Jun 302020

Jon-Emile S. Kenny MD [@heart_lung]

This is the way the world ends
Not with a bang but with a whimper.

-T.S. Eliot

As lightly discussed a few months ago, deploying corticosteroids in COVID-19 began, expectedly, with controversy.  Russell and colleagues submitted a letter arguing against wholesale administration of corticosteroids in patients infected with SARS-CoV-2.  They noted pre-existing literature in MERS-CoV, SARS-CoV and influenza that showed no obvious benefit, and even potential harm [e.g. delayed clearance of viremia], when corticosteroids were administered.  This was countered by Shang and colleagues who argued that the existing data should not preclude ‘low-dose, short-duration’ corticosteroids in COVID-19 [i.e. ≤ 0.5–1 mg/kg per day methylprednisolone or equivalent for less than one week].

The trigger for corticosteroids in COVID-19 at the outset was unclear; should any or all patients receive them?  Those in the intensive care unit?  Those ventilated?  With ‘cytokine storm?’ ARDS?

Thereafter, the Surviving Sepsis Campaign [SSC] guidelines for COVID-19 suggested using corticosteroids in mechanically-ventilated patients only when ARDS is present.

The majority of our panel support a weak recommendation [i.e. suggestion] to use steroids in the sickest patients with COVID-19 and ARDS. However, because of the very low-quality evidence, some experts on the panel preferred not to issue a recommendation until higher quality direct evidence is available.”

As described below, there has since been a handful of enlightening investigations including at least 2 small case series, retrospective observational investigations, quasi-experimental study and, most-recently, a good randomized trial.

Case Series

Kolilekas and colleagues reported a small series of six, consecutive, hospitalized COVID-19 patients with worsening hypoxemia [meeting criteria for ARDS], and indices of hyperinflammatory syndrome [elevated ferritin, C-reactive protein] who received a short course of methylprednisolone, 125mg once daily.  All patients developed ARDS between 8 and 13 days after symptoms onset.  Following initiation of methylprednisolone, oxygen saturations and inflammatory markers improved in all patients and none were intubated.  Only one patient developed hyperglycemia. The small, observational nature of the study precludes any definitive recommendations for steroids, but it is interesting to note their marked improvement in response to methylprednisolone, their hypoxemia on initiation of steroids and the time frame for steroid administration [i.e. at least 8 days into symptoms].

So and colleagues described seven, mechanically-ventilated patients with COVID-19 associated ARDS.  All patients received 1000 or 500 mg of methylprednisolone upon intubation and for three days followed by 1 mg/kg and tapered off.  For 6 of the patients, the time from symptom onset to intubation ranged between 10 and 13 days, while one patient had only 4 days between symptom onset and intubation.  Similar to Koliekas et al. above, all patients had improved oxygenation, inflammatory markers and all were extubated.  Secondary infections related to corticosteroid use were not observed, though hyperglycaemia was reported in five patients and delusions in two patients.

Retrospective Analyses

In mid-March, Wu et al. published a retrospective cohort study of 201 hospitalized COVID-19 patients.  Of the 201 patients, 165 [82.1%] required oxygen in the hospital and methylprednisolone was given to 62 [30.8%] patients; the dose and duration was not explicit.

They evaluated two outcomes: 1. development of ARDS and 2. death among those with ARDS.  As expected, essentially all patients who developed ARDS required supplemental oxygen therapy while 69% of those who did not develop ARDS required oxygen [all received nasal cannula].  Interestingly, amongst those patients who developed ARDS, the administration of methylprednisolone was associated with a lower risk of death [HR, 0.38; 95% CI, 0.20-0.72; P = .003].  Only 10% of patients without ARDS received methylprednisolone; it is unclear if these patients were receiving oxygen or not, nor is it reported if methylprednisolone was associated with lower death rate in the non-ARDS cohort.  Symptom duration before methylprednisolone administration was not reported for any group.

While this retrospective study is hypothesis generating – suggesting that COVID-associated ARDS may benefit from methylprednisolone – it is not conclusive.  As the authors appropriately state, potential bias and residual confounding in an observational study with a small sample size cannot inform firm recommendations.

More recently, Yuan and colleagues published a retrospective analysis of 132 patients with non-severe COVID-19.  Propensity score matching was used to adjust for differences in baseline characteristics between patients who received corticosteroids [n = 74] and those who did not [n = 58].  The average dose of methylprednisolone used was 40-60 mg for about 11 days.

Non-severe COVID-19 pneumonia was defined as cases with radiographic pneumonia, but without the following: 1. respiratory distress, respiratory rate per min ≥ 30; 2. oxygen saturation ≤ 93%; 3. arterial blood oxygen partial pressure/oxygen concentration ≤ 300mmHg  4. progress of chest radiological manifestations > 50% within 24-48 hours.

