key: cord-0921915-c4owzljk
authors: McElvaney, Oliver J.; McEvoy, Natalie L.; Boland, Fiona; McElvaney, Oisín F.; Hogan, Grace; Donnelly, Karen; Friel, Oisín; Browne, Emmet; Fraughen, Daniel D.; Murphy, Mark P.; Clarke, Jennifer; Choileáin, Orna Ní; O’Connor, Eoin; McGuinness, Rory; Boylan, Maria; Kelly, Alan; Hayden, John C.; Collins, Ann M.; Cullen, Ailbhe; Hyland, Deirdre; Carroll, Tomás P.; Geoghegan, Pierce; Laffey, John G.; Hennessey, Martina; Martin-Loeches, Ignacio; McElvaney, Noel G.; Curley, Gerard F.
title: A randomized, double-blind, placebo-controlled trial of intravenous alpha-1 antitrypsin for acute respiratory distress syndrome secondary to COVID-19
date: 2022-03-11
journal: Med (N Y)
DOI: 10.1016/j.medj.2022.03.001
sha: e00af227f241dd5ce16963d82a4bd2a7391a139f
doc_id: 921915
cord_uid: c4owzljk

Background Patients with severe COVID-19 develop a febrile pro-inflammatory cytokinemia with accelerated progression to acute respiratory distress syndrome (ARDS). Here we report the results of a phase 2, multicenter, randomized, double-blind, placebo-controlled trial of intravenous plasma-purified alpha-1 antitrypsin (AAT) for moderate-to-severe ARDS secondary to COVID-19 (EudraCT 2020-001391-15). Methods Patients (n=36) were randomized to receive weekly placebo, weekly AAT (Prolastin, Grifols, S.A.; 120mg/kg), or AAT once followed by weekly placebo. The primary endpoint was the change in plasma interleukin (IL)-6 concentration at one week. In addition to assessing safety and tolerability, changes in plasma levels of IL-1β, IL-8, IL-10 and soluble TNF receptor 1 and clinical outcomes were assessed as secondary endpoints. Findings Treatment with IV AAT resulted in decreased inflammation and was safe and well tolerated. The study met its primary endpoint, with decreased circulating IL-6 concentrations at one week in the treatment group. This was in contrast to the placebo group, where IL-6 was increased. Similarly, plasma sTNFR1 was substantially decreased in the treatment group while remaining unchanged in patients receiving placebo. IV AAT did not definitively reduce levels of IL-1β, IL-8 and IL-10. No difference in mortality or ventilator-free days was observed between groups, though a trend towards decreased time-on-ventilator was observed in AAT-treated patients. Conclusions In patients with COVID-19 and moderate-to-severe ARDS, treatment with IV AAT was safe, feasible and biochemically efficacious. The data support progression to a phase 3 trial, and prompt further investigation of AAT as an anti-inflammatory therapeutic.

Coronavirus disease 2019 (COVID-19) is a global threat to public health. As of October 2021, more than 250 million laboratory-confirmed cases have been documented, with over 5 million deaths.

Patients with severe COVID-19 typically develop a febrile pro-inflammatory cytokinemia with accelerated progression to acute respiratory distress syndrome (ARDS) (1-3), a pathological entity characterized by alveolar epithelial and lung endothelial injury, excessive protease activity and dysregulated airway inflammation (4, 5) that is associated with prolonged invasive mechanical ventilation, prolonged hospitalization and long-term disability (6) .

Circulating concentrations of the master pro-inflammatory cytokine interleukin (IL)-6 have been shown to increase with disease severity and predict outcome in COVID-19, prompting consideration of therapies designed to counteract its well-described pathological effects.

However, blanket inhibition of the cytokine in COVID-19 should be approached with caution, since IL-6 also regulates metabolism, is essential for innate and adaptive immunity and facilitates pathogen clearance (7) (8) (9) . Critically, the physiological and pro-resolution properties of IL-6 are governed by classical signaling via the membrane bound IL-6 receptor (IL-6R) on hepatocytes and select immune cells, while its pathological and pro-inflammatory effects are driven primarily via a process known as trans-signaling (7) (8) (9) . In the latter, cleavage of IL-6R from the cell surface by the metalloprotease and disintegrin ADAM-17 generates a soluble receptor (sIL-6R) capable of binding circulating IL-6 (10). The resultant IL-6/sIL-6R complexes then interact with cell types that would otherwise be unresponsive to the cytokine (7) (8) (9) . In addition to orchestrating IL-6-mediated inflammatory damage, ADAM-17 drives autoimmunity by cleaving J o u r n a l P r e -p r o o f transmembrane tumor necrosis factor (TNF)-α to its active systemically available form (11, 12) , and promotes neutrophil chemotaxis to IL-8 (13) .

A number of nonspecific anti-inflammatories have been investigated in hospitalized patients with COVID-19 (14) (15) (16) (17) (18) (19) (20) (21) . To date, the number of prospective, double-blind trials of candidate therapeutics that have demonstrated benefit has been low, with the majority of positive results coming from open-label studies. Some drugs, such as dexamethasone, have been rapidly integrated into treatment algorithms based on data from large clinical trials (14) . Despite more mixed results (16) (17) (18) (19) (20) (21) , specific anti-cytokine therapiessuch as the anti-IL-6R monoclonal antibody tocilizumabhave also been implemented.

