Inhaled budesonide for COVID-19 in people at higher risk of adverse outcomes in the community: interim analyses from the PRINCIPLE trial

BACKGROUND Inhaled budesonide has shown efficacy for treating COVID-19 in the community but has not yet been tested in effectiveness trials. METHODS We performed a multicenter, open-label, multi-arm, adaptive platform randomized controlled trial involving people aged [≥]65 years, or [≥]50 years with comorbidities, and unwell [≤]14 days with suspected COVID-19 in the community (PRINCIPLE). Participants were randomized to usual care, usual care plus inhaled budesonide (800g twice daily for 14 days), or usual care plus other interventions. The co-primary endpoints are time to first self-reported recovery, and hospitalization/death related to COVID-19, both measured over 28 days from randomisation and analysed using Bayesian models. RESULTS The trial opened on April 2, 2020. Randomization to inhaled budesonide began on November 27, 2020 and was stopped on March 31, 2021 based on an interim analysis using data from March 4, 2021. Here, we report updated interim analysis data from March 25, 2021, at which point the trial had randomized 4663 participants with suspected COVID-19. Of these, 2617 (56.1%) tested SARS-CoV-2 positive and contributed data to this interim budesonide primary analysis; 751 budesonide, 1028 usual care and 643 to other interventions. Time to first self-reported recovery was shorter in the budesonide group compared to usual care (hazard ratio 1.208 [95% BCI 1.076 - 1.356], probability of superiority 0.999, estimated benefit [95% BCI] of 3.011 [1.134 - 5.41] days). Among those in the interim budesonide primary analysis who had the opportunity to contribute data for 28 days follow up, there were 59/692 (8.5%) COVID-19 related hospitalizations/deaths in the budesonide group vs 100/968 (10.3%) in the usual care group (estimated percentage benefit, 2.1% [95% BCI -0.7% - 4.8%], probability of superiority 0.928). CONCLUSIONS In this updated interim analysis, inhaled budesonide reduced time to recovery by a median of 3 days in people with COVID-19 with risk factors for adverse outcomes. Once 28 day follow up is complete for all participants randomized to budesonide, final analyses of time to recovery and hospitalization/death will be published. (Funded by the National Institute of Health Research/ United Kingdom Research Innovation [MC_PC_19079]; PRINCIPLE ISRCTN number, ISRCTN86534580.)

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021

INTRODUCTION
There is an urgent need for effective and safe community-based treatments for coronavirus disease 2019 (COVID-19), especially for older people and those with co-morbidities who are at higher risk of hospitalization and death. 1 Inhaled corticosteroids are widely available, inexpensive and generally safe, and have been proposed as a COVID-19 treatment due to their targeted anti-inflammatory effects in the lungs. 2,3 In animal and human studies, inhaled corticosteroids reduce expression of ACE-2 and TMPRSS2, 4,5 which are used by SARS-CoV-2 for cell entry. 6 Early in the pandemic, the low prevalence of asthma and chronic obstructive pulmonary disease (COPD) among people hospitalized with COVID-19, lead to speculation that inhaled corticosteroids used to treat these conditions may be protective. 2,7 Furthermore, systemic corticosteroids reduce deaths in hospitalized patients with COVID-19 8,9 likely because the hyperinflammatory state is responsible for the subsequent damage from SARS-CoV-2 infection. 10,11 Two large, population-based studies in primary care in the United Kingdom found an increased risk of COVID-19 hospitalization and/or death among people prescribed inhaled corticosteroids, 12,13 although residual confounding by unmeasured disease severity could not be ruled out. An efficacy trial including 146 adults with mild COVID-19 in the community found inhaled budesonide reduced COVID-19 related emergency assessments or hospitalizations. 14 However, thus far, there are no results reported from large effectiveness trials of inhaled budesonide for COVID-19.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04. 10.21254672 doi: medRxiv preprint We therefore aimed to determine whether inhaled budesonide speeds recovery or reduces hospital admission or death from COVID-19 in people at higher risk of an adverse outcome in the community.

