Management of Comedication in Hospitalized COVID-19 Patients With Atrial Fibrillation

To the Editor:

Novel coronavirus disease 2019 (COVID-19) pandemic is currently the biggest threat in global public health where more than 10 million people have been infected worldwide, and over 500,000 people lost their lives due to COVID-19. ¹ The understanding of COVID-19 is evolving, and we observe a predilection of COVID-19 for individuals with comorbidities, especially hypertension and diabetes.^2-4 Therefore, more attention has been given to the management of hypertension or diabetes in COVID-19 patients. While research is rightly focused on these patient populations considering their high prevalence, we tend to forget that many patients with atrial fibrillation (AF) have also suffered from COVID-19 though with a lower prevalence than individuals with diabetes or hypertension. To illustrate, about one-forth (23.1%) of patients who died from COVID-19 in Italy had concomitant AF. ⁵ Also, the Department of Health in New York reported a prevalence of AF at 7.8% among deceased COVID-19 patients. ⁶ Though it is yet to be formally evaluated, the prevalence of AF is likely to be higher in overall COVID-19 patients considering an estimated case fatality rate of 3.4% with COVID-19. ⁷ Whilst it may not be as significant as individuals with hypertension and diabetes, it is expected that many patients with AF could present and hospitalize with COVID-19, and thus management of comedication in this population merits a proper discussion.

In order to reduce the risk of embolization, chronic oral anticoagulation is typically recommended for most patients with AF. Nevertheless, there are concerns specific to hospitalized COVID-19 patients with AF for the continuation of oral anticoagulants. Patients hospitalized with COVID-19 are commonly treated with one or multiple drug therapies including antivirals (lopinavir/ritonavir, remdesivir), interleukin-6 receptor inhibitor (tocilizumab), antibiotics (azithromycin), corticosteroids (methylprednisolone, dexamethasone), antipyretics (paracetamol, nonsteroidal anti-inflammatory drugs), and bronchodilators, though antimalarials (hydroxychloroquine or chloroquine) may have been out of favor. Warfarin is particularly a cause for concerns since it is highly susceptible to drug-drug interaction which could potentially affect optimal control of international normalized ratio (INR) in patients with AF receiving warfarin.

Several case reports described patients who experienced decreases in INR and required increased warfarin doses following initiation of antiretroviral therapy that includes lopinavir-ritonavir.^8-10 The increment in warfarin dose requirements was relatively substantial, with an approximately 4-fold increment noted in one case. ⁸ The mechanism of this possible interaction is uncertain but appears to be due to lopinavir/ritonavir-mediated induction of warfarin metabolism mediated primarily by CYP2C9. ¹¹ Likewise, azithromycin has also been reported to increase the anticoagulant effect of warfarin in several case reports among warfarin-stable patients.^12-15 Increased INR and/or bleeding was noted in each case shortly after (1-3 days) the completion of a short course of azithromycin. Other drug therapies that may interact with warfarin and the associated INR control include systemic corticosteroids^16,17 and paracetamol.^18-21

Though less susceptible than warfarin for experiencing drug-drug interactions, it nonetheless does occur in patients receiving direct oral anticoagulants (DOACs). For instance, in clinical studies summarized in the prescribing information of rivaroxaban, concurrent use of ritonavir increased area under the curve (AUC) of rivaroxaban by 2.5-fold. ²² In fact, prescribing information of DOACs includes recommendations for dose adjustment or discontinuation when combined with drugs that strongly inhibit both CYP3A4 and P-glycoprotein such as lopinavir and ritonavir. ²³ Nevertheless, a subdivision of tablets in an attempt for dose adjustment could cause dose variation which impairs the effectiveness of drug therapy. In addition, while the effect of drug-drug interactions involving warfarin could manifest through INR values, such an effect on DOACs could only manifest through drug level measurements of DOACs or through calibrated anti-Xa levels for factor Xa inhibitors and ecarin chromogenic assay/ecarin clotting time and/or dilute thrombin time for dabigatran. These tests may not be commonly available in many institutions, and yet patients’ conditions could deteriorate in a short time frame considering the hypercoagulable state related to COVID-19. There have been cases of COVID-19 presenting as an ischemic stroke in the early stages of illness. ²⁴ While no study reported as yet, patients with AF may be more likely to develop ischemic stroke considering the elevated baseline risk. ²⁵

In addition to drug-drug interactions, individuals receiving warfarin generally are sensitive to fluctuations in vitamin K intake, and adequate INR control requires close attention to the amount of vitamin K ingested from dietary sources. ²⁶ Since overall dietary intake may be reduced due to current illness or anorexia associated with COVID-19, clinicians may face difficulties in maintaining INR in the therapeutic range even being optimally maintained before hospitalization for COVID-19. ²⁷ Other environmental influences such as current acute febrile illnesses associated with COVID-19 and incident congestive heart failure precipitated by COVID-19 may also increase the risk of supratherapeutic INR. ²⁸ Cohen et al ²⁹ from retrospective chart review noted that approximately half of the patients (non-COVID-19) on chronic warfarin reached supratherapeutic levels of INR greater than 3 during hospital admission. Importantly, roughly 11% of patients reached supratherapeutic INRs at the clinically important threshold of INR ≥ 5.0.