Alpha-1 antitrypsin (AAT) is a 52kDa glycoprotein synthesized primarily in the liver, and the archetypal serine protease inhibitor (22) (23) (24) , acting to protect the airway against damage by neutrophil elastase (NE), an omnivorous protease released by activated or disintegrating neutrophils that is increased in ARDS (22, (25) (26) (27) (28) . Moreover, AAT is a potent anti-inflammatory and immunomodulator, regulating the production and activity of several key pro-inflammatory cytokines, including IL-6, IL-1β, IL-8 and TNF-α (29) (30) (31) (32) (33) (34) , while maintaining the antiinflammatory cytokine IL-10 (35) . In COVID-19, failure of the acute phase AAT response to keep pace with increasing circulating IL-6 concentrations is associated with poor outcome in patients with severe disease requiring ICU admission (3) . Similarly, abrupt cessation of AAT augmentation therapy for patients with a hereditary deficiency of the protein results in increased systemic inflammation and subsequent progression to respiratory failure (34) . Further supporting its potential for use as a COVID-19 therapeutic, AAT also directly inhibits ADAM-17 cleavage activity (12, 13) and TMPRSS2, the priming protease required for SARS-CoV-2 infection (36, 37) . While recent open-label in vivo studies of AAT for hospitalized COVID-19 patients have J o u r n a l P r e -p r o o f shown evidence of an antiviral and anti-inflammatory effect (38, 39) , randomized control trial data are required to support its use in clinical practice.

In this study, based on biological plausibility, we conducted a phase 2, randomized, doubleblind, placebo-controlled trial of intravenous plasma-purified AAT as an anti-inflammatory therapeutic for patients with ARDS secondary to COVID-19. The primary outcome was the change in plasma IL-6 concentration at 7 days after randomization. The secondary outcomes analyzed included the plasma concentrations of IL-1β, IL-6, IL-8, IL-10 and soluble TNF receptor 1 (sTNFR1).

A total of 86 consecutive patients not enrolled in a concomitant clinical trial were screened, with 36 undergoing randomization having satisfied the criteria for entry to the study ( Figure 1 ). Of these 36 patients, 25 were assigned to receive IV Prolastin and 11 were assigned to the placebo group. Of the 25 patients allocated to the treatment group, 3 discontinued the study prior to day 7. Of these 3 individuals, one died on day 6, and two had improved sufficiently for them to be returned to the center that referred them. Although the latter two patients survived beyond 28 days, neither could be included in the final analysis, since the centers they were transferred to were outside the ethical jurisdiction of the trial.

J o u r n a l P r e -p r o o f

Baseline characteristics of the study groups at the time of randomization are available in Table 1 .

The study groups were adequately matched for age, BMI and clinical severity as assessed by PaO2:FIO2 (122.5 +/-40.5 mmHg in the placebo group vs 129.7 +/-38.2 mmHg in the treatment group) and SOFA score (7.8 +/-3.3 vs 7.2 +/-3.4). Patients were receiving lung-protective ventilation with average tidal volumes of 6.3 +/-0.6 ml/kg/IBW in the placebo group and 6.5 +/-0.8 ml/kg/IBW in the treatment group. A large number of patients were either overweight or obese, with an average BMI of 33.4 +/-8.1 kg/m 2 in the placebo group and 35.2 +/-11.0 kg/m 2 in the treatment group. Two-thirds of the total population studied had a BMI ≥30 kg/m 2 . Just over one-third required vasopressors at the time of randomization. Almost three-quarters of patients were receiving dexamethasone, likely a consequence of the dissemination of preliminary results from the steroid arms of RECOVERY and REMAP-CAP on preprint servers while the present study was ongoing.

Patients in both groups had systemic inflammation, with comparable circulating levels of IL-6 (259.9 +/-206.5 pg/mL in the placebo group vs 266.9 +/-206.9 pg/mL in the treatment group), leukocyte count, C-reactive protein (CRP), lactate, D-dimer and fibrinogen (Table 1, Table S3 ).

AAT levels at randomization were identical between the groups (Table 1) .

In patients receiving Prolastin, plasma AAT concentrations were significantly increased 2 days post-infusion (Fig. S1 ). Of note, AAT levels in the treatment group were still increased at day 7 compared to baseline, suggesting that weekly administration may result in a stacking effect. In J o u r n a l P r e -p r o o f contrast, no change in AAT levels was observed in those receiving placebo. At day 2 and day 7, plasma AAT concentrations were significantly higher in patients receiving Prolastin compared to those in the placebo group (both P <0.0001).

Patients receiving Prolastin demonstrated a decrease in circulating IL-6 at day 7 compared to day 0 (day 0: 296.0 +/-219.7 pg/mL, day 7: 217.7 +/-168.7 pg/mL; Fig 

We next investigated changes in other circulating cytokine concentrations in response to IV AAT. The measurement of TNF-α in blood is complicated by its short half-life and rapid turnover. In plasma, concentrations of sTNFR1 act as a reliable surrogate marker for TNF-α levels (12, 40) . As for IL-6, a significant reduction in plasma sTNFR1 concentrations was observed in the treatment group at day 7 (day 0: 4947 +/-2605pg/ml, day 7: 4131 +/-2207pg/ml; (Table   S4 ). Within-week effects on IL-6 and sTNFR1 were also present, with the reduction in plasma levels of these cytokines greatest at day 2 post-infusion, coinciding with peak circulating AAT levels.