TRIAL DESIGN AND OVERSIGHT
We assessed the effectiveness of inhaled budesonide in the UK national, multi-center, primary care, open-label, multi-arm, prospective adaptive Platform Randomised trial of INterventions against COVID-19 In older peoPLE (PRINCIPLE), which opened on April 2, 2020, and is ongoing. The protocol is available as a supplement to this manuscript, and at www.principletrial.org. A "platform trial" allows multiple treatments for the same disease to be tested simultaneously. A master protocol defines prospective decision criteria for dropping interventions for futility, declaring interventions superior, or adding new interventions. 15 Interventions under evaluation in PRINCIPLE have included hydroxychloroquine, azithromycin, 16  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021

PARTICIPANTS
People in the community were eligible if they were aged ≥65 years, or ≥50 years with comorbidities (see trial protocol), and had ongoing symptoms from polymerase chain reaction (PCR) confirmed or suspected COVID-19 (in accordance with the United Kingdom National Health Service definition of high temperature and/or new, continuous cough and/or change in sense of smell/taste), 18,19 which started within the past 14 days. People were ineligible to be randomized to budesonide if they were already taking inhaled or systemic corticosteroids, were unable to use an inhaler, or if inhaled budesonide was contraindicated. 20 Initially, eligible people were recruited, screened and enrolled through participating general medical practices, but from May 17, 2020, people across the UK could enroll online or by telephone. We undertook extensive community outreach to increase recruitment from ethnic minority and socially deprived communities, as these groups have been disproportionally affected by COVID-19.

TRIAL PROCEDURES
Eligible, consenting participants were randomized using a secure, in-house, web-based randomization system (Sortition). When the budesonide group opened on November 27, 2020, the azithromycin, doxycycline and usual care groups were also active. Randomization probabilities were determined using response adaptive randomization via regular interim analyses, which allows allocation of more participants to interventions with better observed outcomes (see Adaptive Design Report). The azithromycin and doxycycline groups had stopped by December 14, 2020, at which point 1:1 allocation between usual care and budesonide . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04.10.21254672 doi: medRxiv preprint occurred, stratified by age, and comorbidity. On March 04, 2021 the colchicine arm opened, with subsequent response adaptive randomization.
Participants were followed up through an online, daily symptom diary for 28 days after randomization, supplemented with telephone calls on days 7, 14 and 28. Participants were encouraged to nominate a trial partner to help provide follow up data. We obtained consent to ascertain healthcare use outcome data from general practice and hospital records. We aimed to provide a self-swab for SARS-CoV-2 confirmatory PCR testing, but capacity issues early in the pandemic meant testing was unavailable for some participants.

TRIAL INTERVENTIONS
Participants received usual care plus inhaled budesonide 800µg twice daily for 14 days (Pulmicort Turbohaler, AstraZeneca), or usual care alone. This breath-actuated inhaler was chosen due to its ease of use for unwell, co-morbid, and potentially frail and older patients, and was either issued by the participant's general medical practitioner (GP), or centrally by the study team and delivered to the participant. Participants in the budesonide arm were also sent a link to a video demonstrating current inhaler use, with further explanation available by telephone support. Usual care in the United Kingdom National Health Service for suspected COVID-19 in the community is largely focused on managing symptoms. 21

PRIMARY OUTCOMES
The trial commenced with the primary outcome of hospitalization or death within 28 days.
However, the proportion requiring hospitalization in the UK 22 was lower than initially . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04.10.21254672 doi: medRxiv preprint expected 23 . Therefore, the Trial Management Group and Trial Steering Committee recommended amending the primary outcome to include a measure of illness duration. 24,25 Duration of illness is an important outcome for patients and has important economic and social impacts. This received ethical approval on September 16, 2020, and was implemented before performing any interim analyses. Thus, the trial has two co-primary endpoints measured within 28 days of randomization: 1) time to first reported recovery defined as the first instance that a participant reports feeling recovered; and 2) hospitalization or death related to COVID-19.

