2020 Volume 84 Issue 10 Pages 1679-1685
The health crisis due to coronavirus disease 2019 (COVID-19) has shocked the world, with more than 1 million infections and casualties. COVID-19 can present from mild illness to multi-organ involvement, but especially acute respiratory distress syndrome. Cardiac injury and arrhythmias, including atrial fibrillation (AF), are not uncommon in COVID-19. COVID-19 is highly contagious, and therapy against the virus remains premature and largely unknown, which makes the management of AF patients during the pandemic particularly challenging. We describe a possible pathophysiological link between COVID-19 and AF, and therapeutic considerations for AF patients during this pandemic.
The global pandemic of coronavirus disease 2019 (COVID-19) has caused more than 1 million infections and a hundred thousand deaths.1 As of 18 May 2020, there were more than 4,500,000 patients infected globally with over 300,000 deaths. The disease is caused by the severe acute respiratory coronavirus 2 (SARS-CoV-2), a single-stranded RNA virus belonging to the family Coronaviridae.2 Although the majority of COVID-19 patients (81%) present with mild illness, more than 15% have developed severe disease to multi-organ decompensation.3 These critical patients requiring intensive care are usually older and have underlying comorbidities (e.g., coronary artery disease, hypertension, diabetes and cerebrovascular disease).4 The severe complications include acute respiratory distress syndrome (ARDS) (29%), acute cardiac injury (12%), and secondary infection (10%).5 COVID-19 patients with cardiac injury have a higher incidence of malignant arrhythmias and mortality rate than those without.6 The incidence of arrhythmias varies in different studies, ranging from 5.9% to 16.7%.4,6
Atrial fibrillation (AF) is the most common arrhythmia associated with aging and a variety of cardiovascular comorbidities, so it is not surprising that new-onset or preexisting AF is frequently observed in COVID-19 patients. The Italian Ministry of Health reported concurrent AF in 19–22% of COVID-19 patients.7,8 In the latest pharmacological trial using remdesivir in 53 hospitalized COVID-19 patients, 3 patients developed new-onset AF (6%).9
The management of AF is complex and includes rhythm and rate control, and prevention of stroke. The COVID-19 pandemic has brought unprecedented and significant challenges in the management of AF patients with COVID-19. Therefore, we review the evidence of COVID-19 and AF from the relevant pathophysiology to clinical management, solicit pivotal issues, and provide provisional recommendations (Table 1). The rationales behind the recommendations are further explained.
| Medical control | |
| Rate control | 1. Avoid the use of verapamil or digoxin with hydroxychloroquine 2. Remdesivir has no interaction with drugs for rate control |
| Rhythm control | 1. Avoid the combination of amiodarone, drondedarone, flecainide and propafenone with hydroxychloroquine or azithromycin 2. Remdesivir has no interaction with drugs for rhythm control |
| Oral anticoagulants | 1. Reduce the dosage of NOACs as hydroxychloroquine or azithromycin would increase NOAC concentration 2. Anticoagulants could be used aggressively for the prevention for venous thromboembolism, but the strategy for stroke prevention still depends on risk score 3. Remdesivir has no interaction with NOACs |
| Cardioversion | 1. Medical treatment and chemical cardioversion is first considered if necessary 2. Electrical cardioversion might be avoided |
| Catheter ablation | |
| Elective ablation | 1. Elective ablation should be avoided 2. Avoid transesophageal echocardiography |
| Urgent ablation | 1. Urgent ablation for severely symptomatic, drug and/or electrical cardioversion refractory, or preexcited AF with syncope or cardiac arrest might be performed with caution and protection 2. Screening test for COVID-19 should be performed before the procedure 3. Use of esophageal temperature probe should be avoided 4. A dedicated catheter laboratory with independent air-supply and ventilation, or negative pressure should be reserved for suspicious COVID-19 cases 5. Hospital-based infection control protocol for ablation procedure should be constructed |
| Follow-up after catheter ablation | |
| 1. In-person clinic visit should be avoided, and instead telehealth/virtual visits could be adopted to minimize unnecessary exposure |
|
| 2. Instead of conventional monitoring device such as 24-hour Holter monitoring, remote monitoring of heart rhythm via mobile device is suggested |
|
AF, atrial fibrillation; COVID-19, coronavirus disease 2019; NOAC, non-vitamin K antagonist oral anticoagulant.