The study was not powered to detect meaningful effects on clinical outcomes such as mortality, but data on these outcomes were collected as part of a safety and feasibility assessment. No difference in mortality was observed between patients who received AAT and those in the treatment group (Fig. S3 ). IV AAT did not significantly reduce the time-to-extubation as assessed at the end of the 28-day study period compared to placebo (Fig. S3) , though the point estimates favored the treatment group, and merit further investigation in a larger study. Similarly, the number of ventilator-free days, SOFA score, PaO2:FIO2, ICU length of stay and hospital J o u r n a l P r e -p r o o f length of stay were numerically improved in patients receiving AAT, without reaching statistical significance ( Table 2, Table S5 ). No increase in secondary bacterial infection was observed in the treatment group, and no rebound effect on safety or clinical outcomes was seen in patients transitioning from AAT to placebo at one week. A summary of clinical outcomes of interest is available in Table 2 .

No AEs or SAEs were considered to be related or probably related to the study drug. One AE (atrial fibrillation in a patient with known paroxysmal AF) was judged to be possibly related to IV Prolastin. One SAE was deemed to be possibly related to IV Prolastin (hypertension persistent for >30 min post-infusion) and resolved without sequelae. No AE or SAE resulted in discontinuation of treatment.

CRP is induced via physiologic classic IL-6 signalingbut not by pathologic trans-signaling or trans-presentationas part of the acute phase response (7) (8) (9) 41) , and serves as an inflammatory biomarker in critical illness. At day 7 post-infusion, patients treated with IV AAT displayed decreased levels of circulating CRP, proportional to the decreases observed for plasma IL-6 ( Fig.   4A ). However, CRP levels post-AAT were still elevated above the normal range (0-5 mg/L), indicating that the classic signalling pathway remained intact in patients receiving IV AAT at 120 mg/kg.

In addition to upregulating the production and release of endogenous AAT during the acute phase response, IL-6 also induces a change in the glycosylation and sialylation of AAT (30) .

This shift, which has previously been described in community-acquired pneumonia, results in the emergence of pro-resolution M0 and M1 AAT glycoforms on serum protein electrophoresis, a phenomenon that is specific for an IL-6-mediated acute phase response via classic signaling (30) .

Monoclonal antibodies against the IL-6 receptor such as tocilizumab do not discriminate between classical or trans-signaling, and therefore abolish both the pathological and physiological effects of the cytokine. Immunofixation of plasma glycoforms from patients in the treatment group by isoelectric focusing gel electrophoresis confirmed the presence of M0/M1 AAT glycoforms ( Fig.   4B ), consistent with preservation of classic signaling. In contrast, when plasma from matched COVID-19 ARDS patients receiving tocilizumab was analyzed, M0/M1 glycoforms were absent, in keeping with inhibition of classic signaling.

Here we present results from a phase 2 randomized, double-blind, placebo-controlled trial of IV plasma-purified AAT for patients with moderate-to-severe ARDS secondary to COVID-19.

Treatment with IV AAT was safe, feasible and biochemically efficacious. Following administration of a single IV infusion of Prolastin, plasma concentrations of the proinflammatory cytokine IL-6 were significantly decreased, an effect mirrored by increases in circulating AAT levels. Levels of sTNFR1 were also substantially decreased at one week in patients receiving Prolastin. Levels of the anti-inflammatory cytokine IL-10 and differential M0/M1 glycosylation of endogenous AAT were preserved, indicating that the decrease in inflammation observed did not come at the cost of pro-resolution mediators.

While the data identify a potential role for AAT as an anti-inflammatory therapeutic in COVID- 19 and ARDS, further studies are required to characterize the COVID-19 and ARDS subphenotypes (42) most likely to benefit from treatment, and the optimal method of administration. We opted for the intravenous route primarily because the study population had systemic inflammation with circulating cytokinemia. However, using aerosolized AAT may also prove effective. There are several potential advantages to aerosolized therapyit limits the potential for volume overload, a relevant consideration in obese patients, and can achieve good deposition in the airways of spontaneously breathing patients (though this is less clear for patients who are intubated and mechanically ventilated) (43) . Data are also available regarding dose equivalencyin previous CF studies for example, the AAT concentration in epithelial lining fluid following a dose of 3 mg/kg was roughly equivalent to that achieved by a 120 mg/kg IV dose (43) .

When designing future studies that incorporate aerosolized therapy, identifying the patient groups most likely to benefit from such an approach represents a key challenge. In COVID-19, aerosolized AAT might prove to be most effective in patients with inflammation mostly confined to the lungs. Although using an airway-directed therapy for a disease that initially takes hold in the respiratory tract seems intuitive, aerosolized AAT therapy may struggle to sufficiently regulate inflammation in patients who have progressed to severe cytokinemia. In these individuals, IV AAT or a combination strategy may be preferable.

