Background: In the most recent European League Against Rheumatism (EULAR) cardiovascular risk reduction recommendations¹, the use of the SCORE algorithm has been advocated as a useful tool to identify an increased 10-year cardiovascular risk of first fatal atherosclerotic event. Even though validated inflammatory disease-specific CV risk score algorithms are still lacking, the EULAR task force advocated the use of a 1.5 multiplication factor for RA, but not clear for other inflammatory diseases.

Objectives: To assess the accuracy of several CV risk algorithms to predict an event and determine its sensibility and specificity.

Methods: A retrospective analysis of Spondyloarthritis (SpA) patients, registered in REUMA.PT, followed in two Portuguese centres was done. We calculated risk prediction algorithms such as Framingham, the American College of Cardiology/American Heart Association (ACC/AHA) risk score and the Systematic Coronary Risk Evaluation (SCORE) for low-risk countries. The adaptation of risk algorithms was done, accordingly to EULAR recommendations. Primary outcome was the first CV event. Discriminatory ability for CV risk prediction was evaluated by the area under the ROC curves. Sensibility and specificity were calculated for low-to-intermediate and intermediate-to-high risk cut-offs. Cut-off values of high risk were defined in 5% for SCORE, 20% for Framingham and ACC/AHA.

Results: 362 patients with SpA were included, 53.9% male (195), with a mean age of 51.1 ± 12.7 years. 67.8% of the patients were HLA B27 positive. Overall, the mean BMI was 26.3± 4.4 Kg/m² and 24.0% of the patients (87) were smokers in their lifetime. The mean of systolic blood pressure was 130± 16.4 mmHg, diastolic blood pressure of 73.5 ± 10.4 mmHg, total cholesterol of 190.1± 37.2 mg/dL and high-density lipoprotein cholesterol of 53.0 ± 14.8 mg/dL. Anti-hypertensive medication was reported in 24.3% of the patients, cholesterol medication in 19.3% and antidiabetic medications in 6.1%. Twenty-five patients (6.9%) presented a cardiovascular event. Patients with a cardiovascular event were older, with higher BMI, prescribed with medication for CV comorbidities and higher diastolic and systolic blood pressure (p<0.05).

Patients were under biologic therapy in 30.9% (112), 16.9% (61) with methotrexate; 16.3% (59) with sulfasalazine and 2.8% (10) with leflunomide; 68.5% (248) prescribed with NSAID and 10.8% (39) with corticosteroids. Area under the ROC in original and adapted scores were equal: 0.709 (95% CI 0.598 to 0.819) for SCORE, 0.805 (95% CI 0.737 to 0.872) for Framingham and 0.776 (95% CI 0.695 to 0.857) for ACC/AHA (Figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

ROC curves for SCORE, Framingham and ACC/AHA

SCORE>1% showed the best sensitivity (96%) but lower specificity. Framingham>20% presented the best specificity (80%) with lower sensitivity (61%). In all cases, specificity raises with higher cut-off with corresponding reduction in sensibility. (Table 1)

Conclusion: A good discrimination between patients with or without CV events has been demonstrated by area under the ROC curve. The adaptation of CV risk algorithms according EULAR recommendations did not provide an improvement in discriminative ability. Overall, the algorithms studied presented a low sensibility or specificity. Better algorithms are needed to correctly assess cardiovascular risk for SpA patients and they should take into consideration the risk associated with the disease.

References: [1]Agca R, Heslinga SC, Rollefstad S, et al. EULAR recommendations for cardiovascular disease risk management in patients with rheumatoid arthritis and other forms of inflammatory joint disorders: 2015/2016 update. Annals of the Rheumatic Diseases 2017;76:17-28.

Disclosure of Interests: None declared.