key: cord-0942161-2a8zv9xl authors: O’Connell, Thomas F.; Bradley, Christopher J.; Abbas, Amr E.; Williamson, Brian D.; Rusia, Akash; Tawney, Adam M.; Gaines, Rick; Schott, Jason; Dmitrienko, Alex; Haines, David E. title: Hydroxychloroquine/Azithromycin Therapy and QT Prolongation in Hospitalized Patients with COVID-19 date: 2020-08-05 journal: JACC Clin Electrophysiol DOI: 10.1016/j.jacep.2020.07.016 sha: aee2bbee40a510721d88114d2e9121301a4bb91b doc_id: 942161 cord_uid: 2a8zv9xl ABSTRACT Background Hydroxychloroquine and azithromycin (HCQ/AZM) are being widely used to treat COVID-19 despite the known risk of QT interval prolongation and unknown risk of arrhythmogenesis in this population. Objective The study aimed to characterize corrected QT (QTc) prolongation in a cohort of hospitalized COVID-19 patients treated with combination HCQ/AZM. Methods A retrospective cohort of COVID-19 hospitalized patients treated with HCQ/AZM was reviewed. The QTc interval was calculated prior to drug administration and for the first 5 days following initiation. The primary end point was the magnitude of QTc prolongation, and factors associated with QTc prolongation. Secondary endpoints were incidences of sustained ventricular tachycardia or ventricular fibrillation and all-cause mortality. Results Among 415 patients receiving concomitant HCQ/AZM, the mean QTc increased from 443±25 msec to a maximum of 473±40 msec (87 (21%) had a QTc ≥500 msec). Factors associated with QTc prolongation ≥500 msec were age (p < 0.001), body mass index <30 kg/m2 (p = 0.005), heart failure (p < 0.001), elevated creatinine (p = 0.005), and peak troponin (p < 0.001). The change in QTc was not associated with death over the short period of the study in a population where mortality was already high (hazard ratio, 0.998, p = 0.607). No primary high-grade ventricular arrhythmias were observed. Conclusions An increase in QTc was seen in hospitalized COVID-19 patients treated with HCQ/AZM. Several clinical factors were associated with greater QTc prolongation. Changes in QTc were not associated with increased risk of death. In December 2019, severe acute respiratory syndrome coronavirus-2, a novel coronavirus which causes coronavirus disease 2019 (COVID- 19) , was first detected in Wuhan, China. 1 Since then, a global pandemic of COVID-19 has emerged with high rates of hospital admission and respiratory failure. 2 While large multi-center clinical trials have begun evaluating potential therapies, 3 several small studies have proposed that repurposed drugs, such as azithromycin and hydroxychloroquine, may offer benefit in the treatment of this disease. 4, 5 The United States Food and Drug Administration has provided emergency use authorization for health care providers to prescribe hydroxychloroquine to hospitalized patients with COVID-19, and it has been used in conjunction with azithromycin. 6 However, both hydroxychloroquine and azithromycin (HCQ/AZM) are known to prolong the electrocardiographic QT interval and could increase risk of drug-induced torsade de pointes and sudden cardiac death in these patients. [7] [8] [9] [10] Because risks associated with QT prolongation in this patient population are largely unknown, the present study aimed to determine the magnitude of QT prolongation and prevalence of serious arrhythmias among patients hospitalized with COVID-19 who were treated with the combination of HCQ/AZM, and determine if QT prolongation was associated with adverse outcomes. This was a retrospective cohort study of consecutive patients enrolled in the Beaumont Health COVID-19 database which was granted a waiver of consent by the Institutional Review Board. All patients 18 years of age and older with proven COVID-19 admitted to Beaumont Hospital -Royal Oak and Beaumont Hospital -Troy between March 13 and April 6, 2020 were screened for study entry. Of this population, those patients who were administered concomitant HCQ/AZM, had at least one day of QTc assessment following drug initiation, and an interpretable baseline ECG were included in the cohort. Demographics, past medical history, relevant laboratory values, and clinical outcomes were recorded (see Appendix 1) . Administration of drugs with known risk or probable risk of inducing torsade de pointes were documented. 