Arrhythmias are the most frequently encountered cardiac complications during pregnancy in patients with and without underlying structural heart disease. The management of arrhythmias can be particularly delicate given the potential drug toxicities and ionizing radiation related to fetal exposure. This case explores the diagnostic dilemma, definitive management, and the supportive literature of the underlying diagnosis associated with a wide complex tachycardia presenting during pregnancy.
A 35-year-old woman, 25 weeks pregnant, was admitted to an outside hospital with a 3-week history of palpitations and dyspnea. Beta blockade and escalating doses of sotalol were attempted without success. Upon transfer to our center, vital signs revealed a heart rate of 140 bpm, blood pressure of 100/60 mmHg, and an oxygen saturation of 94%. The initial ECG (Figure 1) revealed a wide complex tachycardia at 140 bpm, left bundle branch block (LBBB) morphology with transition at V5, predominant inferior axis with positivity in leads I and aVL, and negative in aVR. A 1:1 ventriculo-atrial conduction was apparent. A dose of intravenous adenosine was given, which revealed ventriculo-atrial block with brief termination of the tachycardia (Figure 2) and rapid re-initiation, consistent with an automatic ventricular tachycardia (VT). A rapid bedside echo demonstrated an ejection fraction of 10-15% with a dilated left ventricle and mildly increased right ventricular size. Fetal monitoring revealed a healthy baby in no distress.
The patient’s medical history was unremarkable except for mild obesity, with her current pregnancy (G1P1A0) complicated by gestational pre-diabetes. Family history was relevant for sudden death in her father at age 55 from autopsy-proven myocardial infarction, death from pancreatic cancer in her mother at age 40, and 3 healthy siblings except for uninvestigated palpitations in an older sister. There was no family history of channelopathy, seizure disorder, or stroke.
Course in Hospital
Multiple concerns were present at this time: the patient’s short-term hemodynamic instability, the risk of preterm labor during any potential intervention, the risk of precipitating cardiogenic shock upon conversion of VT to sinus rhythm, and the viability of the fetus in the event of pre-term labor or need for delivery. A decision was made to proceed with Impella (ABIOMED) 3.5L/min insertion and a coronary angiogram, which revealed normal coronaries. After discussion with both the high-risk obstetrics team and the patient, weighing the benefits against the potential for harm to the fetus (specifically, the potential for fetal hypothyroidism), amiodarone was initiated. Transthoracic echo images post Impella are shown in Video 1A, Video 1B, and Video 1C.
After 48 hours, VT persisted and remained sympathetically driven with lower rates at 90-110 bpm. In the setting of a slower VT, it was possible to consider overdriving it with atrial pacing at a rate just above the VT rate. The risk related to the placement of an endocardial electrode was felt to be minimal and would also afford some protection in the event of arrhythmia termination and bradycardia. Active fixation leads were implanted via the internal jugular vein to the atrial and ventricular positions, and attached to an externalized pacemaker. Atrial pacing at 110 bpm was successful in overdriving the VT (Figure 3). The baseline QRS during sinus rhythm revealed an atypical LBBB and QRS duration of 140 ms (Figure 4). Within several days, a repeat echo revealed an improvement in the ejection fraction to 25%; however, a fever with a sudden increase in the white blood cell count prompted removal of the Impella, with a resultant VT rate increase to the 130s despite amiodarone. Although the obstetrics team was ready to deliver the baby, fetal cerebral organogenesis can occur at 32 weeks’ gestation, so other options were explored given that she was currently at 26 weeks.
The decision was made to take the patient to the EP lab, where the procedure was performed under general anesthesia with continuous fetal monitoring and obstetrics present in the room to proceed with urgent C-section if necessary. A 3D mapping system and intracardiac echo were used for an entirely fluoroless procedure. The VT was incessant despite general anesthesia. The activation map confirmed a focal VT, localized to the anterolateral right ventricle just underneath the tricuspid valve (Figure 5). Sufficient catheter force and stability was particularly difficult to achieve in this location, but was possible using a medium-curve bidirectional sheath looping into the right ventricle and pulling up with counterclockwise/clockwise torque. The VT was terminated several times with catheter manipulation, with fortunate spontaneous re-initiation. Radiofrequency energy using an irrigated catheter (30 watts, 43°C) was applied to a location with earliest activation (-20 msec to surface QRS), which was accompanied by a short burst of VT firing prior to termination. Further consolidation lesions were delivered without reappearance of the tachycardia after a waiting period. An endocardial voltage map in sinus rhythm noted low voltage at the base and the lateral RV that extended ‘epicardially’ to the unipolar voltage map; however, given the known limitations of right ventricular voltage mapping, particularly relating to inadequate contact force and vector direction during point procurement and adjudication, there was cautious interpretation.
In sinus rhythm, the HV interval was found to be prolonged at 71 ms, with easily induced episodes of complete heart block after catheter manipulation at the RV septum. This precluded an endomyocardial biopsy which had originally been intended to rule in a diagnosis of myocarditis.
