What is the evidence for using macrolide antibiotics to treat COVID-19?

April 28, 2020

Kome Gbinigie and Kerstin Frie

On behalf of the Oxford COVID-19 Evidence Service Team
Centre for Evidence-Based Medicine, Nuffield Department of Primary Care Health Sciences
University of Oxford

Correspondence to oghenekome.gbinigie@phc.ox.ac.uk

VERDICT
We identified three studies, two in vitro and one in vivo, assessing the use of macrolide antibiotics for the treatment of COVID-19. Each of these studies assessed treatment with azithromycin. The evidence from the in vivo study and one in vitro study suggest a possible synergy between azithromycin and hydroxychloroquine. However, the in vivo study had a small number of participants and was methodologically flawed; the findings must therefore be treated with caution. The two in vitro studies provided conflicting results regarding the activity of azithromycin alone against SARS-CoV-2; one found that azithromycin alone had activity against the virus, whilst the other found anti-SARS-CoV-2 activity only when azithromycin was combined with hydroxychloroquine.

At present, there is insufficient evidence to recommend treatment with macrolides, alone or combined with hydroxychloroquine, for COVID-19 outside of research. Both macrolide antibiotics and hydroxychloroquine can increase the QT interval; combining these drugs may therefore result in cardiovascular harms. Clinicians may wish to use macrolide antibiotics to treat a bacterial super-infection that has complicated COVID-19, in line with local/national treatment protocols.

BACKGROUND
Whilst novel treatments for COVID-19 are being developed, there has been increasing interest in repurposing existing medications for the current pandemic. Macrolides, a class of antibiotic used to treat respiratory, gastrointestinal and skin infections (1), have been considered for this approach. Macrolide antibiotics include azithromycin, clarithromycin, erythromycin and spiramycin (1). According to an online survey of over 6,000 physicians in 30 countries in March 2020 (2), azithromycin was the second most commonly prescribed treatment for COVID-19, and 41% of respondents reported that they had either prescribed azithromycin for COVID-19 or had seen it prescribed for this indication (3). This review evaluates the evidence for the safety and effectiveness of using macrolides to treat COVID-19.

CURRENT EVIDENCE

Methods

We searched PubMed, TRIP, EPPI COVID Living Map, MedRxiv, GoogleScholar and Google on 16^th April 2020. We included in vivo studies assessing the efficacy and/or safety of macrolide antibiotics in the treatment of COVID-19. For inclusion, in vivo studies needed to provide data allowing outcome comparisons between patients who did and did not receive the treatment in question. We also included in vitro studies that assessed the activity of macrolides against SARS-CoV-2.

Results

We identified three studies, two in vitro and one in vivo, that were eligible for this review.

In Vitro Studies

Andreania and colleagues (9) assessed the activity of azithromycin and hydroxychloroquine against SARS-CoV-2, and reported their findings in a pre-print. They tested azithromycin at three different concentrations (2, 5 and 10 µM) against the virus at two multiplicities of infection (MOI – that is, the ratio of virions to host cells). The authors found that azithromycin alone, at any concentration, did not inhibit replication of the virus. However, this was not the case when azithromycin and hydroxychloroquine were combined. At a low MOI of 0.25 and when azithromycin 5 or 10 µM was combined with hydroxychloroquine 5 µM, and at a high MOI (2.5) when azithromycin 10µM was combined with hydroxychloroquine 2 µM, viral replication was inhibited. The authors report that the concentrations of both drugs used in this study are similar to the concentrations that would be found in the lungs in humans.

Contrasting results were found in a second in vitro study by Touret et al (10) published as a pre-print. The authors assessed the in vitro ability of over 1,500 drugs, including azithromycin, to inhibit viral replication at an MOI of 0.002. They found that azithromycin had a half-maximal effective concentration (EC50 – that is, the concentration at which viral RNA increase is inhibited by 50%) of 2.12 µM and a 50% cytotoxic concentration (CC50 – that is, the concentration that results in 50% cell death) of >40 µM. Azithromycin was selective for the virus rather than host cells. The authors therefore concluded that azithromycin could be a candidate drug for COVID-19 treatment.