To project the ability of IV AAT at 120 mg/kg to provide antiprotease protection in the airway, we can extrapolate from prior trials involving this therapy in other conditions (22, 28, 38, (43) (44) (45) (46) (47) (48) (49) , and present day investigations of protease activity in COVID-19 lungs (50) (51) (52) (53) . In a previous study examining IV AAT in patients with CF, IV AAT at 120mg/kg dose was capable of J o u r n a l P r e -p r o o f providing an antiprotease effect (43) . More recently, open-label use of IV AAT at the same dose for CF complicated by severe cytokinemic COVID-19 resulted in decreased NE activity in airway secretions, following a direct AAT/NE binding event (38) . The NE activity levels observed in these CF studies were higher than those reported in non-CF patients with COVID-19-associated ARDS, suggesting that the dose used here is likely to have provided an antiprotease effect not captured by the current study.

Our study is not without limitations. The population studied was small, and this prevents meaningful conclusions regarding clinical outcomes from being drawn. However, when it comes to studies with a biochemical endpoint, a more modest cohort size may in fact be desirable.

Accurate measurement of IL-6 can be affected by a multitude of patient variables including age, obesity, chronic disease and medications such as IL-6 receptor antagonists (9, (54) (55) (56) . Factors relating to processing and handling of samples also stand to pre-analytically influence assay measurements, in particular delays in processing and over-agitation during transit (31, (57) (58) (59) (60) .

Furthermore, the cytokine displays significant diurnal variation, the most conspicuous effect being a trough in the morning (61), making the timing of sampling especially important. In designing the present study, we were conscious of these potential confounders to our primary endpoint and decided to focus on a smaller number of well-matched patients with intensive follow-up and careful handling and processing of samples so as to maximize precision. Indeed, the concentrations of IL-6 reported here are in line with those described in prior studies of COVID-19-associated ARDS that used similar protocols (3, (62) (63) (64) (65) (66) (67) .

Although circulating IL-6 levels have been shown to correlate well with clinical outcomes, interpretation of an elevated IL-6 level in isolation when comparing disease severity, without consideration of the balance between physiological and pathological IL-6 signaling that J o u r n a l P r e -p r o o f cornerstone of management in COVID-19-associated ARDS was an unavoidable consequence of an evolving situation, and reflects the well-documented phenomenon of preprint data shaping policy during the current pandemic (69) . This underscores the difficulty of undertaking clinical trials in COVID-19, and highlights a need to account for the emergence of other novel therapeutics when projecting the sample size required for a phase 3 study.

Indeed, when applying findings such as these, the evolution of COVID-19 itself must be factored in. As new SARS-CoV-2 variants continue to surface, it is possible that the immune response to the subsequent respiratory syndrome will change. The impact of vaccination on the efficiency of immune responses to SARS-CoV-2 should also be consideredin this study, none of the patients studied were vaccinated against COVID-19, whereas at the time of writing, Ireland has a national vaccine coverage of greater than 94%. However, the recent emergence of escape variants capable of infecting triple-vaccinated, healthy people highlights the enduring importance of identifying effective therapeutics for those who go on to develop severe disease despite taking the requisite precautions.

Moreover, the paradox of pandemic medicine is that if a drug is actually found to be beneficial, its supply may be exhausted. The more options available to clinicians, therefore, the better. The data described here support progression to a larger phase 3 trial focusing on clinical endpoints, and prompt further investigation of AAT as an anti-inflammatory therapeutic in critical illness.

The present study demonstrates the anti-inflammatory effect of IV AAT in critically unwell patients with moderate-to-severe ARDS secondary to COVID-19. However, a larger trial is required to determine the effect of this therapy on clinical outcomes such as mortality or ventilator-free survival. Similarly, the data do not clarify the role of AAT in hospitalized patients outside the intensive care unit. While the study groups were matched for age, healthcare system, baseline inflammation, and clinical severity as assessed by PaO2:FIO2 and SOFA score, they were not matched for sex. The population had a variety of races and ethnicities, with an even distribution between groups, but the relatively small sample size meant that differences in outcomes or response to AAT could not be discerned. 

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Prof. Noel G. McElvaney (gmcelvaney@rcsi.ie).

This study did not generate new unique reagents.

Individual participant data (including data dictionaries) will be made available in a manner compliant with European Union general data protection regulations (GDPR). Deidentified individual participant data and raw data derived from human samples that underlie the results reported in this article are deposited in Mendeley. The DOI is listed in the key resources table.

This paper does not report original code. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Characteristics of the study groups at randomization, including total number, age, sex, body mass index (BMI), past medical history, clinical status and physiological parameters are reported in J o u r n a l P r e -p r o o f Table 1 . Participant information on sex, age, ethnicity and race was self-reported. Information on gender and socioeconomic status was not collected. Biochemical characteristics at randomization are available in Table 1 and Table S3 . James' Hospital, both tertiary referral academic medical centers in Dublin.

Patients were recruited to the study using a next-of-kin assent/deferred consent framework developed in conjunction with and approved by the HRCDC and BHEC. In the event that a participant regained capacity, the investigators obtained written informed consent from each subject to continue in the study. This framework, which was designed specifically for the present study, also provided for the use of a remote virtual assent and consent process given the visitor restrictions and patient isolation protocols in place at the time of study commencement.

The study was an investigator-initiated, randomized, double-blind, placebo-controlled, parallelgroup clinical trial. The trial protocol was published prior to completion of enrolment (70) . A planned data and safety interim analysis was conducted after recruitment of the tenth study participant.

