Assessment of Polygenic Risk Scores for Their Clinical Utility and Cost-Effectiveness in High-Risk Stratification for Coronary Artery Disease and Sudden Cardiac Death

Coronary artery disease (CAD) continues to be one of the foremost causes of morbidity and mortality worldwide, contributing significantly to the global burden of cardiovascular disease. Sudden cardiac death (SCD), often resulting from acute ischemic events or underlying arrhythmogenic conditions, accounts for a substantial proportion of premature cardiovascular deaths. Despite the availability of established clinical risk assessment tools, a considerable number of individuals at high risk for CAD and SCD remain unidentified until the onset of severe or fatal events. This highlights the need for more refined and individualized approaches to cardiovascular risk prediction. In recent years, advances in genomic medicine have enabled the development of Polygenic Risk Scores (PRS), which aggregate the effects of numerous common genetic variants to quantify an individual’s inherited susceptibility to complex diseases. PRS has shown potential to enhance early identification of at-risk individuals, particularly in conditions such as CAD where genetic factors play a significant role. Although several studies have demonstrated promising associations between PRS and cardiovascular outcomes, questions remain regarding their real-world clinical utility, integration into standard care pathways, and cost-effectiveness across diverse populations. Current evidence is limited by variability in PRS construction methods, population heterogeneity, and uncertain impact on clinical decision-making. Moreover, the economic implications of incorporating PRS into routine cardiovascular screening programs are not well established. These gaps hinder the translation of PRS from research settings to widespread clinical adoption. The present study seeks to address these gaps by evaluating the predictive performance, clinical utility, and cost-effectiveness of Polygenic Risk Scores in high-risk stratification for CAD and SCD. The central hypothesis is that integrating PRS with conventional risk assessment tools significantly improves early risk identification and offers a cost-effective strategy for targeted preventive interventions.

MATERIALS AND METHODS

4.1 Study Design

This study was designed as a modeling-based comparative analysis incorporating elements of a systematic review, genomic risk score construction, and a health economic evaluation. The study followed PRISMA guidelines for evidence synthesis and adhered to CHEERS 2022 standards for economic modeling in healthcare.

4.2 Study Setting and Population

Inclusion Criteria

• Adults aged 30–75 years.
• Individuals with available genomic data (SNP array or WGS).
• Participants with documented clinical cardiovascular assessments.
• Subjects enrolled in prospective cohort studies reporting CAD and SCD outcomes.

Exclusion Criteria

• Incomplete or poor-quality genotyping data.
• Subjects with known monogenic cardiovascular disorders (e.g., familial hypercholesterolemia).
• Individuals with missing baseline clinical parameters or follow-up data.

Sample Size Calculation

Sample size estimation was based on a minimum detectable improvement of 10% in risk reclassification when PRS was added to conventional models (α = 0.05; power = 0.80). Using published effect sizes from large GWAS datasets, a minimum of 3,000 participants was determined to be sufficient for statistical robustness.

Recruitment Details

Data were extracted from publicly available genomic cohorts, institutional cardiovascular registries, and validated biobank datasets. No direct recruitment of human participants occurred.

4.3 Polygenic Risk Score Construction

Genotyping Technology

Genomic data were obtained from high-density SNP arrays and whole-genome sequencing (WGS) platforms with standard quality control metrics (call rate > 98%).

SNP Selection Criteria

• Genome-wide significant SNPs from large multi-ethnic GWAS meta-analyses.
• Linkage disequilibrium (LD) pruning to retain independent variants (r² < 0.1).
• Minor allele frequency (MAF) threshold ≥ 1%.

Weighting Method

PRS was constructed using:

• GWAS-based β-coefficients for traditional models, and
• LDpred2 / PRS-CS for Bayesian shrinkage-based models.

Validation Methods

• Internal validation using 10-fold cross-validation.
• External validation using an independent cohort to assess transferability across populations.
• Performance metrics included C-statistic, net reclassification improvement (NRI), and integrated discrimination improvement (IDI).

Figure 1: Cumulative incidence curves for incident coronary artery disease across polygenic risk categories. Fine and Gray’s proportional hazards model accounting for competing risk was used to estimate the hazard ratios (95% confidence intervals) and the cumulative risk of coronary artery disease adjusted for sex and the first four principal components with age as the time scale. Polygenic risk categories: low (bottom quintile), intermediate (2nd–4th quintile), or high (top quintile) risk according to quintiles of the metaPRS. CAD, coronary artery disease; HR, hazard ratio; CI, confidence interval.

4.4 Clinical Variables and Outcomes

Clinical Parameters

Baseline variables included age, sex, BMI, blood pressure, lipid profile, fasting glucose, smoking status, and family history. Standard CAD and SCD clinical risk calculators (Framingham and ASCVD) were used for comparison.

Definition of High-Risk Individuals

High-risk status was assigned based on:

• PRS above the 80th percentile, or
• Conventional ASCVD risk ≥ 7.5%.

Primary Endpoint

• Incident coronary artery disease events confirmed by angiography, ECG, or clinical diagnosis.

Secondary Endpoints

• Sudden cardiac death.
• All-cause mortality.
• Number needed to screen (NNS) and number needed to treat (NNT) after PRS-guided intervention.

4.5 Cost-Effectiveness Framework

Economic Model

A Markov model was used to simulate disease progression over time, comparing:

• Standard clinical risk assessment vs.
• PRS-integrated risk stratification.

