Management of pharmacovigilance during the COVID-19 pandemic crisis by the safety department of an academic sponsor: Lessons learnt and challenges from the EU DisCoVeRy clinical trial

Abstract

The current COVID-19 pandemic was an exceptional health situation, including for drug use. As there was no known effective drug for COVID-19 at the beginning of the pandemic, different drug candidates were proposed. In this article, we present the challenges for an academic Safety Department to manage the global safety of a European trial during the pandemic. The National Institute for Health and Medical Research (Inserm) conducted a European multicenter, open-label, randomized, controlled trial involving three repurposed and one-in development drugs (lopinavir/ritonavir, IFN-β1a, hydroxychloroquine, and remdesivir) in adults hospitalized with COVID-19. From 25 March 2020 to 29 May 2020, the Inserm Safety Department had to manage 585 Serious Adverse Events (SAEs) initial notification and 396 follow-up reports. The Inserm Safety Department's staff was mobilized to manage these SAEs and to report Expedited safety reports to the competent authorities within the legal timeframes. More than 500 queries were sent to the investigators due to a lack of or incoherent information on SAE forms. At the same time, the investigators were overwhelmed by the management of patients suffering from COVID-19 infection. These particular conditions of missing data and lack of accurate description of adverse events made evaluation of the SAEs very difficult, particularly the assessment of the causal role of each investigational medicinal product. In parallel, working difficulties were accentuated by the national lockdown, frequent IT tool dysfunctions, delayed implementation of monitoring and the absence of automatic alerts for SAE form modification. Although COVID-19 is a confounding factor per se, the delay in and quality of SAE form completion and the real-time medical analysis by the Inserm Safety Department were major issues in the quick identification of potential safety signals. To conduct a high-quality clinical trial and ensure patient safety, all stakeholders must take their roles and responsibilities.

Abbreviations
• AE
• adverse event
• AIDS
• acquired immuno deficiency syndrome
• CRA
• clinical research associate
• DSMB
• data safety monitoring board
• eCRF
• electronic case report form
• ICF
• informed consent form
• ICU
• intensive care unit
• IMP
• investigational medicinal product
• INF
• interferon
• Inserm
• The National Institute for Health and Medical Research
• IT
• information technology
• MedDRA
• medical dictionary for regulatory activities terminology
• PV
• pharmaco vigilance
• SAE
• serious adverse event
• SAR
• serious adverse reaction
• SARS
• severe acute respiratory syndrome
• SoC
• standard of care
• SOP
• standard operating procedures
• SUSAR
• suspected unexpected serious adverse reaction
• WHO
• World Health Organization

BACKGROUND

In December 2019, a new severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-associated coronavirus disease 2019 (COVID-19) emerged in China, subsequently causing a worldwide pandemic.^1,2 At the beginning of March 2020, COVID-19 had already spread to more than 100 countries and affected more than 100 000 individuals. The WHO declared a COVID-19 pandemic on 11 March 2020.³ Although many drugs, including repurposed drugs, have shown in vitro activity against coronaviruses, no data on efficacy and/or safety in humans were available. WHO recommended researchers around the world to systematically evaluate experimental therapies in randomized controlled trials, and if possible, in large trials that could provide strong evidence for safe and effective treatments. Worldwide research efforts against SARS-CoV-2 initially focused on repurposed drugs that showed broad-spectrum antiviral activity against coronaviruses.^2,4-9 In February 2020, the spread of this outbreak in Europe led the French National Institute for Health and Medical Research (Inserm) to develop a multi-arm randomized, platform-controlled trial: the DisCoVeRy trial. This is a European trial designed to evaluate the clinical and virological efficacy and safety of repurposed antiviral drug, and remdesivir compared with standard of care in adults hospitalized for COVID-19.¹⁰

During the first wave of the pandemic, the disease and its complications were not well characterized and the safety management of these repurposed drugs was very conservative. Moreover, the setting of this trial during the pandemic was very uncommon, with various strains: crowded hospital units, impossibility for Clinical Research Associates (CRAs) to go to hospitals, social and media pressure for results. We present the challenges of managing safety, at the start of this trial, for an academic pharmacovigilance (PV) department: difficulties encountered, hurdles overcome, lessons learnt, and recommendations for the future.