In Vivo Studies

We identified one comparative trial that was suitable for inclusion. Reported in a pre-print, a French study by Gautret and colleagues (11), assessed the clinical outcomes of 20 patients with suspected COVID-19 who were treated with hydroxychloroquine (200mg TDS for ten days). Of these 20 patients, six additionally received azithromycin to prevent bacterial superinfection. Patients receiving the combination treatment had daily ECGs (500mg on Day 1, then 250mg daily for four days) to ensure that their QT interval was not prolonged. They compared the clinical outcomes of patients in the intervention arms to those of 16 control cases. Of note, the patients in the control group were patients who had declined to take part, or were not eligible to take part in the trial. The researchers found that those patients receiving a combination of hydroxychloroquine and azithromycin were significantly more likely to test negative for SARS-CoV-2 on Days 3 to 6, compared with patients receiving hydroxychloroquine alone. On Day 6, 100% of patients in the combined hydroxychloroquine and azithromycin group were virologically cured; this was significantly higher than in patients receiving hydroxychloroquine alone (57.1%) (p<0.001). The researchers argue that this suggests a synergy between azithromycin and hydroxychloroquine. The authors did not report any safety data, stating that this would be published separately.

Limitations of the Identified Studies
We only identified studies assessing azithromycin; it is unclear whether any potential effects of azithromycin would generalise to other macrolide antibiotics. Furthermore, we did not identify any eligible trial assessing the effectiveness of macrolides as standalone treatments; any possible effects of these drugs may therefore be dependent on co-administration with hydroxychloroquine. All three of the included studies were published as pre-prints; these studies have not yet been accepted for publication through a peer-review process and the findings must therefore be treated with caution. During our literature review, we did not identify any studies assessing the safety profile of macrolides in the context of treating COVID-19; the safety of these drugs in this context is therefore unknown.

In vitro studies can never completely reproduce the conditions found in the human body. As a result, the antiviral activity of azithromycin reported in the in vitro studies may not reflect the activity of azithromycin in vivo. Of note, Andreania and colleagues report that the drug concentrations used were similar to those found in human lungs, perhaps aligning the conditions of the study a little more to those found in vivo.

The findings of the in vitro studies were conflicting; Touret et al found that azithromycin alone inhibited SARS-CoV-2 replication, whilst Andreania found that antiviral activity of azithromycin was dependent upon co-administration with hydroxychloroquine. The MOI used by Touret and colleagues was 100-fold lower than that used by Andreania et al, which may account for the difference in their findings.

The methodological limitations of the study by Gautret and colleagues have previously been discussed (12). The study included a small number of participants and was underpowered, which can lead to false positive results (13). This was a non-randomised study; the lack of randomisation can introduce allocation bias. This is, perhaps, highlighted by the unclear reporting of the clinical criteria used by the authors to decide which patients received azithromycin in addition to hydroxychloroquine. Furthermore, one of the six patients receiving azithromycin/hydroxychloroquine tested positive for SARS-CoV-2 on Day 8, having previously tested negative. This highlights possible flaws in the test used, and the need for reporting medium and long-term follow-up data.

MECHANISM OF ACTION
Different theories exist for the possible mechanism of action of macrolides against SARS-CoV-2. In an in vitro study reported in a pre-print, Poschet and colleagues (14) found that treatment of primary CF bronchial epithelial cells with azithromycin at a concentration of 100 µM for 1 hour, or 1 µM for 48 hours, led to an increase in the pH of the trans-Golgi network from 6.1+/-0.2 to 6.7+/-0.1. Treatment of the same cells with azithromycin 100 µM for 1 hour increased the pH of the recycling endosome from 6.1+/-0.1 to 6.7+/-0.2. Both the Golgi network and recycling endosome play important roles in the packaging of proteins into vesicles that are destined for secretion, a process that is exploited by viruses to facilitate their replication and spread. Altering the pH of these organelles may therefore interfere with these intracellular viral activities. The authors also argue that the raised pH of the trans-Golgi network may alter glycosylation of the angiotensin-converting enzyme 2 (ACE 2) receptor, a cell surface enzyme to which SAS-CoV-2 is believed to bind. Glycosylation of the receptor may therefore inhibit SARS-CoV-2 from binding to host cells.

Poschet et al found that incubation of IB3-1 CF cells with 100 µM azithromycin led to a significant reduction in the levels of an enzyme called Furin (p< 0.01) (14). SARS-CoV-2 is believed to possess a furin-like cleavage site in the spike protein (15), the protein that facilitates virus entry into host cells. It is possible that azithromycin interferes with cleavage of the spike protein, preventing viral entry into host cells.

