The Inherited Basis of Coronary Artery Disease

Authors, Journal, Affiliations, Type, DOI

• Heribert Schunkert (Deutsches Herzzentrum München / TU Munich), Pradeep Natarajan (Mass General Brigham / Broad Institute / Harvard), Nilesh J. Samani (University of Leicester / BHF Centre of Research Excellence)
• N Engl J Med 2026;394:576–587 (CME article)
• Review article — no primary data; synthesis of GWAS, clinical trial, and epidemiologic literature
• DOI: 10.1056/NEJMra2405153

Overview

This comprehensive review by three leading CAD geneticists covers the full spectrum of inherited coronary artery disease: monogenic causes (predominantly familial hypercholesterolemia), polygenic risk scores derived from hundreds of GWAS loci, clinical implementation of combined risk prediction, treatment of genetically high-risk individuals, and the translation of genetic discoveries into novel therapeutics. A key theme is that genetic risk — whether monogenic or polygenic — is not destiny: lifestyle, early statin initiation, and PCSK9 inhibition provide proportionally greater benefit in genetically high-risk individuals. Mendelian randomization is used to distinguish causal risk factors (LDL-C, Lp(a), TG, BP, obesity, T2DM) from confounded biomarkers (HDL-C, CRP, vitamin D, uric acid).

Keywords

Coronary artery disease, familial hypercholesterolemia, GWAS, polygenic risk score, Mendelian randomization, PCSK9, lipoprotein(a), LDLR, APOB, genetics, atherosclerosis, therapeutics

Key Takeaways

Monogenic Causes of CAD

Familial Hypercholesterolemia (FH)

• FH is the archetypal monogenic CAD cause; first described by Müller in 1938; HeFH prevalence ~1/250; HoFH ~1/200,000.
• By mid-forties, ~20% of HeFH individuals have developed atherosclerotic conditions.
• Causative genes: LDLR (impaired LDL receptor function), APOB (altered apoB function), PCSK9 (enhanced function → increased LDLR degradation).
• FH phenotype threshold: LDL-C >190 mg/dL adults (>150 mg/dL children); strongest clinical weight goes to tendon xanthomas, personal or family history of premature CAD, xanthelasma.
• Pathogenic variant prevalence: ~0.4% in general population; ~3.5% in LDL-C >190 mg/dL; 5% in probable FH, 24% in definite FH (by clinical criteria).
• Critical point: Even at similar LDL-C levels, persons with a genetically proven FH variant have a 2–3× higher CAD risk than those without the variant — confirming a genetic risk beyond what LDL-C alone captures.
• Genetic diagnosis enables cascade screening of first-degree relatives — currently massively underused worldwide.
• Treatment goal: LDL-C <70 mg/dL for FH with no ASCVD; <55 mg/dL with ASCVD; lipid-lowering benefit is proportionally greater in genetically proven FH than in phenotypic FH.

Other Monogenic Forms (Rare)

• Autosomal recessive hypercholesterolemia: LDLRAP1 variants — impaired LDL receptor adaptor protein.
• Sitosterolemia: ABCG5/ABCG8 variants — impaired plant sterol excretion; premature CAD.
• Pseudoxanthoma elasticum: ABCC6 variants — premature CAD via calcification of elastic fibres.
• Nitric oxide pathway: GUCY1A1 and PDE5A variants — dysfunctional nitric oxide signalling → premature MI.
• Triglyceride regulation: APOA5 rare variants — elevated TG → premature CAD.
• Cholesterol transport: SCARB1 rare variants — severe early-onset CAD.
• These genes are candidates for expanded molecular genetic evaluation panels alongside the standard FH genes (LDLR, APOB, PCSK9).

Polygenic Contribution to CAD

GWAS Discovery

• Since 2007, large-scale GWAS have identified 346 risk loci reaching genome-wide significance (p <5×10⁻⁸) in meta-analyses including >180,000 CAD cases and >1 million total participants.
• Associated variants affect genes expressed across a wide spectrum of tissues — confirming multiorgan pathways in CAD susceptibility.
• Risk alleles with lower frequency, smaller effect sizes, or population-specific distributions require even larger and more diverse cohorts; sequencing studies will fill this gap.