DESCRIPTION OF THE TRIAL Design of the trial

DisCoVeRy is a phase III, open-label, adaptive, randomized, controlled, multicenter trial designed to evaluate the safety and efficacy of repurposed drugs in hospitalized adult patients diagnosed with COVID-19.¹⁰ Initially, the DisCoVeRy study had five arms: (1) a standard of care (SoC) and four investigational arms with repurposed antiviral agents: (2) remdesivir + SoC, (3) lopinavir/ritonavir + SoC, (4) lopinavir/ritonavir associated with IFN-β1a + SoC, and (5) hydroxychloroquine + SoC.

Study treatments

Patients were randomized to one of the five arms (Figure 1).

FIGURE 1. DisCoVeRy study design and treatment schemes.

Safety considerations of the trial

The safety endpoints include the cumulative incidence of any grade 3 and 4 adverse events (AEs); the cumulative incidence of any serious AEs (SAEs); the proportion of patients who prematurely discontinued or temporarily suspended study drug for any reason; and any grade changes in laboratory parameters, measured using the Division of AIDS table. In this study, all grade 3 and 4 AEs and SAEs occurring between the signing of the informed consent form (ICF) and the Day 29 visit (end of study) were documented, recorded, and reported.

Information technology (IT) support The Inserm safety database

All valid SAEs occurring in the trial were captured in the Inserm safety database, AB-Cube® Safety Easy V5. All safety analyses were performed using direct extractions from the Inserm safety database (not from the clinical database) with post hoc reconciliation between the two databases.

Clinical database (electronic case report form, eCRF)

Data were collected in the CleanWeb® eCRF. This software was launched at the end of March 2020.

Inserm's IT tools

The implementation of the trial coincided with the lockdown in France, leading to a generalization of teleworking for all staff members of the Inserm PV Department.

The Inserm PV department

At the beginning of the trial, the Inserm PV Department consisted of: one physician (head of department), one assistant, and six drug safety officers. Another drug safety officer joined the team a few weeks later. During the first wave of the pandemic, healthcare volunteers were mobilized. A simplified training mainly in data entry and Medical Dictionary for Regulatory Activities Terminology (MedDRA) coding was set up; a medical student and a pediatrician benefited from this rapid training program on safety matters.

Other safety stakeholders The independent medical officer

A retired independent physician joined the safety team on a voluntary basis as an external expert and reviewed all SAEs that occurred. This additional and centralized review led to a better global analysis of the SAEs.

The data safety monitoring board (DSMB)

An independent international DSMB was established to preserve the interests of participants to monitor main outcome measures (safety and efficacy) and to supervise the conduct of the trial. The DSMB was convened for a meeting after 100 participants were included in the study, and then after every 200 new inclusions, with a maximum of one meeting per week.

Cause of death adjudication committee

Causes of death were reviewed and classified by an adjudication committee including an infectious diseases specialist (N. Peiffer-Smadja), an ICU specialist (J. Poissy), a pneumologist (C. Andrejak) and the Inserm PV team (A. Diallo and N. Mercier). Adjudication was based on medical records, and all data considered relevant in the eCRF.

Pharmacovigilance department of industrial partners

In the DisCoVeRy trial, investigational medicinal products (IMPs) were kindly provided by pharmaceutical companies as COVID-19 indication was an off-label use. Based on the Safety Data Exchange Agreement, SAEs were transmitted to each partner based on the patient's treatment exposure during the trial.

Gilead Sciences, Inc. provided remdesivir, Sanofi provided PLAQUENIL® (hydroxychloroquine), the Merck Group provided REBIF® (IFN-β1a) and AbbVie provided ALUVIA® and/or KALETRA® (lopinavir/ritonavir).

