Investigating Cardiac Class, Extracardiac Anomalies, and Dysmorphology Patterns Predictive of Mendelian Genetic Disorders in Pediatric Congenital Heart Disease

Abstract

Congenital heart disease (CHD) is the most common class of birth defects, accounting for one-third of all congenital anomalies. There is a need to understand risk factors early in the CHD life course, as half of all mortalities affect infants. Approximately 20-30% of CHD is caused by Mendelian genetic diseases, and genetic risk factors strongly influence health outcomes. However, genetics evaluations are underutilized and unstandardized. We leveraged a clinical program standardizing genetics evaluations for pediatric CHD. Using a test-negative case-control design, we investigated novel phenotypic predictors of patients with Mendelian genetic disorders. In the first study of 588 patients (96% ≤1 year, 20.7% with a genetic diagnosis), we found that dysmorphic status was associated with two-fold increased risk of genetic diagnoses being identified, after adjusting for extracardiac anomalies (ECA) status. In the second work, we developed and applied a novel dysmorphology score for quantifying dysmorphology burden in 1,001 patients (95% ≤1 year, 23.4% with a genetic diagnosis). Using multivariable logistic regression models, we quantified associations between dysmorphology score, ECA status, and genetic diagnoses identified later by genetic testing. A clinical prediction model was developed to improve risk-stratification of patients with genetic disorders more objectively and based on quantification of dysmorphology. Last, we developed a novel method of body region dysmorphology (BRD) classification and investigated how BRD patterns were predictive of cytogenetic and monogenic diagnoses. Dysmorphism of the face, forehead, neck, and hands/feet regions were associated with genetic diagnoses, while adjusting for ECA status. Surprisingly, dysmorphism of the forehead was the strongest predictor of genetic diagnoses, including both chromosomal and monogenic disorders. We found 23.4% (n=234/1001) of patients had genetic diagnoses following standardized testing, and our novel phenotypic predictors, i.e., BRDs and dysmorphology score, will improve identification of rare genetic disorders. However, 8.7-13.1% of apparently isolated/non-dysmorphic patients had genetic disorders identified by genetic testing, highlighting the limitation of phenotype-driven prediction of genetic diagnoses. This work builds a foundation for investigating novel phenotypic associations with genetic causes of CHD, and future research will assess how early recognition of dysmorphology, ECA status, and CHD class predicts other long-term outcomes.