In consideration of the above-mentioned factors for patients receiving oral anticoagulants for AF, we suggest the replacement of oral anticoagulant therapies (warfarin and DOACs) with the easily titratable parenteral low-molecular-weight heparin or unfractionated heparin to avoid the risk of over-/under-anticoagulation whilst in hospital. In fact, the current recommendation by the National Institutes of Health, United States ³⁰ and American Society of Hematology ³¹ is that low-molecular-weight heparin and unfractionated heparin is preferred over oral anticoagulants in hospitalized, critically ill patients because of their shorter half-lives, ability to be administered intravenously or subcutaneously, and fewer drug-drug interactions. Consideration should also be given on the possible necessity of mechanical ventilation during hospitalization among COVID-19 patients, especially those admitted into intensive-care units (ICU). A retrospective case series by Grasselli et al ³² of 1,300 patients with available data on respiratory support observed that 88% of included patients received mechanical ventilation during ICU stay. Parenteral administration of anticoagulant strongly facilitates the antithrombotic treatment in ventilated or intubated patients.

Oral beta-blockers and non-dihydropyridine calcium channel blockers are widely used as primary therapy for rate control in patients with AF. Both beta-blockers and calcium channel blockers decrease the resting heart rate and blunt the heart rate response to exercise. There are no specific contraindications to the use of these rate control agents with regards to COVID-19, as reported by Reynolds et al ³³ in his large retrospective study of 5,894 COVID-19 patients that there was no association of these agents with a severe course of COVID-19 in, but possible decreased risk of severe disease among users of calcium channel blockers though the authors did not segregate between dihydropyridine and non-dihydropyridine calcium channel blockers. Dihydropyridine calcium channel blockers are devoid of negative inotropic effects and thus are not used for pharmacological rate control in patients with AF.

Nevertheless, there may be potential drug-drug interactions between the beta-blocker, bisoprolol, and tocilizumab. Tocilizumab binds to and inhibits the proinflammatory cytokine interleukin-6 (IL-6) which could be released during the course of COVID-19 as part of the inflammatory response. IL-6 has been shown to decrease the gene expression of CYP3A4 by greater than 90% in-vitro. ³⁴ Therefore, concurrent administration of tocilizumab in COVID-19 patients receiving bisoprolol may restore CYP3A4 activity and thus increase the metabolism of CYP3A4 substrates including bisoprolol. This effect may persist several weeks following discontinuation of therapy owing to the long half-life of tocilizumab. Other beta-blockers such as metoprolol and atenolol are of no concerns for such drug-drug interaction since they are not CYP3A4 substrates.

For AF patients who receive digoxin, electrolyte disturbances, particularly hypokalemia that may accompany COVID-19, may place patients at increased risk for digoxin-associated arrhythmias and therefore patients receiving digoxin should undergo monitoring of the serum digoxin level until serum potassium level returns to the normal range. ³⁵ In addition, the drug-drug interaction between ritonavir and digoxin is worth mentioning. In a pharmacokinetic study of 12 healthy volunteers, ritonavir (300 mg twice daily) increased the digoxin (0.5 mg single dose) AUC by 86% and prolonged the half-life by 156%. ³⁶ In another pharmacokinetic study, ritonavir (400 mg twice daily) increased the digoxin (0.5 mg single dose) AUC by 37%. ³⁷ Still in other pharmacokinetic analysis specifically for lopinavir/ritonavir, the use of lopinavir/ritonavir (400 mg/100 mg twice daily) increased the digoxin AUC by 81%. ³⁸ In support of these pharmacokinetic studies, 2 case reports describe patients who experienced elevated digoxin serum concentrations and toxicities after the addition of ritonavir therapy.^39,40 The mechanism of this interaction is likely due to ritonavir-mediated P-glycoprotein inhibition, a transporter responsible for digoxin disposition. According to prescribing information of digoxin, the dose of digoxin should be reduced by 30% to 50%, or the dosing interval lengthened, in the concurrent use with ritonavir. ⁴¹

For patients with AF who require antiarrhythmic drugs to maintain sinus rhythm, the potential for prolonged corrected QT interval (the time duration between the onset of the QRS complex and the end of the T wave) and associated ventricular arrhythmias should be considered when azithromycin is initiated. Patients should undergo continuous telemetry monitoring during treatment where the corrected QT interval in sinus rhythm should not be allowed to exceed 500 ms. ⁴² Identification of QT interval prolongation should prompt discontinuation of both antiarrhythmic drugs and azithromycin. Moreover, we recommend substitution of sotalol, propafenone, dronedarone, and flecainide in patients receiving these antiarrhythmic agents when incident heart failure with reduced ejection fraction is precipitated by COVID-19, due to studies showing an increase in the risk of mortality with these agents in patients with heart failure.^43,44 Amiodarone and dofetilide have been safely used in patients with AF and heart failure or those with a left ventricular ejection fraction less than 35%.^45,46

We acknowledge evidence derived from rigorous clinical trials are lacking with regards to the management of co-medications prescribed for AF in COVID-19 patients. Therefore, the above discussion represents pharmacists’ clinical perspective upon review and synthesis of available literature which aims to facilitate the medication management of this patient population. The aforementioned issues merit the attention of hospital pharmacists involved in the inpatient care activities related to COVID-19 who may encounter AF patients in their daily practice.