Definitions

• Valid SAEs: to be considered valid, registered, and coded in the Inserm safety database, an SAE had to have occurred between D0 (inclusion date) and D29 (last protocol visit) and contain the minimum following information: patient ID, AE description, IMP, and investigator name.
• Reference safety information (RSI): the expectedness of Serious Adverse Reactions (SARs) was defined in accordance with the RSI section contained in the reference document of each IMP

Safety data circuit Adverse event (AE) reporting

All grade 3 and 4 AEs were reported in the eCRF, including laboratory abnormalities and any error in the IMP administration route or dose. All these non-serious AEs were coded by the Inserm PV Department.

Serious adverse event (SAE) reporting

According to the EU regulation, investigators must report all SAEs immediately and no later than 24 h after becoming aware (see Figure 2). Upon receipt of an SAE Form, the Inserm PV Department immediately forwarded the notification to the independent medical officer and when applicable to the industrial partner concerned.

FIGURE 2. Safety and communication data circuit.

The Inserm PV Department was responsible for data entry into the Inserm safety database and for medical evaluation (MedDRA coding, sponsor causality assessment, and expectedness). The Inserm drug safety officer implemented the medical feedback of the independent safety officer. If the SAE was possibly related to the IMP and unexpected according to the RSI section, the SAE was considered as a Suspected Unexpected Serious Adverse Reaction (SUSAR) and had to be promptly reported to competent authorities within the regulatory timeframes. In parallel to this regulatory report, the Inserm PV Department had to notify the sponsor's representative, the DSMB, the principal investigators of each country involved, the independent safety officer and industrial partners (if any). In France, the deadlines for sending the initial report are: without delay after being made aware for fatal and life-threatening SUSARs, and within 15 calendar days for other SUSARs. A follow-up report had to be sent within 8 days after the initial submission.

PROGRESSION OF THE TRIAL AND DIFFICULTIES ENCOUNTERED Context

The DisCoVeRy trial was first authorized by the French regulatory agency (ANSM) on 9March 2020 and the first participant was enrolled in France on 22 March 2020. Due to the pandemic and the urgent need to quickly set up a large-scale clinical trial, different steps were taken in parallel to obtain trial approvals before the eCRF was finalized– a paper CRF was finished and submitted initially. Simultaneously, the lockdown in France started on 17 March 2020 and severely restricted movement, leading to a reorganization of work by the Inserm PV Department. Homeworking was mandatory, the internal communications of the department had to be reorganized via phone calls and the use of telecommunication tools. As the first IT tools were not adapted, the PV Department was faced with frequent difficulties in accessing the network, and mailbox problems. During this period, all members of the PV Department were mobilized in the trial and SAE assessment. This strong mobilization was possible only because during this pandemic most other ongoing clinical trials were temporarily suspended due to the national lockdown.

AE reporting

About 293 non-serious AEs were reported in the eCRF during the first 10 weeks of the trial. The Inserm PV Department was responsible for coding these AEs directly into the eCRF coding module. Unfortunately, this module did not allow the identification of the patient and it was therefore not possible to question the investigators to obtain additional information. The Inserm PV Department identified 61 (20.1%) AEs that should have been considered as serious, including stroke, septic shock or pyelonephritis, without any possibility of contacting investigators to fill out an SAE form.

SAE notification management

The Inserm PV Department received the first SAE reports on 25 March 2020, 3 days after the first inclusion, and a total of 585 initial notifications during the first 10 weeks of the trial (this does not include complementary reports) (see Figure 3). During this period, a drug safety officer had to work more than 10 h/day, including weekends and bank holidays. Indeed, the large number of SAEs received during the first weeks led the Inserm PV Department to work quickly to assess the SAEs and report potential SUSARs to the competent authorities within the legal timeframes. In parallel, a daily clinical review meeting was set up to discuss the SAEs, harmonize MedDRA and to characterize the causal relationship between the event and the IMP. These sponsor assessments were compared with those performed by the independent medical safety officer and led to the reassessment of some cases, but were generally consistent with assessments made by the Inserm PV Department.

FIGURE 3. Number of SAEs (initial and follow-up forms) processed by the Inserm PV Department during the first 10 weeks of the trial.