Macrolide antibiotics are believed to reduce the production of pro-inflammatory cytokines, such as Interleukin-6 and TNF-alpha (4). Poschet et al also found that treatment of CF cells with 1 to 100 µM Azithromycin reduced basal levels of the IL-8 secretion (14). Macrolides may therefore abate the pro-inflammatory state induced by SARS-CoV-2 infection, which can ultimately lead to acute respiratory distress syndrome.

SIDE EFFECTS OF MACROLIDES
Common side effects related to macrolide antibiotics include gastrointestinal upset, dizziness, headache, hearing impairment, insomnia, visual disturbance and skin reactions (16). Uncommonly these drugs can be associated with prolongation of the QT interval, amongst other side effects (16). Care must be taken if macrolides are prescribed to patients who might be predisposed to QT interval prolongation (16). In addition, azithromycin should be used cautiously, or avoided altogether, in patients with severe renal or liver failure (16).

From the identified evidence in this review, we do not know whether the side effects profile of macrolides is different when used in the context of COVID-19.

Recent research has assessed the side effects caused by combining azithromycin and hydroxychloroquine. An analysis of medical records from 956,374 patients identified a significantly increased risk of cardiovascular mortality, chest pain and heart failure in patients who were treated with azithromycin and hydroxychloroquine compared to hydroxychloroquine alone (17). These findings are not specific to the context of COVID-19 disease, but indicate that adverse cardiovascular events are more likely when hydroxychloroquine and azithromycin are combined.

CONCLUSIONS
Given that many clinicians are already using macrolides to treat COVID-19 off-label, without recourse to robust evidence of safety or effectiveness, there is an urgent need for well-conducted, randomised clinical trials in this area. These trials should ideally be double-blinded, and should ensure that safety data is collected and reported. The results of such studies will help to guide clinical practice during this pandemic.

As hydroxychloroquine and macrolide antibiotics can prolong the QT interval, combining these treatments increases this risk. We did not identify rigorous evidence of effectiveness of this combination, and did not identify safety data assessing this combination treatment for COVID-19. We would advise extreme caution to clinicians adopting this approach outside of research studies. We did not identify any trials that assessed the use of macrolides as a standalone treatment, and are therefore unable to determine the safety or efficacy of macrolides alone as a treatment for COVID-19. However, in the context of a suspected bacterial infection that has complicated COVID-19, we recognise that clinicians may wish to prescribe macrolide antibiotics, in line with their local/national antimicrobial guidelines.

Disclaimer: This article has not been peer-reviewed; it should not replace individual clinical judgement and the sources cited should be checked. The views expressed in this commentary represent the views of the authors and not necessarily those of the host institution, the NHS, the NIHR, or the Department of Health and Social Care. The views are not a substitute for professional medical advice.

AUTHORS
Dr Kome Gbinigie MA(Cantab), MB BChir, MRCGP, DRCOG, DfSRH, PGCert(Health Research)
Kome is a General Practitioner and doctoral researcher based at the Nuffield Department of Primary Care Health Sciences, University of Oxford
Corresponding author email: Oghenekome.gbinigie@phc.ox.ac.uk
Dr Kerstin Frie BSc, MSc, DPhil(Oxon)
Kerstin is a postdoctoral researcher in the Health Behaviours team of the Nuffield Department of Primary Care Health Sciences, University of Oxford

ACKNOWLEDGEMENT
We would like to thank Nia Roberts for her help with search terms for the database searches.

SEARCH TERMS
PubMed: (coronavirus*[Title] OR coronovirus*[Title] OR coronoravirus*[Title] OR
coronaravirus*[Title] OR corono-virus*[Title] OR corona-virus*[Title] OR
“Coronavirus”[Mesh] OR “Coronavirus Infections”[Mesh] OR “Wuhan coronavirus”
[Supplementary Concept] OR &quot;Severe Acute Respiratory Syndrome Coronavirus
2&quot;[Supplementary Concept] OR COVID-19[All Fields] OR CORVID-19[All Fields] OR “2019nCoV”[All Fields] OR “2019-nCoV”[All Fields] OR WN-CoV[All Fields] OR nCoV[All Fields] OR “SARS-CoV-2”[All Fields] OR HCoV-19[All Fields] OR “novel coronavirus”[All Fields]) AND (macrolide* OR azithromycin OR spiramycin OR clarithromycin OR erythromycin OR antibiotic*)
Trip: (coronavirus OR covid-19) AND (macrolide* OR azithromycin OR spiramycin OR clarithromycin OR erythromycin)
Medrxiv: (coronavirus OR covid-19 OR “SARS-CoV-2”) AND (macrolide* OR azithromycin OR clarithromycin OR erythromycin OR antibiotic*)
Google Scholar: (coronavirus OR covid-19 OR 2019 nCoV OR 2019-nCov OR WN-cov OR nCoV OR SARS-CoV-2 OR HCov-19) AND (macrolide* OR azithromycin OR spiramycin OR clarithromycin OR erythromycin OR antibiotic*)
Google: (coronavirus OR covid 19 OR SARS-CoV-2) AND (macrolide* OR azithromycin OR clarithromycin OR erythromycin OR spiramycin)

