A multicenter randomized trial to assess the efficacy of CONvalescent plasma therapy in patients with Invasive COVID-19 and acute respiratory failure treated with mechanical ventilation: the CONFIDENT trial protocol

Background The COVID-19 pandemic reached Europe in early 2020. Convalescent plasma is used without a consistent evidence of efficacy. Our hypothesis is that passive immunization with plasma collected from patients having contracted COVID-19 and developed specific neutralizing antibodies may alleviate symptoms and reduce mortality in patients treated with mechanical ventilation for severe respiratory failure during the evolution of SARS-CoV-2 pneumonia. Methods We plan to include 500 adult patients, hospitalized in 16 Belgian intensive care units between September 2020 and 2022, diagnosed with SARS-CoV-2 pneumonia, under mechanical ventilation for less than 5 days and a clinical frailty scale less than 6. The study treatment will be compared to standard of care and allocated by randomization in a 1 to 1 ratio without blinding. The main endpoint will be mortality at day 28. We will perform an intention to treat analysis. The number of patients to include is based on an expected mortality rate at day 28 of 40 percent and an expected relative reduction with study intervention of 30 percent with α risk of 5 percent and β risk of 20 percent. Discussion This study will assess the efficacy of plasma in the population of mechanically ventilated patients. A stratification on the delay from mechanical ventilation and inclusion will allow to approach the optimal time use. Selecting convalescent plasmas with a high titer of neutralizing antibodies against SARS-CoV-2 will allow a homogeneous study treatment. The inclusion in the study is based on the consent of the patient or his/her legal representative, and the approval of the Investigational Review Board of the University hospital of Liège, Belgium. A data safety monitoring board (DSMB) has been implemented. Interim analyses have been planned at 100, 2002, 300 and 400 inclusions in order to decide whether the trail should be discontinued prematurely for ethical issues. We plan to publish our results in a peer-reviewed journal and to present them at national and international conferences. Funding and registration The trial is funded by the Belgian Health Care Knowledge Center KCE # COV201004 Trial registration Clinicaltrials.gov registration number NCT04558476. Registered 14 September 2020—Retrospectively registered, https://clinicaltrials.gov/ct2/show/NCT04558476


Signature pages
A multicenter randomized trial to assess the efficacy of CONvalescent plasma therapy in patients with Invasive COVID-19 and acute respiratory failure treated with mechanical ventilation: the CONFIDENT trial.
Protocol Coordinating Investigator signature page I agree to personally conduct or supervise this study and to ensure that all co-investigators, and study staff assisting in the conduct of this study are informed about their obligations in meeting their commitments.
I will conduct the study in accordance with Good Clinical Practice, the Declaration of Helsinki, and the moral, ethical and scientific principles that justify medical research. The study will be conducted in accordance with all relevant laws and regulations relating to clinical studies and the protection of patients.
I will ensure that the requirements relating to Ethics Committee review and approval are met.
I agree to maintain adequate and accurate records and to make those records available for audit and inspection in accordance with relevant regulatory requirements.
I agree to promptly report to the EC any changes in the research activity and all unanticipated problems involving risks to human subjects or others. Additionally, I will not make any changes in the research without EC approval, except where necessary to ensure the safety of study participants.

Protocol synopsis
Name of Academic promotor/Coordinator/ Sponsor: CHU de Liège / Pr. Benoit Misset Name of test drug/investigational product/device: convalescent plasma from COVID-19 Name of Active Ingredient: human antibodies against SARS-CoV-2 Title of Study: A multicenter randomized trial to assess the efficacy of CONvalescent plasma therapy in patients with Invasive COVID-19 and acute respiratory failure due to treated with mechanical ventilation: the CONFIDENT trial.

