Rare copy number variants (CNVs) and sequence variants (single nucleotide variants (SNVs) and short insertions/deletions (indels)) have been shown to increase the risk of congenital heart disease (CHD). Whether these variants contribute to clinical outcomes remains unknown.1 Genes implicated as causing CHD may lead to pathological remodeling through activation of fetal genes in response to physiologic stress. We sought to determine the association of predicted damaging genetic variation with ventricular dysfunction in a CHD cohort who underwent exome sequencing (ES) at Columbia University Medical Center (CUMC).
IRB approval was obtained and subjects provided informed consent. Over the study period, patients with CHD were invited to participate. We performed a retrospective study of all CHD proband-parent trios with cardiovascular imaging. Outcomes included a) systolic ventricular function on most recent transthoracic echocardiogram (TTE) and b) worsening systolic function between first and last TTE. Data that support our findings are available upon reasonable request. To investigate whether participants carrying at least one predicted damaging de novo variant or pathogenic CNV were more likely to have an impaired ventricular function phenotype, participants were categorized according to the presence of any predicted damaging variant and the presence of the impaired ventricular function phenotype for each outcome. A binomial test was used to evaluate the frequency of damaging variants between clinical phenotypes. We corrected for multiple testing by assessing a penalty for 16 hypotheses (two outcomes, four gene sets, and two patient groups) yielding a significance threshold of 0.00313.
Ventricular function was dichotomized based on historical reads of TTEs. Worsening function was operationalized as change in qualitative function between studies. Reads were validated in 25 randomly selected TTEs by two blinded, independent readers and via cardiac MRI (CMR) when available.
CHD probands were analyzed for de novo variants utilizing a method validated with Sanger sequencing confirmation. Variant calling was performed using Genome Analysis Toolkit. Variants with an allele frequency <10−4 were considered. Variants were classified into predicted damaging de novo variants (DDVs) and non-damaging variants. Predicted DDVs included likely gene disrupting (LGD) variants and predicted damaging missense variants (Dmis). LGD variants were defined as nonsense (stop-gain) or read-through (stop-loss) SNVs, frameshift indels, or variants predicted to be splice altering by SpliceAI (delta score > 0.8). Dmis variants included de novo missense SNVs predicted to be damaging by REVEL2 score > 0.5. CNVs were called from exome sequencing data using XHMM3 and classified according to a suggested best practice pipeline. DDVs were identified in a) all genes; b) genes in the top 25% of expression during murine heart development (high heart expression; “HHE”)4; c) constrained genes (probability of loss-of-function intolerance (pLI) score > 0.5)5; and d) the intersection of HHE and constrained genes.
504 CHD trios underwent ES at CUMC. 183 DDVs and 9 pathogenic/likely pathogenic CNVs were identified in 164 participants. 88 (18%) had decreased ventricular function at their last echocardiogram. 77 (16%) had worsening ventricular function over a mean of 7.7 years. Inter-reader reliability of ventricular function between blinded-readers and historical reads was high (k=0.91, p<0.0001). Participants with decreased function by TTE had a significantly decreased mean EF (60% vs. 53%, p=0.006) by CMR.
Predicted damaging variants in all genes (RR=1.76; 95% confidence interval [CI], 1.40–2.22; p=0.002), HHE (RR=2.28; 95% CI, 1.57–3.29; p=0.001), constrained (RR=2.18; 95% CI, 1.48–3.21; p=0.002) and both HHE and constrained genes (RR=2.76; 95% CI, 1.72–4.41; p=0.0004) were associated with worsening function (Figure [A]). In 123 participants with a single ventricle, DDVs in all genes (RR=2.67; 95% CI, 1.59–4.48; p=0.003), HHE (RR=4.72; 95% CI, 2.02–11.03; p=0.001), constrained (RR=5.00; 95% CI, 2.16–11.59; p=0.0002) and both HHE and constrained genes (RR=8.33; 95% CI, 2.56–27.09; p=9.11E-05) were associated with worsening function (Figure [B]). In single ventricle participants, DDVs were associated with decreased ventricular function at last echocardiogram (Figure [B]) in HHE genes only (RR=4.91; 95% CI, 1.78–13.57; p=0.001).
To assess for confounding, the association of DDVs with worsening ventricular function was determined in a multivariable logistic regression model that included multiple cardiac surgeries, age at last echocardiogram, and lesion (Figure [C]). In the whole cohort, DDVs in all genes, HHE genes, and HHE and constrained genes remained significantly associated with worsening ventricular function. For individuals with a single ventricle, DDVs in all genes, constrained genes, HHE only, and HHE and constrained genes also remained associated with worsening ventricular (Figure [D]).
In our CHD cohort, predicted damaging variants were associated with worsening ventricular function, particularly in those with a single ventricle, and damaging de novo variants in genes that were both constrained and in the top quartile of heart expression had the strongest association with worsening ventricular function. While our study is limited by its reliance on retrospective data, qualitative assessment of ventricular function and proportion of single ventricle patients, these results suggests that damaging variants may impact myocardial function in select individuals with CHD.
Footnotes
Disclosures: None
References:
- 1.Homsy J, Zaidi S, Shen Y, et al. De novo mutations in congenital heart disease with neurodevelopmental and other congenital anomalies. Science 2015;350:1262–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ioannidis NM, Rothstein JH, Pejaver V, et al. REVEL: An Ensemble Method for Predicting the Pathogenicity of Rare Missense Variants. Am J Hum Genet 2016;99:877–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Fromer M, Moran JL, Chambert K, et al. Discovery and statistical genotyping of copy-number variation from whole-exome sequencing depth. Am J Hum Genet 2012;91:597–607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Zaidi S, Choi M, Wakimoto H, et al. De novo mutations in histone-modifying genes in congenital heart disease. Nature 2013;498:220–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lek M, Karczewski KJ, Minikel EV, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature 2016;536:285–91. [DOI] [PMC free article] [PubMed] [Google Scholar]