Patients (n=36) were enrolled between April 20, 2020, and March 18, 2021. Hospitalized patients with a working diagnosis of ARDS secondary to COVID-19 were screened on arrival to the ICU. Eligible patients were aged ≥18 years, had a laboratory confirmed diagnosis of COVID- 19 , and were receiving ventilator support for moderate-to-severe ARDS with PaO2:FIO2 <200mmHg. In all patients diagnosed with COVID-19, SARS-CoV-2 infection was confirmed by RT-PCR of a nasopharyngeal swab specimen.

Patients were excluded if they were more than 96 hours from onset of ARDS, receiving extracorporeal life support, known to be pregnant or breastfeeding, or had a history of active malignancy requiring treatment within the last year, pulmonary embolus on a prior admission within the previous 3 months, chronic kidney disease requiring dialysis, severe chronic liver disease, WHO class III/IV pulmonary hypertension, or major trauma in the preceding 5 days.

Patients were also excluded if they were receiving specific anti-cytokine therapies, had hereditary AAT deficiency (AATD), or had participated in a clinical trial of another immunomodulatory investigational medicinal product (IMP) within 30 days, given the potential confounding effects. A summary of inclusion and exclusion criteria for the study are available in Table S1 .

Plasma-purified AAT (Prolastin) was provided free of charge by Grifols S.A., Spain, on compassionate grounds following a written request by the investigators. A weekly dose of 120mg/kg was chosen given the half-life of IV AAT, and its successful inhibition of airway NE in AAT-deficient individuals (44) . IV administration was preferred to the aerosol route for two J o u r n a l P r e -p r o o f reasons. First, patients with severe COVID-19 frequently display significant systemic inflammation; second, the use of an aerosol-generating device may have introduced a safety hazard by increasing the risk of viral transmission. The placebo used in the study was 0.9% sodium chloride solution for infusion (normal saline).

Prolastin is a sterile, stable, lyophilized preparation of purified human AAT, prepared from pooled human plasma from healthy donors by modification and refinements of the Cohn cold ethanol plasma fractionation technique followed by a purification process. Prolastin contains small amounts of other plasma proteins, which may include IgA, haptoglobin, alpha1-acid glycoprotein, lipoprotein A-1, and albumin. Reconstituted Prolastin contains no preservatives and has a pH of 6.6 to 7.4. At administration, tinted infusion sets were used to ensure adequate masking, since Prolastin may be distinguishable from normal saline by color under certain lighting conditions. Unblinded members of the study teamwho were not involved in the care of trial patients, the entry of outcome data or the statistical analysisprepared the masked infusions at pre-designated sites on each hospital campus prior to transporting them to the bedside.

Unblinded team members were instructed not to reveal the treatment allocation for a study participant unless the participant was subject to emergency unblinding measures; no such instance occurred during the trial. Of note, IV AAT may exhibit a mild color tinge under certain lighting conditions. Furthermore, if unduly agitated, a minorbut visiblefroth may develop on the liquid surface within an infusion bag. For these reasons, all study infusions (Prolastin and placebo) were covered using identical opaque infusion sets designed specifically for the study to J o u r n a l P r e -p r o o f ensure adequate blinding. A similar strategy was used in prior placebo-controlled double-blind RCTs of IV AAT for patients with hereditary AAT deficiency, including the RAPID study.

Randomization was performed using a centralized, computer-generated allocation sequence supervised by the trial statistician and stratified by trial site and age (<50 years and ≥50 years).

Eligible patients were randomly allocated using blocks of size 3 to one of 3 groups: weekly AAT for 4 consecutive weeks, weekly AAT for one week followed by weekly placebo for 3 weeks, or weekly placebo for 4 weeks. This approach aimed to achieve a ratio of patients receiving a single AAT infusion to patients receiving a single placebo infusion of approximately 2:1 at one week.

Similarly, at day 7 post-infusion, the subdivision of the AAT group (one half of the AAT group receiving weekly AAT for a further 3 weeks and the other half switching to placebo) was designed to investigate whether a safety signal would emerge with repeated dosing and to assess the feasibility of weekly infusions. Treatment assignments were concealed from patients, clinicians involved in patient care, blinded investigators, and the committee performing the interim data and safety analysis.

The primary outcome was the change in plasma IL-6 concentration at 7 days after randomization.

The secondary outcomes analyzed included the plasma concentrations of IL-1β, IL-6, IL-8, IL-10 and sTNFR1, a plasma surrogate for circulating TNF-α. Safety and tolerability of Prolastin was defined by the number of adverse events (AEs) and serious adverse events (SAEs) related or possibly related to the active IMP. Changes in circulating AAT levels over the course of the study were also recorded.

Similarly, although the study was not powered to detect meaningful differences in clinical outcomes, data were collected on mortality, ventilator-free days (VFDs), time-to-extubation, sequential organ failure assessment (SOFA) score, pulmonary physiological parameters, development of shock, acute kidney injury and clinical relapse, since this information stands to inform the design of a larger study. Thus, in addition to serving as a phase 2 trial with a biochemical primary endpoint, the present trial doubled as a pilot and feasibility study in advance of a phase 3 trial focusing on clinical endpoints. All outcome definitions appear in Table   S2 .