Ethical and regulatory considerations -
The Principal Investigator at each center will ensure that this study is conducted in agreement with the Declaration of Helsinki as well as the laws and regulations of the country, whichever provides the greatest protection for the patient.
-All study documents (protocol, any protocol amendment, informed consent form and other relevant documents (eg. recruitment advertisements…) will be submitted to the Ethics Committee for formal approval to conduct the study. The decision of the EC concerning the conduct of the study will be made in writing to the promotor. All correspondence with the IRB/IEC should be retained in the Trial Master File.
-The protocol will be approved by the local Ethics Committee, with the Comité d'Ethique Hospitalo-Facultaire de Liège serving as the central Ethics Committee.
-The protocol will be registered in ClinicalTrial.gov before initiation. - The study will be conducted in accordance with legal and regulatory requirements (Belgian law

Primary
The principal objective of the CONFIDENT trial is to assess the efficacy of 2 units (400-500mL) of plasma collected in convalescents of COVID-19 infection with a titer greater or equal to 1/320 (plaque reduction neutralization test 50) neutralizing antibodies against SARS-CoV-2 with the standard of care, as compared to standard of care, to reduce the mortality at day 28 after inclusion of the patients with SARS-CoV-2 pneumonia who require mechanical ventilation. In case of shortage of plasma with a titer greater or equal to 1/320 on the moment of patient enrolment, plasma from convalescent donors with a titer of at least 1/80 will be released for transfusion. The standard of care will be made of therapies currently recommended in international guidelines regarding organ failure support and treatment of secondary events. At the present time, no treatment is directed against SARS-CoV-2 or the inflammatory response characteristic of this specific disease. In case of evolution of the guidelines due to new scientific release during the study period, the standard of care will be adapted in the two treatment groups.

Secondary
The secondary objectives are to assess:

Endpoints
14/55 CONFIDENT : Plasma therapy for mechanically ventilated COVID patients, version 2.0, 15 September 2020 The principal endpoint will be vital status (dead/alive) at day 28 The secondary endpoints will be day-90 mortality number of ventilator-free days at day 28 number of renal replacement therapy-free days at day 28 number of vasopressors-free days at day 28 use of ECMO before day 28 value of the SOFA score at days 7, 14 and 28 [1] (see appendix A).
changes in SOFA scores over the following 7, 14 and 28 days value of the SARS-CoV-2 viral load, expressed as cycle threshold, [2] in the tracheal aspirates -Anxiety-depression evaluation scale at day 90 and at day 365. [4] (see appendix D ).