There was no statistically-significant difference between the two groups, however, the trends implied that corticosteroids were unhelpful.  More patients in corticosteroid group progressed to severe disease than those who did not receive steroids [11.4% versus 2.9%], hospital stay was prolonged by a little more than 3 days and viral shedding was longer in the steroid group.

Experimental Studies

Fadel et al. conducted a single pre-test, post-test quasi-experiment in a multi-centre health system in Michigan.  Adult patients were enrolled if they had confirmed COVID-19 infection, with radiographic evidence of bilateral pulmonary infiltrates, and required oxygen by nasal cannula, high flow nasal cannula [HFNC] – which they classified as ‘moderate’ disease – or mechanical ventilation – classified as ‘severe’ disease.

Over one week patients were treated as ‘standard of care’ which included corticosteroids but not ‘early.’  Thereafter, a one-week protocol was implemented that emphasized early, short-course, methylprednisolone [0.5 to 1 mg/kg/day divided in 2 intravenous doses for 3 days].  Importantly, the early group received methylprednisolone based on the following inclusion criteria: patients with moderate COVID-19 who required 4 liters or more of oxygen per minute on admission, or who had escalating oxygen requirements from baseline.  All patients who required ICU admission were recommended to receive early methylprednisolone.

Corticosteroids were used in 56.8% of the standard of care group and 68.2% in the early corticosteroid group; this was not statistically significant.  Both groups had a similar duration of symptoms of 5-6 days on admission.  Yet, the early group had corticosteroids initiated within 48 hours of presentation more often [12.4% vs. 41.7%], with a shorter median time to initiation [2 vs. 5 days].

Each component of the primary composite endpoint was significantly reduced in the early corticosteroid group.  These outcomes included: escalation to intensive care unit, progression to respiratory failure requiring mechanical ventilation and in-hospital, all-cause mortality.

The RECOVERY collaborative group have recently released a randomized trial comparing dexamethasone 6 mg daily for up to ten days to usual care alone in hospitalized COVID-19 patients. The primary outcome was 28-day mortality.  2104 patients randomly allocated to receive dexamethasone were compared with 4321 patients concurrently allocated to usual care.

In totality 21.6% patients who were randomized to dexamethasone and 24.6% of patients who were allocated to usual care died within 28 days; this was statistically-significant.  Importantly, prior to randomization, subgroups were pre-specified by level of respiratory support.  There was statistically and clinically-significant mortality reduction for patients receiving mechanical ventilation with a number needed to treat [NNT] of 8 patients.  As well, there was significant mortality reduction for those patients requiring oxygen at randomization [NNT = 25].  By contrast, there was no benefit to hospitalized COVID-19 patients not receiving oxygen.


Despite the relatively wide assortment of evidence above [i.e. case series, retrospective, quasi-experimental, and randomized trial] there is seeming consistency in the results.  Patients with more severe pulmonary disease [i.e. at least in need of supplemental oxygen] enjoyed greatest benefit from methylprednisolone or dexamethasone.  These are noteworthy results – as mentioned in the letter by Russell and colleagues, corticosteroid use in other viral respiratory illnesses demonstrated delayed viral clearance and increased mortality. These discordant findings have multiple explanations but may ultimately be mediated by timing of corticosteroid therapy in the disease course.

Pulmonary neutrophilia may be a key mediator of hypoxemia and ARDS; leading to elaboration of cytokines and chemokines within the pulmonary parenchyma.  While viral replication peaks in the second week of illness in SARS-CoV-1, peak shedding in COVID-19 is earlier.  Thus, administering corticosteroids after the first week of symptoms may miss the window of viral replication and only calm down the hyper-inflammatory response in COVID-19.

Of note, the data above suggests that one need not perform complicated calculations to infer ‘cytokine storm’ as a threshold for anti-inflammation.  The benefit of corticosteroids was fairly consistently observed in patients requiring at least supplemental oxygen therapy.  For example, non-severe COVID-19 patients defined by Yuan et al. had an oxygen saturation of 94% or greater and these patients did not appear to benefit from methylprednisolone.  By contrast, in the RECOVERY trial, supplemental oxygen was recommended when oxygen saturation on air fell below 92-94% and these patients did appear to benefit.

While warning readers to avoid speculation and conjecture in this post on the renin-angiotensin system, I will ignore this and wonder: given that angiotensin II – aldosterone stimulation appears to exacerbate pulmonary injury, should we avoid corticosteroids with significant mineralocorticoid activity?  Of note, neither dexamethasone nor methylprednisolone have significant mineralocorticoid activity.  Should angiotensin II be avoided as a vasopressor in COVID-19?  But I digress.

Headpiece filled with straw,


Dr. Kenny is the cofounder and Chief Medical Officer of Flosonics Medical; he also the creator and author of a free hemodynamic curriculum at heart-lung.org

Get our weekly email update, and explore our library of practice updates and review articles.

PulmCCM is an independent publication not affiliated with or endorsed by any organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)


A Brief Overview of Corticosteroids in COVID-19