Association of P2Y12 Polymorphisms With the Risk of Ischemic Stroke Subtypes

To evaluate the association of the purinergic receptor P2Y, G-protein coupled, 12 (P2Y12) gene polymorphisms with susceptibility to different etiological stroke subtypes.

A total of 459 first-ever acute ischemic stroke patients were classified into large-artery atherosclerosis (LAA, n = 163), small-vessel occlusion (SVO, n = 204), and cardioembolism (CE, n = 92) based on the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria. Direct sequencing was used to screen these three stroke subtypes and non-stroke controls for P2Y12 polymorphisms: a T→C transition at 744 nucleotides (nt) downstream of intron 5's start site (i-T744C) and a C→T transition at 34 nt downstream of exon 2's start site (C34T). Based on the results of multivariate logistic analyses, a prediction model was established via a nomogram that incorporated genomic and clinical variables to quantify the risk of LAA stroke.

Significant differences in the P2Y12 i-T744C genotype and allele frequencies were observed between LAA patients and controls. After adjusting for confounding factors, the dominant model (p = 0.009) and additive model (p = 0.023) revealed that the i-T744C polymorphism was significantly associated with increased susceptibility to LAA. No significant associations were found for the SVO and CE stroke subtypes. Moreover, the C34T polymorphism was not an independent factor for any stroke subtype. We further constructed a nomogram prediction model for LAA stroke based on genomic and clinical variables, including age, hypertension, smoking, high-density lipoprotein cholesterol, and the i-T744C polymorphism. This nomogram exhibited satisfactory accuracy and predictive power for LAA stroke, as demonstrated by the area under the curve, calibration plot, and decision curve analysis.

The P2Y12 i-T744C polymorphism may serve as a predictor for LAA stroke. Furthermore, we constructed a genomic-clinical nomogram that may be valuable for predicting LAA stroke risk in the study population.