Background Information and Scientific Rationale
The COVID-19 pandemic started in China in December 2019 and spread to European countries in early 2020, affecting most severely Italy, Spain and France. Belgium was affected during the same wave and first deaths due to COVID-19 were reported in early March 2020. As of September 10, 2020, almost 28 million cases, including 900,000 deaths, have been confirmed over the world and almost 90,000 cases, including 9,900 deaths, have been declared in Belgium [10].
Current data from the Belgian surveillance system indicate that 20% of the hospitalized patients are in ICUs [11]. Hospital deaths occur either in the ICU (13 / 31, 31 % in Liège University Hospital, personal data) or in non-ICU wards for those patients who would not benefit from intensive care as anticipated by the medical team [12]. Based on United Kingdom ICNARC reports, and according to personal observations in Belgium, the COVID-19 ICU patients require mechanical ventilation in 60% of cases, particularly when they are referred directly by the emergency medical service. According to different series, forty to sixty percent of the mechanically ventilated patients are likely to die, most of them when mechanical ventilation is instituted early (30% survival rate) rather than late in the course of the disease (70 % survival rate) [13]. Similar observations have been made in China [14]. Besides the direct cost in human lives, the current situation in European countries has led to a halt in most of medical and scheduled surgical practices for non COVID-19 patients and to a doubling of the number of ICU beds.
The nursing and medical manpower per patient in ICUs has also been doubled due to the intensity of care that patients require with the use of invasive ventilation and the practice of alternate prone positioning for severe ARDS [15].
Several epidemiologic projections suggest that, in the absence of an effective antiviral vaccine or of therapeutic agents, the human cost may reach as many as 45 million deaths until the global population has developed self-immunization. Preliminary data suggest that immunization occurs within 15 to 20 days of illness in most patients [16] but data in the community are still lacking. The potential duration of the current pandemic is largely unknown and epidemics with coronaviruses are likely to recur. A race for developing active therapies and vaccines started since the onset of the pandemic [16]. The potential therapeutic agents are directed against either the virus itself or the mediators of the exacerbated host response observed in the most severe cases [17]. Current proposed treatments are under evaluation and new therapies and vaccines directed against SARS-CoV-2 are at preliminary steps of development [18]. Therapy with plasma collected from convalescent patients has been proposed to provide passive immunization to the patients at risk of developing severe COVD-19 in small pilot uncontrolled studies [19,20]. 16 Kong, 80 received convalescent plasma. Mortality in this group was 12.5% and considered lower than the mortality rate of SARS in Hong-Kong (17%) [21]. Interestingly, patients with a good outcome Prior to treatment, three patients received mechanical ventilation, three received high-flow nasal cannula oxygenation, and two received conventional low-flow nasal cannula oxygenation. Three patients were discharged and the seven others improved and were ready to discharge at the time of publication [20].
In all the published cases of COVID-19 patients, convalescent plasma was associated with a reduction of viral load despite all the patients were receiving various antivirals (including lopinavir/ritonavir or remdesivir). In the study by Shen et al, the Ct value increased in all patients at day 1 post transfusion up to 40 in all patients at day 12 [19]. Duan et al. observed that 10/10 patients reached undetectability 2 to 6 days after the administration of convalescent plasma [20]. A direct antiviral effect of neutralizing antibodies within the convalescent plasma is therefore likely to play a central role in the beneficial effect of the treatment. Nevertheless, it cannot be ruled out that other mechanisms could participate in the effect. It is well known that IV immunoglobulins have a potent anti-inflammatory effect by various mechanisms including engagement of inhibitory FcγR on macrophages and regulatory T cells [23].
Although the amount of IgG present in two units of plasma is clearly lower than the dose of IgG administered during IVIG therapy, such an effect could take place during treatment with convalescent plasma. Other mechanisms could also intervene. Abnormalities of coagulation and complement activation have been described in Covid [24] and are clearly amenable to improvement after plasma administration. In addition, inhibition of ACE2 enzymatic activity by viral infection has been proposed as a central mechanism in the pathogenesis of COVID-19 [25]. Since ACE2 catalytic activity can be 17/55 CONFIDENT : Plasma therapy for mechanically ventilated COVID patients, version 2.0, 15 September 2020 demonstrated in the plasma due to shedding of the ectoenzyme [26], a beneficial antibody-independent effect of plasma therapy cannot be excluded.
Besides rare adverse effects associated with plasma therapy in general, including transfusion-related acute lung injury (TRALI) [27], convalescent plasma therapy raises the issue of potential antibody dependent enhancement (ADE). ADE is the paradoxical exacerbation of an infectious disease due to the appearance of specific antibodies produced by the infected subject. It has been observed in several viral diseases including Dengue, Ebola, HIV, Sars CoV and MERS CoV. In coronavirus infections, the main mechanism involves a facilitation of virus entry in FcγR expressing cells by antibodies directed against the spike protein but with a suboptimal neutralizing affinity [28]. Observations in dengue also suggest that sub-neutralizing antiviral antibodies could also contribute to inhibit IFN-γ secretion and cytotoxic T cell differentiation [29]. Although ADE clearly plays an important role in the pathogenesis of viral infections, it is important to state that it has never been observed after convalescent plasma therapy.
In the Shen's study [19], the donors had been asymptomatic for at least 10 days and had a neutralizing antibody titer higher than 1/40. No precision is given regarding the severity of their symptoms. In the study by Duan et al [20], 40 donors were included. The donor's blood was collected after 3 weeks of illness and 4 days of hospital discharge. 39 donors had a neutralizing antibody titer higher than 1/160.
In a most recent survey, Joyner observed 2 serious adverse events directly related to convalescent plasma administration for COVID-19 in a cohort of 5,000 patients (0.04 %) treated in the USA [30]. In a series of 175 patients with mild COVID-19 from Shangaï, China, Wu et al [31] observed that the titers of neutralizing antibodies assessed at hospital discharge to SARS-CoV-2 varied substantially. In 11 patients with sequential assessments, high titers over 1/256 were measured as early as 10 to 16 days of disease onset.

