Remdesivir treatment and transient bradycardia in patients with coronavirus diseases 2019 (COVID-19).

Dear Editor,

As described recently in this journal, natural coronavirus diseases 2019 (COVID-19) is associated with cardiac arrhythmias; cardiovascular complications were risk factors for severe COVID-19 and poor outcome as well as increased CD4/CD8 ratio, fever, LDH >250 U/l, d-dimer >1000 ng/ml. ,

Remdesivir, a viral RNA-dependent RNA-polymerase inhibitor, was authorized for severe COVID-19 patients without mechanical ventilation but still little is known about its safety except for reports about hepatic disorders and skin reactions. More recently, concern arose about RDV-related cardiac adverse events, especially bradycardia. – The purpose of this study is to describe bradycardia incidence in a group of directly-observed COVID-19 patients and its possible association with RDV.

We retrospectively evaluated all the patients consecutively admitted to our ward with COVID-19 confirmed diagnosis from September 14th to December 14th 2020. We excluded ( i) patients coming from other wards and/or hospitals, (ii) patients that did not receive a complete course of RDV during their stay in our ward, (iii) patients with life expectancy <48 h at admission. The study population was divided into 2 groups: patients who received RDV (cases) and patients who did not receive RDV (controls).

We collected data about demographic and clinical characteristics, laboratory tests, treatments and outcome using an electronic case report form. Heart rate (HR) was measured at least 3 to 6 times daily according to patients’ clinical conditions. RDV was administered as follows: 200 mg as loading dose on Day 1 and 100 mg on Day 2 to 5.

Transient bradycardia was defined as HR <60 bpm in two consecutive measurements or HR <50 bpm in one measurement. To be considered "positive for bradycardia”, cases had not presented other bradycardia episodes before or after RDV administration period. Positive outcome was defined as clinical healing and/or discharge independently of the virological status.

We compared bradycardia incidence between cases and controls and we evaluated risk factors for bradycardia by univariate and multivariate logistic regression. Moreover, as a post hoc analysis, we performed a univariate and multivariate logistic regression analysis about risk factors for mortality.

Frequencies and medians were compared using chi-squared test and Mann–Whitney U test, respectively. Statistical analysis was made using SPSS vers. 23.

We enrolled 141 patients, 62 cases and 79 controls. Table 1 shows patients’ characteristics at admission, treatments and outcomes. Cases and controls are homogeneous at admission except for body temperature >38 °C. Transient bradycardia was observed in 29/62 (46.8%) cases and 22/79 (27.8%) controls (p = 0.023). Univariate and multivariate logistic regression analysis confirmed the association between RDV treatment and bradycardia (OR 2.153, 95%CI 1.052–4.405, p = 0.036) (Table 2 ). All patients were asymptomatic for bradycardia, and we did not observe any cardiac event nor electrocardiographic clinically significant alterations in both groups.

Table 1

Cases’ and controls’ characteristics at admission, treatments and outcomes.

	Study population n = 141	Cases n = 62	Controls n = 79	P
Female sex n (%)	66 (46.8)	32 (51.6)	34 (43)	>0.1
Age years, median (IQR)	69 (56–80)	69 (59–80.75)	67 (53.5–80)	>0.1
CCI median (IQR)	4 (1–5)	4 (2–5)	3 (1–6)	>0.1
CVD n (%)	84 (59.6)	40 (64.5)	44 (55.7)	>0.1
Beta-blockers n (%)	46 (32.6)	19 (30.6)	27 (34.2)	>0.1
Days onset-admissionmedian (IQR)	6 (4–7)	6 (4.25–7)	7 (3–8.5)	>0.1
T ≥ 38°C n (%)	23 (16.3)	16 (25.8)	7 (8.9)	0.011
PaO2/FiO2 <300 n (%)	60 (42.6)	31 (50)	29 (36.7)	>0.1
C-RP >5 mg/dl n (%)	94 (66.7)	46 (74.2)	48 (60.8)	>0.1
C-RP mg/dl median (IQR)	7.8 (3–12.36)	8.93 (5.01–12.92)	6.76 (2.3–11.17)	0.077
Lymphocytes <800/mm3n (%)	56 (39.7)	26 (41.9)	30 (38)	>0.1
Lymphocytes n/mm3 median (IQR)	920 (650–1310)	890 (595–1185)	970 (700–1340)	>0.1
IL-6apg/ml median (IQR)	29.9 (9.05–78.75)	30.25 (13.4–51.025)	28.8 (7.6–86.725)	>0.1
d-dimer >1000 ng/ml n (%)	59 (41.8)	21 (33.9)	38 (48.1)	>0.1
d-dimer ng/ml median (IQR)	789 (473–1399.5)	720.5 (480–1155)	964 (472–1595)	>0.1
CD4/CD8amedian (IQR)	1.9 (1.2–2.65)	2.1 (1.3–3.1)	1.6 (0.975–2.3)	0.069
Bradycardia n (%)	51 (36.2)	29 (46.8)	22 (27.8)	0.023
Steroids n (%)	129 (91.5)	62 (100)	67 (84.8)	>0.1
LMWH n (%)	139 (98.6)	62 (100)	77 (97.5)	>0.1
Exitus n (%)	24 (17)	6 (9.7)	18 (22.8)	0.045

Abbreviations: CCI, Charlson comorbidity index; CVD, cardiovascular diseases; Days onset-admis, days from disease onset to hospital admission; T, temperature; C-RP, C-reactive protein; IL-6, interleukin 6; LMWH, low molecular weight heparin.

