Phenotypes and genotypes in a cohort of children with single-ventricle CHD

Abstract

Objective:

CHD is known to be associated with increased risk for neurodevelopmental disorders. The combination of CHD with neurodevelopmental disorders and/or extra-cardiac anomalies increases the chance for an underlying genetic diagnosis. Over the last 15 years, there has been a dramatic increase in the use of broad-scale genetic testing. We sought to determine if neurodevelopmental disorders in children with single-ventricle CHD born prior to the genetic testing revolution are associated with genetic diagnosis.

Methods:

We identified 74 5–12-year-old patients with single-ventricle CHD post-Fontan procedure. We retrospectively evaluated genetic testing performed and neurodevelopmental status of these patients.

Results:

In this cohort, there was an overall higher rate of neurodevelopmental disorders (80%) compared to the literature (50%). More of the younger (5–7-year-old) patients were seen by genetic counsellors compared to the older (8–12-year-old) cohort (46% versus 19% p value = 0.01). In the younger cohort, the average age of initial consultation was 7.7 days compared to 251 days in the older cohort. The overall rate of achieving a molecular diagnosis was 12% and 8% in the younger and older cohorts, respectively; however, the vast majority of did not have broad genetic testing.

Conclusion:

The minority of patients in our cohort achieved a genetic diagnosis. Given a large increase in the number of genes associated with monogenic CHD and neurodevelopmental disorders in the last decade, comprehensive testing and consultation with clinical genetics should be considered in this age range, since current testing standards did not exist during their infancy.

CHD is the most common congenital anomaly, affecting approximately 1% of live births. ¹ CHD can be categorised by complexity of the anatomical defect. ² CHD with single-ventricle physiology includes hypoplastic left heart syndrome, double-inlet left ventricle, tricuspid atresia truncus arteriosus, and transposition of the great arteries. ³ In 2017, the incidence of CHD with single-ventricle physiology was 1.89 per 1000. ⁴ CHD with single-ventricle physiology requires early neonatal intervention with a total of three heart surgeries before 5 years of age. Estimates suggest that up to 50% of all those undergoing neonatal cardiac surgery are subsequently diagnosed with neurodevelopmental disorders, a continuum of phenotypes that include developmental delays, intellectual disability, attention-deficit hyperactivity disorder, and autism. ⁵ CHD often co-occurs with extra-cardiac phenotypes, including other structural birth defects and/or, most commonly, neurodevelopmental disorders. ^6–8 Neurodevelopmental disorders affect 5% of otherwise typical children, 10% of all patients with CHD, and 50% of patients with severe CHD such as CHD with single-ventricle physiology. ^3,9

Studies have shown that genetic aetiologies for CHD include chromosomal aneuploidy, copy number variants, and monogenic causes. ^10,11 Clinical genetic testing suggests that copy number variants contribute to 15–25% of CHD. Furthermore, pathogenic variants in known highly penetrant monogenic genes contribute to approximately 10% of CHD. ¹² While CHD and neurodevelopmental disorders in some children may be attributable to an identifiable genetic syndrome, neurodevelopmental disorders are also independently associated with CHD, especially in cases of cyanotic heart lesions or if open-heart surgery is required as is the case with CHD with single-ventricle physiology. ^3,13 Consistent developmental screening and referral to therapies are widely recommended for this patient population. ^7,9 Many studies have revealed how patients with CHD can be affected by a spectrum of neurodevelopmental disorders, including cognitive impairment learning delay, motor delay, and psychosocial vulnerabilities. ^13–16

The yield of genetic testing in individuals with single-ventricle physiology is not established, and it is unknown whether the combination of neurodevelopmental disorders with isolated, single-ventricle CHD is associated with increased likelihood for a genetic diagnosis, given that neurodevelopmental disorders are independently associated with cyanotic CHD requiring neonatal surgery. From 2010 to 2017, genetic testing evolved rapidly with the advent of large gene panels and, later, exome sequencing. In a study evaluating the expansion of genetic testing, 14,000 more genetic tests were available clinically from March 2014 to August 2017 with an average of 10 new genetic tests being brought to market daily. ¹⁷ However, most of these multi-gene panels (51% of policies) were not covered by insurance in 2015. ¹⁸ Furthermore, exome sequencing became clinically available in 2011 but was not covered by the five largest insurance companies prior to 2015. ¹⁹ Historically, karyotype and SNP microarray were first-line genetic testing for patients with CHD. ¹² Algorithms have been created recently to perform genetic testing on infants with CHD. ^20,21 The EHRA/HRS/APHRS/LAHRS expert guidelines support genetic testing in infants with CHD including consideration for exome sequencing as the results can be re-analysed in the future. ²² Having a comprehensive genetic testing approach for infants in the cardiac ICU has been noted in multiple studies to allow for better management and treatment for these patients. ^20,21,23