Hypothesis
Our hypothesis is that passive immunization with plasma collected from patients having contracted After fulfilling the inclusion criteria, patients will be included by the local investigator in each participating center and randomized through a standard randomization system (UCL clinical trial center, Pr Laterre) into 2 arms (1:1) to receive 2 units (400-500 mL in total) of convalescent plasma plus standard of care from 2 different donors preferably (intervention group) or control treatment made of standard of care alone.

Description of population -
The study population will be patients with SARS-CoV-2 pneumonia associated with acute respiratory failure and requiring mechanical ventilation

Inclusion criteria
We will include patients with the following criteria:  Personal data sharing strategies must comply with national and EU data protection rules.

Exclusion criteria
The following criteria for donor eligibility will be applied additional to national legal requirements: 1. a asymptomatic donor with prior diagnosis of COVID-19 documented by a laboratory test or a clinical picture with radiological confirmation of COVID-19.
2. at least 28 days will have passed since full recovery and disappearance of the symptoms.
Blood establishments might modify the timing of plasma collection when findings regarding the timing of optimal and maximal antibody production in those who have recovered from COVID-19 become available and is confirmed acceptable by AFMPS/FAGG. Plasma collection: Collection, processing and storage Donors will ideally donate plasma by plasmapheresis, but where apheresis is not sufficient to supply enough plasma, whole blood can also be collected, with plasma separation in the blood establishment.
Eligible plasmapheresis donors are allowed, according to Belgian Blood legislation, to donate a total volume of 650 ml per session, 2 liters per month and 15 liters per year. The inter-donation interval is at least one week. Plasma obtained by plasmapheresis and processed will be split before freezing into 2-21/55 CONFIDENT : Plasma therapy for mechanically ventilated COVID patients, version 2.0, 15 September 2020 3 separate units (e.g. 3x200 ml). Final products will be specifically labelled as COVID-19 Convalescent Plasma and stored in a dedicated location. The processing that is routinely used for pathogen reduction by Methylene Blue will be applied according to standard practice in the blood establishment.
The production process and product characteristics will comply with current Belgian Blood legislation.
For this trial, volumes will be between 200 and 250 ml. The product can be stored for 36 months at a temperature below-25°C or for 3 months at a temperature between -18°C and -25°C.

Data collected from donors
The following data will be collected: Plasma sample from all donors will be stored in tube with gel for that purpose and for testing of antibody specificities.
Blood establishments will determine SARS-CoV-2 neutralizing antibody titers at or prior to the first donation to allow the selection of donors with high titers. The neutralizing antibody titers will be repeated with a maximal interval of one month. Samples of donations done in-between will be stored for later testing.
The Blood Establishment will qualify donations from donors with neutralizing antibody titers greater or equal to 1/320 as appropriate for this study. In case of shortage of plasma with a titer greater or equal to 1/320 on the moment of patient enrolment, plasma from convalescent donors with a titer of at least 1/80 will be released for transfusion.
If an adequate correlation between neutralizing activity and Elisa antibody testing for SARS-CoV2 were to be demonstrated by a validation report to be approved by the sponsor, this assay could replace the test for neutralizing antibodies. Additional archive samples of the donated plasma will be saved for reference studies e.g. frozen aliquots from plasma samples taken at the time of donation.
When the measured neutralizing activity in the collected plasma is considered too low, the plasma will be made available for other use.

Convalescent Plasma and Dosing Regimen
Patients in the intervention arm will get 2 units (400-500 mL in total) of convalescent plasma. A single unit has a volume that ranges from 200mL to 250mL. Determination of ABO compatibility will follow local procedures and plasma will be transfused within six hour after thawing. Plasma will be provided by the 22/55 CONFIDENT : Plasma therapy for mechanically ventilated COVID patients, version 2.0, 15 September 2020 local blood bank. Convalescent plasma will be stored in a limited number of places. Products can only be shipped to the local site only when the patient is randomized.