The patient remained in sinus rhythm thereafter, with an atypical LBBB (Figure 6), and amiodarone was replaced by nadolol. An echo revealed an impressive increase of the LVEF to 50%. She went on to deliver a healthy baby girl at 38 weeks’ gestation, without evidence of amiodarone-induced neonatal hypothyroidism. A post-partum cardiac MRI revealed the presence of a nonischemic cardiomyopathy, LVEF 49%, with a mildly dilated LV. Mild structural anomalies were noted, including regional non-compaction with increased trabeculation (Figure 7A), and hypertrophy of the anterolateral papillary muscle with a thin papillary muscle base and divided inferolateral papillary muscle heads (Figure 7B). Borderline high native myocardial T1 and a normal T2 were consistent with diffuse mild fibrosis. At 6 months post procedure, the patient is asymptomatic, her ECG remains unchanged, and she has maintained her LVEF.
The initial ECG reveals a wide complex tachycardia of atypical LBBB morphology. Ventriculo-atrial activity is apparent as ‘notching’ immediately post QRS, particularly noticeable in lead V1. Adenosine given during the tachycardia results in ventriculo-atrial block (15th to 18th QRS, Figure 2), confirming a diagnosis of VT. Adverse effects to the fetus would not be expected with adenosine, given that it is unlikely the drug will reach the fetal circulation because of adenosine’s short half-life.1 The VT morphology suggested a right ventricular origin potentially originating at the moderator band, septal papillary muscle or tricuspid valve annulus given the late precordial transition and strong positivity in leads I and aVL, and discrepancy in the inferior leads. However, a deeply negative aVR more strongly suggested a lateral tricuspid valve annular origin.
Given the acute hemodynamic instability, the location of the VT, and the atypical LBBB during sinus rhythm, the diagnosis pointed towards a myocarditis or a pre-existing undiagnosed cardiomyopathy. The diagnosis of post-partum cardiomyopathy was excluded given this was the patient’s first pregnancy. Genetic testing sent during the patient’s hospital course revealed a missense heterozygous mutation in the Protein Kinase AMP-Activated Non-Catalytic Subunit Gamma 2 (PRKAG2), suggesting a possible clinical diagnosis of glycogen storage cardiomyopathy, but was classified as a variant of unknown significance (VUS).
PRKAG2 syndrome is an autosomal dominant metabolic heart disease characterized by left ventricular hypertrophy (LVH), conduction abnormalities, and ventricular preexcitation2; however, other less classic presentations exist without the presence of preexcitation.3 The prevalence of PRKAG2 syndrome is approximately 0.23-1% in patients with suspected hypertrophic cardiomyopathy (HCM).3 The mutations modify the tridimensional structure of 5’-AMP-activated protein kinase (AMPK), thereby impairing myocardial glycogen uptake.4 This results in cardiac hypertrophy, improper glycogen accumulation, ion channel dysfunction, and altered atrio-ventricular septation.5 A systematic review of all reported PRKAG2 mutations, including 193 patients, noted that a trend toward certain phenotypic features was associated with specific mutations.6 C.905G>A (Arg302Glen) and c.1463A>T (Asn488Ild) were the most common mutations with 110 and 40 cases (57% and 21%, respectively).6 C.905G>A was significantly more likely to be associated with preexcitation, a higher frequency of syncope, and rate of pacemaker implantation, whereas c.1463A>T patients had a higher frequency of LVH.
In our patient, the absence of affected family members prevents a definitive diagnosis due to the inability to demonstrate segregation. However, our patient’s findings were particularly interesting and could represent an atypical presentation given the presence of unexplained eccentric hypertrophy, conduction disease, and an arrhythmia originating from the annulus. Collaboration with another center following a family with the same PRKAG2 VUS revealed a form of HCM, with no clear sudden cardiac death history and no known sustained supraventricular or ventricular arrhythmias. Investigations regarding the patient’s siblings are currently underway, and consideration for implantation of a biventricular ICD will depend on her clinical course independent of the genetic finding.
This case highlights the intricacies of managing an arrhythmia in a pregnant patient. The involvement of a multidisciplinary team for decision-making and treatment was paramount. Finally, the importance of keeping an open mind to alternative diagnoses was particularly helpful for both acute and long-term management.
Disclosures: The authors have no conflicts of interest to report regarding the content herein.
- Qasqas SA, McPherson C, Frishman WH, Elkayam U. Cardiovascular pharmacotherapeutic considerations during pregnancy and lactation. Cardiol Rev. 2004;12(4):201-221.
- Banankhah P, Fishbein GA, Dota A, Ardehali R. Cardiac manifestations of PRKAG2 mutation. BMC Med Genet. 2018;19(1):1.
- Pöyhönen P, Hiippala A, Ollila L, et al. Cardiovascular magnetic resonance findings in patients with PRKAG2 gene mutations. J Cardiovasc Magn Reson. 2015;17:89.
- Ha AC, Renaud JM, Dekemp RA, et al. In vivo assessment of myocardial glucose uptake by positron emission tomography in adults with the PRKAG2 cardiac syndrome. Circ Cardiovasc Imaging. 2009;2(6):485-491.
- Zaha VG, Young LH. AMP-activated protein kinase regulation and biological actions in the heart. Circ Res. 2012;111(6):800-814.
- Porto AG, Brun F, Severini GM, et al. Clinical spectrum of PRKAG2 syndrome. Circ Arrhythm Electrophysiol. 2016;9(1):e003121.