Since the onset of the pandemic, the role for corticosteroids (CS) as a therapy for COVID-19 has evolved. Initially, there was reluctance to use oral corticosteroids (OCS) outside of COVID-19-related sepsis or acute respiratory distress syndrome (ARDS). This was in keeping with community-acquired pneumonia (CAP) guidelines (Metlay JP, et al.Am J Respir Crit Care Med. 2019; 200:e45-e67) and reflected concerns that OCS might worsen outcomes in viral pneumonias. At my hospital, the reluctance to use OCS was extended to inhaled corticosteroids (ICS), with early protocols advising cessation in patients with COVID-19.
In fairness, the hesitation to use ICS was short-lived and reflected attempts to provide reasonable guidance during the early pandemic data vacuum. Over time, OCS therapy has gained acceptance as a treatment for moderate-to-severe COVID-19. On top of this, the relationship between COVID-19 and asthma has proved to be complicated. It seemed intuitive that asthmatics would fair worse in the face of a highly transmissible respiratory pathogen. Data on COVID-19 and asthma provide a mixed picture, though. It also appears that the interaction varies by phenotype (Zhu Z, et al. J Allergy Clin Immunol. 2020;146:327-329).
Improvements with OCS and the complicated interaction between COVID-19 and asthma led some to speculate that ICS, the primary treatment for asthma, may actually be protective. There is biologic plausibility to support this concept. Generally, we’ve seen a variety of immunomodulators show efficacy against moderate or severe disease. Specific to ICS, data have shown a down-regulation in COVID-19 gene expression and reduction in proteins required by the virus for cell entry. This includes a reduction in the evil, much maligned ACE-2 receptor (Peters M, et al. Am J Respir Crit Care Med. 2020;202:83-90).
Like much with COVID-19, the initial asthma phenotype and ICS data were observational and hypothesis- generating, at best. More recently, a series of randomized trials has tested the effects of ICS in patients with milder forms of COVID-19. The data are promising and are worth a thorough review by all physicians caring for COVID-19 outside of the hospital.
The STOIC trial (Ramakrishnan S, et al. Lancet Respir Med. 2021;9:763–772) randomized 146 patients to budesonide via dry powder inhaler (DPI), 800 ug twice per day (BID), versus usual care. The primary outcome was clinical deterioration, defined as presentation to acute or emergency care or need for hospitalization. There was a number of secondary outcomes designed to assess time-to-recovery, predominantly by self-report via questionnaires. The results were nothing short of spectacular. There was a significant difference in the primary outcome with a number-needed to treat (NNT) of only 8 to prevent one instance of COVID-19 deterioration. A number of the secondary outcomes reached significance, as well.
The PRINCIPLE trial, only available in preprint form (https://tinyurl.com/mr4cah7j), also randomized patients to budesonide via ICS vs usual care. PRINCIPLE is one of those cool, adaptive platform trials designed to evaluate multiple therapies simultaneously that have gained popularity in the pandemic era. These trials include predefined criteria for success and futility that allow treatments to be added and others to be dropped. The dosage of budesonide was identical to that in STOIC, and, again, it was delivered via DPI. By design, patients were older with co-morbidities, and there were two primary outcomes. The first was a composite of hospitalization and death, and the second was time to recovery.
The PRINCIPLE preprint is only an interim analysis. There were 751 and 1,028 patients who received budesonide and usual care, respectively. Time to recovery was significantly shorter in the budesonide group, but budesonide failed to meet their prespecified criteria for reducing hospitalization/death. The authors noted that the composite outcome of hospitalization or death did not occur at the rates originally anticipated, presumably due to high vaccination rates. This may have led to type II error.
In a third trial published online in November (Clemency BM, et al. JAMA Intern Med. 2021;10.1001/jamainternmed.2021.6759), patients were randomized to 640 micrograms per day of the ICS ciclesonide. Delivery was via metered-dose inhaler (MDI) for a total duration of 30 days. Unlike the STOIC and PRINCIPLE trials, this one wasn’t open label. It was blinded and placebo-controlled. The investigators found no difference in their primary outcome, time to resolution of symptoms. Ciclesonide did reduce the composite secondary outcome of ED visits or hospital admissions. The number needed to treat was 23.
Please indulge me while I overreact. It seems we’ve got a positive signal in all three. In the era of the Omnicron variant and limited health resources, a widely available therapy that curtails symptoms and prevents acute care visits and hospitalizations could have a tremendous impact. It doesn’t require administration in a clinic and, in theory, efficacy shouldn’t be affected by future mutations of the virus.
A more sober look mutes my enthusiasm. First, as the authors of the ciclesonide article note, open-label trials tracking subjective outcomes via self-assessment can be prone to bias. The ciclesonide trial was double-blinded and didn’t find a difference in time to symptom resolution, only the two open-label trials did. Second, the largest study (PRINCIPLE) didn’t show a difference in escalation of care.
Given, they defined “escalation” as hospitalization or death, and vaccines and patient selection (enrolled only outpatients with mild disease) made proving a statistical reduction difficult. However, in the text they state there wasn’t an improvement in “health care services use” either. In essence, the largest trial showed no change in escalation of care, and the trial with the best design did not show reduction in symptoms.
Although three randomized trials are enough for the inevitable meta-analysis that’ll be published soon; don’t expect it to shed much light. Combining data won’t be particularly helpful because the PRINCIPLE trial is larger than the other two combined, so its results will dominate any statistical analysis of combined data. Not to worry though – there are several more ICS COVID-19 trials underway (NCT04355637, NCT04331054, NCT04193878, NCT04330586, NCT04331054, NCT04331470, NCT04355637, NCT04356495, and NCT04381364). Providers will have to decide for themselves whether what we have so far is sufficient to change practice.
Dr. Holley is Program Director, Pulmonary and Critical Care Medicine Fellowship; and Associate Professor of Medicine USU, Walter Reed National Military Medical Center, Bethesda, Maryland. He also serves as Section Editor for Pulmonary Perspectives®.