Polygenic Risk Score (PRS)

• Individual SNPs have tiny effect sizes — not useful alone; but aggregated into a PRS, they create a Gaussian distribution of genetic risk in the population.
• Top 5% of CAD PRS → 3–5× the CAD risk of those in the middle quintiles.
• PRS is independent of family history of CAD — adds risk information beyond clinical history.
• PRS for CAD includes all informative variants (not just genome-wide significant ones) weighted by their association sizes, and performance continues to improve with larger GWAS.

Combined Clinical + PRS Risk (Most Actionable Framework)

• PRS most informative when layered on a clinical risk assessment: relative risk from PRS multiplies absolute risk from clinical score.
• If baseline clinical risk is small, PRS implications are also small (UK Biobank: 1 event prevented per 5,750 screened if applied broadly to 45–70-year-olds).
• Intermediate-risk adults (5–10% 10-year risk): PRS reclassifies ~10% to high risk (2× the event rate) — preventing 1 additional event per 340 screened (~7% of all events). This is the highest-yield use case for CAD PRS.
• Young adults (age 40): Clinical risk scores identify only ~1 in 4 of those who will have a major cardiovascular event. A 40-year-old man with no risk factors in the highest quintile PRS has 30–40% risk of CAD by age 70 vs 10% for the lowest quintile. PRS provides earlier, actionable risk stratification.
• Premature CAD: High PRS predicts recurrent events; clinical actionability somewhat less clear than in primary prevention; may inform therapy intensification when equipoise exists.

PRS Limitations

• Predominantly derived from European-ancestry populations — reduced predictive validity in non-European ancestries (actively being addressed with multi-ancestry GWAS).
• No single-consensus CAD PRS; no agreed reporting standards.
• Not yet incorporated into current clinical guidelines (ESC 2021, AHA PREVENT equations).
• Cost-effectiveness, implementation, and insurance coverage gaps remain.

Treatment of Persons at High Genetic Risk

FH — Intensive LDL Lowering

• Lipid-lowering provides greater absolute benefit in FH than in other primary prevention groups.
• Children with FH + pathogenic variant + persistent hypercholesterolemia: statin to reduce LDL-C <135 mg/dL (European guidelines); LDL-C <70 mg/dL target in adults.
• HoFH: life-threatening, often presenting in childhood; requires specialised centre with evinacumab, lomitapide, and lipoprotein apheresis.

High Polygenic Risk — Lifestyle and LDL Lowering

• Healthy lifestyle nearly counterbalances a high CAD PRS in observational studies.
• Post hoc analyses of primary and secondary prevention trials consistently show greater absolute AND relative risk reductions from statins and PCSK9 inhibitors in high-PRS vs average-PRS patients:
  • JUPITER (rosuvastatin primary prevention): high-PRS patients derive greater statin benefit.
  • FOURIER (evolocumab): high-PRS patients have greater benefit from PCSK9 inhibition.
  • ODYSSEY OUTCOMES (alirocumab): high-PRS associated with greater clinical benefit.
  • Mega 2015 Lancet analysis across multiple statin trials: consistent greater absolute and relative benefit with high polygenic risk.
• In one randomised trial (single site), disclosing PRS led to a reduction in MACE over ~10 years.

Translating Genetic Architecture into Therapeutics

Validated Drug Targets from Genetics

• LDLR variants → statins: Elucidation of LDLR as the FH gene led to the understanding that up-regulating hepatic LDL receptor reduces LDL-C → statin development.
• PCSK9 rare disruptive variants → PCSK9 inhibitors: Rare LOF variants in PCSK9 reduce LDL-C and CAD risk → alirocumab, evolocumab, inclisiran.
• NPC1L1 variants → ezetimibe: Variants in the ezetimibe target gene associate with reduced LDL-C and reduced CAD risk — genetic validation of the drug's mechanism.
• ACLY variants → bempedoic acid: Variants in the bempedoic acid target gene (ACLY) associate with reduced LDL-C and CAD risk — genetic validation.
• LPA (apolipoprotein(a)): The strongest non-LDL GWAS signal; LPA genetic variation determines serum Lp(a) levels (highly heritable); causal for CAD. Agents targeting Lp(a) (RNA-based: olpasiran, muvalaplin, lepodisiran) are in cardiovascular outcomes trials.

