Abstract
Objective
This study aimed to develop and validate a prediction model combining clinical indicators and pharmacogenomic markers for predicting clopidogrel-related high on-treatment platelet reactivity (HPR) after approximately 7 days of dual antiplatelet therapy in patients with coronary artery disease (CAD), and to identify its influencing factors to support individualized antiplatelet therapy. The model is intended for use before platelet function testing results are available, to identify patients at high risk of HPR who may benefit from early testing or genotype-guided therapy adjustment.
Methods
A total of 380 CAD patients were retrospectively enrolled and divided into a training set (
n
= 266) and a validation set (
n
= 114) at a 7:3 ratio. HPR was defined as ADP-induced platelet aggregation ≥40%. We screened predictive variables via univariate analysis and LASSO regression, then identified independent factors using multivariate logistic regression. Four prediction models were established and compared. Model performance was assessed by area under the receiver operating characteristic curve (AUC), calibration curve, decision curve analysis and SHAP values.
Results
Seven core predictive variables were identified. Multivariate logistic regression analysis revealed that
CYP2C19
functional deletion allele carrier status (OR = 4.341,
P
P
= 0.007), acute coronary syndrome (OR = 2.445,
P
= 0.014), age (OR = 1.053,
P
= 0.006), and platelet count (OR = 1.014,
P
P
= 0.001) and hemoglobin (OR = 0.952,
P
CYP2C19
genotype exerted the greatest predictive impact.
Conclusion
This study developed and internally validated a clinical-pharmacogenomic prediction model for clopidogrel-related HPR in CAD patients. The model demonstrated promising performance; however, external validation in multicenter prospective cohorts is required before clinical implementation.