The SAE forms were poorly detailed (missing data), resulting in numerous queries to the investigators: 520 queries were sent during the lockdown period. These particular conditions of missing data and lack of description of the SAEs made it very difficult to evaluate the SAEs, particularly on the causal role of IMP.

From March 25, 2020 to May 29, 2020, 585 initial SAE notification forms and 396 follow-up reports were received. Of these notifications, 74 (12.6%) were not valid (e.g., SAE started after D29) and 47 (8.0%) occurred in patients without a valid written informed consent form.

At the same time, the investigators were overwhelmed by the management of patients suffering from COVID-19 and therefore did not have time to answer queries and document the SAEs. The clinical staff was also very busy with several clinical studies on COVID-19 and working conditions were poor. The exhaustive reporting of all SAEs represented too much information to report to the sponsor in this pandemic context.

Another major point of discussion with clinicians concerned the reporting of SAEs for patients hospitalized in ICU. Indeed, in ICUs, all patients were monitored closely, and some laboratory parameters could fluctuate from day to day without any major clinical consequences. Some other clinical AEs were identified by investigators as common in COVID-19 patients (e.g., respiratory failure). In this case, investigators did not understand why they should notify the sponsor of respiratory failure immediately. Inserm, as the sponsor, tried to discuss with the French competent authority to lighten the reporting of SAEs using a predefined list of SAEs that do not require immediate reporting to the sponsor.

Impossibility of real-time and on-site monitoring

In clinical studies, pharmacovigilance is directly impacted by on-site monitoring. During lockdown, this monitoring was very difficult to implement due to restrictions on access to hospitals (notably due to a“White Plan” in France). At the beginning of the trial, the sponsor tried alternatives, including remote monitoring, but this needed time to be able to be implemented. The first remote monitoring took place on April 24, 2020, followed by 54 other monitoring visits set up in 11 different centers during the French lockdown.

The COVID-19 data generated from the DisCoVeRy trial were regularly analyzed for the evaluation of the safety and efficacy of repurposed drugs. The lack of implementation of real-time monitoring may have led to under-reporting of SAEs, although SAE reporting was in the end very acceptable. Another problem identified was the reporting of SAEs for participants with non-conformity of written consent (n = 47).

eCRF

In this pandemic situation, it was very important to have a quickly operational eCRF, which led to some concessions in the development of the trial. The main concessions were the absence of automatic alerts when an investigator modified a notification form or answered to queries, and the possibility for the investigator to delete a form without reason and without an automatic alert to the sponsor. These lack of alerts resulted in the non-receipt of updated forms by the sponsor. Another important concession was related to the non-serious AE coding module, which did not permit a non-serious AE to be associated with a participant ID, it was impossible to have the severity and to question investigators.

In parallel, at the beginning of the trial, when the eCRF was not ready, the Inserm PV Department received the first SAE in paper forms (n = 43). Subsequently, when the eCRF was ready, all these paper forms were entered in the eCRF and the Inserm PV Department had to check that the correct data had been entered and whether new information had been added.

In addition, the DisCoVeRy team chose to have a very complete eCRF with a large collection of data such as pharmacokinetic and pharmacodynamic data. The completion of the eCRF was identified, hindsight posteriori, as a significant source of additional workload for the clinical teams.

DSMB meeting

During the first months of the trial, the DSMB met on 21 March, 20 April, 27April, and 11 May. The DSMB members were informed that all data transmitted to them were not monitored.

To perform better statistical analysis, all safety data were extracted from the Inserm safety database. This decision had important consequences, including the importance of a real-time quality control of all SAEs and the harmonization of their coding. Regarding the causes of death, these were most often multifactorial. Causes of death were reviewed and classified by the adjudication committee.

Moreover, the DSMB members notified the Inserm PV Department, in April, of under-reporting of SAEs for patients included in the SoC arm. This under-reporting could have led to a wrong interpretation of the safety analysis profile. The sponsor therefore had to remind to the investigators of the importance of reporting SAE, even if patients were randomized to the SoC arm.