Plasma was obtained by centrifugation of whole venous blood at 250 x g for 5 minutes at room temperature. Blood samples were obtained at the bedside under sterile procedures and transferred immediately to the laboratory for processing. The time from sample acquisition to completion of processing was capped at 30 minutes. Cytokine measurements were undertaken by a blinded investigator not involved in the clinical care of the patients. Similarly, results were not shared with the treating physicians so as not to bias or influence the clinical outcomes assessed. To minimize the potential for inter-assay variability, plasma supernatants were stored immediately J o u r n a l P r e -p r o o f at -80 o C, with samples thawed and assayed on communal ELISA plates en bloc, with a selection of samples run on every plate as points of reference.

AAT levels were measured centrally at Beaumont Hospital by immunoturbidimetric assay (71) .

Regular blood sampling was undertaken immediately before the first dose (day 0), 2 days postdose to coincide with peak circulating AAT levels (day 2) (28), and at 7 days post-dose (day 7), with further sampling at days 14, 21 and 28. AAT protein phenotypes in plasma were determined by immunofixation of glycoforms via isoelectric focusing gel electrophoresis prior to analysis (72) . Plasma levels of IL-1β, IL-6, IL-8, IL-10 and sTNFR1, were measured by ELISA (all R&D systems, Minneapolis MN, USA) in accordance with the manufacturer's instructions.

Although plasma-purified AAT has an established safety profile in patients with hereditary AAT deficiency (AATD) (1-4), it has not been extensively studied in critically unwell populations. For patients recruited prior to the interim analysis, adverse event (AE) and serious adverse event (SAE) reporting was undertaken in line with guidelines specified by the HPRA for studies involving novel COVID-19 therapeutics. Continuation of the study beyond this probationary period was contingent on a favorable safety assessment as part of the interim analysis.

AEs and SAEs were reported to day 28. Events expected in this critically ill population were not reported as AEs unless considered to be related to the study drug, unexpectedly severe or frequent or atypical for a patient with ARDS. In addition, and in accordance with HPRA 

At each study visit, clinical characteristics, process variables and outcome data were inputted into electronic case report forms (eCRFs). AEs and SAEs were identified by treating clinicians and documented in real time and reported to the study sponsor. The relatedness of AEs or SAEs to the study drug was adjudicated by two blinded consultant physicians of requisite experience.

Trial data were monitored at the sites (including consent and source data verification) by independent monitors according to a pre-specified monitoring plan and centrally by pre-assigned staff at each study center.

During the study period, all individuals with a history of recent travel to a designated high-risk or endemic area, recent close contact with a confirmed case, or symptoms in the presence of clinical 

Results are reported as absolute numbers or means and standard deviations, as appropriate.

Categorical variables are summarized as counts and percentages. No imputation was made for missing data. The primary efficacy analysis was on an intention to treat basis. Changes in the levels of inflammatory biomarkers within each patient group were analyzed using paired t-tests for normally distributed data and a nonparametric paired Wilcoxon signed-rank test in the event of data failing the test for normality. Changes between groups were analyzed using unpaired ttests for normally distributed data and nonparametric Mann-Whitney testing for non-normally distributed data. Kaplein-Meier plots were also used to explore time-to-mortality and time-toextubation over the 28 day follow-up period. The statistical analysis plan was approved prior to completion of the study. An independent data monitoring committee (DMEC) performed an interim data and safety analysis after recruitment of the 10th trial subject. Analyses and graphing of data were conducted using Stata 13.0 and GraphPad Prism 8.0 software for Windows. A value of P <0.05 was considered statistically significant.

The recruitment sample size was chosen based on prior studies evaluating the primary biochemical endpoint (5-7) and previous broad recommendations for similar studies pilot studies (8) (9) (10) . Since the present study served only as a pilot study with respect to clinical outcomes, a formal sample size calculation was not undertaken regarding these endpoints.

With the exception of the initial AE/SAE reporting framework required, no substantial feasibility issues arose. No doses of study drug or placebo were delayed and none were missed. In 10 of the infusions administered over the course of the study, the infusion time came within 30 minutes of exceeding the allocated cut-off of 3 hours. In each case, the patient had a BMI ≥30 kg/m 2 . In addition to a dysregulated immune response, obese patients with COVID-19 are more likely to require ICU support, and are also more likely to have comorbidities when compared to patients of equivalent age with a normal BMI. A standard 1g vial of Prolastin is reconstituted in 40ml of sterile water before usegiven the dosing regimen of 120 mg/kg/week, this equates to a total volume of 480ml per infusion for a 100kg person. Some of the patients on this study were as heavy as 200kg. Furthermore, given the set time allocated for infusion, larger volumes could not be spaced out under the study protocol. In clinical practice, this may risk volume overload in some patients and necessitate the use of diuretic therapy. At recruitment, venous access for each patient was established via a multi-lumen central venous catheter. However, following discharge of a patient to the ward, we relied on peripheral venous access for drug administration. Failure of J o u r n a l P r e -p r o o f peripheral venous access required temporary suspension of infusion on 3 occasions while access was established elsewhere.

