Development and validation of a diagnostic model for symptomatic large artery atherosclerotic stenosis: integrating lipid metabolism, coagulation, and inflammatory biomarkers.

BACKGROUND

Early diagnosis of symptomatic large artery atherosclerotic stenosis (sLAAS) is crucial for therapeutic decision-making. ​Reliance on imaging for sLAAS diagnosis generates substantial obstacles in resource-constrained settings.​ This challenge ​prompted​ the development of an accessible diagnostic model using routinely available clinical parameters, ​enabling sLAAS diagnosis when advanced imaging is unavailable.

METHODS

This study included acute ischemic stroke (AIS) patients with imaging-confirmed large artery atherosclerotic stenosis (LAAS). Biomarkers of sLAAS initially filtered by least absolute shrinkage and selection operator (LASSO) regression ​were further assessed​ using multivariable logistic regression. We subsequently developed a clinically applicable diagnostic tool based on a nomogram, and validated​ the model's performance in a separate cohort.

RESULTS

A total of 318 AIS patients with LAAS were included in this study. Multivariable logistic regression confirmed lipoprotein-associated phospholipase A2 (Lp-PLA2), triacylglycerol (TG), fibrinogen (FIB), and aggregate index of systemic inflammation (AISI) as independent biomarkers of sLAAS. The nomogram exhibited strong diagnostic performance, with the area under the receiver operating characteristic curve (AUC-ROC) of 0.860 in the training set and 0.837 in the validation set. Both calibration curves and decision curve analysis (DCA) validated the robust performance of the model.

CONCLUSIONS

We developed a diagnostic model for sLAAS ​that integrates lipid metabolism, coagulation, and inflammatory biomarkers. This model ​enables accurate sLAAS identification, demonstrating specific value in​ resource-constrained settings.