This study provided a comprehensive description of neurodevelopment, extra-cardiac anomalies, and diagnostic genetic testing performed in a cohort of children with single-ventricle CHD who were born prior to 2017. This time frame was chosen because of a change in genetic testing practices at Cincinnati Children’s Hospital Medical Center; additionally, these patients are now old enough to have had detailed neurodevelopmental phenotyping. We hypothesise that children born before the widespread use of next-generation diagnostic testing may have unrecognised genetic diagnoses. Our goal was to describe which patients had testing, their phenotype and diagnoses, and if further genetic testing should be considered.

Materials and methods

We performed a retrospective chart review study which was deemed exempt from IRB oversight (IRB protocol #2021-0391). The study cohort was identified through query of a REDcap database of post-Fontan patients seen at Cincinnati Children’s Hospital Medical Center. Patients were included if they were 5–12 years of age at the time of the database query (born prior to 2017), had their initial cardiac operation before day of life 30, and were receiving regular follow-up with cardiology (defined as having at least one visit between August 2020 and July 2021). Deceased patients were excluded if death occurred prior to the age of 5 years as this precluded collection of neurodevelopmental phenotype in our target age range of 5–12 years. This age range was chosen to ensure adequate and accurate phenotype information was available, particularly with respect to neurodevelopmental disorders. Patients with chromosomal aneuploidies were also excluded because this population is typically recognised and diagnosis confirmed early; ^12,23 our goal was to focus on a patient population with more subtle presentation who are presumed higher risk of going undiagnosed. For eligible patients, we collected CHD phenotype, neurodevelopmental disorders phenotype, extra-cardiac anomalies, evaluation by geneticist or genetic counsellor, genetic testing performed, and genetics follow-up. Neurodevelopmental disorders phenotype included gross motor delay, fine motor delay, speech delay, global developmental delay, intellectual disability, attention-deficit hyperactivity disorder, and autism. The cardiac neurodevelopmental clinic includes evaluation by developmental and behavioural paediatricians, therapists, and psychologists employing neuropsychiatric testing, therefore providing objective documentation of patients’ neurodevelopmental status. For patients who were not seen in this clinic, neurodevelopmental status was inferred based on documentation in the patients’ most recent cardiology clinic note and/or problem list in the electronic medical record. All information was reported through Excel (Supplementary Tables 1 and 2).

The final study cohort was divided into two groups based on age (5–7 years and 8–12 years at the time of chart review), and comparisons were made between the two groups. This breakdown was chosen because 2015, the birth year dividing the groups, represented a turning point in availability of genetic testing. We compared genetic testing performed, diagnostic yield of genetic testing, genetics evaluation, CHD phenotype, neurodevelopmental disorders phenotype, and extra-cardiac phenotype between the two groups. The associations between categorical clinical/phenotype variables and abnormal genetic testing were tested using 2 × 2 cross tables. Pearson’s χ² testing was used when all values in the cross table were 5 or greater. When at least one value was less than 5, the Fisher exact two-tailed test was used. Unadjusted p values were tabulated. Reported p values used a threshold of < 0.05 for statistical significance. Statistical tests were performed using SISA (https://www.quantitativeskills.com/sisa/).

Results

Cohort description

The final study cohort included 74 patients with 26 between ages 5 and 7 years (19 male, 7 female) and 48 between ages 8 and 12 years (29 male, 19 female) (Fig. 1). The majority of the patients were non-Hispanic and White (Table 1). The most common CHD in both cohorts was hypoplastic left heart syndrome (7/26, 27% and 20/48, 42%, respectively) followed by double-outlet right ventricle (4/26, 15.3%, and 5/48, 10%, respectively) (Table 2).