Post-hoc analyses did not demonstrate a differential response to therapy in study participants <50 years compared to those ≥50 years (Table S7 ). Perhaps the most robust data available regarding the effect of sex on the response to AAT comes from the two largest clinical trials of IV AAT augmentation therapy (RAPID, its open-label extension study RAPID-OLE, and EXACTLE), and their respective subgroup analyses (11) (12) (13) (14) . None of these studies demonstrated a differential response to AAT between sexes. Our trial was no different in this regard. In patients receiving IV AAT, no significant sex differences were observed regarding changes in cytokine levels following therapy (Table S7) .

Given that AAT levels measured at one week post-infusion were elevated compared to preinfusion levels, we examined whether an effect on total protein levels, and by extension a theoretical effect on oncotic pressure, would be observed over time in patients receiving weekly IV AAT compared to patients who transitioned from IV AAT to placebo at day 7 (Table S9) . A statistically significant difference in total protein concentration did not emerge until day 28.

Although it would seem unlikely that weekly dosing would result in a clinically significant effect on oncotic pressure given the short half-life of IV AAT and the offsetting effect of exogenous AAT administration (and its anti-inflammatory effects) on endogenous AAT production and generation of other acute-phase plasma proteins, interpretation of the protein levels reported here J o u r n a l P r e -p r o o f is complicated by several factors. In patients with severe inflammation and ARDS due to COVID-19, hypoalbuminemia is common, not only because of hepatic injury and/or impaired hepatic perfusion, but also because of a downregulated production of albumin by the hepatocyte in favor of other acute phase proteins. Patient attrition in the cohorts also presents a challengesome of these patients died, some recovered well enough to be transferred back to their referring center, and others improved sufficiently to be discharged home. Further study is therefore 

One or more of the authors of this paper self-identifies as a member of the LGBTQ+ community.

While citing references scientifically relevant for this work, we also actively worked to promote gender balance in our reference list. The author list of this paper includes contributors from the location where the research was conducted who participated in the data collection, design, analysis, and/or interpretation of the work.

J o u r n a l P r e -p r o o f Tables   Table 1 J o u r n a l P r e -p r o o f Data presented as number (%) or mean +/-SD. All variables assessed at day 28 unless stated otherwise.

SOFAsequential organ failure assessment PaO2:FIO2ratio of partial pressure of oxygen in arterial blood to the fraction of inspired oxygen ICUintensive care unit * number of patients with an event over duration of study J o u r n a l P r e -p r o o f

Risk Factors Associated With Acute Respiratory Distress Syndrome and Death in Patients With Coronavirus Disease

Severe Covid-19

Acute Respiratory Distress Syndrome

Acute respiratory distress syndrome

Clinical characteristics and day-90 outcomes of 4244 critically ill adults with COVID-19: a prospective cohort study

Interleukin-6: designing specific therapeutics for a complex cytokine

Recent insights into targeting the IL-6 cytokine family in inflammatory diseases and cancer

Interleukin-6: obstacles to targeting a complex cytokine in critical illness

Proteolytic Origin of the Soluble Human IL-6R In Vivo and a Decisive Role of N-Glycosylation

Modulation of TNF-alpha-converting enzyme by the spike protein of SARS-CoV and ACE2 induces TNF-alpha production and facilitates viral entry

The circulating proteinase inhibitor alpha-1 antitrypsin regulates neutrophil degranulation and autoimmunity

alpha-1 Antitrypsin regulates human neutrophil chemotaxis induced by soluble immune complexes and IL-8

Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report

Repurposed Antiviral Drugs for Covid-19 -Interim WHO Solidarity Trial Results

Effect of Tocilizumab vs Usual Care in Adults Hospitalized With COVID-19 and Moderate or Severe Pneumonia: A Randomized Clinical Trial

Effect of Tocilizumab vs Standard Care on Clinical Worsening in Patients Hospitalized With COVID-19 Pneumonia: A Randomized Clinical Trial

Tocilizumab plus standard care versus standard care in patients in India with moderate to severe COVID-19-associated cytokine release syndrome (COVINTOC): an open-label, multicentre, randomised, controlled, phase 3 trial

Coalition covid-19 Brazil VII. Effect of tocilizumab on clinical outcomes at 15 days in patients with severe or critical coronavirus disease 2019: randomised controlled trial

Tocilizumab in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial

Replacement therapy of alpha 1-antitrypsin deficiency. Reversal of protease-antiprotease imbalance within the alveolar structures of PiZ subjects

Structure and variation of human alpha 1-antitrypsin

Isolation, properties, and complete amino acid sequence of human secretory leukocyte protease inhibitor, a potent inhibitor of leukocyte elastase

The role of neutrophil elastase in chronic inflammation

Elastolytic activity in pulmonary lavage fluid from patients with adult respiratory-distress syndrome

High bronchoalveolar levels of tumor necrosis factor and its inhibitors, interleukin-1, interferon, and elastase, in patients with adult respiratory distress syndrome after trauma, shock, or sepsis

Replacement therapy for alpha 1-antitrypsin deficiency associated with emphysema

Anti-inflammatory and immunomodulatory properties of alpha1-antitrypsin without inhibition of elastase

Glycosylation Repurposes Alpha-1 Antitrypsin for Resolution of Community-acquired Pneumonia

Alpha-1-antitrypsin is an endogenous inhibitor of proinflammatory cytokine production in whole blood