Identity of Investigational Product (s)
Intervention arm: 2 units (400-500 mL in total) of convalescent plasma. The treatment will be administered within 24 hours of inclusion in the study, through a venous line over at least 2 hours. (to prevent potential adverse effect of fluid overload).
Control arm: no plasma

Selection of Doses in the Study
Interventional group: 2 units (400-500 mL in total) of convalescent plasma with at least 1/320 PRNT50 of neutralizing antibodies. Each unit will preferably come from two different donors. Control group: no plasma.

No blinding
We chose not to blind the treatments because the principal endpoint (day 28 mortality) is an objective one. No particular drug-drug interactions are anticipated.

Treatment Compliance
The patients included will be totally dependent from the organ supply they will receive for life support in the intensive care unit. The treatment will be administered in a single administration. The issue of compliance is therefore not applicable.

Availability of the study product
Prior to request informed consent from the patient, his/her representative or a impartial witness (see below), the investigator will contact the CCC St-Luc. The CCC St Luc will check the eligibility of the patient with the local investigator or his/her representative for the study. The CCC St Luc will check the 23/55 CONFIDENT : Plasma therapy for mechanically ventilated COVID patients, version 2.0, 15 September 2020 website of the Red-Cross to ascertain that the convalescent plasma of the patient's ABO group at a titer of 1/320 is available for delivery to the site. In case of shortage of plasma with a titer greater or equal to 1/320 on the moment of patient enrolment, plasma from convalescent donors with a titer of at least 1/80 will be released for transfusion. If the patient is eligible and the plasma is available, the informed consent will be obtained according to the procedure indicated in the "participant consent" paragraph.

Participant consent
The eligible patients will be unable to speakdue to tracheal intubation -and most often likely not able to understand the information about the benefits and risks of the trialdue to sedation and/or confusion due to COVID-induced sepsis. Their relatives will likely be only contacted by phone according to the guidelines of the participating ICU, which most frequently preclude access of visitors to the hospitals due to the pandemic aspect of COVID-19.
The capability to be informed and to provide their consent will be assessed for all potential patients. If the written consent of the patient cannot be obtained, it will be asked from a relative / legal representative of the patient over the phone (preferably through video conference) and the form will be sent for signature to the relative and returned by the him/her by email. In the absence of written consent due to incapacity of the patient or to impossibility to have a contact with the relative / legal representative, the investigator will have to obtain the written approval from an impartial witness. The witness can be a physician from another department of the hospital.
In case of inclusion without the written consent of the patient and in case of recovery, his/her consent to continue to participate and/or to use his/her personal data will be requested.

Process of randomization
The process of randomization will be performed through a standard randomization system (UCL clinical trial center, Pr Laterre) either directly through the eCRF system or by the CCC St Luc. The randomization process will be stratified according to the delay from tracheal intubation to inclusion ("<= 48 hours"; "between 48 hours and 5 days"). A response to telephone calls for any issue regarding the inclusion criteria will be available 24/7 (see "availability of the study product" paragraph).
The treatment (intervention versus control) will be allocated in 2 arms in a 1:1 ratio. The randomization table will be made of permutated block size provided by the statistician.

Screening
The principal investigator of each participating center will organize the detection of potentially eligible

Assessment visits in case of discharge before day 90
In case of acute care hospital discharge before day 90, the day 90 visit will be performed by phone call.

Laboratory tests
The laboratory tests collected are based on the descriptive series of critically ill patients already published on COVID-19, mostly from China and Italy, and on the usual tests collected during the treatment of acute respiratory distress syndrome (ARDS) associated with infection or sepsis.

Premature discontinuation of trial treatment
The treatment will be delivered as a single administration of convalescent plasma. In case of severe adverse reaction during the administration (specifically hemodynamic shock), the study plasma will be stopped, and the adverse event described and collected.
If for any reason the treatment is either not administered or prematurely stopped, the patient will remain in the trial for the intention to treat analysis.
No cross-over will be allowed. in case of hospital discharge, and by phone for day 365 endpoint.

