A 79-year-old woman with a medical history of hypertension, hyperlipidaemia and hypothyroidism presented with progressive dyspnoea and bilateral lower extremity oedema. On admission, she was found to have new-onset AF with rapid ventricular response and a heart rate of 157 BPM (Figure 1). Chest radiography revealed bilateral interstitial oedema and small pleural effusions, and B-type natriuretic peptide was markedly elevated (13,173 pg/ml, normal ≤540 pg/ml). Rate control therapy was initiated with IV diltiazem, followed by digoxin. Transthoracic echocardiography showed a severely reduced left ventricular ejection fraction of 21%, consistent with newly diagnosed heart failure with reduced ejection fraction, likely tachycardia-induced, and the diltiazem was discontinued. She was started on low-dose metoprolol tartrate 12.5 mg twice daily. Transoesophageal echocardiogram-guided cardioversion successfully restored sinus rhythm. Her heart rate variability post-cardioversion ranged from 55 to 65 BPM.
A post-cardioversion ECG showed a Bazett-corrected QT interval of 430 ms (Figure 2). The patient was started on an oral loading dose of amiodarone a few hours after her cardioversion was successful. However, while preparing for hospital discharge and off telemetry (6 hours after her cardioversion), she suffered a witnessed cardiac arrest due to pulseless VF. Return of spontaneous circulation was achieved with defibrillation and one cycle of CPR. Review of telemetry prior to her cardiac arrest was notable for a QTc of 447 ms (Figure 3). Post-resuscitation, her rhythm was junctional bradycardia (Figure 4). At the time of arrest, her potassium level was 3.9 mmol/l (normal 3.6–5.2 mmol/l), magnesium level was 1.9 mg/dl (normal 1.7–2.3 mg/dl) and liver/kidney function was stable at baseline. No drug–drug interactions were identified.
A second arrest occurred in the intensive care unit with pulseless ventricular tachycardia, requiring CPR and defibrillation. Of note, her QTc was 629 ms after the first arrest and 600 ms after the second arrest, respectively. Emergent coronary angiography showed non-obstructive coronary disease. A temporary intra-aortic balloon pump and Swan–Ganz catheter were placed. Electrolytes remained within normal limits, and post-arrest echocardiogram confirmed persistent left ventricular ejection fraction at 25% with severe global hypokinesis. Her β-blocker was discontinued.
Telemetry revealed frequent ventricular ectopy and torsades de pointes (TdP), associated with early afterdepolarisations (EAD)-type premature ventricular contractions (Figure 5). A temporary transvenous right ventricular pacemaker was inserted to facilitate overdrive pacing and suppress ventricular ectopy. Amiodarone was discontinued, and lidocaine was initiated to shorten the QT interval. However, the patient reverted to AF with rapid ventricular response on day 6 of her admission, and she was restarted on amiodarone with concurrent ventricular overdrive pacing. With this treatment, sinus rhythm was not restored, but there was no recurrence of ventricular arrhythmias. She then underwent a second successful transoesophageal echocardiogram-guided cardioversion on day 15.
Given the patient’s episodes of VF and ventricular tachycardia (VT) without reversible pathology, she met the criteria for secondary prevention of sudden cardiac death. Due to difficulty controlling her rapid ventricular response, atrioventricular node ablation was also planned. With her low ejection fraction and anticipated frequent ventricular pacing, a biventricular ICD was implanted and programmed with a lower rate limit of 90 BPM. She was discharged home in a stable condition on guideline-directed medical therapy for heart failure with reduced ejection fraction, including metoprolol succinate, losartan, spironolactone and a loop diuretic. Amiodarone was continued at a maintenance dose. Outpatient follow-up was arranged with electrophysiology for possible atrioventricular node ablation and ongoing rhythm monitoring.
This case illustrates tachycardia-induced cardiomyopathy complicated by malignant ventricular arrhythmias and possible drug-induced QT prolongation or bradycardia-induced VT/VF. Rhythm control in AF, along with careful QT surveillance, is crucial for patients in preventing life-threatening arrhythmias in heart failure with reduced ejection fraction.
Discussion
We present a case of TdP ventricular tachycardia associated with EADs that was precipitated by prolongation of the action potential duration of ventricular myocardial cells in combination with QT interval prolongation. This occurred following cardioversion for rapid AF and after initiation of amiodarone, an antiarrhythmic medication with IKr blocking effects.1 During repolarisation, there is a net outward flow of positive ions that begins to increase after phase 0 of the action potential and ultimately brings the membrane potential back to its resting diastolic level. This net outward current is a combination of outward potassium currents (such as IKs and IKr) and opposing inward currents (such as INa and ICaL) that occur during phases 1, 2 and 3 of repolarisation.2,3
When the outward potassium currents are reduced, or the inward sodium and calcium currents are increased, or both, repolarisation slows. This delay can allow for EADs, voltage oscillations that take place during the repolarisation phase. EADs are more likely to occur at slower heart rates or longer cycle lengths, which extend the action potential duration, slow repolarisation and give more time for inward currents that cause them to be activated. Conversely, at shorter cycle lengths, repolarisation accelerates, and the action potential duration shortens, reducing the likelihood of EAD formation.4
In our patient, the extremely fast ventricular rate before cardioversion and the slower heart rate afterwards both played an important role. A rapid ventricular rate during AF leads to excessive calcium entry into ventricular cells via L-type calcium channels. After cardioversion, when the heart rate slows down, this excess calcium is removed by the Na–Ca exchanger, which, in turn, increases intracellular sodium levels. When this occurs in the setting of delayed repolarisation due to an IKr-blocking drug (such as sotalol or amiodarone), it can promote further development of EADs that can trigger TdP.
