Investigators reported that plasma cell-free DNA (cfDNA) and cardiomyocyte-derived cfDNA (cfDNA-CM) can be isolated and quantified from stable patients with hypertrophic cardiomyopathy (HCM). The study used whole genome bisulfite sequencing (WGBS) and methylation-pattern tissue deconvolution to assign cfDNA fragments to cardiomyocyte origin. The authors framed the primary goal as testing assay feasibility and exploring relationships between cfDNA-CM and clinical characteristics in a chronic HCM cohort.
The study team compared 57 HCM plasma samples with 10 acute myocardial infarction (AMI) plasma samples. Data indicated that total cfDNA yield and the cfDNA-CM fraction were higher in AMI controls than in the HCM cohort (reported p values 0.003 and 0.002, respectively), which the investigators cited as supporting the assay's ability to detect injury-related cardiomyocyte DNA release. The HCM cohort's median cfDNA yield was 12.22 ng with an interquartile range of 7.77 to 18.94 ng.
Within the HCM group, authors observed a wide distribution of cardiomyocyte-derived fraction. The cfDNA-CM fraction ranged from 0% to 6.4% of total cfDNA across samples. The report does not provide per-patient longitudinal sampling or time-since-symptom-onset data; the source did not report those methodological details. The presence of measurable cfDNA-CM in chronic, clinically stable patients was the central empirical finding.
When the HCM cohort was stratified into tertiles by cfDNA-CM fraction, the investigators detected associations between higher cfDNA-CM and cardiometabolic features. Higher tertiles correlated with increased body mass index (BMI) (p = 0.001) and a higher prevalence of BMI ≥ 30 kg/m2 (p < 0.001). The prevalence of obstructive sleep apnea was also greater in higher cfDNA-CM tertiles (p = 0.038). The authors noted that these associations point toward a relationship between cardiomyocyte DNA release and cardiometabolic comorbidity burden in this cohort.
The study team reported a numerical trend toward higher New York Heart Association (NYHA) functional class and more frequent left ventricular outflow tract (LVOT) obstruction among patients in higher cfDNA-CM tertiles; the manuscript does not document those comparisons as statistically significant. Conventional circulating markers of wall stress and myocardial injury—NT-proBNP and troponin, respectively—did not differ across cfDNA-CM tertiles (p = 0.83 and p = 0.29). Authors interpreted that lack of concordance as indicating cfDNA-CM may reflect a distinct domain of myocardial health or injury not captured by these standard biomarkers.
The investigators used WGBS libraries and short-read sequencing to profile cfDNA methylation, then applied tissue deconvolution to estimate cardiomyocyte contribution. The source states the higher yields in AMI samples support assay validity; specific library preparation parameters, sequencing depth, or bioinformatic thresholds are not provided in the preprint. Data underlying the article are not publicly available because of privacy and ethical restrictions. The corresponding author may share anonymized subsets on reasonable request, subject to a data-use agreement and institutional review board (IRB) approval. The research received OHSU IRB oversight, and the authors confirmed participant consent procedures were followed.
The authors acknowledge feasibility but do not report detailed methodological parameters in the preprint; readers cannot assess sequencing depth, fragment-size distributions, or exact methylation markers used from the available text. The analysis is cross-sectional and derived from a single enrolled cohort of stable HCM patients with AMI samples as a comparator; the source did not provide longitudinal outcome data or external validation in independent HCM cohorts. The absence of association with NT-proBNP and troponin leaves unresolved whether cfDNA-CM offers incremental prognostic value or simply reflects cardiometabolic comorbidity. Data access is restricted and any further analysis requires institutional approvals per the authors' data-sharing statement.
The preprint was posted October 13, 2026. The authors declared no competing interests. The manuscript does not contain trial registration identifiers, and no prospective interventional study details were reported.