33/55
Vital status, dependence on the ventilator and clinical and biological data to calculate the SOFA score will be collected from the patient's record. Viral load, expressed as cycle threshold, will be assessed from tracheal aspirates or naso-pharyngeal swabs and RT-PCR assessment by the usual technique of the routine laboratory. Adverse events possibly related to study intervention will be collected prospectively, according to definitions of iatrogenic events, transfusion-related events and nosocomial infections by the investigator and the clinical research assistant in charge of the study. Scales assessing functional state, psychological state and quality of life will be collected at day 90 visit and over the phone at day 365 visit by a clinical research assistant dedicated to this task by the coordinating center. The adverse events will be collected and severity rating according to Belgian guidelines.

Appropriateness of Measurements
All the efficacy or safety assessments will be standard and consistent with international guidelines.

Primary Efficacy Variable
The primary endpoint used to determine efficacy will be mortality at day 28. and 5=fatal. Relatedness to the plasma therapy will be scored as follows: 1=not related, 2=unlikely related, 3=possibly related, 4=probably related and 5=definitely related.

Safety Reporting
-A serious adverse event (SAE) is any experience with a significant hazard and includes any event that: is fatal is life-threatening (places the patient at immediate risk of death) requires or prolongs hospitalization is permanently or significantly disabling/incapably is a (new) malignancy is a congenital anomaly/birth defect is a known or suspected overdose

Safety reporting procedures
AEs at least possibly related to the plasma infused are to be reported in the CRF.
Any AE should be documented and followed until it is resolved. New AE appearing after the start of a subsequent disease treatment that is not part of this protocol should not be reported anymore.
Adverse events (as specified above) and Serious adverse events must be reported from the initiation of treatment until day 28 after treatment, excepted death that has to be reported until the primary endpoint is reached.
SAEs must be recorded on the Serious Adverse Event Report form, and reported by e-mail to sponsor within one working day (24 hours) of discovery.

Confidentiality and data handling
-Data will be collected at each center on the eCRF. The investigator of each participating center will be responsible for entering these data into the central database with the help of a clinical research assistant. All statistical analyses will be performed centrally in Liège.
-Patient data will remain confidential, but paper and/or electronic medical records may be reviewed for trial purposes by authorized individuals other than the treating physicians that are involved in the conduct of the trial.All patients are assigned a unique patient number.
This number will be used in any publication that describes the research that the patient has participated in, thereby preventing identification of the patient. All research records are maintained in a locked room, and access to electronic files are restricted by a password.
Data protection according to the EU General Data Protection Regulation (2016/79 of April 27th, 2016) is described in details in each site agreement (Data protection agreement)

