The combination of hydroxychloroquine [HCQ] and azithromycin [AZI] has been advocated as a potential treatment of Covid-19 infection based on the assumption of a viral load reduction in patients [1]. This resulted in a widespread use of this therapeutic regimen worldwide despite the lack of convincing data supporting its efficacy and perhaps more worryingly, its safety, particularly in critically ill patients[1].
The combination of HCQ + AZI has been shown to be associated in the general population with a safety concern related to QT segment prolongation, drug-induced torsades de pointes [TdP] and sudden cardiac death[2]. This concern is all the more present in critically ill patients presenting with cardiovascular comorbidities [which constitute the main risk factor in Covid-19 patients], electrolyte disturbances and hypoxia. Indeed, severe QT segment prolongation ≥ 500 ms, a known marker of high risk of malignant arrhythmias, has been reported to occur in up to 36% of ICU patients, treated with HCQ + AZI for Covid-19 infection in a recent observational study[3]. The risk of severe QT segment prolongation could be even higher in critically ill Covid-19 patients presenting with SARS-CoV-2 induced myocardial damage.
This is a retrospective observational single center study involving critically ill mechanically ventilated SARS-CoV-2 patients who were admitted in a tertiary Belgian ICU at the CHU de Liège during the first four weeks of Covid-19 pandemic in March and April 2020. This study aims to compare the influence of HCQ alone and HCQ combined with AZI on QT segment prolongation in patients presenting with PCR-proven Covid-19 pneumonia. The Institutional Review Board waived the need for consent to use prospectively collected clinical data and the study was appointed the serial number 2020 − 214.
In the HCQ-group [n = 13] and HCQ + AZI - group [n = 10], HCQ was given at a dose of 400 mg twice daily on the first day, followed by 200 mg twice daily for 4 days. In the HCQ + AZI- group, AZI was given at a dose of 500 mg per day for 5 days. The peak value of high sensitivity cardiac troponin I [cTnI] during the first 3 days of ICU stay was used as a marker of SARS-CoV-2 induced myocardial injury. The corrected QT segment interval [QTc] using the Bazett formula was prospectively recorded, daily, in the 23 patients. The treatment was planned to be prematurely stopped in case of QTc ≥ 500 ms. Potassium and magnesium levels were kept at the upper limit of the normal range [4 mmol/l and 1 mmol/l, respectively]. Data, expressed as median [interquartile range [IQR]] and numbers [percentage] were compared using Wilcoxon matched-pairs signed rank or Fisher exact tests, as appropriate.
Demographic and clinical data are summarized in Table 1.
Table 1
Demographic and clinical data according to treatment allocation (N = 23).
| |
HCQ – group
n = 13
|
HCQ + AZI – group
n = 10
|
|
|
Age (y) (med; IQR)
|
59 [50–65]
|
65 [61–67]
|
p = 0.29
|
|
Female sex (N)
|
7
|
4
|
p = 0.68
|
|
SAPS 2
|
29 [27–33]
|
38 [34–42]
|
p = 0.01
|
|
Pafi(mmHg),(med;IQR)
|
151 [120–165]
|
121 [95–166]
|
p = 0.56
|
|
cTnI(ng/l),(med; IQR)
|
59 [18–135]
|
50 [24–105]
|
p = 0.98
|
cTnI = high sensitivity cardiac troponin I (nl value < 30 pg/l).
There were no differences between the two groups. Cardiac arrhythmias and QTc segment measurements according to treatment are presented in Table 2.
Table 2: QTc measurements and cardiac events during treatment.
All patients underwent echocardiography and none of them was deemed to have viral myocarditis or acute ischemic heart disease [data not shown]. Treatment had to be interrupted in 10/23 pts [43,5%] because of QTc prolongation ≥ 500 ms: 7/10 pts [70%] in the HCQ + AZI group versus 3/13 pts [23%] in the HCQ group [P = 0.04]. Because of early discontinuation of treatment, the median duration of treatment in the HCQ + AZI group was significantly shorter than in the HCQ group: 2 [1–4] days versus 5 [5–5] days, respectively; P = 0.02.
None of the patients in the HCQ + AZI group received additional medication known to increase the QT segment interval; on the opposite, two patients in the HCQ group presenting with QTc ≥ 500 ms received domperidone in one case and ondansetron in the other case. The median value of baseline [pre-treatment] QTc was similar in the two groups: 400 [337–461] ms in the HCQ + AZI group versus 400 [370–422] ms in the HCQ group [P = 0.90].
The QTc increased significantly from baseline to the end of treatment in all patients: from 400 [354–460] ms to 440 [393–504] ms; [P = 0.02] and this was mainly observed in patients treated with HCQ + AZI who had a significant increase in QTc [from 400 [337–461] ms to 462 [388–510] ms; P = 0.02] compared to patients treated with HCQ alone [400 [370–422] ms to 426 [388–450] ms; P = 0.28].
None of the patients experienced torsade de pointes nor sudden cardiac death. However, four episodes of atrial fibrillation [AF] and one episode of non-sustained self-resolving ventricular tachycardia occurred in 5 pts of the HCQ + AZI group [50%] versus two episodes of AF in 2 pts of the HCQ group [15%]; [P = 0.17].
Three patients of the HCQ + AZI group eventually died [30%] versus 4 patients in the HCQ group [31%]. No death was deemed attributable to malignant arrhythmia.