Figure 1. Flow diagram of patients evaluated (N = 74). Rectangles represent if seen by genetics or what testing is performed. Ovals represent diagnoses from this testing.

Table 1. Demographics.

Table 2. CHD and NDD phenotype of cohort.

Neurodevelopmental evaluations and phenotypes

The majority of the cohort (24/26, 92% of the younger group and 42/48, 88% of the older group) had been seen at least once in the cardiac neurodevelopmental clinic at Cincinnati Children’s Hospital Medical Center.

Overall, 22/26 (85%) of the younger cohort and 33/48 (73%) of the older cohort had a diagnosis of neurodevelopmental disorder (Table 2). More specifically in the younger group, 14/26 (54%) had global developmental delay, 4/26 (15%) had gross motor delay, 3/26 (12%) had speech delay, and 4/26 (15%) did not have any developmental delays. In the older group, 29/48 (60%) had global developmental delay, 5/48 (10%) had gross motor delay, 1/48 (2%) had speech delay, and 13/48 (27%) did not have any developmental delay.

Genetic evaluations

Of the 26 patients in the younger group, 11 (42%) were seen by a geneticist while 12 (46%) were seen only by a cardiovascular genetic counsellor with an average age of 7.7 days at initial evaluation (Table 1). Subsequently, 23/26 (89%) had genetic testing recommended by geneticist or genetic counsellor with 17 patients completing testing which was comprised of microarrays, gene panels, and exome sequencing. The patients who were not evaluated by a geneticist or genetic counsellor did not have any genetic testing ordered. SNP microarray was performed in 17/26 (65%) of patients, resulting in identification of two diagnostic copy number variants including 17p12 deletion and 20q13.2 duplication. Three patients had heterotaxy gene panels (CFC1, FOXH1, NODAL, and ZIC3) performed (3/15, 20%) which did not identify any causal variants. Exome sequencing was performed on one patient (1/15, 7%) which revealed a de novo pathogenic variant in KMT2D (c.10394dupG, p. Pro3466Thrfs*2) resulting in the diagnosis of Kabuki syndrome seen in this hypoplastic left heart syndrome patient. The majority of those who had genetic testing recommended were not seen again by genetics after initial evaluation in the cardiac ICU, and no further genetic testing was completed (Table 1 and Fig. 1).

Twelve (46%) of the 26 younger patients had CHD, neurodevelopmental disorders, and extra-cardiac anomalies. These extra-cardiac anomalies included multiple organ systems of which neurologic and skeletal systems were the most commonly affected (Fig. 2). Three (25%) of these 12 patients had a documented genetic diagnosis. In contrast, none of the 14 patients with isolated CHD or CHD plus neurodevelopmental disorders had a documented genetic diagnosis. The children with extra-cardiac phenotypes (structural birth defects and/or neurodevelopmental disorders) had more testing, both initial SNP microarray and further testing with panels and exome sequencing, than those with fewer symptoms (Table 3).

Figure 2. Percentage of patients with certain Anomalies black bars represent percentage of patients with isolated cardiac defects. Dark grey bars represents percentage patients with cardiac and NDD. Light grey bars represents percentage of patients with cardiac, NDD, and ECA.

Table 3. Genetic testing performed and diagnostic yield.

Of the 48 patients in the older group, 23 (48%) were seen by a geneticist while 9 (19%) were seen only by a cardiovascular genetic counsellor (Table 1), with an average age of 251 days at initial genetic evaluation. All patients who saw a geneticist or cardiovascular genetic counsellor had genetic testing recommended, and 29 patients completed testing which comprised of karyotypes, microarrays, and gene panels. The patients who were not evaluated by a geneticist or genetic counsellor did not have any genetic testing ordered. Three other patients were offered microarray but declined. Among the 29 patients tested, 6 had normal karyotypes and 25 had normal chromosomal microarrays. Three had abnormal chromosomal microarrays with diagnostic abnormalities identified (4q28.3 deletion, 16p11.2 deletion, and 22q11.2 duplication). One patient with heterotaxy had an array which identified a 14.0 Mb region of homozygosity 7p21.1p15.1 which includes DNAH11. Subsequent DNAH11 sequencing identified a homozygous pathogenic variant (c.4348C > T, p. Arg1450*) consistent with a diagnosis of primary ciliary dyskinesia. Thirteen patients (13/25, 52%) were tested with sequencing panels which included heterotaxy (CFC1, FOXH1, NODAL, and ZIC3), Noonan syndrome, or autism/intellectual disability genes. Three patients had multiple gene panels performed. From these panels, no causal variants were found. Exome sequencing was recommended and ordered for a single patient but was denied by insurance in 2013. The majority of the patients who had an initial genetics evaluation did not have any follow-up visit, and no further genetic testing was completed (Table 1 and Fig. 1).