Notes: a, n = 83, 49 patients and 34 controls.

Table 2

Univariate analysis and multivariate logistic regression assessing risk factors for transient bradycardia and for mortality.

Transient Bradycardia	Univariate analysis		Multivariate analysis
	OR (95% CI)	p	OR (95% CI)	p
Female sex	1.62 (0.807–3.256)	0.175
Age	1.025 (1.000–1.051)	0.048	1.016 (0.986–1.047)	0.291
CVD	2.1 (1.011–4.362)	0.047	1.628 (0.643–4.125)	0.304
Beta-blockers	1.385 (0.671–2.86)	0.378	0.935 (0.389–2.244)	0.88
T ≥38°C	1.075 (0.421–2.742)	0.88
Lymphocytes	1 (0.999–1)	0.402
C-RP	1.027 (0.982–1.074)	0.246
d-dimer	1 (1-1)	0.934
RDV therapy	2.277 (1.13–4.588)	0.021	2.153 (1.052–4.405)	0.036
PaO2/FiO2 <300	1.161 (0.577–2.337)	0.676
Mortality
Female sex	0.405 (0.156–1.049)	0.063	0.267 (0.072–0.997)	0.049
Age	1.138 (1.072–1.208)	<0.001	1.138 (1.046–1.238)	0.003
CCI	1.63 (1.281–2.074)	<0.001	0.977 (0.636–1.499)	0.914
Days onset-admission	0.799 (0.679–0.94)	0.007	0.759 (0.603–0.956)	0.019
PaO2/FiO2<300	1.523 (0.62–3.738)	0.359
Pneumonia	1.029 (0.273–3.875)	0.966
T ≥38°C	1.447 (0.479–4.373)	0.655
Lymphocytes	0.999 (0.998–1)	0.256
C-RP	1.082 (1.024–1.143)	0.005	1.11 (1.022–1.206)	0.013
IL-6a	1 (0.999–1.002)	0.58
d-dimer	1.0001 (1–1.0003)	0.055	1 (1-1)	0.78
CD4/CD8	0.62 (0.326–1.179)	0.145
LMWH administration	0.198 (0.012–3.286)	0.259
Steroids administration	5.806 (0.332–101.45)	0.228

Abbreviations: OR, odds ratio; CI, confidential interval; CVD, cardiovascular diseases, T, body temperature; C-RP, C-reactive protein; RDV, remdesivir; CCI, Charlson comorbidity index; IL-6, interleukin 6; LMWH, low molecular weight heparin.

Mortality was higher in the control group (22.8% vs 9.7%, p = 0.045). To better understand this finding, we performed a univariate and multivariate logistic regression for the assessment of risk factors for mortality. Although RDV still represented a protective factor at the univariate analysis, this finding was not confirmed at the multivariate model (OR 0.3, 95%CI 0.086–1.049, p = 0.059) (Table 2). Age and elevated C-RP, instead, were risk factors for mortality, while female sex and a longer time from disease's onset to admission as protective ones (OR 0.827, 95%CI 0.701–0.976, p = 0.025). In particular, patients with negative outcome were admitted to the hospital after a median of 4 days (interquartile range [IQR] 2–6.25 days) from the onset of the symptoms, while patients with a positive outcome after 6 days (IQR 4–8 days) with p = 0.047.

Our study described transient bradycardia as a very common finding in patients administered with RDV (incidence about 47%). With this study, we confirmed the previously published data by our group, in which 60% of patients treated with RDV had transient bradycardia. Some other authors also described this association in case reports and small reviews. , How RDV could provoke bradycardia is still unknown. A possible explanation comes from the similarity between a nucleotide triphosphate metabolite of RDV and adenosine triphosphate (ATP). ATP has a negative chronotropic and dromotropic activity by an adenosine-mediated pathway; this mechanism was hypothesised to be also used by RDV's metabolite. ,

We found an elevated incidence of bradycardia. On the other hand, Touafchia and colleagues in a recently published extensive article reported only 94 cases of bradycardia out of 2603 RDV side effects reports. Moreover, we did not observe any severe adverse event while in the above-mentioned study, 80% were serious, and 17% were fatal. These substantial differences may lie in the design of the studies. While we directly observed every enrolled patient, Touafchia and colleagues evaluated “just” a big amount of reports of possible RDV associated adverse events. Theoretically, the indirect observation might be considered as a sort of selection bias that could lead to underestimating the incidence of bradycardia. At the same time, the small sample size of our study was probably responsible for the lack of severe adverse events.

In the first analysis, we found RDV administration significantly associated with positive outcome. This finding made mandatory an additional post hoc analysis that did not confirmed RDV association with reduced mortality, while the threshold of significance was reached by other factors such as C-RP, age, female sex. These data were consistent with the literature. Since patients receiving RDV had a greater clinical improvement and a faster time to recovery, they did not have benefits in terms of mortality.

This study had some limitations, especially the limited sample size and the retrospective design. However, it was performed in a real-life setting and provided directly-observed data that could possibly improve the clinical management of COVID-19 patients.

In conclusion, RDV is associated with a quite high incidence of transient bradycardia. Clinicians should be aware of this frequent adverse event in order to provide appropriate care to COVID-19 patients. A serial electrocardiographic control during RDV administration could be suggested to avoid severe cardiologic adverse events.