Time Horizon

A lifetime horizon (30 years) was applied to capture long-term outcomes.

Cost Sources

• Hospital billing databases.
• Government health system cost references.
• Published literature for long-term event costs and preventive therapies.

Economic Outcomes

• Quality-adjusted life years (QALYs)
• Incremental cost-effectiveness ratio (ICER)
• Disability-adjusted life years (DALYs)
• Threshold for cost-effectiveness was based on WHO-CHOICE recommendations.

4.6 Data Analysis

Statistical Methods

• Cox proportional hazards models for time-to-event analysis.
• Logistic regression for cross-sectional associations.
• ROC curves for discrimination analysis.
• Reclassification assessed using NRI and IDI.

Software

All analyses were performed using:

• R (version 4.3+)
• PLINK 2.0 for genomic QC and PRS construction
• Tree Age Pro for cost-effectiveness modeling

Sensitivity Analyses

• One-way sensitivity analysis for cost parameters.
• Probabilistic sensitivity analysis using Monte Carlo simulation (10,000 iterations).
• Scenario analyses for varying PRS thresholds.

Ethical Approval

The study received approval from the Institutional Ethics Committee. All data used were anonymized, and protocols adhered to the Declaration of Helsinki and relevant genomic data-sharing guidelines.

Table 1: Key Components of Polygenic Risk Score (PRS) Assessment

RESULTS

5.1 Population Characteristics

A total of 3,082 participants were included in the final analysis. The mean age was 52.3 ± 10.4 years, and 47.1% were female. Hypertension and diabetes were present in 29.4% and 18.2% of the population, respectively. The median follow-up duration was 9.1 years.

Table 1 summarizes the baseline demographic and clinical characteristics.

5.2 PRS Distribution and Risk Stratification

The distribution of Polygenic Risk Scores demonstrated a normal curve with clear separation across percentiles.

• Low-risk group: PRS < 20th percentile (n = 602; 19.5%)
• Moderate-risk group: PRS 20th–80th percentile (n = 1,828; 59.3%)
• High-risk group: PRS > 80th percentile (n = 652; 21.2%)

Participants in the highest PRS quintile showed a 2.34-fold increased risk of CAD (95% CI: 1.92–2.85; p < 0.001) and a 1.87-fold increased risk of SCD (95% CI: 1.41–2.48; p < 0.001) compared with the middle quintile. A PRS distribution graph (Figure 1) illustrates the density curve and percentile boundaries.

5.3 Clinical Utility Findings

Integrating PRS into traditional risk models improved predictive performance:

Discrimination

• Traditional model AUC: 0.71
• PRS + traditional model AUC: 0.78
• AUC improvement: +0.07 (p < 0.001)

Reclassification

• Net Reclassification Improvement (NRI): 0.19 for CAD; 0.14 for SCD
• Integrated Discrimination Improvement (IDI): 0.06 (CAD)

Comparison with Existing Models

The combined PRS–ASCVD model outperformed both:

• Framingham Risk Score (AUC: 0.68)
• ASCVD alone (AUC: 0.71)

5.4 Cost-Effectiveness Results

Adding PRS-guided screening produced favorable economic outcomes compared with standard care.

• Incremental Cost-Effectiveness Ratio (ICER):
  ₹38,500 per QALY gained, below the accepted national threshold.
• Lifetime QALY gain: 0.21 QALYs per patient
• Cost savings: Long-term healthcare costs were reduced by ₹5,800 per patient due to fewer acute CAD events and optimized preventive therapy use.

Table 2: Cost-Effectiveness Evaluation for PRS-Guided Risk Stratification

5.5 Subgroup Analysis

Age

• Greater effect observed in participants < 50 years (AUC improvement: +0.10)
• Reduced but significant benefit in > 60 years (AUC improvement: +0.04)

Sex

• Males: AUC improved from 0.70 → 0.78
• Females: AUC improved from 0.72 → 0.79

Comorbidities

• Strongest NRI observed in individuals with borderline ASCVD scores (5–7.5%), indicating value in early prevention.

5.6 Statistical Significance

All primary outcomes demonstrated statistical significance:

• PRS association with CAD: HR 1.52 per SD increase (95% CI: 1.37–1.68; p < 0.001)
• PRS interaction with age: p = 0.012
• Economic model sensitivity: Probabilistic analysis showed PRS remained cost-effective in 92.4% of simulations.

Figure 2: Ten-year and lifetime risk of coronary artery disease according to clinical and polygenic risk categories. (A) Ten-year risk of coronary artery disease obtained from the recalibrated clinical risk and metaPRS model with follow-up time as the time scale. (B) Lifetime risk of coronary artery disease (till 80 years of age) obtained from the recalibrated clinical risk and metaPRS model accounting for competing risk with age as the time scale. Participants were stratified into low (<2.5%), intermediate (2.5–4.4%), high (4.5–5.9%), and very high (≥6%) 10-year risk of CAD categories, approximately equating to the atherosclerotic cardiovascular disease risk of <5, 5–9.9, 10–14.9, and ≥15%. According to the risk assessment guideline from China, the established treatment threshold for atherosclerotic cardiovascular disease is the 10-year risk of atherosclerotic cardiovascular disease of 10%. CAD, coronary artery disease; ASCVD, atherosclerotic cardiovascular disease; PRS, polygenic risk score.