Furthermore, 20/48 (42%) of the older group had CHD, neurodevelopmental disorders, and other extra-cardiac congenital anomalies. These congenital anomalies affected multiple body systems with the most common being pulmonary and gastrointestinal (Fig. 2). Of these patients, 3/20 (15%) were found to have a genetic diagnosis through microarray compared to none of the isolated CHD and 1/20 (5%) in the CHD plus neurodevelopmental disorders group. The children with extra-cardiac phenotypes (structural birth defects and/or neurodevelopmental disorders) had more testing, both initial SNP microarray and panels, than those with less symptoms (Table 3).

Overall, the diagnostic rate of microarray was similar in the younger patients (2/23, 9%) and older patients (3/32, 9%). A monogenic aetiology was confirmed in 1/15 (7%) of the younger group and 1/26 (4%) in the older patient group. Diagnostic yield was higher in patients with combination of CHD, neurodevelopmental disorders, and ECA, with 6/32 (19%) of these patients, across all ages, achieving a diagnosis (Table 3). Among the patients with CHD + neurodevelopmental disorders, only 16/30 (53%) were tested and testing was limited to microarray in the younger cohort, while three received panel testing in the older cohort. However, the majority of patients with isolated CHD did not receive genetic testing: only 3 of 4 with isolated CHD were tested, all in the younger cohort (Table 3).

Discussion

This study provided a comprehensive description of neurodevelopment, extra-cardiac anomalies, and diagnostic genetic testing performed in a cohort of children with single-ventricle CHD who were born prior to 2017. Furthermore, all patients underwent neonatal cardiac surgery and had functional single-ventricle morphology, putting them at an expected higher risk versus commonly reported figures that include less severe cardiac morphology and a higher proportion of non-neonatal first cardiac surgery. The study compared, secondarily, two different age groups (5–7 and 8–12 years old) to describe their single-ventricle CHD, neurodevelopmental disorders, and extra-cardiac anomalies phenotype as well as genetic testing offered as there were drastic changes in genetic evaluation and testing over the course of those 7 years. Interestingly, within our cohort of 5–7 year olds and 8–12-year-olds with CHD with single-ventricle physiology, we saw a higher rate of neurodevelopmental disorders affecting 57/74 (77%) compared to reported 50% of children with severe CHD requiring cardiac intervention. ⁹ This increase is possibly due to the fact that 24/26 (92%) of the younger patients and 42/48 (88%) of the older patients were seen in the cardiac neurodevelopmental clinic. Evaluation was performed by developmental and behavioural paediatricians, psychologists, and therapists, and there was neuropsychological testing to evaluate for neurodevelopmental disorders. Not all patients have access to these types of clinics, but paediatricians and paediatric cardiologists should have high suspicion for neurodevelopmental disorders in these patients and have a low threshold to refer to therapies and developmental and behavioural paediatricians.