Angiotensin-converting enzyme 2 (ACE2) converts angiotensin I (Ang I) and angiotensin II (Ang II) into the biologically active peptides Ang-(1–9) and Ang-(1–7), respectively,10,11 which provides counter-regulatory effect of ACE in the renin-angiotensin-aldosterone system (RAAS). ACE2 could be a functional receptor and cellular entry point for SARS-CoV-2 and SARS-CoV to invade target cells in the heart, lungs, and kidney where prominent expression of ACE2 occurs.12,13 Although the current evidence remains insufficient, SARS-CoV-2 may directly injure the myocardium by binding to ACE2.14 Previous data about SARS-CoV might provide a theoretical mechanism for COVID-19 because SARS-CoV and SARS-CoV-2 have similar affinity for ACE2.12,13 The downregulation of myocardial ACE2 expression after SARS-CoV infection leads to excessive accumulation of Ang II and causes myocardial injury, remodeling and even adverse cardiac outcomes.15 This might underlie AF arrhythmogenesis in COVID-19. Atrial ACE2 expression would decrease the levels of transforming growth factor-β1 (TGF-β1) and collagen deposition.16 ACE2 might be able to regulate the cardiac action potential.17 Laboratory data have also demonstrated that ACE2 metabolites, Ang-(1–7), could modulate electrophysiological characteristics and calcium hemostasis in atrial tissue and pulmonary vein cardiomyocytes.18,19 Downregulation of ACE2 in COVID-19 might increase AF vulnerability and its perpetuation. In animal studies, treatment with ACE inhibitors (ACEIs) and angiotensin-receptor blockers (ARBs) might have increased the expression of ACE2 receptor.20 Therefore it was hypothesized that the risk for COVID-19 infection could be increased and linked to poor outcomes in patients receiving ACEIs/ARBs. However, the latest data suggest these drugs should be continued in COVID-19 patients with hypertension or heart failure (HF) as the outcomes and risk for COVID-19 infection seemed neutral even under treatment with ACEIs/ARBs.20
Cytokine Storms and AFSARS-CoV-2 triggers an immune response with a cytokine storm syndrome.21 In the majority of severe cases there are elevated levels of infection-related biomarkers and inflammatory cytokines, such as tumor necrosis factor α (TNF-α), interleukin (IL)-6, IL-1β, IL-2, IL-7, IL-10, granulocyte colony-stimulating factor, granulocyte macrophage colony-stimulating factor, interferon-γ-induced protein, monocyte chemoattractant protein 1, fibroblast growth factor, vascular endothelial growth factor, and macrophage inflammatory protein 1-α.5,14,21 A cytokine storm might trigger a violent immune system response, and cause ARDS and hypoxemia, which indirectly damage myocardial cells.5,14,21,22 Additionally, some cytokines have direct effects on cardiomyocytes. These factors might induce atrial electrical and structural remodeling, and lead to AF occurrence.
Although the electrophysiological effects of cytokines remain incompletely understood, distinctive cytokines have been shown to induce AF in bench work and translational research.23,24 TNF-α increases AF vulnerability and exerts direct effects on atrial structural and electrical remodeling.25,26 TNF-α induced triggered activity in pulmonary vein cardiomyocytes, which was attributed to increased activity of the sodium-calcium exchanger (NCX) and impaired sarcoplasmic reticulum (SR) ATPase.25 Abnormal Ca2+ leakage from the SR could also be induced by TNF-α.27 Both TNF-α and IL-1β impair cardiac contractility,26–28 and contribute to arrhythmogenesis through mechano-electrical feedback. IL-6 reduces cardiac connexins and promotes electrical remodeling during acute inflammation.29 Different cytokines, including TNF-α, IL-1β, fibroblast growth factor, and vascular endothelial growth factor, all increase cardiac fibrosis and underlie arrhythmogeneity.30,31 Activation of inflammasome NLRP3-related signals was hypothesized after SARS-CoV-2 infection,32 and increased activity of the NLRP3 inflammasome has been similarly observed in AF patients, which might be associated with macrophage recruitment, fibrosis, and myocardial dysfunction in the heart.33,34 Several therapeutic trials have been launched on the basis of these hypotheses. Hydroxychloroquine (HCLQ), which is a derivative of chloroquine and a potential therapeutic candidate for COVID-19, could reduce the activity of the NLRP3 inflammasome in inflammatory cells.35 The immunosuppressive agent, tocilizumab (an IL-6 inhibitor), was approved by the US Food and Drug Administration for a phase 3 trial in severely ill COVID-19 patients in order to reduce the complications of the infection.
