Hydroxychloroquine is ineffective as Post-Exposure Prophylaxis.
Early in the COVID-19 pandemic, there was great excitement around the use of chloroquine (CQ) and its derivative, hydroxychloroquine (HCQ), as potential therapies. Both medications have shown in vitro activity against SARS-CoV, the virus responsible for the 2005 SARS epidemic, and SARS-CoV-2, the virus responsible for COVID-19 (Vincent 2005, Yao 2020). According to Yao et al, “the drugs can change the pH at the surface of the cell membrane and thus, inhibit the fusion of the virus to the cell membrane.
It can also inhibit nucleic acid replication, glycosylation of viral proteins, virus assembly, new virus particle transport, virus release and other processes to achieve its antiviral effects.” In late March, the FDA issued an emergency approval for the use of these medications for patients admitted to the hospital with COVID-19, basing the approval on “limited in-vitro and anecdotal clinical data”(Lenzer 2020).
Numerous trials were initiated to test the use of HCQ, but despite the antiviral and immunomodulatory properties demonstrated in vitro, clinical studies have been largely disappointing. The RECOVERY trial, studying the use of HCQ and other therapies on 6,000 patients in the U.K. concluded that “there is no beneficial effect of hydroxychloroquine in patients hospitalized with COVID-19” and has stopped enrolling participants in the HCQ arm of the study (Torjesen 2020).
Rosenberg et al published a retrospective, multi-center cohort of 1,438 patients in New York, in which they assessed patients who were treated with HCQ and azithromycin, HCQ alone, azithromycin alone and neither drug. They found that no treatment arm was associated with improved in-hospital mortality (Rosenberg 2020).
While there is mounting evidence that neither CQ nor HCQ are effective in COVID-19 treatment, successful management of this pandemic relies on not just treatment of the infected, but also the halting of transmission. A drug that functioned as post-exposure prophylaxis (PEP) could markedly decrease viral spread.
A small, nonrandomized, non-controlled cohort study was performed in Korea after a large COVID-19 exposure event. HCQ was given to 211 individuals daily during a 14-day quarantine after exposure. They found no significant adverse events and at the end of quarantine, all subjects had negative COVID-19 PCR (Lee 2020). Given that there was no control group however, the efficacy of this PEP regimen remains in question. Enter the NEJM study, “A Randomized Trial of Hydroxychloroquine as Postexposure Prophylaxis for COVID-19.”
This was a randomized, double-blind, placebo-controlled trial in the U.S. and parts of Canada attempting to answer the question of whether HCQ could prevent symptomatic COVID-19 infection in patients exposed to SARS-CoV-2. The primary endpoint was the development of symptomatic illness confirmed by a positive molecular assay or, if testing was unavailable, by presence of COVID-19 related symptoms.
Researchers looked at a variety of secondary endpoints including: incidence of hospitalization for COVID-19 or death, incidence of PCR-confirmed SARS-CoV-2 infection, incidence of COVID-19 symptoms, incidence of discontinuation of the trial intervention, severity of symptoms at days 5 and 14, and adverse events. Based on data from Burke et al, researchers estimated symptomatic COVID-19 infection would develop in approximately 10% of close contacts with COVID-19 exposure (Burke 2020).
Adults aged 18 years or older were recruited via social media outreach and traditional media platforms. They qualified for enrollment if they had a household or occupational exposure to a person with confirmed COVID-19 infection at a distance of less than 6 feet, for more than 10 minutes. There were a number of exclusion criteria including hospitalization and known prolonged QT interval. Patients were sorted into either high- or moderate-risk exposure.
The high-risk exposure group had an exposure in which they were wearing neither a face-mask nor eye protection. The moderate-risk exposure group had an exposure in which they were wearing a face mask, but no eye protection. Participants had to enroll within three days after exposure (with medication to start within four days after exposure) as the objective was to start PEP before the 5- to 6-day median incubation period of the virus.
Researchers recruited 821 asymptomatic patients and randomized them to receive either HCQ or matching placebo. Dosing of HCQ was 800 mg once, followed by 600 mg six to eight hours later, followed by 600 mg every 24 hours for four days. Of those enrolled, 66.4% (545/821) were healthcare workers, 87.6% (719/821) had a high-risk exposure (no eye protection or face mask) and 12.4% (102/821) had a moderate-risk exposure (face mask, but no eye protection) while 60% of participants reported wearing no PPE whatsoever during their initial exposure.
Participants were followed for 14 days. Overall, new COVID-19 infection developed in 13.0% (107/821) of patients with no significant difference between the HCQ and placebo groups — 11.8% (49/414) in the HCQ group and 14.3% (58/407) in the placebo group. There was one hospitalization in each group and there were no reported arrhythmias or deaths.
