Safety and efficacy of sirolimus in hospitalised patients with COVID-19 pneumonia

Abstract

Background:

There is a critical need to develop novel therapies for COVID-19.

Methods:

We conducted a phase 2, multicentre, placebo-controlled, double-blind, randomised trial; hospitalised patients with hypoxemic respiratory failure due to COVID-19 and at least one poor prognostic biomarker, were given sirolimus (6 mg on Day 1 followed by 2 mg daily for 14 days or hospital discharge, whichever happens first) or placebo, in a 2:1 randomization scheme favouring sirolimus. Primary outcome was the proportion of patients alive and free from advanced respiratory support measures at Day 28.

Results:

Between April 2020 and April 2021, 32 patients underwent randomization and 28 received either sirolimus (n = 18) or placebo (n = 10). Mean age was 57 years and 75 % of the subjects were men. Twenty-two subjects had at least one co-existing condition (Diabetes, hypertension, obesity, CHF, or asthma/COPD) associated with worse prognosis. Mean FiO₂ requirement was 0.35. There was no difference in the proportion of patients who were alive and free from advanced respiratory support measures in the sirolimus group (n = 15, 83 %) compared with the placebo group (n = 8, 80 %). Although patients in the sirolimus group demonstrated faster improvement in oxygenation and spent less time in the hospital, these differences were not statistically significant. There was no between-group difference in the rate of change in serum biomarkers such as LDH, ferritin, d-dimer or lymphocyte count. There was a decreased risk of thromboembolic complications in patients on sirolimus compared with placebo.

Conclusions:

Larger studies are warranted to evaluate the role sirolimus in COVID-19 infection.

1. Introduction

The coronavirus disease 2019 (COVID-19) global pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS CoV-2) has resulted in unprecedented mortality and morbidity. There is a critical need to develop novel therapies to mitigate the pathophysiological effects of COVID-19. Modulating signalling downstream of the mechanistic target of rapamycin (mTOR) pathway, which plays a crucial role in the regulation of fundamental cellular processes, such as anabolism, autophagy, immune response, and oxidative stress [1] has been suggested as a potentially useful strategy to combat the effect of COVID-19 and other viral illnesses [2,3].

Sirolimus (rapamycin) is an oral mTOR inhibitor that is widely used to prevent rejection in organ transplant recipients and in other diseases with dysregulation of PI3K/AKT/mTOR pathway such as tuberous sclerosis complex, lymphangioleiomyomatosis (LAM), vascular malformations and certain malignancies [4]. mTOR inhibitors possess a variety of immunomodulatory and antiviral properties. They can enhance virus-specific CD8 T cell response [5], upregulate interferon-induced antiviral responses [6], and inhibit replication of a myriad of viruses including the closely related Middle East respiratory syndrome coronavirus [7,8]. Treatment with sirolimus attenuated lung injury and delayed mortality in murine models of influenza [9,10] and was associated with improved oxygenation and reduced time on mechanical ventilation in a single centre study of critically ill patients with H1N1 pneumonia [11]. Sirolimus can inhibit coronavirus replication by inhibiting the nucleocapsid protein [12,13], and targeting the RNA-dependent RNA polymerase [14]. In addition to the direct antiviral effects, sirolimus can inhibit the expression of proinflammatory cytokines, such as IL-6 and IL-1β, and help suppress the cytokine storm that has been implicated in the progression of COVID-19 [15,16]. Lastly, sirolimus was identified as a leading drug repurposing candidate to target COVID-19 by multiple machine learning and network-based analyses [17–24].

To further assess the safety and efficacy of sirolimus in patients with COVID-19, we designed and conducted the Sirolimus Treatment in Hospitalised Patients with COVID-19 Pneumonia (SCOPE) trial (NCT04341675). Our hypothesis was that treatment with sirolimus will be well tolerated and lead to improved clinical outcomes in hospitalised patients with COVID-19.

