Clinical characteristics

The following family-based case series represents three independent case presentations for evaluation and management of CHDs and conduction disease. Due to limited intra-family communication, the relationships of these individuals were not known prior to completing the genetic testing. The individuals derive from a family with significant history of cardiac disease. Refer to Figure 1 for the pedigree, with reference to each case below as noted in pedigree nomenclature.

Figure 1. Case series family pedigree. Dark-shaded symbols indicate verified phenotypes and grey-shaded symbols indicate unverified phenotypes likely associated with NKX2.5-related disease. Acronyms: ASD=atrial septal defect, AVB=atrioventricular block, CHD=congenital heart disease, CHF=congestive heart failure, L-SVC=persistent left superior vena cava, LVNC=left ventricular non-compaction, NOS=not otherwise specified, SCD=sudden cardiac death, VT=ventricular tachycardia. Based on family variant segregation and family relationships, several relatives can be determined to have the NKX2.5 variant (in obligate fashion). Note: (+) indicates genetic testing confirmation; otherwise, obligate carriers for the NKX2.5 are noted.

Case 1. The index case proband, which prompted genetic investigation, initially presented to electrophysiology clinic in 2021 after her primary cardiologist retired (Individual IV-1, Fig. 1). She is currently a 10-year-old female; status – post-surgical repair of a large secundum atrial septal defect at a local hospital at age 2. She additionally had a persistent left superior vena cava. She was previously noted to have a family history of CHDs and relatives with pacemakers, though details were unclear and genetic testing was not completed by previous providers (Fig. 1). She also had first-degree atrioventricular block. At age 9, a Holter monitor showed a non-sustained run of monomorphic ventricular tachycardia that has been managed on nadolol. Cardiac MRI and echocardiogram were unremarkable aside from the persistent left superior vena cava, and exercise treadmill testing was normal aside from occasional ectopic atrial rhythm. During the initial visit with our electrophysiology clinic, the family disclosed the recent death of a paternal half-sibling with CHD (Case 2 below; Individual IV-2 in Fig. 1) leading to genetic counselling and testing.

Genetic testing was completed through a commercial clinical genetic testing laboratory (Invitae, Inc.). Analysis identified a pathogenic variant in NKX2.5 [NM_004387.3], denoted as c.437C > A (p. Ser146*). This variant occurs in exon 2 and creates a premature stop signal not expected to cause nonsense-mediated RNA decay but would truncate the last 179 amino acids of the NKX2.5 protein. The homeobox domain and the NK2 domains would be disrupted by this variant. This variant is absent from large population databases (ExAC/gnomAD), and other C-terminal variants in NKX2.5 have been reported in NKX2.5-associated disease. ^{Reference Guntheroth, Chun, Patton, Matsushita, Page and Raskind9} Based on the type of variant and consistency with the patient’s phenotype, this variant was interpreted clinically as pathogenic. There were additional variants of uncertain significance in the following genes, though these are not felt to be clinically significant based on current review: CBL, CHD7, and DSP (see Supplement 1).

After confirmation of the NKX2.5 variant in the setting of non-sustained VT and family history of sudden death, discussion regarding implantable cardioverter-defibrillator took place with the family. Given unclear benefit of implantation, the decision was made to implant a loop recorder for long-term rhythm monitoring. There has been no further arrhythmia to date. Following the identification of the NKX2.5 variant, the family history was reassessed, and family risk counselling was provided. Due to family dynamics, early cascade gene testing was challenging to complete.

Case 2. This individual (Individual IV-2, Fig. 1) had been previously evaluated at our institution, but the relationship with the proband (Case 1) was not known in prior evaluations. She presented to our centre in 2017 from her primary cardiologist for consideration of pacemaker placement due to an underlying diagnosis of first- and second-degree atrioventricular block as well as third-degree atrioventricular block in the setting of her secundum atrial septal defect (surgically repaired). The atrial septal defect was first diagnosed at age 7 and closed surgically by a local hospital, and she was noted to have sinus node dysfunction of unclear cause. The patient had been lost to follow-up for 4 years before presenting to a new local cardiologist.