SOLUTIONS AND PROSPECTS

At the beginning of the trial, there was insufficient knowledge about COVID-19 disease: there was no consensus on the management of disease-related events and the impact of comorbidities was unknown. The disease and its complications were not well established. Three of the four IMPs used in the DisCoVeRy trial were repurposed, so the safety profiles of these drugs were a priori well known, but the pathophysiological context could modify their safety profiles.

The use of repurposed drugs in these conditions could lead to the identification of new toxicities. Therefore, the same vigilance as for the drugs in development, was required. The Inserm PV Department was responsible for the management and evaluation of all SAEs occurring in the trial. The continuous flow of reported SAEs needed a quick safety evaluations. The main challenge was to find the time to get an overview of the safety profile of each IMP.

In the future, sponsors must anticipate the massive flow of SAEs to be evaluated and facilitate and accelerate the staff recruitment process. Another important challenge in the future will be to identify specific drug toxicities through close monitoring and immediate reporting to the sponsor. The disease-related events are not requiring immediate notification should be identified with a periodic review by the sponsor and the DSMB. For hospitalized patients, including ICU, it will be crucial to identify protocol specified exempt non-serious and serious events.

Accordingly, it will be crucial for the competent authorities relax the regulations on reporting in a pandemic context, particularly for repositioned drugs. Finally, real-time monitoring should be simplify, flexible, and targeting essential data. Even if in a pandemic context, there were published paper unmonitored data, it will always be more consistent to draw conclusions from monitored data.

Filling the eCRF was a major hurdle for investigator sites. Facing an unknown disease, maximum data collection seemed a reasonable choice, but this led to increased workload for the clinical teams. In the future, it will be preferable to develop a pragmatic eCRF; it will always be possible to plan sub-trials to obtain more data.

All these difficulties were accentuated by the frequent IT problems.

All these points contributed to the decrease in the efficiency of the Inserm PV Department. In the future, sponsors should anticipate these problems.

To reduce some of the difficulties encountered in this pandemic crisis, the Inserm PV Department should have anticipated by drafting standard operating procedures (SOPs) for the management of multi-country trials and by developing a risk management plan on the setting up of clinical trials during a health crisis. In the future, sponsors should prepare templates of protocols, eCRFs, and general SOPs dedicated to trials in an epidemic context.

CONCLUSION

During a pandemic crisis, Sponsors have to raise awareness among the investigators of the importance of the pharmacovigilance issues and implement prompt monitoring of the informed consent form. The management of the pharmacovigilance during a pandemic of new pathogen should be clearly included in preparedness plans at national and international levels.

Conducting a clinical trial during a pandemic crisis is challenging: the right balance must be found between collecting a lot of data and taking into account the lack of time for the investigators to provide these data.

Nomenclature of Targets and Ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2019/20 (Alexander et al., 2019).

AUTHOR CONTRIBUTIONS

N. Mercier and A. Diallo were in charge of data curation and accessed and verified the data. N. Mercier and A. Diallo wrote the original draft of the manuscript, which was reviewed and edited by N. Mercier, A. Diallo, M. Hites, F. Mentré, D. Costagliola and Y. Yazdanpanah. All authors contributed to refinement of and approved this manuscript. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.