Many investigational drugs have been proposed for COVID-19. Safety alerts due to drug side effects, including long QT interval and ventricular arrhythmias, were raised for HCLQ and azithromycin.36 This red flag should be emphasized when these drugs are prescribed together with AF medications for rate and rhythm control. AF medications include β-blockers, non-dihydropyridine (DHP) calcium-channel blockers (CCB), digoxin and antiarrhythmic agents.37 The use of these medications increases the risk for serious side effects such as bradycardia, HF, long QT interval, or torsades de pointes. Therefore, a guideline of medical treatment for AF has been provided by international consensus.37 Drug-drug interaction between AF and COVID-19 medications might share a common metabolism and therefore side effects (Table 2). In addition to an evaluation of baseline renal function, liver function and ECG parameters when initiating medications, thoughtful consideration of drug interactions, with adjustment, and monitoring, is critical for patient safety and therapeutic efficiency.
| Remdesivir | Hydroxychloroquine | Azithromycin | |
|---|---|---|---|
| Rate control drugs | |||
| β-blockers | |||
| Atenolol | – | – | – |
| Bisoprolol | – | – | – |
| Metoprolol | – | – | – |
| Propranolol | – | – | – |
| CCB | |||
| Diltiazem | – | – | – |
| Verapamil | – | ↑ | – |
| Others | |||
| Digoxin | – | ↑↑ | – |
| Rhythm control drugs | |||
| Amiodarone | – | ↑↑↑ | ↑↑↑ |
| Dronedarone | No available data | ↑↑↑ | ↑↑↑ |
| Flecainide | – | ↑↑↑ | ↑↑ |
| Propafenone | – | – | ↑↑↑ |
| Oral anticoagulants | |||
| Apixaban | – | ↑ | ↑↑ |
| Dabigatran | – | ↑↑ | ↑↑ |
| Edoxaban | – | ↑↑ | ↑↑↑ |
| Rivaroxaban | – | ↑ | ↑↑ |
| Warfarin | – | – | – |
↑↑↑, Potential substantially increased exposure of the medications; these drugs should not be prescribed together. ↑↑, Potential moderately increased exposure of the medications; dosage adjustment or close monitoring may be required. ↑, Potential mildly increased exposure of the medications; the interactions are weak. –, No significant effects. Data about drug-drug interactions presented in this table were adopted from the work by the Liverpool Drug Interactions Group (https://www.covid19-druginteractions.org).64 CCB, calcium-channel blockers; COVID-19, coronavirus disease 2019.
Most β-blockers can be safely used with investigational COVID-19 medications. Verapamil is P-glycoprotein (P-gp) related and metabolized by CYP3A4-type cytochrome P450-dependent enzymes (Table 3). Because HCLQ has an effect on the modulation of the hyperpolarization-activated current If in the sinoatrial node and atrioventricular node cells, co-administration of verapamil and HCLQ may cause bradycardia and conduction disturbance.38 Digoxin is well-known for its narrow therapeutic range.39 Moreover, investigational COVID-19 medications that consist of the substrate of CYP3A4 and P-gp, such as HCLQ, may potentially increase the digoxin level and its toxicity, resulting in conduction disturbance.40 Therefore, ECG monitoring for bradycardia and conduction disturbance should be considered.
