Inflammation-associated combined vascular and ventricular phenotype in rheumatoid arthritis is incompletely captured by conventional cardiovascular risk scores

Abstract

Rheumatoid arthritis (RA) is associated with increased cardiovascular (CV) risk driven by both traditional risk factors and chronic systemic inflammation. Conventional CV risk scores may not adequately reflect subclinical structural CV involvement in this population. To characterize vascular and ventricular phenotypes in RA and evaluate their association with systemic inflammation and estimated CV risk. In this cross-sectional study, 195 patients fulfilling 2010 ACR/EULAR RA criteria and without prior CV disease underwent carotid ultrasound and transthoracic echocardiography. Carotid plaque was defined as a focal wall thickening >50% relative to adjacent wall, or a focal protrusion > 1.5 mm. Left ventricular (LV) geometry was classified according to ASE/EACVI criteria. Patients were categorized into four phenotypes: no structural involvement, isolated carotid plaque, isolated LV remodeling (abnormal geometry), and combined phenotype. High estimated CV risk was defined as QRISK3 ≥20%, EULAR-adjusted Framingham BMI ≥20%, and SCORE2 High/Very High per ESC 2021 thresholds. Framingham BMI risk estimates were multiplied by 1.5 according to EULAR recommendations. Multivariable logistic regression models compared each phenotype with the no structural involvement group, adjusting for age, sex, hypertension, and log-transformed C-reactive protein (CRP). Among 195 RA patients, 53 had no structural involvement, 30 isolated plaque, 58 isolated remodeling, and 54 (28%) combined phenotype. Patients with the combined phenotype were older than other groups (61 ± 7.3 vs 51 ± 9.9, 55 ± 10.4, and 53 ± 8.8 years, p < 0.001). CV comorbidities and RA characteristics were similar across groups. High estimated CV risk was identified in 33% of patients with combined phenotype by both SCORE2 and Framingham BMI. Accordingly, approximately two-thirds of patients with combined phenotype were not classified as high risk. In isolated phenotypes, Framingham BMI identified the highest proportion of high-risk patients in the remodeling group (15%), while SCORE2 identified the highest proportion in the plaque-only group (14%). This discordance persisted despite comparable distributions of traditional CV risk factors across groups. In multivariable analysis, isolated plaque was independently associated with hypertension (OR 3.28, 95% CI 1.09–10.15, p = 0.035). Isolated remodeling showed no independent associations. The combined phenotype was independently associated with higher log-CRP (OR 1.47, 95% CI 1.03–2.09, p = 0.031) and age (OR 1.16 per year, p < 0.001). The coexistence of carotid plaque and LV remodeling identifies an inflammation-associated combined vascular and ventricular phenotype in RA. However, most patients with combined subclinical involvement were not classified as high risk by conventional scores. Imaging-based phenotyping may complement CV risk stratification in RA.