Carotid ultrasonography improves residual risk stratification in guidelines-defined high cardiovascular risk patients

Abstract

Background The clinical value of carotid atherosclerosis markers for residual risk stratification in high atherosclerotic cardiovascular disease (ASCVD) risk patients is not established. Purpose In the present study we aimed to derive and validate optimal values of markers of carotid subclinical atherosclerosis improving risk stratification in guidelines-defined high ASCVD risk patients. Methods We consecutively analysed high or very high ASCVD risk patients from a cardiovascular (CV) prevention registry (n=751, derivation cohort) and from the Atherosclerosis Risk in Communities (ARIC) study (n=2,897, validation cohort). Baseline ASCVD risk was defined using the 2021 European Society of Cardiology (ESC) guidelines (clinical ESCrisk). Intima-media thickness (IMT) excluding plaque, average maximal (avg.maxWT), maximal wall thickness (maxWT) and number of sites with carotid plaque were assessed. As endpoint of the study was defined the composite of CV death, acute myocardial infarction (MI) and revascularization after a median of 3.4 years in both cohorts and additionally for 16.7 years in the ARIC cohort. Results MaxWT >2.00mm and avg.maxWT >1.39mm provided incremental prognostic value, improved discrimination and correctly reclassified risk over the clinical ESCrisk both in the derivation and the validation cohort (p<0.05 for net reclassification index, integrated discrimination index and Delta Harrell's C index). MaxWT <0.9mm predicted very low probability of CV events (negative predictive value = 97% and 92% in the derivation and validation cohort, respectively). These findings were additionally confirmed for very long-term events in the validation cohort. Conclusion Integration of carotid ultrasonography in guidelines-defined risk stratification may identify very high risk patients in need for further residual risk reduction or at very low probability Funding Acknowledgement Type of funding sources: None.