| Medications that are P-glycoprotein substrates or inhibitors | |
| β-blockers | Bisoprolol (substrate/inhibitor), propranolol (substrate/inhibitor) |
| CCBs | Diltiazem (substrate/inhibitor), verapamil (substrate/inhibitor) |
| Antiarrhythmic agents | Amiodarone (inhibitor), dronedarone (inhibitor), propafenone (inhibitor) |
| Oral anticoagulants | Apixaban (substrate), dabigatran (substrate), edoxaban (substrate), rivaroxaban (substrate), warfarin (substrate/inhibitor) |
| Others | Digoxin (substrate) |
| COVID-19 drugs | Azithromycin (substrate/inhibitor) |
| Medications that are CYP3A4 substrates | |
| CCBs | Diltiazem, verapamil |
| Antiarrhythmic agents | Amiodarone |
| Oral anticoagulants | Apixaban, rivaroxaban |
| COVID-19 drugs | Hydroxychloroquine, azithromycin |
AF, atrial fibrillation; CCB, calcium-channel blocker; COVID-19, coronavirus disease 2019.
HCLQ is notoriously known for drug-induced QT prolongation.36,41–43 Moreover, azithromycin also induces QT prolongation.36 The administration of antiarrhythmic agents, including amiodarone, dronedarone, flecainide and propafenone, with HCLQ or azithromycin may further increase the risk of QT prolongation and induce ventricular tachyarrhythmia or sudden cardiac death (Table 2). Therefore, ECG monitoring of the QT interval is important. Drugs should be either withheld or removed if the QT interval is >500 ms, as generally recommended. The increased plasma levels of antiarrhythmic agents while being used with these antiviral drugs would potentially increase the risk of malignant arrhythmias and sudden cardiac death.
In addition to ventricular tachyarrhythmia, co-administration of antiarrhythmic agents and investigational COVID-19 medications could increase the blood concentration of Class Ic and Class III antiarrhythmic drugs, and expose the patients to risk of atrial proarrhythmic events. The increased blood concentrations of Class Ic and III antiarrhythmic agents, including propafenone and amiodarone, would change the conduction properties of atrial tissue. For example, persistent atrial flutter may occur with increased atrial refractory period.44–46
COVID-19 is associated with a hypercoagulable state.47,48 COVID-19 patients with acute respiratory failure present with a severe hypercoagulability rather than consumptive coagulopathy. Fibrin formation and polymerization may predispose to thrombosis and correlate with a worse outcome.47 The incidence of venous thromboembolism could be as high as 25% in COVID-19 patients with severe pneumonia.49 In a different cohort, the composite outcome of symptomatic acute pulmonary embolism, deep vein thrombosis, ischemic stroke, myocardial infarction, or systemic arterial embolism was 31%.50 The stroke rate in COVID patients is around 1–2%,50,51 but the incidence of stroke is not as high as that of venous thromboembolism. Evidence is still needed to establish a causal relationship between stroke and COVID-19. The International Society on Thrombosis and Haemostasis and the American Society of Hematology recommend prophylactic low-molecular-weight heparin for venous thromboembolism in all hospitalized COVID-19 patients in the absence of any contraindications.52 However, supporting evidence for prophylactic anticoagulant therapy for stroke in AF patients with COVID-19 remains sparse. We do not know whether anticoagulants should be administered more aggressively in COVID-19 patients with AF, even under low CHA2DS2-VASc score (0 for males and 1 for females). The hypercoagulable state in severe COVID-19 disease is frequently observed with concurrent multi-organ failure, which would also significantly increase the risk of bleeding.
Interactions Between Oral Anticoagulants (OACs) and COVID-19 TreatmentsStroke prevention with OACs, including vitamin K antagonist (warfarin) and non-vitamin K antagonist OACs (NOACs), is the cornerstone of managements of AF. However, before initiation of OACs for AF patients infected with COVID-19, it is necessary to check platelet count, coagulation profile (e.g., prothrombin time and activated partial thromboplastin time), and renal and liver functions, which are probably affected by severe infection. Importantly, these parameters should be monitored and followed up closely during the disease course, and the dosages of OACs should be adjusted when indicated.