SECONDARY OUTCOMES
Secondary outcomes include a rating of how well participants feel ("How well are you feeling today? Please rate how you are feeling now using a scale of 1 -10, where 1 is the worst you can imagine, and 10 is feeling the best you can imagine"), time to sustained recovery (date participant first reports feeling recovered and subsequently remains well until 28 days), binary outcome of early sustained recovery (reports feeling recovered within the first 14 days from randomization and remains recovered until day 28), time to initial alleviation of symptoms (date participant first reports all symptoms as minor or none), time to sustained alleviation of symptoms, time to initial reduction of severity of symptoms, contacts with health services, hospital assessment without admission, oxygen administration, Intensive Care Unit admission. mechanical ventilation, adherence to study treatment and the WHO-5 Well-Being Index 26 . We included secondary outcomes capturing sustained recovery due to the often recurrent nature of COVID-19 symptoms.

STATISTICAL ANALYSIS
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04.10.21254672 doi: medRxiv preprint Sample size calculation and statistical analysis are detailed in the Adaptive Design Report and the Master Statistical Analysis Plan . We justify sample sizes by simulating the operating characteristics of the adaptive design in multiple scenarios, which explicitly account for response adaptive randomization, early stopping for futility/success and multiple interventions. In brief, for the primary outcome analyses, assuming a median time to recovery of nine days in the usual care group, approximately 400 participants per group would provide 90% power to detect a 2day difference in median recovery time. Assuming 5% hospitalization in the usual care group, approximately 1500 participants per group would provide 90% power to detect a 50% reduction in the relative risk of hospitalization/death.
The first co-primary outcome, time to first self-reported recovery, was analyzed using a Bayesian piecewise exponential model regressed on treatment and stratification covariates, and included parameters for temporal drift. The second co-primary outcome, hospitalization/death, was analyzed using a Bayesian logistic regression model regressed on treatment and stratification covariates. The primary outcomes were evaluated using a "gate-keeping" strategy to preserve the overall Type I error of the primary endpoints without additional adjustments for multiple hypotheses. The hypothesis for the time-to-first-recovery endpoint was evaluated first, and if the null hypothesis was rejected, the hypothesis for the second co-primary endpoint of hospitalization/death was evaluated. In the context of multiple interim analyses, the master protocol specifies that each null hypothesis is rejected if the Bayesian posterior probability of superiority exceeded 0.99 for the time to recovery endpoint and 0.975 (via gate-keeping) for the hospitalization/death endpoint. Based on trials of antibiotics for lower respiratory tract . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04.10.21254672 doi: medRxiv preprint infection, 27 a minimum of 1.5 days difference in median time to first report of recovery, and 2% difference in hospitalization/mortality were pre-specified as clinically meaningful.
At the beginning of the trial, due to difficulties with community SARS-CoV-2 PCR testing early on in the pandemic in the United Kingdom, participants with suspected COVID-19 were included in the primary analysis population, irrespective of confirmatory testing. When testing became more readily available, the Trial Steering Committee recommended restricting the primary analysis population to those with confirmed COVID-19, and this change was included in protocol version 7.1 on February 22, 2021 and approved on March 15, 2021, before any interim budesonide results were disclosed to the Trial Management Group, and before those interim results had been reviewed the Trial Steering Committee. Therefore, the pre-specified primary analysis population is defined as all eligible SARS-CoV-2 positive participants randomized to budesonide, usual care, and other interventions, from the start of the platform trial until the dataextraction for this updated interim analysis, on March 25 th , 2021. Because this population includes participants randomized to usual care before the budesonide group opened, the primary time to recovery analysis model includes parameters to adjust for temporal drift in the study population, which may occur due to changes in circulating SARS-CoV-2, usual care including vaccination, or the pandemic situation, as well as changes in the inclusion/exclusion criteria over time. The inclusion of parameters for temporal drift in the hospitalization/death models are currently being implemented, as pre-specified in the Adaptive Design Report version 3.5 which was approved on March 11, 2021, prior to review of the interim analysis by the Data Monitoring and Safety and Trial Steering Committees which led to unblinding of the budesonide arm. As this change has not yet been completed, we do not present secondary analyses of the . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04.10.21254672 doi: medRxiv preprint hospitalization/death results in the overall study population in this pre-print, as this population contains a large proportion of participants randomized before the budesonide arm opened, and is therefore more susceptible to temporal drift. This pre-specified secondary analysis, accounting for temporal drift, will be presented in the final report.. We also conducted secondary analyses for time to recovery among all study participants irrespective of SARS-CoV-2 status, and prespecified sensitivity analyses using the concurrent randomized population; defined as all participants who were eligible for budesonide and randomized to budesonide or usual care during the time period when the budesonide arm was active, important because participants already using steroid inhalers, and therefore may have had asthma or COPD, were excluded from randomization to the budesonide arm.
Analysis of the secondary outcomes, and pre-specified sub-group analyses, were conducted on the concurrent randomization and eligible SARS-CoV-2 positive population. Secondary time-toevent outcomes were analyzed using Cox proportional hazard models, and binary outcomes were analyzed using logistic regression, adjusting for comorbidity status, age, and duration of illness.

