Prognostic Stratification of Familial Hypercholesterolemia Patients Using AI Algorithms: A Gender-Specific Approach

Abstract

Background: Familial Hypercholesterolemia (FH) is the most prevalent autosomal dominant disorder, affecting about 1 in 200-250 individuals, with an estimated 30 million patients globally. It is the leading cause of early and aggressive coronary artery disease (CAD). Objective: To developed an artificial intelligence (AI) and machine learning (ML) algorithm for cardiovascular risk stratification in a FH population, emphasizing sex-specific differences and model explainability. Methods: We analyzed patients with genetically confirmed FH or a score greater than 8 on the Dutch Lipid Clinics Network (DLCN) criteria from the National Registry of the Spanish Atherosclerosis Society, including individuals enrolled from January 2010 to December 2017. The model utilized a comprehensive dataset incorporating family history, clinical characteristics, laboratory results, genetic data, imaging studies, and lipid-lowering treatment details. Eighty percent of the population was allocated for training the AI algorithm, while the remaining 20% was used for testing and 70/30 population for internal validation. A Histogram-based Gradient Boosting Classification Tree was used. The stability of the AI system was assessed using K-fold cross-validation. Shap methodology analyzed the influence of different variables by sex. Youden's J statistic established the optimal cutoff point for identifying very high cardiovascular risk. Results: A total of 1.764 patients were included (51.8% women), among whom 264 experienced major adverse cardiovascular events (MACE), with 8% being women. Notably, 52% experienced a cardiovascular event before age 50, and 37% had subclinical atherosclerosis. The final model incorporated 82 variables, achieving metrics of precision for MACE accuracy (0.92), recall (0.89), F1 score (0.91), and ROC (0.88; 95% CI, 0.85-0.90), revealing significant sex-based differences. Women showed a lower association with MACE compared to men, although this effect diminished with the inclusion of multiple variables, particularly in younger women. In the model, age, GGT levels, and subclinical disease significantly impacted risk for women, while year of birth, age at initiation of statin treatment and HbA1c levels were more influential for men. The optimal risk threshold was 0.25 for the association with of MACE Conclusion: AI-ML algorithms are promising tools for enhancing vascular risk stratification in patients with FH, revealing critical sex-based differences.