Clinical Impact of Rare Variants in Racial and Ethnic Minorities with Atrial Fibrillation

Atrial fibrillation (AF) remains the most common clinical arrhythmia. Although aging and acquired comorbidities contribute to most cases, genetic variation is associated with AF development and adverse outcomes. Rare, likely pathogenic or pathogenic (LP/P) arrhythmia or cardiomyopathy gene variants have been identified in up to 10% of individuals with early-onset (EO) AF of predominantly European ancestry, and are associated with increased mortality¹. Recent clinical guidelines reflect these observations, advising genetic testing in select patients developing AF before age 45². However, racial and ethnic minorities are underrepresented in most genomic datasets, and whether these prognostic relationships apply to non-White individuals in a real-world clinical cohort is unknown. To address this gap, we performed whole-exome sequencing (WES) in non-Hispanic Black (NHB) and Hispanic/Latinx (H/L) individuals with AF from an urban academic medical center, identifying rare LP/P arrhythmia/cardiomyopathy variants. We then examined the association between rare LP/P variants and a composite of mortality, hospitalization, and stroke, hypothesizing that rare LP/P variants are associated with adverse outcomes in NHB and H/L individuals with AF.

Adults 18 years or older at AF diagnosis were enrolled by written informed consent between August 25, 2015, and May 19, 2019 (University of Illinois Chicago Institutional Review Board #2015–0681, Figure 1A). WES followed methods established by the National Heart, Lung, and Blood Institute (NHLBI) Trans-Omics for Precision Medicine (TOPMed) Atrial Fibrillation Study (available at https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs001062.v5.p2). Genes were identified from cardiomyopathy and arrhythmia panels (Invitae, Ambry Genetics, GeneDx; current as of March 16, 2025), and analysis was limited to those with strong/definitive gene-disease associations per ClinGen Gene-Disease Validity Framework (https://search.clinicalgenome.org/kb/gene-validity, accessed October 14, 2025; Figure 1B). Pathogenicity of variants in the canonical transcript with minor allele frequency ≤1% was graded by an automated platform (Franklin, Genoox, Palo Alto, CA, USA) verified by manual review. The primary outcome was a composite of all-cause mortality, hospitalization with primary discharge diagnosis of AF or decompensated heart failure (HF), or stroke. Vital status was queried from the United States Social Security Administration Limited Access Death Master File, with censoring on February 23, 2023. Risk associated with LP/P status was modeled using a Cox proportional hazards regression model, adjusting for age at AF diagnosis, sex, race-ethnicity, and left ventricular ejection fraction (LVEF). The proportional hazards assumption was assessed using scaled Schoenfeld residuals and inspection of residual plots.

Figure: Rare variants in non-Hispanic Black and Hispanic/Latinx individuals with atrial fibrillation (AF).

A) Study flowchart with exclusion criteria, number of clinical events, and duration of follow-up. B) Genes analyzed from commercial testing panels with strong or definitive evidence supporting gene-disease linkage. Genes in bold denote those in which rare likely pathogenic/pathogenic (LP/P) variants were identified in this cohort. Asterisk denotes genes in which variant is present in European genetic ancestry group in gnomAD. C) Clinical characteristics of the AF cohort stratified by LP/P variant carrier status. Data were obtained closest to AF diagnosis and are reported as number (%) except where noted. Fisher’s exact test compared categorical variables, and the Kruskal-Wallis test evaluated continuous and ordinal variables. Missing data were as follows: estimated glomerular filtration rate (4 total, 2 LP/P carriers), left ventricular end diastolic diameter (4 total, 1 LP/P carrier), left atrial size (2 total, 0 LP/P carriers), left atrial diameter (6 total, 1 LP/P carrier). Early-onset was defined as diagnosis before 65 years of age. D) Proportion of variants identified classified by strong or definitive association to cardiomyopathy or arrhythmia syndrome based on ClinGen framework (DCM = dilated cardiomyopathy, HCM = hypertrophic cardiomyopathy, ARVC = arrhythmogenic right ventricular cardiomyopathy, CPVT = catecholaminergic polymorphic ventricular tachycardia, SQTS = short QT syndrome, LQTS = long QT syndrome). E) Proportion of variant carriers stratified by age category at AF diagnosis. F) Cumulative incidence of composite outcome of all-cause mortality, stroke, and hospitalization with primary discharge diagnosis of AF or heart failure, stratified by LP/P variant status. Adjusted hazard ratio (aHR) and 95% confidence interval (CI) were obtained from Cox proportional hazard models including LP/P variant presence, age at AF diagnosis, sex, race-ethnicity, and left ventricular ejection fraction (LVEF).