8 The pharmacodynamics of QTc prolongation after initiation of HCQ/AZM was assessed over the 5 days of drug administration in a subgroup of patients. Patients with QTc measurements at baseline, day 1, days 2 and/or 3, and days 4 and/or 5 were included in this analysis. Those patients with incomplete data due to missing or technically inadequate tracings were excluded from this analysis. Patients with proven or suspected COVID-19 infection were prescribed HCQ/AZM at or soon after hospital admission at the discretion of the admitting physician. The standard dose regimens prescribed were hydroxychloroquine 400 mg twice daily for two doses then 200 mg twice daily for 4 days and azithromycin 500 mg once followed by 250 mg daily for 4 days. Patient care decisions were made by the hospital clinicians, and no attempt was made by study personnel to direct their care. Decisions regarding early termination of hydroxychloroquine and/or azithromycin (in some cases due to excessive prolongation of the QTc interval) were at the sole discretion of the physicians supervising care. An institutional policy was created during the COVID-19 pandemic to balance safe monitoring for QT interval prolongation in patients prescribed HCQ/AZM while also limiting ECG technician exposure to infected patients and preserving personal protective equipment ( Figure 1 ). Per institutional policy, a baseline corrected QT interval (QTc) was measured from a 12-lead ECG prior to initiation of HCQ/AZM. If the patient had an ECG on file within 30 days of drug initiation, that QTc interval value could be used as the baseline measure. Due to limits on telemetry availability during the COVID-19 pandemic, institutional guidelines recommended no further QTc monitoring for patients with baseline QTc intervals less than 440 milliseconds (msec) unless there was another indication for telemetry monitoring present. The institutional policy recommended that patients with QTc intervals 440 msec or greater were to be placed on telemetry for daily QTc and arrhythmia monitoring during drug initiation. Multiple repeat 12lead ECGs to monitor the QTc interval were discouraged. It was at the discretion of the attending physician whether to initiate or discontinue hydroxychloroquine, azithromycin, and/or telemetry. Available lead II telemetry monitor strips and ECGs were reviewed for 5 consecutive days while the patient was on combination drug therapy. The day 1 QTc measured on telemetry was defined as the first day after initiation of both HCQ/AZM. The QTc data were only recorded when patients were on combination therapy. Those without an interpretable baseline ECG or available telemetry/ECG monitoring for at least 1 day after drug administration were excluded from analysis ( Figure 2 ). The ECGs and telemetry recordings were analyzed off line. The QT intervals were measured manually using the tangent method and dividing by the square root of the average RR interval (Bazett's formula) to derive the QTc. 11 For patients with intraventricular conduction delays (paced rhythms or bundle branch block), a modified QTc was calculated using the formula: modified QTc = (QT -(QRS -120msec)) /√RR. 12 The primary study endpoint was to quantitate the magnitude of QTc prolongation observed in a hospitalized cohort of patients with COVID-19 treated with both HCQ/AZM, and to identify risk factors in this patient population for significant QTc prolongation. Secondary endpoints were the incidences of sustained ventricular tachycardia or ventricular fibrillation, and the relationship of QTc prolongation to in-hospital sustained ventricular tachyarrhythmias and mortality during the study period. Cause of death was adjudicated by review of the resuscitation records from all patients with attempted resuscitation, and the reviewers of these data were blinded to the QTc data. Patients who died without monitoring because of transfer to "comfort care only" status were defined as non-arrhythmic deaths. High-grade ventricular arrhythmias were defined as ventricular tachycardia or ventricular fibrillation greater than 30 seconds in duration or requiring urgent medical intervention. Secondary ventricular tachyarrhythmias observed during the course of cardiac resuscitation subsequent to a bradycardic or pulseless electrical activity arrest were not designated as high-grade ventricular arrhythmic events. All authors participated in multiple aspects of the trial including data collection (TFO, CJB, AR, AMT, RG, JS), data compilation and analysis (all), statistical analysis (AD), and manuscript preparation (TFO, CJB, AEA, BDW, AR, AMT, AD, DEH). Repeated measures of 10% of the QTc data were performed, and the average intraclass correlation coefficient was 0.946. Descriptive analyses were performed to assess differences between groups. Continuous variables were expressed as means and standard deviations. Categorical variables were expressed as numbers and percentages. The differences between the groups were evaluated using the twosample t-test for continuous