When she presented to our clinic, Holter monitoring showed first-degree atrioventricular block with periods of second-degree and third-degree atrioventricular block as long as six seconds. No specific cardiovascular symptoms, however, were described. An echocardiogram also demonstrated a persistent left superior vena cava to coronary sinus with no right superior vena cava (similar to Case 1 above), as well as left ventricular non-compaction (Fig. 2). Given her lack of symptoms we completed an exercise treadmill test which showed non-conducted beats with elevated sinus. Left ventricular non-compaction was identified on cardiac MRI, and after discussion with the family, we implanted a dual-chamber epicardial pacemaker. This surgery was uncomplicated though post-operatively she developed incessant atrial flutter not amenable to cardioversion requiring flecainide antiarrhythmic therapy. No genetic evaluation nor genetic testing was completed prior to hospital discharge. The family did not present for scheduled post-operative follow-up in pacemaker clinic. We planned to continue flecainide and warfarin treatment for six months post-surgery and attempted care coordination with local cardiologists after multiple no-shows to our clinic. About one year after pacemaker implant, she presented to local emergency services after cardiac arrest and could not be resuscitated. A post-mortem device interrogation by a local device representative showed ventricular tachycardia events at the time of arrest (although we requested records, they were never sent). No post-mortem genetic testing or molecular autopsy was available.

Figure 2. Clockwise from upper left – baseline ECG in Case 1 showing PR prolongation > 220 msec; non-sustained VT on Holter in Case 1; echocardiography of LV non-compaction in Case 2; and Case 2 atrial flutter with ventricular pacing during mode switch to DDI.

Case 3. This 19-year-old family member (Individual III-5, Fig. 1) has been followed jointly by her local cardiologist and our centre since birth prior to the others. Her relationship with the first two individuals was not connected until about seven months following the evaluation and genetic testing of Case 1. Upon admission to our institution in the setting of recent onset of syncope, the relationship between Case 3 and Case 1 was made.

She was diagnosed with a large atrial septal defect and valvar pulmonary stenosis at birth and underwent transcatheter septal defect device closure at 8 months of life. She had neonatal supraventricular tachycardia though this spontaneously resolved. At age 6, she had symptoms of chest pain and fluttering with an ECG showing a severe PR prolongation and intra-ventricular conduction delay. In the setting of her paternal family history of pacemakers and sudden cardiac arrest of unknown cause, a Holter showed a wide-complex tachycardia suspicious for aberrant supraventricular tachycardia vs. ventricular tachycardia. She underwent electrophysiologic study in 2009, though the findings from this study are not available. At age 7, due to progressive fatigue, evidence of rare second- and third-degree atrioventricular block on Holter monitor, and an unresolved family history of CHDs and malignant arrhythmia, it was decided to place a dual-chamber epicardial pacing system. She continued to demonstrate palpitations and have evidence of atrial and ventricular ectopy prompting initiation of atenolol treatment, which was taken irregularly due to dizziness. Device evaluations showed ventricular pacing percentages as high as 70% during this time. At age 12, she remained symptomatic, showing evidence of AV node dysfunction, and continued palpitations with no recorded arrhythmias, and non-medicinal treatment for anxiety lead to improvement. In the following years, device evaluations showed periodic episodes of non-sustained supraventricular ectopy not associated with symptoms and ventricular pacing < 20%. She remained on atenolol treatment by personal choice rather than discontinuing to assess for her underlying burden or transitioning to a different agent. At 14 years old, she experienced an atrial lead fracture resulting in dizziness. We revised her pacing system to a dual-chamber transvenous system. She continued to ventricularly pace around 15% with minimal evidence of atrial or ventricular arrhythmia such that atenolol was stopped with no evidence of increased arrhythmia. In the following years, she had progressively decreased ventricular pacing from 8–15% and rare episodes of non-sustained atrial tachycardia and wished to remain off medication. In the setting of the COVID-19 pandemic, she did not present for scheduled follow-up evaluation for nearly two years with remote monitoring showing continued rare atrial ectopy. She experienced an episode of unexplained syncope at school, which prompted transfer to our hospital for observation. The team elected to complete a comprehensive cardiomyopathy/arrhythmia/CHD gene panel instead of targeted NKX2.5 analysis given the distance in familial relationship.

The same pathogenic NKX2.5 variant found initially in Case 1 (c.437C > A; p. Ser146*) was identified in Case 3. The results, however, were relatively more complex, with the laboratory reporting an additional likely pathogenic variant in SCN5A (c.4978A > G; p. Ile1660Val). Based on in-depth clinical review, this variant is currently interpreted as an inconclusive variant of uncertain significance that will be re-reviewed annually. ^{Reference Behr, Dalageorgou and Christiansen10,Reference Glazer, Wada and Li11} There were additional variants of uncertain significance in the following genes: SCN5A, ACTN2, and KCNQ1. Based on current review, these are not felt to be clinically significant at this time but will be reassessed regularly in clinic (Supplement 1).

Family History Summary. Figure 1 summarises key findings about the family. Notably, based on family genetic testing and relationships, we can conclude that individuals III-1, II-1, and II-4 also have the NKX2.5 variant; this increases the likelihood that Case 2 also had the NKX2.5 variant even though death preceded confirmatory genetic testing.