health records, we first searched for potential CV events using diagnosis and procedure codes, and then adjudicated events through a medical record review process conducted by a cardiologist. Due to the uncertainty in accuracy of self-reported CV events, we evaluated CV events from claims and adjudicated electronic medical records in primary analyses (narrow definition) and integrated self-reported events as a third data source (broad definition) in a secondary analysis. We defined the follow-up period as time from enrollment until first CV event or last study visit. For the narrow CV definition, we additionally censored participants with self-reported CV events at the study visit prior to their self-reported event. We truncated follow-up at a maximum of 10 years, restricted to participants aged 20-80 years at enrollment, and excluded participants with self-reported CV events at the baseline study visit. We used multiple imputations with fully conditional specification to impute missing measures for smoking status, CDAI, and mHAQ. Model performance was assessed by evaluating discrimination using the C-statistic and calibration through visual inspection of calibration plots comparing predicted versus observed risks. Additionally, model performance was evaluated by dichotomizing predicted 10-year risk as <7.5% vs. ≥7.5% and <10% vs. ≥10%. Results: We identified 1,332 participants in BRASS without evidence of CV events at baseline. The mean (SD) age of the cohort was 54 (13) years, 16% were male gender, 8% were current smokers, 6% had diabetes, 22% had hyperlipidemia, and 30% had hypertension. Additionally, 37% had an RA duration of <5 years, mean (SD) CDAI was 18 (15.3), and 71% were rheumatoid factor positive. Mean (SD) follow-up was 81 (41) months [approximately 6.8 (3.4) years]. We observed 71 (5%) participants with a CV event using the narrow definition and 166 (12%) with the broad definition. We observed a c-statistic of 0.659 for the broad CV event definition and 0.810 for the narrow definition (Table 1). Calibration curves (Figure 1) indicated slight overprediction of CV risk using the narrow definition, with tighter calibration around lower CV-risk deciles. Conversely, calibration curves indicated slight underprediction of risk using the broad definition, with less precise calibration than observed with the narrow definition. Using a cut-point of <7.5% versus ≥7.5%, we observed sensitivity (0.542 vs. 0.789), specificity (0.714 vs. 0.708), positive predictive value (PPV, 0.212 vs. 0.132), and negative predictive value (NPV, 0.916 vs. 0.984) for the broad and narrow definitions, respectively. Results were similar for a cut-point of 10% predicted risk. Conclusion: In this cohort of individuals with established RA the ERS-RA generally demonstrated good discrimination and calibration for cardiovascular events defined using validated algorithms in claims and EHR; however, its precision was lower when self-reported CV events (broad definition) were included. REFERENCES: [1] Solomon DH, Greenberg J, Curtis JR, Liu M, Farkouh ME, Tsao P, Kremer JM, Etzel CJ. Derivation and internal validation of an expanded cardiovascular risk prediction score for Rheumatoid Arthritis: A Consortium of Rheumatology Researchers of North America Registry Study. A&R 2015; 67(8):1995-2003. Figure 1Calibration Curves, Broad and Narrow CV Definitions. Table 1Model Performance of the ERS-RA for Broad and Narrow CV Definitions.MetricCV Event DefinitionsBroad definitionNarrow definitionNumber of patients with a CV event16671Observed %12.55.3Predicted %31.831.8Absolute difference, %19.426.5C-statistic0.6590.810Predicted 10-year risk: <7.5% vs. ≥7.5% predicted riskParticipants with ≥7.5% predicted risk, n (%)424 (31.8)424 (31.8)Sensitivity, %54.278.9Specificity, %71.470.8Positive Predictive Value, PPV, %21.213.2Negative Predictive Value, NPV, %91.698.4Predicted 10-year risk: <10% vs. ≥10% predicted riskParticipants with ≥10% predicted risk, n (%)307 (23.1)307 (23.1)Sensitivity, %41.663.4Specificity, %79.679.2Positive Predictive Value, PPV, %22.514.7Negative Predictive Value, NPV, %90.597.5Abbreviations: CV=cardiovascular; Broad CV definition=events from Medicare/Medicaid claims, electronic health records and/or self-reported. Narrow CV definition=events from Medicare/Medicaid and/or electronic health records. Acknowledgements: The authors would like to thank all the participating investigators and patients in the BRASS study who contributed data for this study. Disclosure of Interests: Misti L. Paudel: None declared, Katherine Liao UCB and Merck, Jon Giles Pfizer, AbbVie, Eli Lilly, Novartis, Merck, Genentech, and Sana, Joan Bathon AbbVie and Merck, Hongshu Guan: None declared, Brendan Everett: None declared, Leah Santacroce: None declared, Nancy A Shadick BMS, Amgen, Eli Lilly, Mallinckrodt, Sanofi-Regeneron, and Crescendo Biosciences, Michael E. Weinblatt Canfite; Inmedix; Scipher, Aclaris; Amgen; Anaptysbio; Bristol Myers Squibb; Biohaven; Deep Cure; Gilead Ignite; Johnson and Johnson; Lilly; Lifordi; Matchpoint; Novartis Prometheus; Rapt; Rani; Revolo; Sanofi; Scipher; Sci Rhom; Set Point; Surf Therapeutics; ZuraBio, Bristol Myers Squibb; Aqtual; Abbvie; Janssen, Pamela Rist: None declared, Daniel H. Solomon AMGEN: Contract for gout and cardiovascular disease; Janssen: Contract for digital health technology for psoriatic arthritis; CorEvitas: Contract for RA epidemiology research. © The Authors 2025. This abstract is an open access article published in Annals of Rheumatic Diseases under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Neither EULAR nor the publisher make any representation as to the accuracy of the content. The authors are solely responsible for the content in their abstract including accuracy of the facts, statements, results, conclusion, citing resources etc.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.