Role of inflammation in respiratory tract infections

Alpha1-Antitrypsin Deficiency

Consequences of Abrupt Cessation of Alpha1-Antitrypsin Replacement Therapy

Alpha1-antitrypsin, old dog, new tricks. Alpha1-antitrypsin exerts in vitro anti-inflammatory activity in human monocytes by elevating cAMP

Alpha 1 Antitrypsin is an Inhibitor of the SARS-CoV-2-Priming Protease TMPRSS2

Alpha-1 antitrypsin inhibits TMPRSS2 protease activity and SARS-CoV-2 infection

Alpha-1 antitrypsin for cystic fibrosis complicated by severe cytokinemic COVID-19

Therapeutic Application of Alpha-1 Antitrypsin in COVID-19

Tumor necrosis factor soluble receptors in patients with various degrees of congestive heart failure

Therapeutic strategies for the clinical blockade of IL-6/gp130 signaling

Latent Class Analysis Reveals COVID-19-related Acute Respiratory Distress Syndrome Subgroups with Differential Responses to Corticosteroids

Aerosol alpha 1-antitrypsin treatment for cystic fibrosis

The Biological Effects of Double-Dose Alpha-1 Antitrypsin Augmentation Therapy. A Pilot Clinical Trial

Intravenous augmentation lung density in severe alpha1 antitrypsin deficiency (RAPID): a randomised, double-blind, placebo-controlled trial

Unusual Acute Sequelae of alpha1-Antitrypsin Deficiency: A Myriad of Symptoms With One Common Cure

Biochemical efficacy and safety of monthly augmentation therapy for alpha 1-antitrypsin deficiency

Alpha-1 Antitrypsin Therapy in Cystic Fibrosis and the Lung Disease Associated with Alpha-1 Antitrypsin Deficiency

Long-term efficacy and safety of alpha1 proteinase inhibitor treatment for emphysema caused by severe alpha1 antitrypsin deficiency: an open-label extension trial (RAPID-OLE)

Neutrophil extracellular traps infiltrate the lung airway, interstitial, and vascular compartments in severe COVID-19

Proteinase release from activated neutrophils in mechanically ventilated patients with non-COVID-19 and COVID-19 pneumonia

SARS-CoV-2-triggered neutrophil extracellular traps mediate COVID-19 pathology

Protease-antiprotease imbalance in patients with severe COVID-19

Visceral and subcutaneous adipose tissue volumes are cross-sectionally related to markers of inflammation and oxidative stress: the Framingham Heart Study

Raised interleukin-6 levels in obese patients

Signalling role of adipose tissue: adipokines and inflammation in obesity

Cytokine assays: an assessment of the preparation and treatment of blood and tissue samples

Influence of collection and separation of blood samples on plasma IL-1, IL-6 and TNF-alpha concentrations

Measurement of cytokines in clinical samples using immunoassays: problems and pitfalls

The complex pattern of cytokines in serum from patients with meningococcal septic shock. Association between interleukin 6, interleukin 1, and fatal outcome

Diurnal Variation of Circulating Interleukin-6 in Humans: A Meta-Analysis

An inflammatory cytokine signature predicts COVID-19 severity and survival

Impaired type I interferon activity and inflammatory responses in severe COVID-19 patients

Immune responses to SARS-CoV-2 infection in hospitalized pediatric and adult patients

Severe immunosuppression and not a cytokine storm characterizes COVID-19 infections

Cytokine profile in plasma of severe COVID-19 does not differ from ARDS and sepsis

The role of IL-6 in host defence against infections: immunobiology and clinical implications

Rapid Translation of COVID-19 Preprint Data into Critical Care Practice

A randomised, double-blind, placebocontrolled, pilot trial of intravenous plasma purified alpha-1 antitrypsin for SARS-CoV-2-induced Acute Respiratory Distress Syndrome: a structured summary of a study protocol for a randomised

Serum levels and genotype distribution of alpha1-antitrypsin in the general population

Evaluation of a new Sebia isoelectrofocusing kit for alpha 1-antitrypsin phenotyping with the Hydrasys System

We thank the patients and their families for engaging with this project, the Health Products Regulatory Authority of Ireland for their input regarding the study protocol, and the staff of Beaumont Hospital, who underwent training and engaged in this studywhich was not directly fundedon a goodwill basis, for logistical assistance throughout. We note in particular the 

Treatment options for patients with severe COVID-19, particularly those who progress to acute respiratory distress syndrome (ARDS), are limited. ARDS is a highly inflammatory state hallmarked by airway damage, respiratory failure and increased mortality. Alpha-1 antitrypsin (AAT) is an anti-inflammatory protein produced by the liver and present the bloodstream. We investigated the use of AAT purified from the blood of healthy donors as a therapeutic option for patients with COVID-19-associated ARDS. Treatment with AAT resulted in decreased inflammation at one week, was safe and well tolerated, and did not interfere with patients' ability to generate their own protective response to COVID-19. The results suggest a potential role for AAT in the treatment of COVID-19-associated ARDS and other inflammatory diseases. Circulating levels of interleukin-6 and other pro-inflammatory mediators were decreased at one week in the treatment group, identifying a potential anti-inflammatory therapeutic for critical illness.