Cohort characteristics are presented in Figure 1C. Among LP/P carriers, history of HF and family history of HF were more frequent, and mean left ventricular end diastolic diameter was higher. There were no statistically significant differences in other comorbidities, echocardiographic measurements, or treatments. Nineteen LP/P variants were identified in 18 patients, with TTN (n=3) the most common gene identified, followed by KCNH2 and PKP2 (n=2 each; Figure 1B). One subject carried an LP/P variant in both TTN and RYR2. Variants in dilated cardiomyopathy genes were most frequent (23.1%, Figure 1D). LP/P variants were detected in a higher proportion of individuals <65 years at AF diagnosis, though not statistically significant (9.5% vs. 3.7%, p=0.088, Figure 1E). During follow-up, the primary outcome occurred in 165 individuals, with Kaplan-Meier estimated cumulative incidence of 58.7% (95% CI 51.6%−64.7%) at the median (interquartile range [IQR]) follow-up time of 6.7 (4.3–9.8) years. After multivariable adjustment, LP/P status was associated with increased hazard for the composite outcome (adjusted hazard ratio [aHR] 2.00, 95% CI 1.16–3.47, p=0.013, Figure 1F). Decline in LVEF was associated with modestly increased hazard (per 10% decrease, aHR 1.15, 95% CI 1.02–1.30, p=0.018).

Our findings in a real-world, ethnically diverse cohort complement recent works characterizing the prevalence and clinical impact of rare variants in AF. Previously, our group identified LP/P variants from 60 candidate genes in 7% of ethnic minority probands with AF³. Analyzing a broader arrhythmia/cardiomyopathy panel, we identified LP/P variants in 7% of participants overall, and 9.5% in participants with EOAF, approximating the 10% reported in individuals of European descent¹. We also observed that LP/P status in NHB and H/L individuals was associated with a two-fold increased risk of adverse outcomes. This is consistent with single-center data showing an association between LP/P variants and all-cause mortality¹, though it contradicts a UK Biobank analysis, which found no association between LP/P variants and a composite of stroke, HF hospitalization, or death⁴. Healthy volunteer bias and differences in variant filtering and adjudication of clinical outcomes may partially explain these discrepancies. Notably, a recent pooled cohort analysis of Thrombolysis in Myocardial Infarction (TIMI) participants demonstrated that cardiomyopathy LP/P variants were associated with increased mortality, incident HF, and HF hospitalization⁵. Further research is needed to support these findings in other minority groups (i.e., Asian/Pacific Islander, Native American, and Alaska Native), and to assess gene subgroups, individual clinical outcomes, incident cardiomyopathy, and LVEF-variant interactions. Nonetheless, these results further support the potential prognostic value of rare deleterious variants in individuals with AF across race and ethnicity.

Nonstandard Abbreviations and Acronyms:

AF

atrial fibrillation

LP/P

likely pathogenic/pathogenic

EO

early-onset

WES

whole exome sequencing

NHB

non-Hispanic Black

H/L

Hispanic/Latinx

HF

heart failure

LVEF

left ventricular ejection fraction

aHR

adjusted hazard ratio