The combination of bradycardia-induced repolarisation abnormalities, drug effects on ion channels and the arrhythmogenic substrate of diseased myocardium synergistically increases the risk of ventricular arrhythmias, especially TdP and VT. Another differential for ventricular tachycardia in our case is catecholaminergic polymorphic VT, an inherited arrhythmia syndrome characterised by adrenergically mediated bidirectional and/or polymorphic ventricular tachycardia, most often presenting in children and adolescents, but also seen in adults.5 The syndrome is most commonly caused by pathogenic variants in the cardiac ryanodine receptor gene, with rarer involvement of other genes, such as CASQ2. Catecholaminergic polymorphic VT is a significant cause of autopsy-negative sudden cardiac death in the young, often triggered by physical or emotional stress in the absence of structural heart disease or baseline ECG abnormalities.
A large, retrospective multicentre study by Tovia-Brodie et al. noted an incidence of 0.2% of ventricular arrhythmias post-electrical cardioversion for AF among 11,897 patients.6 Congestive heart failure, hypertension and use of QT-prolonging drugs were found in 74% of patients who developed arrhythmias, and all these risk factors were noted in our patient. Additionally, the median time to index event was 28.5 hours, and our patient developed her first event 30 hours after cardioversion. In their study, recurrent arrhythmias occurred in 39% of patients, 13 ± 15 hours after the index ventricular arrhythmia event.6 It is important to note that awareness of this condition represents an opportunity to avoid premature closure, as her initial event may have been attributed to cardiogenic shock from myocardial stunning post-resuscitation.
The vast majority of ventricular arrhythmias are related to QT prolongation, resulting from marked and sudden slowing of heart rates, treatment with Class III antiarrhythmic drugs or other QT-prolonging medications and the ventricular repolarisation remodelling effect of electrical cardioversion.7 Previous studies have shown that the period shortly after electrical cardioversion for AF is more susceptible to QT prolongation caused by antiarrhythmic drugs, due to a steeper QT-RR slope during this post-cardioversion phase compared with the period before cardioversion.8 Interestingly, the telemetry machine algorithm missed the patient’s QT prolongation, further highlighting the role of careful data interpretation at the bedside.
The American Heart Association states that QTc measurements from bedside monitors (telemetry) should not be considered equivalent to, or used interchangeably with, standard 12-lead ECGs for serial comparison, due to differences in lead configuration, signal quality and algorithmic variability. Automated and semiautomated telemetry systems may introduce both fixed and proportional bias, and studies have shown only moderate correlation and poor agreement with 12-lead ECGs, as automated systems may not reliably identify the end of the T wave or correct for heart rate accurately, particularly in AF or at high heart rates.9 While ambulatory ECGs can provide long-term QT dynamics, there is significant technical variability, including challenges in T-wave end identification, lead selection, and the influence of autonomic tone, circadian variation and medications. There is also a lack of standardisation in measurement protocols, and individual QTc values can differ widely between Holter and standard ECG, limiting clinical utility for diagnostic purposes.10
It is worth noting that most patients who undergo electrical cardioversion for AF undergo a short observation period, and many are discharged on the same day. A recent study by Younis et al., in which patients with AF underwent continuous 7-day Holter monitoring beginning just before electrical cardioversion, showed that a dynamic and significant QT prolongation was observed in 43% of patients, with peak QT durations occurring at an average of 44 hours post-cardioversion.11 This raises concern that some arrhythmias may occur at home following discharge, which may translate to a higher rate of post-cardioversion ventricular arrhythmias than described in studies.6
Our patient was initially paced with temporary pacing, increasing the heart rate and preventing bradycardia or pause-dependent TdP. Despite post-cardioversion dispersion of repolarisation being a potentially reversible state, an ICD was implanted for secondary prevention, given her reduced ejection fraction and potential for recurrent ventricular arrhythmias. While the decision to implant an ICD in such cases is complex and requires multidisciplinary discussion, we elected to implant the ICD, as she continued to have short-long-short runs of premature ventricular contractions, and EADs that only temporised with pacing at a faster heart rate. In a retrospective analysis, Tovia-Brodie et al. showed that 48% of patients had a permanent pacemaker or ICD implanted, and two patients, both without a permanent pacemaker or ICD, had died due to malignant ventricular arrhythmias within 72 hours of the index event.6 Once the patient had an ICD implanted, we were safely able to resume amiodarone for its beneficial antiarrhythmic effects.
Conclusion
EADs and polymorphic ventricular tachycardia can develop after electrical cardioversion of AF, particularly with rapid ventricular response and with the use of antiarrhythmic medications. This potentially life-threatening phenomenon can occur later than same-day discharge that many centres implement. Further large-scale studies are necessary to identify predictors of ventricular arrhythmias after electrical cardioversion and patient subgroups at risk, for which extended monitoring post-cardioversion may be warranted.
Clinical Perspective
- Torsades de pointes ventricular tachycardia can occur after electrical cardioversion for AF, especially when QT-prolonging antiarrhythmic drugs, such as amiodarone, are used. This is often due to early afterdepolarisations triggered by delayed repolarisation, excessive intracellular calcium and bradycardia post-cardioversion.
- The risk of ventricular arrhythmias is heightened in patients with congestive heart failure, hypertension and those on QT-prolonging drugs, with events often occurring 24–48 hours post-cardioversion.
- Standard telemetry may fail to detect QT prolongation, emphasising the need for vigilant bedside assessment.
- Given the reversible, but high-risk, nature of repolarisation abnormalities, ICD implantation or extended monitoring should be considered in high-risk patients to prevent sudden arrhythmic death.