2. Materials and methods

2.1. Study design

SCOPE was a phase II, randomised, double blind, placebo-controlled study. Patients had to satisfy the following criteria for inclusion: 1) age ≥18 years with signed informed consent, 2) confirmed COVID-19 pneumonia by SARS-CoV-2 molecular testing and presence of multifocal pulmonary infiltrates suggestive of viral pneumonia on chest X-ray and/or computed tomography (CT) scan, 3) hypoxia as defined by oxygen saturation less than 92 % while breathing ambient air or the need for supplemental oxygen, and 4) presence of at least one of the following biomarkers: a) serum ferritin ≥500 μg/l, b) lactate dehydrogenase (LDH) ≥250U/L, c) d-dimer ≥1 μg/L, or d) lymphopenia as defined by absolute lymphocyte count <1000/μL. Key exclusion criteria included inability to provide consent, requiring advanced respiratory support (high flow oxygen ≥15 L/min, non-invasive or invasive mechanical ventilation) at the time of enrolment, on chronic immunosuppression for other underlying medical conditions, enrolment in another interventional clinical trial, known allergy or hypersensitivity to sirolimus, and pregnant or lactating during the study.

SCOPE was conducted at the following sites: University of Cincinnati (UC) Medical Centre, UC Health Westchester Hospital (WCH), and Loyola University Medical Centre (LUMC). The study was funded by the University of Cincinnati College of Medicine and Pfizer Inc. The study was reviewed and approved by the Institutional Review Boards at UC and LUMC (UC IRB# 2020–0337 and LUMC IRB# 213545). The authors report no conflict of interest. An independent Data and Safety Monitoring Board (DSMB) was convened and met approximately every three months to ensure the safety of study participants. All subjects, or their legally authorized representatives, provided written informed consent prior to the conduct of any study related activity.

2.2. Outcome measures

Our main objective was to determine if treatment with sirolimus can improve clinical outcomes in hospitalised patients with COVID-19. The primary endpoint of our study was progression to death or respiratory failure requiring advanced support measures at day 28 following randomization. Advanced respiratory support measures were defined as the need for either high flow oxygen ≥15 L/min or positive pressure ventilation (non-invasive or invasive). Secondary endpoints included all-cause mortality during hospitalization and 28 days post discharge, increase in total lymphocyte count, normalization of biomarkers (ferritin, LDH, and d-dimer), time to hospital discharge, and drug safety profile.

Prior to starting study drug, baseline demographic and clinical variables were collected along with vital signs, amount of supplemental oxygen required, concomitant medications, past medical history, and laboratory biomarkers such as absolute lymphocyte count, ferritin, LDH, d-dimer, renal and liver function tests. After study initiation, chart reviews were conducted daily to assess vital signs, determine changes in clinical status and oxygenation indices, concomitant medications and laboratory parameters. Study specific biomarkers (absolute lymphocyte count, serum ferritin, d-dimer and LDH) were measured at baseline and then at days 3, 7 and 14.

2.3. Study drug

The study drug (sirolimus) was provided by the drug manufacturer, Pfizer. The tablet was over-encapsulated and matching placebo capsules were created by the UC Investigational Drug Services Pharmacy. In case a patient progressed to require advanced respiratory support and was unable to swallow the study capsules, provisions were made to continue the study drug as scheduled with the liquid formulation of sirolimus and matching placebo via feeding tube.

Study subjects were randomised to sirolimus versus placebo in a 2:1 manner. Patients were randomised according to a computer-generated permuted block stratified on age: <50 years, 50–64 years, and ≥65 years. In addition, randomization was also stratified for each study site. Study subjects as well as study coordinators and investigators were blinded to the assignment. Sirolimus was given as a 6 mg oral loading dose on day 1 followed by 2 mg daily for a maximum treatment duration of 14 days or until hospital discharge, whichever happened sooner. Investigational new drug (IND) designation for studying sirolimus in this trial was obtained from the United States Food and Drug Administration: IND 149969; IND Sponsor: N. Gupta. Treatment related adverse events (AEs) were tabulated using the National Cancer Institute’s Common Terminology Criteria for Adverse Events version 5.

