Convolutional neural networks (CNNs) have emerged as a novel method for evaluating heart failure (HF) in adult electrocardiograms (ECGs). However, such CNNs are not applicable to pediatric HF, where abnormal anatomy of congenital heart defects plays an important role. ECG-based CNNs reflecting neurohormonal activation (NHA) may be a useful marker of pediatric HF. This study aimed to develop and validate an ECG-derived marker of pediatric HF that reflects the risk of future cardiovascular events. Based on 21,378 ECGs from 8324 children, a CNN was trained using B-type natriuretic peptide (BNP) and the occurrence of major adverse cardiovascular events (MACEs). The output of the model, or the electrical heart failure indicator (EHFI), was compared with the BNP regarding its ability to predict MACEs in 813 ECGs from 295 children. EHFI achieved a better area under the curve than BNP in predicting MACEs within 180 days (0.826 versus 0.691, p = 0.03). On Cox univariable analyses, both EHFI and BNP were significantly associated with MACE (log10 EHFI: hazard ratio [HR] = 16.5, p < 0.005 and log10 BNP: HR = 4.4, p < 0.005). The time-dependent average precisions of EHFI in predicting MACEs were 32.4%-67.9% and 1.6-7.5-fold higher than those of BNP in the early period. Additionally, the MACE rate increased monotonically with EHFI, whereas the rate peaked at approximately 100 pg/mL of BNP and decreased in the higher range. ECG-derived CNN is a novel marker of HF with different prognostic potential from BNP. CNN-based ECG analysis may provide a new guide for assessing pediatric HF.

Background: Many prenatal factors are reported to be associated with congenital heart defects (CHD) in offspring. However, these associations have not been adequately examined using large‐scale birth cohorts. Methods and Results: We evaluated a data set of the Japan Environmental and Children's Study. The primary outcome was a diagnosis of CHD by age 2 years. We defined the following variables as exposures: maternal baseline characteristics, fertilization treatment, maternal history of diseases, socioeconomic status, maternal alcohol intake, smoking, tea consumption, maternal dietary intake, and maternal medications and supplements up to 12 weeks of gestation. We used multivariable logistic regression analysis to assess the associations between various exposures and CHD in offspring. A total of 91 664 singletons were included, among which 1264 (1.38%) had CHD. In multivariable analysis, vitamin A supplements (adjusted odds ratio [aOR], 5.78 [95% CI, 2.30–14.51]), maternal use of valproic acid (aOR, 4.86 [95% CI, 1.51–15.64]), maternal use of antihypertensive agents (aOR, 3.80 [95% CI, 1.74–8.29]), maternal age ≥40 years (aOR, 1.59 [95% CI, 1.14–2.20]), and high maternal hemoglobin concentration in the second trimester (aOR, 1.10 per g/dL [95% CI, 1.03–1.17]) were associated with CHD in offspring. Conclusions: Using a Japanese large‐scale birth cohort study, we found 6 maternal factors to be associated with CHD in offspring.

Background: Variants in the LZTR1 (leucine‐zipper‐like transcription regulator 1) gene (OMIM #600574) have been reported in recessive Noonan syndrome patients. In vivo evidence from animal models to support its causative role is lacking.Methods: By CRISPR‐Cas9 genome editing, we generated lztr1‐mutated zebrafish (Danio rerio). Analyses of histopathology and downstream signaling were performed to investigate the pathogenesis of cardiac and extracardiac abnormalities in Noonan syndrome.Results: A frameshift deletion allele was created in the zebrafish lztr1. Crosses of heterozygotes obtained homozygous lztr1 null mutants that modeled LZTR1 loss‐of‐function. Histological analyses of the model revealed ventricular hypertrophy, the deleterious signature of Noonan syndrome‐associated cardiomyopathy. Further, assessment for extracardiac abnormalities documented multiple vascular malformations, resembling human vascular pathology caused by RAS/MAPK activation. Due to spatiotemporal regulation of LZTR1, its downstream function was not fully elucidated from western blots of adult tissue.Conclusion: Our novel zebrafish model phenocopied human recessive Noonan syndrome and supported the loss‐of‐function mechanism of disease‐causing LZTR1 variants. The discovery of vascular malformations in mutants calls for the clinical follow‐up of patients to monitor for its emergence. The model will serve as a novel platform for investigating the pathophysiology linking RAS/MAPK signaling to cardiac and vascular pathology.