REFERENCES

1. British National Formulary. Macrolides Overview [Available from: https://bnf.nice.org.uk/treatment-summary/macrolides.html.
2. Sermo. Sermo Methodology [Available from: https://www.sermo.com/methodology/.
3. Sermo. Breaking Results: Sermo’s COVID-19 Real Time Barometer Study. Wave I: March 25th – 27th. 2020 [Available from: https://public-cdn.sermo.com/covid19/c8/be4e/4edbd4/dbd4ba4ac5a3b3d9a479f99cc5/wave-i-sermo-covid-19-global-analysis-final.pdf.
4. Min JY, Jang YJ. Macrolide therapy in respiratory viral infections. Mediators Inflamm. 2012;2012:649570.
5. Miyamoto D, Hasegawa S, Sriwilaijaroen N, Yingsakmongkon S, Hiramatsu H, Takahashi T, et al. Clarithromycin inhibits progeny virus production from human influenza virus-infected host cells. Biol Pharm Bull. 2008;31(2):217-22.
6. Tran DH, Sugamata R, Hirose T, Suzuki S, Noguchi Y, Sugawara A, et al. Azithromycin, a 15-membered macrolide antibiotic, inhibits influenza A(H1N1)pdm09 virus infection by interfering with virus internalization process. J Antibiot (Tokyo). 2019;72(10):759-68.
7. Bosseboeuf E, Aubry M, Nhan T, Pina J, Rolain J, Raoult D, et al. Azithromycin Inhibits the Replication of Zika Virus. J Antivir Antiretrovir. 2018;10:6-11.
8. Arabi YM, Deeb AM, Al-Hameed F, Mandourah Y, Almekhlafi GA, Sindi AA, et al. Macrolides in critically ill patients with Middle East Respiratory Syndrome. Int J Infect Dis. 2019;81:184-90.
9. Andreania J, Le Bideaua M, Duflota I, Jardota P, Rollanda C, Boxbergera M, et al. In vitro testing of Hydroxychloroquine and Azithromycin on SARS-CoV-2 shows 1 synergistic effect 2. lung. 2020;21:22.
10. Touret F, Gilles M, Barral K, Nougairède A, Decroly E, de Lamballerie X, et al. In vitro screening of a FDA approved chemical library reveals potential inhibitors of SARS-CoV-2 replication. bioRxiv. 2020.
11. Gautret P, Lagier J-C, Parola P, Meddeb L, Mailhe M, Doudier B, et al. Hydroxychloroquine and azithromycin as a treatment of COVID-19: results of an open-label non-randomized clinical trial. International journal of antimicrobial agents. 2020:105949.
12. Gbinigie K, Frie K. Should chloroquine and hydroxychloroquine be used to treat COVID-19? A rapid review. BJGP open. 2020.
13. Dumas-Mallet E, Button KS, Boraud T, Gonon F, Munafo MR. Low statistical power in biomedical science: a review of three human research domains. R Soc Open Sci. 2017;4(2):160254.
14. Poschet J, Perkett E, Timmins G, Deretic V. Azithromycin and ciprofloxacin have a chloroquine-like effect on respiratory epithelial cells. bioRxiv. 2020.
15. Coutard B, Valle C, de Lamballerie X, Canard B, Seidah NG, Decroly E. The spike glycoprotein of the new coronavirus 2019-nCoV contains a furin-like cleavage site absent in CoV of the same clade. Antiviral Res. 2020;176:104742.
16. British National Formulary. Azithromycin [Available from: https://bnf.nice.org.uk/drug/azithromycin.html.
17. Lane JC, Weaver J, Kostka K, Duarte-Salles T, Abrahao MTF, Alghoul H, et al. Safety of hydroxychloroquine, alone and in combination with azithromycin, in light of rapid wide-spread use for COVID-19: a multinational, network cohort and self-controlled case series study. medRxiv. 2020.