Population
The first participant was randomized on April 2, 2020. Enrolment into the budesonide group participants who provided some follow up data and were randomized to inhaled budesonide (n = 751), usual care alone (n = 1028), and other treatment groups (n = 643). To protect the integrity of the platform trial and other interventions, we only provide descriptive summaries of participants randomized to budesonide and usual care. The average age (range) of participants was 62.8 (50 -100) years, of which 2474 (83.2%) had co-morbidities. The median (interquartile range) was 6 (4 to 9) days from symptom onset. Baseline characteristics were comparable between the two groups (Tables 1 and S1).
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021 In the preliminary primary Bayesian analysis of hospitalizations (which do not currently account for temporal changes in hospitalizations, but will do so in the final analysis, as pre-specified in the Adaptive Design Report version 3.5), among those who had the opportunity to contribute . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021 proportion of usual care participants randomized before the budesonide arm opened, and is therefore more sensitive to changes hospitalization rates over time. The final, complete analysis will include parameters for time in the Bayesian primary outcome models, as pre-specified in the Adaptive Design Report version 3.5.

Secondary outcomes
Analysis of secondary outcomes, using the concurrent randomization and eligible SARS-CoV-2 positive population, showed evidence of benefit with budesonide in the daily score of how well participants felt over 28 days (Table 2 and Figure S1), the WHO-5 Wellbeing Index, early sustained recovery, time to sustained recovery, ( Table 2). There was no clear evidence of differences in both participants reported or GP reported healthcare services use between groups (Table 2).
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04. 10.21254672 doi: medRxiv preprint In the prespecified subgroup analyses, there was no clear evidence that symptom duration prior to randomization, baseline illness severity score, age or comorbidity modified the effect of budesonide on time to first reported recovery ( Figure 3). Regarding serious adverse events, two participants reported hospitalizations unrelated to COVID-19, both in the budesonide group.