variables and Fisher's exact test for categorical variables. A multivariate logistic regression analysis was conducted evaluating the impact of key demographic and clinical factors associated with QTc ≥500 msec. A second multivariate analysis was conducted to evaluate the impact of key demographic and clinical factors associated with QTc prolongation as well as changes in the QTc interval on mortality. These analyses were performed using the Cox proportional hazards model. QTc changes from baseline to Days 1 through 5 were included as a time-varying covariate in the second analysis. The hazard ratio, 95% confidence interval and p-value were computed from this model to characterize the influence of each factor on mortality. A two-sided p-value less than 0.05 was considered statistically significant. No formal adjustments for multiplicity were applied. The analyses were performed using R software (Version 3.6.1) (R Foundation for Statistical Computing, Vienna, Austria). A total of 586 patients were admitted during the study period with a proven diagnosis of COVID-19 and treated with combination therapy of HCQ/AZM. Of those patients, 415 met the inclusion criteria, and 171 were excluded because of incomplete or technically limiting ECG or clinical data ( Figure 2 ). The demographics of the included and excluded patients are listed in Table 1 . The excluded group was younger, had shorter baseline QRS duration, and had shorter QTc intervals (attributable to the institutional policy that ECG telemetry monitoring was not required in patients with baseline QTc intervals <440 msec). In the included study population, the baseline QTc interval was 443±25 msec (range 365 -539 msec). The QTc progressively increased coincident with administration of HCQ/AZM ( Figure 3A ). The maximum QTc observed over the course of monitoring for all patients receiving the drugs was 473±40 msec (range 372 -693 msec), the magnitude of QTc prolongation compared to the baseline tracing was 30 ±39 msec (range 69 -249 msec). The QTc was ≥500 msec after drug administration in 87 patients (21%). Notably, during the hospitalization, 255 patients (61%) received additional medications with known risk of torsade de pointes (Table 1S in the Supplementary Appendix). Factors associated with QTc prolongation ≥500 msec are presented in Table 2 . Changes in the QTc over the 5-day prescription duration were assessed in a subset of 137 patients with QTc measurements at baseline prior to medication administration and on days 1 through 5. In this subgroup, the average time to maximum QTc was 2.9±1.4 days. The change in the QTc from baseline increased progressively with time ( Figure 3B ). Throughout the study period, no primary high-grade ventricular arrhythmias were observed in any patient in the included or excluded groups. Of the entire treatment population, there were a total of 111 deaths in hospital. Twenty-six of 171 (15%) excluded patients died, and 85 of 415 (21%) in the included population died. Of those 85 patients, 32 patients had pulseless electrical activity or bradycardia at the initiation of the resuscitation and 53 patients died off monitor because "do not resuscitate" orders were placed. The change in QTc interval from baseline was greater in patients who died during hospitalization compared to those who survived to hospital discharge or study termination ( Figure 3B ). Data from the proportional hazards model for time-varying QTc changes with concurrent mortality for the subset of 137 patients with complete QTc data and the overall population of 415 patients are presented in Table 3 . Despite the association of QTc prolongation with mortality, the only independent predictor of mortality was age. In response to the poor outcomes observed in patients with COVID-19 infections and lack of proven curative therapies, clinicians have turned to unproven remedies like HCQ/AZM, even though this drug combination is known to prolong the QT interval and has an unknown risk of proarrhythmia in this disease. The present study was performed to quantitate the magnitude of QT prolongation in a cohort of hospitalized COVID-19 patients treated with HCQ/AZM and to identify QT-prolonging risk factors in this population. The QTc was found to increase over the 5day HCQ/AZM course with the average change in the baseline to maximum QTc measuring 30 msec. This is the first study in a COVID population to identify that older age, lower BMI, higher admission creatinine, higher peak troponin, and a history of congestive heart failure independently predicted potentially hazardous QTc prolongation ≥500 msec. However, despite the significant increases in QTc intervals in these patients, no high-grade ventricular arrhythmias were observed and changes in QTc were not associated with increased likelihood of dying. This suggests that the actual risk of torsade de pointes in this setting is very low, or that the prescribing physicians were appropriately adjusting therapies in patients in whom excess QT prolongation was observed. Hydroxychloroquine is a widely prescribed anti-malarial and anti-rheumatic medication 13, 14 and azithromycin is a commonly prescribed antibiotic. 15 They are both currently being prescribed, often in combination, as an off-label treatment for COVID-19. 6 Both medications increase the QTc interval by blocking the KCNH2-encoded hERG potassium channel. [16] [17] [18] Prior to the COVID-19 pandemic, most of the clinical evidence of hydroxychloroquine's QT prolonging properties was limited to case reports. 19, 20 Azithromycin's QT-prolonging properties had been better studied with a large case-control study demonstrating that hospitalized patients treated with azithromycin were more likely to have severe QT prolongation. 21 The present study confirms evidence of the QT-prolonging properties of HCQ/AZM, and extends these results to a COVID-19 population. The QTc progressively and significantly increased during HCQ/AZM administration similar to other studies in COVID-19 populations. [22] [23] [24] [25] Confirmation of QTc prolongation in this population better informs physicians of this important side effect of HCQ/AZM that will necessitate QTc assessment prior to drug initiation, and QTc monitoring during treatment for many patients. Identifying QTc prolonging risk factors will be important before prescribing HCQ/AZM and when deciding on a monitoring strategy. Genetics, medications, underlying medical conditions, and other metabolic factors predispose patients to QT prolongation which can lead to torsade de pointes. One out of 2000 people are genetically predisposed to have a long QT interval due to congenital long QT syndrome. 10 Hospitalized patients are particularly susceptible to drug-induced QT prolongation due to concomitant proarrhythmic risk factors, and those on non-antiarrhythmic QT prolonging medications have twice the risk of in-hospital cardiac arrest. This risk further increases with prescription of multiple QTc prolonging medications. 26, 27 While no QTc value can predict the onset of torsade de pointes with certainty, data from the congenital long QT syndrome and druginduced torsade de pointes populations have shown patients with QTc >500 msec are at increased risk. 26 The United States Food and Drug Administration has discouraged use of HCQ/AZM outside of clinical trials or hospitalized settings with monitoring 28 , and cardiovascular leaders have issued guidance to clinicians to consider withholding and withdrawing HCQ/AZM for a QTc>500 msec. 29 Over 20% of the patients in our cohort had a QTc ≥500 msec after HCQ/AZM administration. Cohort characteristics which predicted QTc prolongation were similar to those predicting significant QT interval prolongation in a validated risk score model in hospitalized patients. 30 Elevated troponin was independently associated with QTc prolongation ≥500 msec which does not rule out ischemia as a mechanism for QTc prolongation in this cohort. 31 While other known precipitants of QTc prolongation such as hypokalemia and hypomagnesemia were not significant in our cohort, this is likely attributable to an institutional policy that recommended careful electrolyte monitoring in these patients. Despite careful monitoring of patients on HCQ/AZM, over 60% of the cohort was administered at least one additional medication during hospitalization that increased the risk of torsade de pointes. This speaks to the medical complexity of this patient population, but also suggests inadequate awareness of the QT prolonging effects of many common medications. Throughout the study period, no primary high-grade ventricular arrhythmias were observed. Over 20% of the study population died which highlights the high mortality of hospitalized COVID-19 patients who suffer significantly from severe respiratory illness. 2 While positive changes in QTc from baseline were associated with increased morality in univariate analysis, this was not observed in multivariate analysis. QTc prolongation may be a marker of sicker patients as evidenced by the comorbid conditions and laboratory abnormalities predicting QTc prolongation ≥500 msec. Given the magnitude of QTc prolongation on HCQ/AZM, larger studies are warranted to investigate the prevalence of arrhythmias and mortality if this remains a therapy for COVID-19. While the world waits for the results of larger trials, 3 proposed systems to safely monitor the QTc 29 will remain necessary and cannot be ruled out as an explanation for the results of this study. There were several limitations to the study. Although a specific dosing schedule was recommended for all patients, the decisions when and how to prescribe HCQ/AZM were deferred to the prescribing physicians. Due to the retrospective nature of the study, it was not possible to definitively determine whether HCQ/AZM was discontinued by clinicians in response to QTc assessments. Retrospective collection of some data and technical limitations prevented acquisition of QTc intervals from all patients on all days, so the associations between magnitude of QT prolongation and clinical and outcome data are based upon incomplete information. The initial QTc interval was obtained from a 12-lead ECG but subsequent QTc measurements were almost exclusively obtained from a telemetry monitoring system. This policy was instituted by the health care system in an attempt to balance monitoring for QTc interval prolongation while also protecting staff and limiting use of personal protective equipment. Although the death rate was high, the rate of high-grade ventricular arrhythmias was very low. The rate of appropriate ICD therapy or death in patients with congenital long QT syndrome and QTc intervals >500 ms was reported as 35% over five years, 32 but the duration of follow up in the present study was brief and it was not adequately powered to assess the association of druginduced QT prolongation and mortality. In addition, the results of QT monitoring certainly affected clinical decisions thus confounding outcome data. Also, it is possible that the COVID-19 infection could directly prolong the QTc independent of drug effects. Unfortunately, the high utilization of HCQ/AZM in these hospitals precluded analysis of a drug-free control arm. Finally, the cohort studied was comprised of hospitalized patients, and the results may not apply to non-hospitalized patients or prophylactic therapy regimens. Because decisions when and how to prescribe HCQ/AZM were deferred to the prescribing physicians. It is possible that the 586/822 patients prescribed HCQ/AZM were the sickest patients. Hospitalized patients with COVID-19 treated with HCQ/AZM had a significant and progressive increase in QTc during combination drug therapy. Several risk factors identified patients at risk of severe QTc prolongation. Despite this finding, the risk of serious arrhythmias during this brief observation period was low. Competency in Medical Knowledge -HCQ/AZM are both known to prolong the QT interval and are associated with a risk of torsade de pointes. Addition of these drugs to other medications with known QT prolonging effects can further increase patient risk. Competency in Patient Care and Procedural Skills -Patients with COVID-19 infection are susceptible to cardiac injury and significant mortality. The value of HCQ/AZN in treating this infection is unknown, but their effects on QT interval prolongation are significant. Even though risk of serious arrhythmia in these patients is low, it may be increased with the use of these drugs. Prospective randomized trials of HCQ/AZM treatment in patients with COVID-19 infections are needed to assess possible benefit and compare those benefits to the small but important risk of excess QT interval prolongation and drug-induced torsade de pointes. Tacrolimus 5 Ondansetron 114 Tamoxifen 1 Propofol 123 Tizanidine 1 Sotalol 5 Tolterodine 8 Tramadol 26 Venlafaxine 3 Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia Covid-19 in critically ill patients in the Seattle region -Case series Race to find COVID-19 treatments accelerates Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus (2019-nCoV) in vitro Evaluation of Ebola virus inhibitors for drug repurposing Association of treatment with hydroxychloroquine or azithromycin with in-hospital mortality in patients with COVID-19 in New York State Guidance for cardiac electrophysiology during the coronavirus (COVID-19) pandemic from the Heart Rhythm Society COVID-19 Task Force; Electrophysiology Section of the American College of Cardiology; and the Electrocardiography and Arrhythmias Committee of the Council on Clinical Cardiology Azithromycin and the risk of cardiovascular death Urgent guidance for navigating and circumventing the QTc prolonging and torsadogenic potential of possible pharmacotherapies for coronavirus disease 19 (COVID-19) Intra-and interreader variability in QT interval measurement by tangent and threshold methods in a central electrocardiogram laboratory A novel method for correcting QT interval for QRS duration, predicts all-cause mortality 2019 update of the EULAR recommendations for the management of systemic lupus erythematous Hydroxychloroquine: An old drug with new relevance Diagnosis and treatment of adults with communityacquired pneumonia: An official clinical practice guideline of the Drug-induced inhibition and trafficking disruption of ion channels: Pathogenesis of QT abnormalities and drug-induced fatal arrhythmias Hydroxychloroquine reduces heart rate by modulating the hyperpolarization-activated current If: Novel electrophysiologic insights and therapeutic potential Electrophysiologic studies on the risks and potential mechanism underlying the proarrhythmic nature of azithromycin Chronic hydroxychloroquine use associated with QT prolongation and refractory ventricular arrythmia Suspected hydroxychloroquine-associated QT-interval prolongation in a patient with systemic lupus erythematous Risk evaluation of azithromycin-induced QT prolongation in real-world practice The QT interval in patients with COVID-19 treated with hydroxychloroquine and azithromycin The effect of chloroquine, hydroxychloroquine and azithromycin on the corrected QT interval in patients with SARS-CoV-2 infection Assessment of QT intervals in a case series of patients with coronavirus disease 2019 (COVID-19) infection treated with hydroxychloroquine alone or in combination with azithromycin in an intensive care unit Risk of QT interval prolongation associated with use of hydroxychloroquine with or without concomitant azithromycin among hospitalized patients testing positive for coronavirus disease 2019 (COVID-19) Prevention of torsade de pointes in hospital settings: A scientific statement from the American Heart Association and the American College of Cardiology Foundation Inhospital cardiac arrest is associated with use of non-antiarrhythmic QTc prolonging drugs FDA cautions against use of hydroxychloroquine or chloroquine for COVID-19 outside of the hospital setting or a clinical trial due to risk of heart rhythm problems. U.S. Food and Drug Administration Consideration for drug interactions on QTc in exploratory COVID-19 (Coronavirus Disease 2019) Treatment. Circulation Development and validation of a risk score to predict QT interval prolongation in hospitalized patients Prolongation of the QTc interval is seen uniformly during early transmural ischemia Who are the long-QT syndrome patients who receive an implantable cardioverter-defibrillator and what happens to them?: Data from the European Long-QT Syndrome Implantable Cardioverter-Defibrillator (LQTS ICD) Registry The Beaumont Health COVID-19 database was established to calculate all demographics, clinical characteristics, comorbid conditions, medications, and outcomes of patients tested for COVID-19. The data from the registry was abstracted through automated reports generated through a multi-platform database query tool from Beaumont Health's electronic medical record. Demographics abstracted from the registry for the purpose of this study were age, sex, race, and body mass index. Comorbid conditions abstracted were hypertension, diabetes mellitus, coronary artery disease, chronic obstructive pulmonary disease, chronic kidney disease, tobacco use, cancer, history of atrial fibrillation prior to hospitalization, and congestive heart failure. Congestive heart failure was defined as a previous admission for congestive heart failure or a diagnosis of congestive heart failure in the electronic medical record. Laboratory values abstracted were the first creatinine, aspartate aminotransferase, alanine transaminase, and lactic acid. The inpatient medication administration record for each patient was abstracted from the registry and compared with a list of 145 medications available in the United States which are known or suspected to increase the risk of torsade de pointes. 8 All ECG intervals, telemetry strip intervals, death dates, discharge dates, maximum troponin laboratory values, potassium and magnesium laboratory values immediately prior to azithromycin and hydroxychloroquine initiation, azithromycin and hydroxychloroquine initiation and completion dates, and occurrences of sustained ventricular tachycardia, non-sustained ventricular tachycardia, and ventricular fibrillation were abstracted from the electronic medical record by clinicians trained to abstract the needed data into SharePoint. All occurrences of ventricular tachycardia and ventricular fibrillation were reviewed by a Board Certified Cardiac Electrophysiologist. A random 10% sample of QTc intervals were re-abstracted to ascertain agreement and monitor calibration.