2.4. Sample size estimation

In a study evaluating 201 hospitalised patients with COVID-19 in China, 165 (82 %) patients required supplemental oxygen. Among the patients needing supplemental oxygen, 67 (40 %) patients progressed to require advanced respiratory support during their hospital stay [25]. Considering the fact that our study inclusion criteria mandated the need for supplemental oxygen and the presence of at least one additional poor prognostic biomarker, we anticipated that 50 % of our control group would progress to require advanced respiratory support during their hospitalization. We postulated that treatment with sirolimus would reduce the need for advanced respiratory support to 25 %. The goal of this pilot study was to determine a meaningful treatment difference in order to conduct a larger confirmatory trial. Given the limited data on treatment responsiveness in patients with COVID-19 at the time of our study design, we examined various enrolment sizes commensurate with planning a pilot study. Specifically, we undertook a recommended approach for pilot clinical studies, which is to not perform hypothesis testing with a significance level set at the traditional value of 0.05 [26–28]. The confidence intervals for the treatment difference were calculated based on demonstrating a reduction in the proportion of subjects needing advanced respiratory support from 0.50 in the placebo group to 0.25 in the sirolimus group. Various levels have been proposed as optimal thresholds for the tolerance of type I error in the context of pilot studies, with the highest being p = 0.25 [27]. We chose the type I error rate of 0.20 to denote significant treatment effect between the treatment and placebo arms [28]. A sample size of 30 patients randomized in a 2:1 fashion in favour of sirolimus would allow us to estimate the treatment differences with 80 % confidence based on a binomial two-proportion test.

2.5. Data analysis

The analysis was based on an intention to treat design. All subjects that received study drug were included in the safety and efficacy analysis set. Baseline cohort characteristics and AEs were summarized using descriptive statistics. Continuous variables are presented as mean with standard deviation (SD). Categorical variables are presented as frequencies and percentages. The differences between treatment arms were assessed using Wilcoxon test for continuous variables and Fisher’s exact test for categorical variables. The proportion of patients who progressed to respiratory failure requiring advanced support measures was determined for both the sirolimus and the placebo groups, followed by an assessment of the difference in proportions and the 80 % confidence intervals (CI). Sample size calculations and outcome analyses were performed using R version 4.2.

3. Results

3.1. Participants

Between April 2020 to April 2021, 32 patients admitted with COVID-19 pneumonia were enrolled into the trial after obtaining informed consent, of whom 4 were excluded prior to receiving study medication, 18 received sirolimus and 10 received placebo (Fig. 1). Mean age of the participants was 57 years and 75 % (21/28) of the subjects were men. More than half of the study subjects (16/28, 57 %) were white, 7 (25 %) were black and 5 (18 %) belonged to other races. Twenty-two subjects (79 %) had at least one co-existing condition (Diabetes Mellitus, hypertension, obesity (body mass index >30 kg/m²), congestive heart failure, or asthma/COPD) known to be associated with increased risk of disease progression [29]. The mean fraction of inspired Oxygen (FiO₂) requirement at enrolment was 0.35 with mean SaO₂/FiO₂ ratio of 287. Twenty patients (71 %) received remdesivir and 23 patients (82 %) received systemic corticosteroids. Nine patients (32 %) received convalescent plasma. None of the patients received anti-IL6 therapies or Janus kinase inhibitors. Mean time to randomization was 2 days from admission and 5 days from the initial positive test. Baseline demographics and clinical characteristics were well balanced between the sirolimus and placebo groups (Table 1).

Fig. 1.

Flow Diagram of the study.

Table 1.

Baseline demographic and clinical characteristics of our cohort.