In this entire cohort, 55/74 (74%) were evaluated by a geneticist or a genetic counsellor. Strikingly, when comparing these two cohorts, more of the younger patients were seen by genetic counsellors compared to the older cohort (46 and 19%, respectively with p value = 0.01). In the younger cohort, the average age of initial consultation was 7.7 days compared to 251 days in the older cohort. This reveals that genetics evaluation and expertise were utilised sooner following the recommendations at the time to have genetics evaluation and microarray performed. Furthermore, previous studies have revealed that genetic panels with known CHD and neurodevelopmental disorders genes should be offered to all families with CHD to allow for diagnosis of monogenic aetiology of disease which microarray and karyotypes will miss. ²⁰ At the time of initial genetic evaluation for much of our cohort, large genetic panels including CHD-specific genes, Noonan syndrome panels, or autism/intellectual disability panels were not performed as this specific testing was not available. Prior to 2015, most private insurance companies also were not covering multigene panels, and exome sequencing was not covered by any private insurance companies. ^18,19 Therefore, genetic testing was particularly limited in the older cohort. Looking through a contemporary lens, we believe these patients should be followed up in genetics and given the opportunity for comprehensive genetic testing with the increased likelihood of making a diagnosis where CHD and neurodevelopmental disorders co-exist. In addition, since neurodevelopmental disorders cannot be evaluated or diagnosed perioperatively or at a young age, follow-up is necessary to expand genetic testing based on the phenotype of the patient at later time points. Also, genetics follow-up is necessary to allow for gene discovery or more information to be collected about specific variants found in patients. For example, the heterotaxy panel in the older cohort only evaluated four genes (CFC1, FOXH1, NODAL, and ZIC3) compared to current panels assessing 30+ genes. In addition, a VUS can be re-interpreted based off new data years after original molecular testing was performed and can be changed in classification from benign through pathogenic. In this cohort, a previously classified VUS in CFC1 (c.433G > A) has been reclassified as benign in two patients. On the other hand, one patient in the older cohort had a VUS in CDH2 (c.2629_2630delAG, p. Ser877Stop) reclassified as likely pathogenic (Supplementary Tables 1 and 2).

The American College of Medical Genetics and Genomics has developed clinical guidelines recommending that patients with at least one congenital anomaly or developmental delay have exome or genome sequencing as a first- or second-tier diagnostic test. ²⁴ Therefore, patients with CHD with or without neurodevelopmental disorders or other anomalies should be offered this testing. As exemplified by our cohort in which only 1 out of 74 5–12-year-olds with single-ventricle CHD had exome sequencing, children with single-ventricle CHD in this age range are less likely to have been offered genetic testing as recommended by current standards. Therefore, these patients should be evaluated or re-evaluated by a geneticist or genetic counsellor to discuss the options of more genetic testing including exome sequencing. Even if the patient had been seen by genetics as an infant with a normal microarray or limited-gene panel, they should be re-evaluated due to the advent of new testing options with more genes evaluated as well as a better understanding of different variants in known genes.

Our study does have several limitations. It is a single-center retrospective case series which analysed existing data, some of which could be incomplete. Our centre sees high-acuity patients, and there may therefore be selection bias in our relatively small sample size. Finally, our biggest limitation is that we were not able to directly prove our hypothesis, which would require sequencing the untested and undiagnosed patients. However, previous studies have found that de novo variants account for ∼ 28% of patients with CHD, NDD, and ECA. ²⁵ In our study, only 19% (6/32) of those with CHD, NDD, and ECA achieved a genetic diagnosis (CNVs and monogenic). Therefore, we can infer that at least 10% of these individuals in our study may have an undiagnosed genetic condition. Future directions would include sequencing these patients and evaluating for highly penetrant monogenic aetiologies of their CHD, NDD, and ECA.

Conclusion

Our experience suggests considerable under-investigation of current 5–12-year-old patients with single-ventricle CHD in both the provision of entry level and more advanced genetic testing. With the increased recognition of neurodevelopmental disorders in children with severe CHD as well as the increase in the number of genes with monogenic aetiology, de novo variants, and copy number variants, re-evaluation by genetics should be considered for this population. With the expansion of genetic testing along with better insurance coverage, these children may not have had the opportunity to obtain the same genetic testing and counselling now available. Given the high association of neurodevelopmental disorders with single-ventricle CHD, the association between genetic diagnosis and CHD with neurodevelopmental disorders, and lack of ability to distinguish whether genetics and/or surgery underlie neurodevelopmental disorders in this population, it is particularly important for these children to be re-evaluated. Paediatricians and paediatric cardiologists are likely in the best position to refer these children and their families back to clinical genetics and/or genetic counsellors so that they can receive the most up-to-date recommended care.