Statistical Analysis
-Primary outcome The primary outcome of the trial is mortality at day 28. Based on prior reports from China, Italy and UK, and consistent with personal observations, we anticipate that the mortality at day 28 will be around 40 %. As no comparative study has yet been published, the reduction that we anticipate is empiric. We consider that a one third relative reduction in 28-day mortality is realistic and clinically relevant. As this reduction might be greater, we plan to perform an interim analysis for ethical reasons. With a two-sided α-risk of 0.05 and a β-risk of 0.20, the minimal number of patients to include to reach significance for a one third relative reduction in 28-day mortality (an absolute reduction of 13.5% day-28 mortality in the intervention group assuming a day-28 mortality of 40 % in the control group) is 250 in the control group and 250 in the active group (total number of patients = 500). This number allows for four interim analyses to be performed in addition to the final analysis, respectively after assessment of 100, 200, 300, 400 and 500 patients. Considering that others trials in COVID-19 may result in modification in mortality and therefore in standard of care, we anticipate that the number of patients that is necessary to address our hypothesis may vary, and will have to be adapted depending on the results of the interim analyses.
-Statistical analysis plan Results will be analysed in intention to treat and expressed as means and standard deviations (SD) for quantitative variables with a Normal distribution and as medians and 25th to 75th percentiles for skewed distributed quantitative variables. Qualitative variables will be expressed using counts and percentages.
Mean values between groups will be compared by Student t-test but also by the non-parametric Wilcoxon test when normality assumption is not fulfilled. Proportions will be compared by the chi-square 38/55 CONFIDENT : Plasma therapy for mechanically ventilated COVID patients, version 2.0, 15 September 2020 test. Appropriate multivariate analyses will be performed to assess the relationship between outcome and patients' characteristics and clinical features.
For the primary endpoints, vital status at day 28 in all groups will be calculated with corresponding 95% CI.
Comparison between groups will be tested using chi-square test.
Regarding the secondary endpoints, quantitative values between groups will be compared by Student t-test or non-parametric Wilcoxon for the SOFA score (at different time points), viral load of SARS-CoV-2 (at different time points), circulating CRP (at different time points), ADL score at day 90, anxietydepression scale at day 90, sero-conversion. Poisson regression analysis will be considered to analyse number of ventilator-free days up to day 28. Comparison of proportions will be applied to compare location of patient at day 90, as well as presence of various adverse events. If descriptive analysis revealed the presence of potential confounding factors, appropriate multivariate analysis will then be considered.
The two-sided level of significance used in this trial is 5%. Statistical analyses will be carried out using SAS (version 9.4 for Windows) statistical package.
-Interim analysis plan Due to ethical considerations of the current situation, we propose to conduct a group-sequential design with several interim analyses. The objective of including interim analyses in the study design is to allow the possibility of stopping the study early if a clinically relevant benefit is already established from a statistical point of view. Conversely, the interim analyses also allow the possibility of an early stopping of the study if the observed benefit, at the time of the interim analysis, is so small as to make it extremely unlikely that the study would succeed, were it continued until full accrual. Finally, the interim analyses will also include a futility stopping rule for the SOC group, which may appear to be unjustified if mortality in that group appears too large to be acceptable.
Specifically, the interim assessment of the SOC group will be based on the numbers of patients and of deaths shown in Table 1. The recommendation will be made to stop accruing patients in the SOC group if the number of deaths observed in this group suggests a true mortality rate in excess of 0.4. The rule for DSMB recommendation shown in Table 1 will not be binding. boundary to stop the study for extreme efficacy, If the interim analyses take place with increments of exactly 100 patients, and assuming a mortality of 40% in the control group, the recommendations to stop the study for futility or for efficacy are summarized in Table 2. Of note, the trial will reach statistical significance if the absolute difference in 28-day mortality is 9% or larger, which might still be a clinically worthwhile difference. The rules for DSMB recommendations shown in Table 2 will not be binding.

Changes in the Conduct of the Study or Planned Analyses
-Provision for adaptive design changes at the time of interim analyses Because of the considerable uncertainties about the outcomes that will be observed in the three arms of the trial, the DSMB will be allowed to recommend design changes based on the observed data at the times of the interim analyses. These changes may include an adaptation of the sample size or the patient selection. Any such changes are adaptive in nature and will be implemented using appropriate methods to control the impact of these changes on the operating characteristics of the trial (especially the probability of type I error).

Protocol Amendements
If amendments to the protocol (modifying sense or objectives or modifying the undergone constraints or the risks incurred by the subjects) turn out to be necessary, they will be submitted to the opinion of the

Protocol Deviations
All deviations related to study inclusion or exclusion criteria, conduct of the trial, patient management or patient assessment should be described.

Data Management Responsibilities
The study will be monitored by a DSMB that will include international members not involved in this protocol and a statistician. The DSMB will review the safety data at least every 6 months and at each interim analysis.