Warfarin It is known that warfarin has multiple drug-drug and drug-food interactions. However, there have not been any significant interactions between warfarin and remdesivir or HCLQ (Table 2). For patients being initiated with warfarin, the international normalized ratio (INR) should be monitored every day until the therapeutic range (INR 2–3) is achieved. Thereafter, the INR could be followed up when clinically indicated.
NOACs NOACs can directly inhibit thrombin (dabigatran) or factor Xa (rivaroxaban, apixaban and edoxaban) and are prescribed more commonly for stroke prevention in AF.53 NOACs have been recommended as the first-line medication unless contraindications are present.37 Different from warfarin, NOACs have less drug-drug or drug-food interactions and are prescribed at a fixed dose following individual dosage criteria without routine monitoring of the drug concentration or anticoagulation activity.54 However, NOACs still have drug-drug interactions with some medications. An important interaction mechanism for all NOACs is significant gastrointestinal resecretion over a P-gp transporter after absorption in the gut. Competitive inhibition of this pathway will result in increased plasma levels.54 As well, CYP3A4-dependent elimination is relevantly involved in the hepatic clearance of rivaroxaban and apixaban. Therefore, strong CYP3A4 inhibition or induction may affect plasma concentrations.54 Based on these concepts, COVID-19 medications that compete with or inhibit P-gp or CYP3A4 would result in increased plasma levels of NOACs.
As Table 3 shows, HCLQ is a substrate of CYP3A4, and therefore may pontentially increase the plasma concentrations of NOACs. Generally, NOACs could be co-administered with HCLQ with either the usual dosage (apixaban or rivaroxaban) or dosage adjustment (dabigatran or edoxaban). However, marcolides (e.x. erythromycin or clarithromycin or azithromycin) are moderate P-gp competitors and strong CYP3A4 inhibitors, and may increase the exposure to NOACs by 15–20% for dabigatran, 60% for apixaban, 90% for edoxaban and 34% (erythromycin) / 54% (clarithromycin) for rivaroxaban.54 Therefore, if a HCLQ+azithromycin regimen is adopted for AF patients with COVID-19, the NOACs should be co-administered with caution.
Remdesivir, a nucleotide analog prodrug that inhibits viral RNA polymerases, provides great hope for the treatment of COVID-19 infection.9 It is expected to have no significant effects on the metabolism or elimination of any of the 4 NOACs, and could be co-administered with them.
Future DirectionsAlthough the routine monitoring of the plasma concentrations of NOACs and anti-factor Xa activity is generally not necessary in daily practice, the plasma concentrations of NOACs have been shown to be significantly associated with bleeding and ischemic events.55,56 If measurement of plama concentrations of NOACs is available in the institutions caring for COVID-19 infected patients, monitoring should be considered when drug-drug interactions between NOACs and medications for COVID-19 are a concern. More clinical experience of concomitant use of OACs and COVID-19 therapies is necessary.
Catheter ablation for AF patients with any active infection is generally contraindicated, and is similarly applied to COVID-19 patients.57 Medical treatment to control ventricular rate or rhythm with adequate anticoagulation based on the risk score will be the preferred managment.57 Electrical or chemical cardioversion is a reasonable alternative for refractory arrhythmias. COVID-19 fears put interventional laboratories in lockdown. Asymptomatic carriers or mildly symptomatic COVID-19 patients remain highly contagious and could be unexpectedly admitted. During an active pandemic, the hospitalization of these patients would significantly increase the risk for group infection within hospitals. As most cases of AF ablation are not urgent, it is suggested that these procedures be postponed for several weeks or months during the active pandemic to avoid unexpected outbreaks in hospitals.58 However, some electrophysiology (EP) procedures that are considered urgent or emergent for diseases with substantial risk of clinical decompensation, hospitalization, or death might not be avoidable. These include AF, atrial flutter, or atrioventricular nodal ablation, which is hemodynamically significant, severely symptomatic, drug and/or cardioversion refractory, or preexcited AF with syncope or cardiac arrest.58
Electrical Cardioversion for AF in Acute HFAF can induce HF due to the loss of atrial systole, and irregular and/or rapid ventricular conduction.59 Therefore, electrical cardioversion to restore sinus rhythm might improve HF. However, in COVID-19 patients, acute HF is highly attributed to cytokine storm instead of AF-induced hemodynamic changes. AF and HF are both harmful consequences of a cytokine storm. Electrical cardioversion might have a minor role in treatment, and instead, immunological treatment might be associated with recovery of left ventricular systolic function.60 However, the procedure of electrical cardioversion might increase the risk to healthcare workers. For example, aerosol-generating procedures during general anesthesia are needed for infected patients.61 The potential complications from electrical cardioversion, including ventricular fibrillation due to unsynchronized shock, heart block, bradycardia, transient myocardial dysfunction, and hypotension, would lead to unavoidable urgent resuscitation.61 All of these increase the infection risk for caregivers. AF recurrence after successful cardioversion is not uncommon in patients with HF.59 Therefore, we suggest medical treatment should be maximized, including adequate rate control, anticoagulation, or diuresis or inotropic support. Prompt electrical cardioversion should be avoided.
