Non-isolated tetralogy of fallot (TOF+): exome sequencing efficacy and phenotypic expansions

Abstract

Tetralogy of Fallot (TOF) is the most common cyanotic congenital heart defect (CHD). TOF may present in isolation or in conjunction with one or more non-cardiac congenital anomalies or neurodevelopmental disorders (TOF+). Uncertainty regarding the efficacy of various genetic testing strategies, and an incomplete understanding of the genetic causes of TOF+, may lead to hesitancy in recommending genetic testing, particularly, clinical exome sequencing (cES). Here, we analyzed cES data from 131 individuals with TOF+. A definitive or probable diagnosis was made for 31 individuals, yielding a diagnostic rate of 23.6% (31/131). One individual received three diagnoses. Commercially available CHD panels would have detected only 27.3% (9/33) to 63.6% (21/33) of the diagnoses made by cES. We then used a machine learning approach to identify four genes for which there is sufficient evidence to support a phenotypic expansion including TOF: DVL3, MED13L, PUF60, and MEIS2. Since chromosomal microarray analysis (CMA) has been reported to have a diagnostic efficacy of 10-20% in individuals with TOF, we conclude that cES should be considered for all individuals with TOF+ for whom a molecular diagnosis has not been established by CMA. We also conclude that TOF represents a low penetrance phenotype associated with genetic syndromes caused by pathogenic variants in DVL3, MED13L, PUF60, and MEIS2.

Fig. 1. Generating and validating TOF-specific rank annotation scores for all RefSeq genes.

A A previously published machine learning algorithm was trained using 53 genes known to cause TOF in humans. Receiver operating characteristic (ROC) style curves were generated based on a leave-one-out cross-validation study analysis performed for each knowledge source (colored lines). The area under the omnibus curve (black) indicates the ability of the algorithm to identify genes in the training set more effectively than chance (diagonal black line). B Box plots showing the algorithmically generated TOF-specific rank annotation scores for the TOF training genes and the four candidate genes, DVL3, MED13L, PUF60, and MEIS2, for which there was sufficient evidence to support a phenotypic expansion involving TOF (Table 1). The median rank annotation scores of these groups—99.6% and 82.5%, respectively—were greater than what would be expected by chance alone (50%; dotted line).