Study Data, Data Ownership and Data Sharing with KCE
After the completion of the study the Sponsor will transfer the pseudonymized study data set to KCE.
KCE will request approval from the competent chamber of the Information Security Committee to have the relevant study data linked with IMA data by a trusted third party (TTP, eHealth platform) using the patient national number.
The patient information and consent includes wording that the national number will be recorded on site by the investigator for later data linkage. The patient information and consent will also include that in case the patient is randomized, it is planned that a trusted third party (TTP, eHealth platform) will receive and use the national number to link with IMA administrative data. This data linkage is planned to obtain a more complete data set that will be used for the analysis of effectiveness and cost-effectiveness of the intervention by KCE.
KCE and Sponsor have entered into a research agreement detailing the roles and responsibilities of each party, as well as other legal aspects of this collaboration, including the right to use and access of KCE to the Study Data.
"Background" means any intellectual property (IP), data, materials, information owned or controlled by the Sponsor or a Site, and required to run this Study. Sponsor will identify such Background including the legal restrictions of which Sponsor or Sites are aware that may affect the use of the Background for the purpose of the Study or the rights granted to KCE under this Agreement.
The Study Data consist of this protocol, including amendments, the electronic forms for data capture, including the annotations and guidance for use, the electronic database of the pseudonymized clinical and non-clinical data collected using data capture, including the log of changes from data entry to 42/55 CONFIDENT : Plasma therapy for mechanically ventilated COVID patients, version 2.0, 15 September 2020 database lock, study reports based on these pseudonymized data, and any data or reports generated at a later stage, eg based on exploratory analyses or stored samples.
"Foreground" means any Study Data, and any tangible biological, chemical and physical material and inventions, that are generated, acquired, discovered, conceived, developed, created, exemplified or derived as a result of carrying out the Clinical Study, whatever its form or nature, whether it can be protected or not, as well as any Foreground IP. Sponsor acknowledges that the main purpose of the research performed under this Agreement is to generate results that will serve the general public interests, and specifically the interests of the patients and public healthcare decision making bodies, and, therefore, undertakes not to exploit the Foreground in any way that is or could be detrimental to such interests. According to the Belgian law, the principal investigator is responsible for any damage, even without fault, caused directly or indirectly to the participant by the experimentation. Therefore, a specific trial insurance will be provided for patients treated in Belgium through the CHU of Liège.
 The financial agreement between the promotor, the investigator and the Institution,

Dissemination of Results and Publication Policy
The results will be submitted for presentation to international societies of critical care and/or infectious diseases and for publication to an international scientific medical journal. The authorship will include the members of the writing and/or steering committee, the investigators from centers having recruited at least 10 patients, a representative of the Belgian Red Cross and transfusion services, the biostatistician of the study and all scientists whose contribution will have been substantial. Pseudonymised Data from this study can be used by the funder (KCE) (for more details, please check the research agreement signed between sponsor and KCE) or similar public healthcare research institutes in Europe for further analyses, for example to determine whether one of the treatments studied provides better value in metaanalyses.
The data required by the European requirements will be shared according to https://ec.europa.eu/health/blood_tissues_organs/covid-19_en, and the trial will be registered at https://ec.europa.eu/health/sites/health/files/blood_tissues_organs/docs/guidance_plasma_covid19_e n.pdf.

Archiving
The CHU de Liège, more specifically the Department of Intensive Care provides a secure locked office for the archiving of all study documents related to the study as in the Trial Master File (CRF, ICF, Source documents,…) during at least 25 years (as to Circular 536/2014, Art.58).
As to the Data Processing Agreement: Patient data will be collected by personnel of the clinical department where patient is taken care of, all members of hospital staff.
Data will be stored in electronic format in the main database at CHU de Liège for 25 years, Department of Intensive Care.
Informed consent forms signed by patient (or legal representative or impartial witness) and physician, as well as all trial documentation, be it in paper or electronic format, shall be stored in the clinical department where patient is taken care of in a secure, locked office. A copy of the signed consent form will also be scanned in the electronic patient medical file of the respective hospital.
All trial documentation communicated to the sponsor, be it in paper or electronic format, shall be stored in the Department of Intensive Care at CHU de Liège in a secure, locked office.

Study Report
The final report will be written after statistical analyses and interpretation, preferably within 3 months after the last visit of the last patient. The submission for publication of the main results should be available within one month after day 28 of the last patient included. This short schedule is motivated by the importance of the global crisis induced by the COVID-19 pandemic.