Emerging Genetic Targets Under Investigation

• ANGPTL3: Rare LOF variants lower TG and reduce CAD risk → evinacumab (anti-ANGPTL3 mAb, LDLR-independent LDL-C lowering in HoFH) and experimental agents.
• ANGPTL4, APOC3: Genetic LOF → lower TG and CAD risk → APOC3 inhibitors (volanesorsen, olezarsen, plozasiran) and ANGPTL4 targeting in development.
• SVEP1: Gain-of-function amino acid switch → enhanced inflammation in vascular smooth muscle → increased CAD risk; therapeutic target under investigation.
• ADAMTS7: Antibodies and vaccination against ADAMTS7 show anti-atherosclerotic and anti-neointima effects in animal models.
• NLRP3 inflammasome: Genetic analyses prioritised causal inflammatory features (later confirmed in clinical trials — colchicine COLCOT, LoDoCo2).
• SH2B3 (LNK): Common amino acid switch → excess neutrophil extracellular traps (NETs) + arterial thrombosis → increased CAD risk.
• CCM2, TLNRD1: GWAS signals converging on cell-specific signalling in atheroprotective endothelial cell pathways.
• NOS3, GUCY1A1, PDE5: Nitric oxide signalling variants → increased CAD risk + hypertension; PDE5 inhibitor observational data suggests possible benefit but trial data lacking.

Mendelian Randomization — Causal Inference for CAD Risk Factors

Validated Causal Risk Factors (MR positive)

• LDL-C (LDLR locus)
• Triglycerides (LPL, APOC3, ANGPTL loci)
• Hypertension (multiple BP loci)
• Obesity/BMI
• Lipoprotein(a) (LPA locus)
• Type 2 diabetes
• Shorter height (causal)
• Habitual alcohol consumption (causal for CAD risk)

Deprioritised Biomarkers — NOT Causal by MR (despite epidemiological correlation)

• HDL-C: Strong epidemiologic correlation with CAD; multiple HDL-raising drug trials (niacin, CETP inhibitors) failed. MR confirms HDL-C is NOT causally protective — genetic variants modulating HDL-C show no convincing association with CAD risk.
• C-reactive protein (CRP): Epidemiologic correlation; but MR (IL6R MR Consortium, Zacho 2008) shows genetically elevated CRP does not cause CAD — CRP is a marker, not a mediator.
• Vitamin D: Mendelian randomization studies do not support a causal role of vitamin D in CAD.
• Uric acid: Mendelian randomization shows no causal role despite strong epidemiologic association with CAD and cardiometabolic disease.

Limitations of the Document

• Review article — no primary data; all evidence is synthesised and interpreted by the authors.
• PRS evidence primarily from European-ancestry cohorts; conclusions may not generalise to all populations.
• No single-consensus CAD PRS exists; performance metrics reported from multiple different PRS tools with different methods.
• Translation of genetic discoveries into therapies (ADAMTS7, SVEP1, SH2B3) is largely preclinical — clinical validation lacking.
• Long-term cost-effectiveness analyses of CAD PRS implementation are modelling studies, not real-world evidence.

Key Concepts Mentioned

• concepts/Familial-Hypercholesterolemia — monogenic CAD; LDLR/APOB/PCSK9; 2-3× excess risk beyond LDL-C
• concepts/Polygenic-Risk-Score — CAD PRS framework; top 5% = 3-5× risk; combined clinical+PRS risk
• concepts/GWAS-Cardiac-Genetics — 346 CAD loci; genetics-to-therapeutics pipeline; Mendelian randomization
• concepts/Dyslipidemia-Management — LDL-C, TG, Lp(a) as validated causal targets; HDL-C deprioritised by MR
• concepts/Lipoprotein-a — strongest non-LDL CAD GWAS signal; LPA genetic variation; emerging RNA agents

Key Entities Mentioned

• entities/Chronic-Coronary-Disease — primary prevention in genetically high-risk individuals; genetics-guided therapy intensification

Wiki Pages Updated

• wiki/sources/inherited-basis-cad-nejm-2026.md — created
• wiki/concepts/Familial-Hypercholesterolemia.md — other monogenic forms added; 2-3× genetic risk beyond LDL-C added; contradiction updated
• wiki/concepts/Polygenic-Risk-Score.md — dedicated CAD PRS section added; genetics-to-therapeutics; MR findings
• wiki/concepts/GWAS-Cardiac-Genetics.md — CAD GWAS section added (346 loci); MR-validated/deprioritised targets
• wiki/sourceindex.md — new entry added
• wiki/wikiindex.md — updated