DISCUSSION
This interim analysis from a platform, randomized trial involving participants in the community with COVID-19 at increased risk of an adverse outcome, found that inhaled budesonide shortened time to first self-reported recovery. Our findings were consistent across various measures that capture symptom severity, sustained recovery, wellbeing. There was also no evidence of differential effects among pre-specified sub-groups by age, symptom duration, symptom severity and co-morbidity. Due to decreases in hospital admissions associated with the United Kingdom lockdown and vaccination program, 28,29 and yet to be completed 28 day follow up data for some participants, the current interim analysis is not powered for the COVID-19 related hospital admission and/or death outcome. The final analysis of the effects of inhaled budesonide will be made available once all participants randomized to inhaled budesonide have completed 28 day follow up.
We identified eight ongoing randomized controlled trials of inhaled corticosteroids as treatment for COVID-19, 30 and one completed study (the STOIC study). In the STOIC phase 2, open-label trial among adults aged 18 and over with mild, suspected COVID-19 in the community, 146 participants were randomly assigned to inhaled budesonide 800µg twice a day until symptoms resolved, or usual care. 14 The primary outcome of COVID-19 related urgent care or emergency . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021 We did not combine different types of healthcare utilization into one outcome in the PRINCIPLE trial. Secondary outcomes in STOIC also favored budesonide over usual care with respect to time to self-reported recovery, symptom persistence at day 14, and resolution of fever.
PRINCIPLE is the largest randomized trial to date to evaluate inhaled budesonide for community treatment of COVID-19 and adds to the evidence base supporting this therapeutic agent following the earlier Phase 2 findings. 31 Furthermore, recent transcriptome analysis indicates that budesonide, alongside others which include dexamethasone and ciclesonide is a novel therapeutic immunomodulatory target. 32 Several randomized trials have demonstrated the benefits of systemic corticosteroids for treatment of people hospitalized with COVID-19. 8,9 Our findings are immediately relevant for clinical practice as they suggest that early treatment in the community with inhaled corticosteroids is effective at speeding recovery, which has important benefits for patients and wider society. While global access to vaccines continues to be scaled up, inhaled budesonide is readily available in many primary care settings and is included in the World Health Organization List of Essential Medicines. 33 .
Strengths of our analysis include the pragmatic trial design which allows for efficient evaluation of multiple interventions as they would be used in the community, the evaluation of budesonide as a standalone, early treatment and the focus on patients at higher risk of complications. We included patients with suspected COVID-19 but without PCR confirmed SARS-CoV-2 infection in our secondary analyses as this reflects community testing conditions early in the UK . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04.10.21254672 doi: medRxiv preprint pandemic, and limited SARS-CoV-2 testing may necessitate early empirical treatment in many other community and low resource hospital settings. Given the variation in PCR testing sensitivity, particularly if self-administered, some participants will have had false negative tests. [34][35][36][37] We conducted an open label study to determine whether the addition of budesonide to usual care benefitted patients, rather than to assess benefit of budesonide compared to a placebo.
The pragmatic study design therefore does not allow us to determine mechanism of effect or the degree to which a placebo effect influenced outcomes. However, we found no evidence of a placebo effect in evaluations of other treatments in this trial platform. 16,17 In conclusion, in this updated interim analysis, inhaled budesonide improved time to recovery by a median 3 days when used to treat COVID-19 in people at higher risk of adverse outcomes in the community.

Writing group contributions
CCB and FRDH had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. CCB and FDRH decided to publish the

Acknowledgments
We thank the patients who participated in this study. We also thank the many health, and social care professionals and who contributed. The PRINCIPLE trial platform is led from the Primary

Data Sharing
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04.10.21254672 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021
† Data on ethnicity were collected retrospectively via notes review before July 2020 ‡ E.g. angina, heart attack, heart failure, atrial fibrillation, valve problems § Such as Ramipril, Lisinopril, Perindopril, Captopril or Enalapril ll Well-being is measured using the WHO well-being index which includes 5 items relating to well-being measured on a five point scale. A total score is computed by summing the scores to the five individual questions to give a raw score ranging from 0 to 25 which is then multiplied by 4 to give the final score from 0 representing the worst imaginable well-being to 100 representing the best imaginable well-being. ‡ All secondary outcome analyses were conducted on the concurrent randomization and eligible analysis population in participants with SARS-CoV-2 positive analysis population, but restricted to those in the inhaled budesonide and usual care group only. § Estimated relative risk derived from a logistic regression model adjusted for age, comorbidity at baseline, duration of illness, at baseline, with 95% confidence interval. ll Estimated hazard ratio derived from a Cox proportional hazard model adjusted for age, comorbidity at baseline, duration of illness, at baseline, with 95% confidence interval. ¶ Mixed effect model adjusting age, comorbidity, duration of illness at baseline, and time. Participant was fitted as a random effect. WHO well-being score was also adjusted for the score at baseline # Relative risk adjusted for age, comorbidity at baseline, duration of illness at baseline ** Relative risk without adjustment of baseline covariates due to low event rate. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 12, 2021. ; https://doi.org/10.1101/2021.04.10.21254672 doi: medRxiv preprint Figure S1 Estimated mean and 95% confidence interval of daily rating of feeling well over the 28 days follow-up by treatment arm using data extracted on 25 th March 2021 (Concurrent Randomisation and Eligible Analysis population in participants with SARS-CoV-2 positive) . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.