Key result
Machine learning adding RNA profiling improves diastolic heart failure classification to ~94% auROC.
Why the study?
Pediatric cardiomyopathy is a major cause of diastolic heart failure, yet its underlying mechanisms remain unclear.
Cohort (n=78)
Abnormal lipid signaling and specific transcriptomic signatures define a distinct molecular phenotype of pediatric diastolic dysfunction, suggesting that targeting lipid metabolism (e.g., with semaglutide) may be a promising therapeutic avenue.
Supports lipid signaling investigation in pediatric diastolic HF; leaves open clinical adoption of the ML classifier.
Pediatric cardiomyopathy is a major cause of diastolic heart failure; yet, its mechanisms remain unclear. Global myocardial transcriptomic and blood lipidomic profiling revealed a distinct metabolic signature of diastolic dysfunction marked by dysregulated lipid signaling. A machine learning model using these gene markers accurately classified diastolic dysfunction across cardiomyopathy subtypes. Lipidomic changes included excess saturated lipids and impaired oxidation that correlated with myocardial gene expression. Induced pluripotent stem cell cardiomyocytes from patients with diastolic dysfunction exhibited lipid accumulation, and mitochondrial dysfunction that was rescued by semaglutide. These findings define a new molecular phenotype of diastolic dysfunction and point to abnormal lipid signaling as a promising therapeutic target in childhood cardiomyopathy.
No takes yet. Share an insight, caveat, or question.
Turinsky et al. (2026) conducted a cohort in Pediatric cardiomyopathy (DCM, HCM, RCM, ACM) (n=78). Diastolic dysfunction (and its associated transcriptomic/lipidomic signature) vs. Patients without diastolic dysfunction was evaluated on Classification accuracy (auROC) for diastolic heart failure using a model combining clinical parameters and RNA expression. A machine learning model incorporating clinical and RNA expression features classified diastolic heart failure with an auROC of 94.1%, outperforming clinical features alone (auROC 78.5%) and identifying a signature of abnormal lipid signaling.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: