Hydroxychloroquine-azithromycin, doubase C, and QTc prolongation in congolese patients with COVID-19: Myth or reality?

Abstract

BACKGROUND

QTc interval prolongation with an increased risk of torsade de pointes (Tsd) has been described in coronavirus disease 2019 (COVID-19) patients treated with hydroxychloroquine (HCQ) and azithromycin (AZI) in Western countries. In the DR Congo, few studies have evaluated the safety of this association or proposed new molecules.

AIM

To determine the incidence of QTc prolongation and Tsd in COVID-19 patients treated with HCQ-AZIs vs doubase C (new molecule).

METHODS

In present randomized clinical trial, we have included patients with mild or moderate COVID-19 treated with either HCQ-AZI or doubase C. Electrocardiogram (ECG) changes on day 14 of randomization were determined based on pretreatment tracing. Prolonged QTc was defined as ≥ 500 ms on day 14 or an increase of ≥ 80 ms compared to pretreatment tracing. Patients with cardiac disease, those undergoing other treatments likely to prolong QTc, and those with disturbed ECG tracings were excluded from the study.

RESULTS

The study included 258 patients (mean age 41 ± 15 years; 52% men; 3.4% diabetics, 11.1% hypertensive). Mild and moderate COVID-19 were found in 93.5% and 6.5% of patients, respectively. At baseline, all patients had normal sinus rhythm, a mean heart rate 78 ± 13/min, mean PR space 170 ± 28 ms, mean QRS 76 ± 13 ms, and mean QTc 405 ± 30 ms. No complaints suggesting cardiac involvement were reported during or after treatment. Only four patients (1.5%) experienced QTc interval prolongation beyond 500 ms. Similarly, only five patients (1.9%) had an increase in the QTc interval of more than 80 ms. QTc prolongation was more significant in younger patients, those with high viral load at baseline, and those receiving HCQ-AZI (P < 0.05). None of the patients developed Tsd.

CONCLUSION

QTc prolongation without Tsd was observed at a lower frequency in patients treated with HCQ-AZI vs doubase C. The absence of comorbidities and concurrent use of other products that are likely to cause arrhythmia may explain our results.

Key Words: QTc prolongation, COVID-19, Doubase C, HCQ-AZT, Sub-Saharan Africa

Core Tip: During the coronavirus disease 2019 (COVID-19) pandemic, western studies have shown the ineffectiveness and cardiotoxic effects of hydroxychloriquine and azithromycin (HCQ-AZI). In Africa, the heart toxicity of HCQ-AZI is little reported by cardiologists, and some African countries used these two molecules to treat COVID-19 patients. In the present study, we have evaluated the occurrence of prolongation of the QTc interval and torsade de pointes torsade in Congolese COVID-19 patients who received HCQ-AZI compared to doubase C, a herbal medicine with broad-spectrum antiviral activity.

INTRODUCTION

The first coronavirus disease 2019 (COVID-19) cases were reported in Wuhan, China in December 2019[1]. COVID-19 then spread very quickly across the world and was declared a pandemic on March 11, 2020 by the World Health Organization (WHO)[2]. COVID-19 curative treatment has been controversial. Some countries, particularly in Africa, have used hydroxychloroquine (HCQ) or chloroquine (CQ) alone or in combination with azithromycin (AZI) without much solid evidence; however, in general, COVID-19-related mortality in Africa was not as high as in countries that did not use this treatment[3].

In addition to adverse effects such as allergies and retinal damage, HCQ and CQ can cause cardiac arrhythmia and prolong the QTc interval[4-8]. This risk of QTc prolongation increases when AZI is added to the treatment and can potentially develop into torsade de pointes (Tsd), thus increasing patient mortality[4-8]. Series published in the West, including that of Bessière et al[9] in France, found that more than 35% of patients had a prolongation of the QTc interval. Healthcare teams have raised concerns about the use of HCQ and AZI. Therefore, electrocardiogram (ECG) must be monitored regularly during treatment[10]. In several hospitals in the West, HCQ was no longer used in the treatment of COVID-19 patients[11].

Although the cardiac side effects of HCQ/CQ have been reported in patients with COVID-19 in Western countries, few studies have been conducted in sub-Saharan Africa (SSA), particularly in the DR Congo, where HCQ/CQ has been previously used to treat malaria or common rheumatological diseases such as lupus erythematosus and rheumatoid arthritis. This is also the case for other molecules used to treat severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection during the COVID-19 pandemic in SSA. However, their efficacy and safety have not been evaluated within the population. In 2021, a randomized clinical trial was conducted in Kinshasa to compare the efficacy and safety of HCQ-AZI versus doubase C, a plant-based medicine, in the treatment of COVID-19[12]. We assessed the risk of cardiac toxicity associated with these two molecules because we did not have local data.

This clinical trial aimed to determine the incidence of QTc prolongation, its predictors and the consequences (Tsd and other arrhythmia) in COVID-19 patients (mild and moderate cases) followed at Kinshasa University Hospital (KUH).

MATERIALS AND METHODS

Study design

Several aspects of the general methodology of this study have been published in the WHO Clinical Trials Registry[12]. We collected data from the KUH COVID-19 treatment center (CTCO) and focused on assessing the cardiac safety of the two treatments from May 2021 to January 2022.

Doubase C is a large-spectrum antiviral derived from two plants, Uvaria brevistipita and Haroungana madasgascariensis. The treatment was given over 7 d depending on body weight: 2 tablets 3 times per day if weight < 80 kg; 3 tablets 3 times a day between 80-99 kg and 4 tablets 3 times a day if weight ≥ 100 kg[12].

At the start of the pandemic in DR Congo, HCQ-AZI was the reference protocol: HCQ was prescribed at a dose of 200 mg 3 times a day for 10 d while AZI was given over 5 d, i.e. 500 mg the first day, then 250 mg for 4 d[12,13]. In addition to the basic treatment, all patients received zinc sulfate (20 mg once a day), vitamin C (500 mg once a day) and vitamin D (400 IU three times per day) for 10 d. Diabetic and hypertensive patients continued their usual treatment. Oxygen was indicated when O2pSa was less than 95%[13].

Sampling

Inclusion criteria: Patients had to be at least 18 years old, have mild or moderate COVID-19, sign informed consent, receive one of two medications according to the clinical trial protocol, and have an ECG examination before the treatment, and then after 7 and 14 d.

Exclusion criteria: Pregnant women; patients with cardiac, kidney, or liver disease; asymptomatic, severe, or critical COVID-19; and all patients who had a prolonged QTc interval on ECG before treatment.

Patients who presented with complications (hematological, renal, hepatic, cardiac, or neurological) left the study and were treated in the hospital. In the event of prolongation of the QTc interval during the study, treatment was immediately stopped, and the patient was followed up in the Cardiology Department.

Sample collection

Parameters unrelated to this study are described in detail elsewhere[12]. For this study, we used the following variables of interest: age, sex, history of diabetes, hypertension, current treatment, COVID-19 treatment, vital signs at each visit, days 1, 7, and 14 of randomization (consciousness, blood pressure, heart and respiratory rates, and temperature), oxygen saturation, and ECG results (on days 1 and 14).

Electrocardiogram: The MAC ECG 600 device made it possible to perform examinations with twelve leads at a speed of 25 mm/s. The protocols were carried out by two cardiologists in a consensual manner. The device gave measurements of the QTc, QT, QRS and PR intervals. Lead D2 or V5 was used to calculate the QTc interval. When the heart rate fluctuated between 60 and 100 beats/min, the Bazett formula was applied and when it was less than 60 beats/min or more than 100 beats/min, the Fredericia formula was used[14]. In case of bundle branch block, the JTC formula: [QT- (QRS-120 ms)]/√RR was used[15].

Operational definitions

The clinical form of COVID-19 was based on the WHO classification[16]. The baseline SARS-CoV-2 viral load was determined using the CTE and CTN2[17]. The QTc (lead D2 or V5) was defined as the duration from the beginning of the QRS complex to the end of the T-wave upon its return to baseline. Prolonged QTc was defined as QTc interval ≥ 470 ms for women and ≥ 450 ms for men[14]. QTc requiring discontinuation of treatment was defined by a prolonged QTc ≥ 500 ms[8] or by a QTc increase > 80 ms[18].

Endpoint

The endpoint was ECG change on day 14 of randomization (QTc ≥ 500 ms or QTc increase > 80 ms).

Ethics approval

The present study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Ethics Committee of the Kinshasa School of Public Health (ESP/CE/038/2021). All data were fully anonymized before being accessed.

Statistical analyses

After verification, the data were analyzed using SPSS Statistics version 25.0. Categorical variables are presented as absolute and relative frequencies (%) while continuous variables are presented as mean ± SD. The Kolmogorov-Smirnov test was used to check whether the data distribution was Gaussian. The Student's t test was used to compare the means between two groups. The MacNemar test was used to compare paired data. The statistical significance threshold was set at P < 0.05.

RESULTS

Baseline characteristics of patients

261 patients were randomly assigned to two arms: 123 (47%) were treated with HCQ-AZI and 138 (53%) with doubase C. Among them, 93.5% of patients had mild and 6.5% had moderate COVID-19. The mean patient age was 41 ± 15 years and the sex ratio was 1.1. The average patient BMI was 26.1 ± 5.4 Kg/m², and the average partial oxygen saturation was 96.9 ± 2.0%. Patients with a medical history of hypertension, diabetes, HIV/AIDS, asthma, or tuberculosis accounted for 11.1%, 3.4%, 0.4%, 2.3%, and 0.4%, respectively. The averages of hemoglobin (13.6 ± 2.7 g/dL), creatinine (0.97 ± 0.30 mg/dL), AST (28 ± 12 IU/L) and ALT (25 ± 16 IU/L), were not statistically different between the two arms (Table 1).

Table 1 Baseline characteristics of patients.

Variables	Whole group, n = 261	Doubase C, n = 138	HCQ-AZI, n = 123	P value
Age	41.0 ± 14.8	40.5 ± 15.3	41.5 ± 14.2	0.581
Men/women	137/134	64/74	73/50	0.036
BMI	26.1 ± 5.4	26.4 ± 5.7	25.6 ± 5.0	0.249
O2pSa	96.9 ± 2.0	97.0 ± 1.7	96.9 ± 2.3	0.449
Hypertension	29 (11.1)	18 (13.0)	11 (8.9)	0.569
Diabetes	9 (3.4)	7 (5.1)	2 (1.6)	0.197
HIV	1 (0.4)	1 (0.7)	0	0.386
Asthma	6 (2.3)	3 (2.2)	3 (2.4)	0.744
Tuberculosis	1 (0.4)	1 (0.7)	0	0.383
Hemoglobin (g/dL)	13 ± 3	13.7 ± 3.2	13.5 ± 2.2	0.673
Creatinine (mg/dL)	0.97 ± 0.30	0.96 ± 0.31	0.99 ± 0.25	0.420
ALAT (UI/L)	25 ± 16	23 ± 14	27 ± 18	0.064
ASAT (UI/L)	28 ±12	29 ± 14	26 ± 11	0.141
K+ (mmol/L)	4.2 ± 2.0	3.9 ± 1.0	3.8 ± 0.11	0.167
Ca2+ (meq/L)	1.17 ± 0.11	1.15 ± 0.11	1.18 ± 0.11	0.430

The results are presented as means and standard deviations or as absolute frequencies (percentages). BMI: Body mass index; HIV: Human immunodeficiency virus; ALAT: Alanine aminotransferase; ASAT: Aspartate aminotransferase; HCQ-AZI: Hydroxychloriquine and azithromycin.

Table 2 shows that the various ECG parameters were normal at the initiation of treatment. The mean QTc was 405 ± 30 ms.

Table 2 Baseline characteristics of electrocardiograms.

ECG parameter	Participants, n = 261	Doubase C, n = 138	HCQ-AZI, n = 123	P value
Sinus rhythm	261 (100)	123 (100)	138 (100)
Heart rate (/min)	78 ± 13	79 ± 13	77 ± 13	0.420
P wave (ms)	114 ± 10	113 ± 12	115 ± 8	0.495
PR space (ms)	170 ± 28	172 ± 30	168 ± 26	0.331
QRS complex (ms)	76 ± 13	76 ± 14	76 ± 12	0.839
T-wave (ms)	193 ± 32	193 ± 30	194 ± 34	0.679
QTc interval (ms)	405 ± 30	406 ± 31	401 ± 29	0.193
Sokolow- Lyon (mm)	24.5 ± 8.9	25.0 ± 10.1	23.1 ± 8.0	0.611

HCQ-AZI: Hydroxychloriquine and azithromycin.

Patient complaints and changes in ECG pattern during treatment

During treatment, none of the patients experienced disease worsening. No patient was admitted to the intensive care unit and none died. Only three patients did not undergo ECG on day 14. They preferred not to attend follow-up meetings because they felt well.

Among the 258 patients who visited on the 14^th-day appointment, none of the complaints suggesting cardiac involvement, such as dyspnea, shortness of breath, palpitations, lipotymia, or sudden death, were reported. Table 3 shows that the mean QTc interval increased from 411 ± 41 to 418 ± 37 ms and the PR space from 169 ± 25 to 177 ± 3.6 ms.

Table 3 Electrocardiogram changes after treatment (day 14 of randomization).

Variable	Days 14-1	Day 1	Day 14	P value
Heart rate (/min)	0.820	78 ± 13	79 ± 12	0.343
P wave (ms)	1.010	114 ± 10	115 ± 13	0.314
PR space (ms)	8.232	170 ± 28	177 ± 36	0.001
QRS complex (ms)	0.469	76 ± 13	76 ± 13	0.560
QTc interval (ms)	8.835	405 ± 30	413 ± 35	< 0.001
T-wave (ms)	3.867	193 ± 32	197 ± 42	0.197
Sokolow–Lyon (mm)	0.004	24.5 ± 8.9	24.5 ± 8.5	0.990

Table 4 shows that only four patients (1.5%), all from the HCQ-AZI group, presented a prolongation of the QTc interval beyond 500 ms. Similarly, only five patients (1.9 %) in the HCQ-AZI arm had a more than 80 ms increase in the QTc interval (Table 4).

Table 4 Proportion of patients who experienced prolongation of QTc after treatment.

Treatment	QTc, ≥ 500 ms		QTc, < 500 ms		P value	QTc increase, 80 ms		No QTc irncrease		P value
	n	%	n	%		n	%	n	%
HCQ-AZI	4	3.3	117	96.7	0.047	5	4.1	116	95.9	0.021
Doubase C	0	0	137	100		0	0	137	100

HCQ-AZI: Hydroxychloriquine and azithromycin.

As shown in Table 5, QTc interval prolongation was greater in younger patients, those with a high viral load, and those who received HCQ-AZI. No cases of Tsd were observed throughout the study.

Table 5 Risk of electrocardiogram pattern changes (QTc prolongation) in each subgroup.

Variables	QTc prolongation		P value
	Day 1, n = 261	Day 14, n = 258
Men	0	1	0.500
Women	0	4	0.061
Age < 40 yr	0	5	0.031
Obesity	0	2	0.247
Normal and overweight	0	3	0.124
CtN2 or CtE < 33	0	5	0.030
HCQ-AZI	0	5	0.030

CtE: Cycle threshold value of envelope genes of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2); CtN2: Cycle threshold value of nucleoprotein genes of SARS-CoV-2; HCQ-AZI: Hydroxychloriquine and azithromycin.

DISCUSSION

QTc interval prolongation under HCQ is linked to the blockade of the HERG potassium channel, which is involved in the final phase of repolarization. HERG channel blockade prolongs ventricular repolarization and can lead to Tsd[19,20]. HCQ can also act on sinus nodes and cause bradycardia[20]. Macrolide antibiotics such as AZI block the HERG channel[21]. In the present study, a few patients treated with HCQ-AZIs experienced QTc interval prolongation; however, the incidence was significantly lower than that in studies conducted in Western countries[6-9], and the abnormality was not associated with cardiac complications. A meta-analysis that selected 13 studies reported incidences of QTc interval prolongation varying between 0 and 35% in patients with COVID-19 who received HCQ or CQ in combination or not with AZI[22]. Six studies were conducted in France, four in the United States, one in Italy, one in Holland, and one in Brazil. The authors did not report the proportion of Black patients in these studies.

Several factors may explain the differences between studies, such as the threshold used to define QTc interval prolongation (≥ 500 ms, increase of 60 ms or 80 ms compared to the baseline value, and other different thresholds), the clinical state of the patients (critical patients in intensive care units or patients with less severe disease), and the concomitant use of other medications likely to prolong the QTc interval. In addition to medications, electrolyte disorders, including hypokalemia and hypomagnesemia, are the most frequently encountered risk factors. The role of hypocalcemia is less obvious, as low serum calcium levels prolong the QTc interval[23]. Cardiac disease and acute stroke are also risk factors[23]. The absence of major comorbidities, severe and critical forms of COVID-19, concurrent use of other medications, and lower doses could explain the low incidence of QTc prolongation in our patients treated with HCQ-AZIs.

The absence of major cardiac complications such as Tsd with HCQ-AZI is consistent with incidence varying between 0 and 1.11%[22]. Studies reporting cases of sudden death deserve to be reevaluated because they included intensive care patients who presented with several comorbidities and were administered very high doses of HCQ[6-9]. Our study is, to the best of our knowledge, the only study to have evaluated the safety of doubase C. Being a new drug, studies with larger sample sizes must be conducted to confirm that the product does not prolong the QTc interval and has no risk of Tsd.

In addition to the HCQ-AZI combination, viral load and age were associated with QTc interval prolongation. The signs of a prolonged QTc interval can occur under specific conditions such as exercise, stress, emotion, or auditory stimulation[24]. In hindsight, we know that there are no cardiac arrhythmias specific to COVID-19, but a particular discrepancy between heart rate and temperature can be observed in a few patients. Therefore, the heart rate may be slower than expected in relation to temperature[25]. This phenomenon has also been observed in other infectious diseases such as typhoid fever. Hypoxemia can also play a role in explaining arrhythmias. With regard to age, we should expect a much higher QTc frequency in elderly patients than in patients aged 10 and 30 years, as observed for certain congenital forms of prolonged QTc[24].

Despite the interest in this study, some limitations are worth mentioning. ECG was not performed daily, which may have reduced the frequency of detected abnormalities. In some studies, the ECG was performed on the day after drug administration. In our study, for doubase C, after 7 d of administration of the last dose; in the HCQ-AZI arm, HCQ was taken from days 1 to 10, AZI from days 1 to 5, and the ECG check was performed on day 14. However, HCQ has a long duration of action, as it may be taken weekly for some indications[26]. Indeed, at a dose of 200 mg, the plasma half-life of HCQ is 30 d[24]. The half-life of AZI is shorter at, 2-4 d[27]. The small sample size is also a limitation of this study. Regarding COVID-19 treatment in the two arms, the duration differed (7 and 10 d). However, this choice was made to comply with the two therapeutic strategies already in use in the country.

CONCLUSION

Contrary to studies published in Western countries, the risk of QTc interval prolongation was present in Congolese patients with COVID-19 and treated with the HCQ-AZI combination but at a low frequency and without Tsd. This risk was not observed for doubase C. The absence of comorbidities and the concomitant use of other products that are likely to cause arrhythmia may explain our results.

ACKNOWLEDGMENTS

The authors thank the staff of Kinshasa University Hospital CTCO and Mister Mihigo Bashengezi of the CREPPAT laboratory for providing documentation on doubase C. The government has provided HCQ and AZI tablets to all patients with COVID-19. We would like to thank Professor Pierre Zalagile Akilimali for carrying out the more complex statistical analyses, and Professor Augustin Okwe Nge and Selain Kabula Kasereka for completing certain statistical analyses. We would also like to thank Yannick Bazitama Munyeku for their correction.