COVID19 breaking news, August 28, 2020: New Belgian large-cohort study shows hydroxychloroquine significantly reduces mortality

Dr. John Campbell, a retired British nursing school instructor and textbook author whose videoblog on COVID19 I have been following diligently, highlights a new observational study from Belgium that caused him to change his mind about hydroxychloroquine: he was always skeptical but now believes it is effective.

Hydroxychloroquine, evidence of efficacy

The original paper “Low-dose Hydroxychloroquine Therapy and Mortality in Hospitalized Patients with COVID-19: A Nationwide Observational Study of 8075 Participants” by L. Catteau et al., can be read here as a postprint (i.e., an accepted manuscript in press following peer review): https://doi.org/10.1016/j.ijantimicag.2020.106144

Allow me to highlight some quotes from the actual paper:

[…]Early during the amplification phase of the epidemic in Belgium, and pending results of clinical trials, off-label administration of a low-dose regimen of HCQ sulphate in monotherapy (400mg twice on day 1, followed by 200mg twice a day from day 2 to 5, i.e. a total dose of 2400 mg) was recommended as an acceptable immediate treatment option for hospitalized COVID- 19 patients [8].  This guidance, officially released  on March, 13th, was  based  on  the  following considerations:  (1) HCQ was the  only  drug  with  demonstrated in  vitro effect against  SARS-CoV-2 available in Belgium at that time; (2) HCQ exhibited a superior in vitro antiviral effect in comparison to CQ,  likely explained  by  the  higher  intracellular drug  accumulated  concentrations [9];  (3)  limited pharmacokinetic data suggested that the selected dosage should have sufficient antiviral activity [10]; (4) chronic administration of HCQ  for rheumatological disorders has not been associated  with major safety  signals since  decades  of  use;  (5) restricting  HCQ  use to well-selected COVID-19 patients monitored at hospitals appeared as  a reasonable risk/benefit compromise considering the  well-known dose-dependent  cardiotoxicity of the  drug;  (6) it  was  advised  to Belgian  hospitals  to administer  this off-label  regimen whenever  possible within  clinical  studies.[…]
Patients treated  either  with  HCQ  alone  and  supportive  care (HCQ  group)  were  compared  to  patients  treated  with  supportive  care  only  (no-HCQ  group) using  a competing  risks  proportional  hazards  regression  with  discharge  alive  as  competing  risk,  adjusted  for demographic and clinical features with robust standard errors. […] Results: Of 8075 patients with complete discharge data on 24th of May and diagnosed before the 1st of May, 4542 received HCQ in monotherapy and 3533 were in the no-HCQ group. Death was reported in 804/4542 (17.7%) and 957/3533 (27.1%), respectively. In the multivariable analysis, the mortality was lower in  the  HCQ  group compared to  the no-HCQ  group (adjusted hazard  ratio  [HR] 0.684,  95% confidence  interval [CI] 0.617–0.758). Compared to the  no-HCQ group,  mortality  in the HCQ  group was reduced both in patients diagnosed ≤ 5 days (n=3975) and > 5 days (n=3487) after symptom onset (adjusted HR 0.701, 95% CI 0.617–0.796 and adjusted HR 0.647, 95% CI 0.525–0.797, respectively).

The senior authors are affiliated with Ghent University, but the sample is nationwide, namely: all hospital admissions with COVID19 in Belgium until May 24 for which proper documentation on admission and discharge could be secured, and who had been given either HCQ or just “standard of care” (i.e., patients who got other experimental treatments were excluded, whether or not they also received HCQ).

So what do the authors speculate about the mechanism?

Antiviral efficacy of HCQ in humans has been poorly studied so far with adequate methods. Questions have also been raised whether safe HCQ dosages are sufficient to reach antiviral activity in target pulmonary cells [34]. Translating in vitro data into in vivo drug concentration in tissue appears particularly challenging for HCQ, as plasma concentrations do not appear to be a reliable surrogate [35]. Preprint studies in animal models (non-humans primates and Syrian hamsters) also suggest that HCQ has no antiviral efficacy [36,37]. Clinical efficacy might however be mediated through immunomodulatory mechanisms [7], preventing the progression toward severe disease with over-inflammatory responses by dampening the cytokine storm [38]. HCQ has indeed been shown to decrease the production of pro-inflammatory cytokines, both ex vivo and in lung explant model [5,39,40]. In the same line, use of low-dose dexamethasone (one of the RECOVERY arm) was recently reported to significantly decrease mortality in COVID-19 patients requiring oxygen [28]. Also, HCQ has been suggested to have some anticoagulant properties that may be beneficial in preventing thrombotic events in complement to low-molecular weight heparin [41].

The authors also note that concerns about cardiotoxicity refer to much higher dosage regimes (12,000 mg over 10 days). I know anecdotally from a friend who got such high doses as standard treatment for acute malaria that this was not a pleasant experience. The low dosage regime considered here is normally associated with the management of autoimmune diseases like lupus and rheumatoid arthritis; anecdotally, friends and acquaintances have been taking such for years without serious adverse events. As Paracelsus so memorably wrote in 1538, “All things are poison and nothing is without poison; only the dose causes something not to be a poison” (Alle Dinge sind Gift, und nichts ist ohne Gift, allein die Dosis macht dass ein Ding kein Gift ist.)