Characteristic	All Patients (n = 28)	Sirolimus (n = 18)	Placebo (n = 10)
Age
Mean (years ± SD)	57 ± 14	57 ± 15	57 ± 14
Distribution; number (%)
<40 years	4 (14 %)	3 (17 %)	1 (10 %)
40–64 years	16 (54 %)	10 (56 %)	5 (50 %)
≥65 years	9 (32 %)	5 (28 %)	4 (40 %)
Gender; number (%)
Female	7 (25 %)	5 (28 %)	2 (20 %)
Male	21 (75 %)	12 (72 %)	8 (80 %)
Race; number (%)
Black	7 (25 %)	5 (28 %)	2 (20 %)
White	16 (57 %)	10 (56 %)	6 (60 %)
Others	5 (18 %)	3 (17 %)	2 (20 %)
Ethnic group; number (%)
Hispanic or Latino	4 (14 %)	3 (17 %)	1 (10 %)
Not Hispanic or Latino	24 (86 %)	15 (83 %)	9 (90 %)
Co-existing condition; number (%)
Diabetes Mellitus	7 (25 %)	5 (28 %)	2 (20 %)
Hypertension	17 (61 %)	11 (61 %)	6 (60 %)
Obesity (BMI ≥30)	17 (61 %)	10 (56 %)	7 (70 %)
CHF	2 (7 %)	1 (6 %)	1 (10 %)
Asthma/COPD	3 (11 %)	2 (11 %)	1 (10 %)
Presence of risk factors*
1 or more	22 (79 %)	14 (78 %)	8 (80 %)
2 or more	19 (68 %)	12 (67 %)	7 (70 %)
3 or more	8 (29 %)	5 (28 %)	3 (30 %)
Days to randomization; mean ± SD
From admission	2 ±1	2 ±1	2 ±1
From positive test	5 ±4	5 ±4	6 ±3
Severity of illness at randomization; mean ± SD
FiO2 requirement	0.35 ± 0.12	0.35 ± 0.11	0.36 ± 0.12
SaO2/FiO2 ratio	287 ± 63	290 ± 58	281 ± 72
Concomitant treatment; number (%)
Steroids	23 (82 %)	15 (83 %)	8 (80 %)
Remdesivir	20 (71 %)	13 (72 %)	7 (70 %)
Convalescent plasma	9 (32 %)	7 (39 %)	2 (20 %)

^*Presence of age ≥65 years, DM, HTN, Obesity, CHF and/or asthma/COPD.

3.2. Primary outcome

There was no significant difference in the primary outcome of patients without progression to death or the need for advanced respiratory support measures in the two groups; 15 patients (83 %) in the sirolimus group compared with 8 patients (80 %) in the placebo group (Table 2, Fig. 2). The point estimate (80 % CI) for the difference in proportions between the sirolimus and placebo arms was 3.3 % (−17.73 %–27.13 %). One patient in each group died during the hospital stay, both of whom had required advanced respiratory support. We did not adjust for baseline demographics and other comorbidities since they were fairly balanced after randomization. Further exploratory analyses would not yield valid results given the small sample size.

Table 2.

Key study outcomes.

Outcome	Sirolimus	Placebo	Difference (80 % CI)
Patients alive and not needing advanced respiratory support measures by Day 28	15 (83 %o)	8 (80 %)	3.3 % (−17.7 %–27.1 %)
Mortality; number (%)	1 (6 %)	1 (10 %)	−4.4 % (−26.9 %–12.6 %)
Need for advanced respiratory support
Total	3 (17 %)	2 (20 %)	−3.3 % (−27.1 %–17.7 %)
High flow nasal cannula (≥15 L/min)	1 (6 %)	1 (10 %)
Non-invasive or invasive mechanical ventilation	2 (11 %)	1 (10 %)
Length of hospital stay in days; mean ± SD
Total hospital stay	9.9 ± 12.2	12.9 ± 11	− 2.96 (−9.08 to 3.16)
Length of stay from randomization	7.9 ± 12.5	11 ± 11	−3.06 (−9.28 to 3.17)

Fig. 2.

Kaplan Meier Curves of the probability of not requiring advanced respiratory support in patients on sirolimus and placebo.

3.3. Secondary outcomes

The average (mean ± SD) length of hospital stay was shorter in the sirolimus group compared with the placebo group: 9.9 ± 12.2 days in the sirolimus group vs 12.9 ± 11 days in the placebo group, although the difference was not statistically significant (Difference of 2.96 days; 80% CI: −9.08 to 3.16 days) (Fig. 3). 5 patients in the placebo arm and 8 patients in the sirolimus arm were discharged on supplemental oxygen. Oxygenation during hospitalization tended to improve faster in the sirolimus group compared with the placebo group, although the difference was not statistically significant; the mean ± SD daily increase in SaO₂/FiO₂ was 26.11 ± 49.94 in the sirolimus group compared to 17.43 ± 50.50 in the placebo group (p = 0.67). There were no significant differences in the rate of change of other biomarkers such as LDH, ferritin, d-dimer (Fig. 4) or absolute lymphocyte count (data not shown).

