Relative bradycardia in patients with coronavirus disease 2019 (COVID-19)

Abstract

To the Editor, Arrhythmias and other cardiovascular symptoms in patients with COVID-19 are frequently reported and are likely associated with infection-related myocarditis, ischemia, and/or systemic proinflammatory stimulation (“cytokine storm”). In a case series including 138 hospitalized patients, 17% (and 44% of the patients admitted to the intensive care unit) had an (unspecified) arrhythmia [1]. Moreover, medications used in COVID-19 patients may increase arrhythmic risk. We have observed fever plus relative bradycardia, i.e. an inappropriately low heart rate response to increased body temperature [2], in several hospitalized COVID-19 patients. Our primary objective was to assess the prevalence of relative bradycardia in patients with COVID-19. We retrospectively reviewed the electronic medical records of the first 174 patients with confirmed COVID-19 (detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) by RT-PCR from nasopharyngeal swabs) admitted to the University Hospital Basel, Switzerland, from February 27, 2020 to April 15, 2020. During this period, symptomatic inpatients without contraindications were routinely treated with lopinavir/ritonavir for 5-7 days and hydroxychloroquine for two days. In addition, patients with severe disease received tocilizumab. An ECG was routinely performed on admission in all patients and again on the third day of hospitalization in all patients on treatment. We included in our analysis only patients primarily admitted to the ward (162 patients), and excluded further 52 patients for the following reasons: treatment with heart rate lowering agents (e.g. beta blockers, non-dihydropyridine calcium channel blockers, amiodarone, digoxin) and conditions associated with bradycardia (e.g. hypokalemia <3.0 mmol/L) (33); non-sinus rhythm on ECG (7); missing ECG (1); refusal of consent (11). Relative bradycardia was defined as a heart rate < 90/min and concomitant fever (tympanic temperature ≥ 38.3°C), measured at least twice within 24 hours. If more measurements met these criteria, we included the measurements with the highest body temperature. The local ethical board approved the study (EKNZ 2020-00769). 110 patients with COVID-19 (median age: 59 years, males: 60%) were evaluated for bradycardia. 71 of the 110 patients (64%) had fever during hospitalization. 40 patients had relative bradycardia (36% of all COVID-19 patients and 56% of COVID-19 patients with fever). Relative bradycardia occurred a median of 9 days (IQR: 6-11) after symptoms onset. Moreover, 38 of the 110 patients (34%) had at least once a heart rate of < 60/min during the hospital stay irrespective of the body temperature (18 of the relative bradycardia group, 20 without relative bradycardia). None of the 110 patients had a QT-prolongation. The temperature-heart rate relationship in patients with relative bradycardia is reported in Figure 1 . Figure 1 Temperature-heart rate relationship in patients with COVID-19 and a relative bradycardia (for each patient the two values with the highest temperature ≥ 38.3°C and a heart rate < 90 bpm measured in the consecutive 24 hours are reported) (red). Physiologically appropriate temperature-heart rate relationship (blue), adapted from [2]. Temperature-heart rate relationship in patients with COVID-19 without relative bradycardia (for each patient the two values with the highest temperature are reported) (green). Dotted lines represent linear regression between temperature and heart rate. RB: relative bradycardia. Figure 1 Patients with relative bradycardia were significantly older (median age: 62 years) and presented with significantly higher maximal temperatures (median: 39.3 °C) compared to patients with fever and an appropriate heart rate response (49 years; 38.7 °C). Otherwise the two groups did not differ significantly regarding off-label drug treatment, oxygen therapy or laboratory findings. The clinical outcome (intensive care unit admission, intubation, death) was similar in patients with fever and relative bradycardia (20%,18% and 3% of patients respectively), and in patients with fever and appropriate heart rate response (19%, 13% and 6%). We found that applying a conservative definition, 56% of hospitalized COVID-19 patients with fever had relative bradycardia. A recently released study of 54 Japanese patients with mild to moderate COVID-19 used a broader definition (not requiring the presence of fever or a minimal temperature) and also showed that relative bradycardia was a common characteristic [3]. Typically, the heart rate increases by about 10/min for each Fahrenheit degree increase in body temperature above 101 °F (38.3 °C). The appropriate heart rate with a body temperature of 38.3 °C is about 110/min [2]. The term relative bradycardia describes the failure of the heart rate to rise when body temperature is elevated. Many infectious and non-infectious causes of relative bradycardia in febrile patients have been described (e.g. typhoid fever), but the pathogenesis of this phenomenon is still unknown. Direct pathogen effects on the sinoatrial node and effects of inflammatory cytokines are among proposed mechanisms [4]. Interestingly, interleukin-6 (IL-6) is the cytokine reported to exhibit the strongest correlation with depressed heart rate variability, which in turn may predict relative bradycardia [4]. On the other hand, IL-6 appears to play also an important role in the “cytokine storm” caused by SARS-CoV-2. A recent report described the development of sinus bradycardia in 8 of 26 patients with severe COVID-19 pneumonia (without mentioning the presence of fever). The authors postulated an inhibitory effect of SARS-CoV-2 on sinus node activity [5]. In conclusion, relative bradycardia is a frequent clinical feature of COVID-19, occurring in 56% of febrile patients hospitalized on our hospital’s wards. This fact should be taken into account while evaluating febrile patients in the context of a possible SARS-CoV-2 infection. Relative bradycardia in non-critically ill patients does not appear to be associated with a worse clinical outcome. Authors’ contributions GC: Conceptualization, Methodology, Investigation, Formal analysis, Writing - Original Draft; MO: Formal analysis, Writing – Reviewing and Editing; AE, MS: Writing – Reviewing and Editing; SB: Conceptualization, Methodology, Supervision, Writing - Original Draft and Reviewing and Editing. Transparency declaration The authors declare no conflicts of interest. No external funding was received for this work.