General Practice After the COVID-19 PandemicExcessive delay can threaten a patient’s prognosis even for catheter ablation on a regular basis to palliate drug-refractory arrhythmias. Therefore, re-initiation of regular procedures with adequate protection and infection control is recommended after the pandemic is under control. The setup of hospital-based infection control protocols for ablation procedures will be critical to avoid outbreaks in hospitals.62 The procedure and preparation for AF ablation will still mostly comply with current guidelines.57 However, several recommendations are made to prevent viral transmission as adequacy. There is no consensus whether universal screening should be applied to all patients before interventional procedures. However, the test for COVID-19 should be performed, especially if recent travel history, contact history, fever after high-risk exposure, or typical symptoms are highly suggestive of COVID-19 infection. All admitted patients and their companions should declare their health condition, travel, and contact history. Onsite visitors are not avoided, and the stay in hospital for patients and visitors should be minimized to decrease the risk of nosocomial infection. Instead, virtual video visits via cell phone or webcam are strongly suggested. A dedicated catheter laboratory with independent air-supply and ventilation or negative pressure, should be reserved for suspicious COVID-19 cases,58 and in Japan, split operation is strictly implemented. All staff are divided into 2–3 independent working shifts without personal contact during daily activities. High-flow oxygen, intubation, noninvasive positive pressure ventilation, or transesophageal echocardiography are generally not recommended because they increase the chance of virus spreading and transmission to the operators. In addition to personal protective equipment of medical personnel, patients should also wear face masks during hospital stay and procedures. CT scanning or intracardiac echocardiography could be considered to exclude atrial thrombus rather than transesophageal echocardiography before AF ablation procedure. The use of esophageal temperature probe during ablation should be avoided. Either the procedure time or hospital stay should be minimized.58
Practice for Confirmed or Suspicious COVID-19 PatientsThe recommendations for the pre-/post-AF procedure and catheter laboratory and caring units can be similarly applied from the perioperative recommendations.63 Patients with confirmed or suspected COVID-19 infection should be scheduled as the last case of the day in the dedicated laboratory and followed by extensive sterilization of the environment and/or instruments. To minimize the transport of infected patients, elective intubation of highly selective patients in the intensive care unit or in a negative ventilation pressure room should be performed prior to entering the EP laboratory.58
Follow-up After Catheter AblationCurrent guidelines suggest all patients who undergo catheter ablation for AF should be followed up for a minimum of 3 months following the ablation procedure, and then every 6 months for at least 2 years. A personal visit with 12-lead ECG and more intense monitoring is also recommended. However, the strategy of follow-up during or after the COVID-19 pandemic will change.57 For example, in-person clinic visit should be avoided whenever possible, and instead, telehealth/virtual visits could be used to minimize unnecessary exposure.58 Inspection of incision sites after device implantation or catheter ablation can be managed via telehealth utilizing video conference or mobile photography. Instead of conventional monitoring devices such as 24-hour Holter monitoring, remote monitoring of heart rhythm via a mobile device is suggested to decrease onsite visits.
None.