Fig. 3.

Length of hospital stay in days for patients on placebo and sirolimus treatment arm. Box plot shows median and interquartile range. The circles denote individual patients who were discharged alive and the triangles denote the patients who died during hospitalization.

Fig. 4.

Longitudinal trends of oxygenation status and biomarkers in our study.

A-C, Change in blood Oxygen Saturation (SpO₂)/Fraction of Inspired Oxygen (FiO₂) ratio with overall slope (A) and individual trends in patients in placebo (B) and sirolimus group (C).

D-F, Change in Lactate dehydrogenase (LDH) with overall slope (D) and individual trends in patients in placebo (E) and sirolimus group (F).

G-I, Change in Ferritin with overall slope (G) and individual trends in patients in placebo (H) and sirolimus group (I).

J-L, Change of D-Dimer with overall slope (J) and individual trends in patients in placebo (K) and sirolimus group (L).

3.4. Post-discharge follow up

The study protocol required a 4-week post hospital discharge follow up phone call to determine outcomes following the hospitalization. In addition to the two deaths, 5 patients (2 in the placebo arm and 3 in the sirolimus arm) were lost to follow up after 1 month of hospital discharge. The remaining 21 patients were alive and reported improvement in their symptoms. Three out of 5 patients (60 %) in the placebo arm and 5 out of 8 patients (63 %) in the sirolimus arm who were discharged on supplemental oxygen were successfully weaned off oxygen on follow up.

3.5. Adverse events

A total of 27 AEs occurred in the 28 study subjects (Table 3). Mild diarrhea was the most common non serious adverse event, noted in three participants in the sirolimus group. Two patients from the sirolimus group also experienced bacterial infection: one patient was found to have Stenotrophomonas maltophilia in sputum which was collected prior to initiation of sirolimus, and another patient was found to have Escherichia coli urinary tract infection. Three patients developed thromboembolic disease post admission (2 pulmonary embolism (PE) and 1 deep venous thrombosis (DVT)), all 3 were in the placebo group (difference between sirolimus and placebo group of −30 % (80%CI: −55.17 % to −11.46 %). The majority of AEs were mild and resolved within the study period.

Table 3.

Adverse events.

	Sirolimus	Placebo
Total adverse events	19	8
Patients experiencing any adverse event	9 (50 %)	5 (50 %)
Adverse events
Diarrhea	3	0
Cytopenia	2	0
Increased cell counts	2	0
Elevated transaminases	1	1
Bacterial infection	2	0
Abnormal serum electrolytes	1	0
Abdominal pain	1	0
Epistaxis	0	1
Fluid overload	1	0
Thromboembolism	0	3
Serious adverse events
Readmission	2	0
Need for advanced respiratory support measures	3	2
Death	1	1

Serious AEs (SAEs) were well matched in the two groups (6 SAEs in 18 subjects on sirolimus and 3 SAEs in 10 subjects on placebo). The majority of the SAEs were deemed to not be related to drug; rather were deemed to be secondary to progression of underlying illness. There was one death in each group that was attributed to progression of COVID-19. Two patients required re-admission, both in sirolimus group. One patient with underlying Parkinson’s disease was admitted with respiratory failure suspected secondary to aspiration pneumonia and required mechanical ventilation. Another patient was admitted with worsening hypoxia and dyspnoea suspected secondary to exacerbation of congestive heart failure and pneumonia. Both patients survived the re-admission and were discharged.

4. Discussion

In this phase II, double-blind, placebo-controlled, multi-centre randomised trial among high-risk adults hospitalised with COVID-19, we found that while treatment with sirolimus was safe and well tolerated, it was not associated with reduced risk of progression to advanced respiratory support measures or death. To the best of our knowledge, this is the first controlled clinical trial exploring the role of sirolimus in patients with COVID-19. Even though the primary outcomes in this study were not statistically significant; there was a trend towards improvement in oxygenation, reduced hospital length of stay, and decreased risk of thromboembolic complications amongst patients on sirolimus. Thus, we believe that larger studies are warranted to further evaluate the role sirolimus in COVID-19.