There are a number of strengths to this study. First, it asks a clinically and epidemiologically relevant question; one that is important from a public health standpoint and also personally to us as health care workers at great risk for exposure. The intervention and placebo groups were well balanced and the study used a pragmatic approach to diagnosis given the limited availability of COVID-19 testing in the US.
If COVID-19 testing was not available, four infectious disease specialists who were unaware of which trial group a participant was enrolled within, reviewed symptomatic patient cases to generate a consensus with respect to whether the condition met the case definition of COVID-19.
Second, randomization and blinding appear successful. Randomization occurred at research pharmacies and participants did not show a reliable ability to identify which treatment arm they were enrolled in. Next, the dosing regimen of HCQ had a sound scientific basis, being chosen based on pharmacokinetic simulations to achieve plasma concentrations above the SARS-CoV-2 in vitro half maximal effective concentration for 14 days. And lastly, the loss to follow up was low, with only 11.7% (96/821) of the participants not completing the follow up survey at 14 days.
There are a few limitations that may impact the applicability of this study. First, the authors planned the study to look for a 50% relative reduction in cases which may have been overly ambitious. A smaller reduction would still have been clinically important but would have required a much larger study. Second, recruitment was primarily through media which may have led to a bias in enrolled patients.
Third, the required exposure history (< 6 ft for > 10 minutes) is subjective and susceptible to recall bias. Next, the data regarding symptoms and SARS-CoV-2 test positivity were by patient report allowing for recall bias. Additionally, most patients were unable to be formally tested for SARS-CoV-2 due to lack of testing availability during the study period. Although patients were diagnosed with COVID-19 infection based on symptoms, this lack of testing could have had an impact on results.
While not necessarily limitations, there are a few other aspects of this study that are worth mentioning. The researchers performed an intention-to-treat analysis, which reflects real world conditions. A per-protocol data set would also be useful to see if a difference exists when the regimen is closely followed. Initially, the research group specified a second primary outcome: ordinal scale of COVID-19 disease severity that was changed to overall change in disease severity over 14 days among those who are symptomatic at baseline. This data does not appear in the manuscript, rather the outcome and data presented in the paper is entirely for participants who were asymptomatic at baseline.
For any PEP medication to be worth using, the benefits would need to outweigh the risks. For CQ and HCQ specifically, the cardiac risks have garnered significant attention. While the authors note that there were no serious adverse reactions or arrhythmias, QT prolongation may be present without the researchers knowing as study participants didn’t all get ECGs. Additionally, the study sample is too small to make robust safety comments.
Lastly, patients in this trial were generally younger and healthier than those at risk for severe COVID-19. The researchers may have seen different results in an older cohort or in those with more comorbid disease.
Overall, this is a high-quality randomized control trial, asking a clinically relevant question. There is no evidence that HCQ is effective as post-exposure prophylaxis in reducing the development of symptomatic COVID-19 infection. With an absence of clinical equipoise, it is unclear if additional studies are warranted. Despite early excitement, HCQ does not seem to have stood up to rigorous testing, either as a therapy or post-exposure prophylactic for COVID-19. Perhaps, our scientific efforts are now best focused elsewhere.
- Vincent MJ, Bergeron E, Benjannet S, et al. Chloroquine is a potent inhibitor of SARS coronavirus infection and spread. Virol J 2005;2:69. PMID 16115318
- Yao X, Ye F, Zhang M, et al. In vitro antiviral activity and projection of opti- mized dosing design of hydroxychloro- quine for the treatment of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Clin Infect Dis 2020 March 9 (Epub ahead of print). PMID 32150618
- Lenzer J. Covid-19: US gives emergency approval to hydroxychloroquine despite lack of evidence. BMJ2020;369:m1335. doi:10.1136/bmj.m1335 PMID 32238355
- Torjesen I. Covid-19: Hydroxychloroquine does not benefit hospitalised patients, UK trial finds. BMJ2020;369:m2263. doi:10.1136/bmj.m2263 PMID 32513810
- Rosenberg ES, Dugort EM, Udo T, et al. Association of treatment with hydroxychloroquine or azithromycin with in-hospital mortality in patients with COVID-19 in New York state. JAMA 2020; May 11. DOI: 10.1001/jama.2020.8630 PMID 3239228
- Lee SH, Son H, Peck KR. Can post- exposure prophylaxis for COVID-19 be considered as an outbreak response strat- egy in long-term care hospitals? Int J An- timicrob Agents. Epub 2020 Apr 17. PMID 32305587
- Burke RM, Midgley CM, Dratch A, et al. Active monitoring of persons exposed to patients with confirmed COVID-19 — United States, January–February 2020. MMWR Morb Mortal Wkly Rep 2020;69: 245-6. PMID 32134909