A key aspect of developing therapeutics to combat viral infections relates to the timing of drug administration following viral infection. For instance, most antiviral therapies aimed at inhibiting viral replication are most efficacious if administered soon after contracting the infection [30]. Given that on average the study drug in our trial was started after 5 days of positive test results, it plausible that we missed the optimal treatment window for therapeutic efficacy of sirolimus in COVID-19. Moreover, a murine study of influenza pneumonia suggested that the treatment benefit of sirolimus was likely driven by suppressing the mTOR activation in airway epithelial cells and the reduction in spread of the influenza virus from conducting airways to the alveolar epithelium [10]. The subjects in our study population already had evidence of viral spread to the alveolar epithelium by virtue of their diagnosis of pneumonia and the need for supplemental oxygen. We submit that future treatment trials of mTOR inhibitors in patients with COVID-19, or other viral infections, should aim to begin treatment earlier in the disease course with the primary goal to prevent the development of pulmonary involvement. The potential role of mTOR inhibitors in prevention and early treatment of COVID-19 has also been suggested by other studies [6, 31,32].

COVID-19 infection has been linked to hypercoagulable state and is associated with an increased risk of complications such as DVT and PE [33,34]. While the exact mechanisms of the development of a hypercoagulable state in patients with COVID-19 remain incompletely understood, PI3K/AKT signalling pathway driven platelet activation and expression of coagulation factor has been postulated as a possible mechanism of this effect [35]. We observed three cases of thromboembolic events in the placebo arm, compared to none in the sirolimus arm, suggesting a potential benefit of sirolimus in preventing thromboembolic complications in COVID-19.

The overall safety of sirolimus use in patients with COVID-19 in our study has implications for patients on mTOR inhibitors for other reasons, such as patients with LAM. In a recently published retrospective analysis conducted across the worldwide LAM clinics, the overall outcomes following COVID-19 were similar in LAM patients on sirolimus compared with patients not on sirolimus [36]. There was no difference in the humoral response to SARS-CoV-2 vaccination in LAM patients on sirolimus compared to those not on sirolimus [37,38]. The use of mTOR inhibitors as part of the immunosuppressive regimen was associated with better prognosis following COVID-19 infection in kidney transplant patients [39]. Taken together, these results suggest that patients requiring the use of sirolimus for other indications may not need to interrupt therapy in the wake of COVID-19 infection.

Our study has several noteworthy limitations. The overall sample size for our study was small and limits the derivation of concrete clinical conclusions. Despite the high-risk study population, significantly fewer number of patients progressed to require advanced respiratory support measures than we anticipated, likely due to the interval development and widespread use of medications such as remdesivir and corticosteroids [40,41]. This caused our study to be underpowered in assessing the true treatment effect and might have led to a type II error in the interpretation of the study results. Furthermore, use of remdesivir and steroids may also have mitigated the therapeutic effects of sirolimus. We were unable to ascertain long-term outcomes following hospital discharge in approximately 20 % of our population. Thromboembolic complications were evaluated on clinical grounds as opposed to prospective screening of all subjects. The vaccination status of the patients was not tracked, although we anticipate that most patients were unvaccinated as majority of the patients were recruited in 2020, prior to the introduction of COVID-19 vaccination in the United States. The major strengths of our study include a rigorously conducted, randomized, placebo-controlled, double blind, study design; the inclusion of a select group of patients with COVID-19 associated early respiratory failure and high risk for progression to advanced respiratory failure; well-balanced treatment and placebo groups with regards to baseline characteristics and prevalence of risk factors; and the use of background standard of care clinical treatment of COVID-19.

5. Conclusions

Sirolimus treatment was safe and well tolerated but was not associated with reduced risk of disease progression in hospitalised patients with COVID-19 pneumonia. The results might have been influenced by the timing of treatment initiation and the evolving COVID-19 treatment paradigms. Larger studies investigating the safety and efficacy of mTOR inhibitors in patients with COVID-19, ideally starting early in the disease course, are warranted to definitively assess the utility of this approach.