ACKNOWLEDGMENTS

We thank the Data Safety Monitoring Board for their invaluable support and help. This article would not have been possible without the teams involved in the DisCoVeRy study and designated as the DisCoVeRy study group: liste des noms Séverine Gibowski, Christelle Paul, Joséphine Balssa, Elise Landry, Anaïs Le-Goff, Solange Lancrey-Javal, Lea Levoyer, Laurence Moachon, Vida Terzic, Mélanie Figarella, Nicolas Pittoni-Minvielle, Anaïs Boston, Joe Miantezila-Basilua, Pauline Ginoux, Blandine Gautier, Adele Wegang-Nzeufo, Alain Caro, Alexandra De Lemos, Alexandra Seux, Alice Verdier, Ambre Champougny, Anne Sophie Martineau, Anne-Lise Varnier, Aude Le Breton, Audrey Langlois, Axel Duquenoy, Beatrice Mizejewski, Benlakhryfa Assia, Bercelin Maniangou Zonzeka, Boudjoghra Nour El Houda, Brigitte Mugnier, Brigitte Risse, Camille Chevalier, Camille Harpon, Camille Jourdan, Caroline Dubois-Gache, Celina Pruvost, Charlotte Cameli, Cherifi Asma, Chloe Bureau, Christelle Back-Laufenburger, Christelle Lucas, Christina Schiano, Christine Pintaric, Coline Perrier, Collette Camille, David Simo, Dominique Lagarde, Edit Faillet, Elina Haerrel, Elodie Donet, Eric Marquis, Euma Fortes Lopes, Fabrice Bouhet, Florence Le Cerf, Gabriel Huguenin, Gerome Quention, Hend Madiot, Isabelle Calmont, Isabelle Gaudin, Isabelle Pacaud, Issraa Osman, Jean-Loup Devassine, Jeremy Tobia, Justine Rousseaux, Leslie Reyrolle, Lorrie Lafuente, Lydie Antoine, Lyna Gouichiche, Malek Ait Djoudi, Manon Pelkowski, Marcellin Bellonet, Marianne Maillet, Marie Diesel, Marie Granjon, Marie Laure Stupien, Marie-Jose Aroulanda, Marie-Jose Ngo Um Tegue, Marielle Simon, Marine Bou, Marine Douillet, Marion Ghidi, Maxence Passageon, Melanie Grubner, Morgane Heberle, Murielle Mejane, Pietro Todessayi, Rachida Yatimi, Robin Pinilla, Sabine Camara, Sandra Marchionni, Sandrine Vautrat, Shervin Fonooni, Solange Trehoux, Sophie Tallon, Stephanie Flasquin, Stephanie Lejeune, Stuart Byrom, Sylvie Grandmange, Traikia Chaima, Valerie Opderbeck, Veronique Pelonde Perimee, Victoria Mouanga, Volanantenaina Fanomezantsoa, Dr Pochesci Alessia, Nassera Aouali, Alexandra Keusching, Brigitta Elsasser, Dr David Grimaldi, Dr Filipa Luz, Martha Colban, Ruben E. Keane, Radka Troníčková, Katerina Nerušilová, Agnieszka Wiesner, Irene Garcia.

FUNDING INFORMATION

The trial is funded by grants from the European Commission (EU-Response, Grant 101015736), Programme Hospitalier de Recherche Clinique (PHRC-20-0351) (Ministry of Health), from the DIM One Health Île-de-France (R20117HD) and from REACTing, a French multidisciplinary collaborative network working on emerging infectious diseases.

CONFLICT OF INTEREST STATEMENT

No author declared a conflict of interest in relation to the submitted work. C. Delmas reports receipt of drugs from Gilead Sciences, Inc., Sanofi, Merck Group, AbbVie for Inserm, outside the submitted work. J. Saillard reports receipt of drugs from Gilead Sciences, Inc., Sanofi, Merck Group, AbbVie for Inserm, outside the submitted work. M. Dumousseaux reports receipt of drugs from Gilead Sciences, Inc., Sanofi, Merck Group, AbbVie for Inserm, outside the submitted work. S. Le Mestre reports receipt of drugs from Gilead Sciences, Inc., Sanofi, Merck Group, AbbVie for Inserm, outside the submitted work. A. Ferrane reports receipt of drugs from Gilead Sciences, Inc., Sanofi, Merck Group, AbbVie for Inserm, outside the submitted work. C. Burdet reports consulting fees from Mylan, Da Volterra, outside the submitted work; participation on a Data Safety Monitoring Board or Advisory Board for 4Living Biotech, outside the submitted work. F. Ader reports grants from The French Health Ministry, European Community, outside the submitted work. M. Hites reports grants from the Belgian Centre for Knowledge (KCE), the Fonds Erasme-COVID-Université Libre de Bruxelles, outside the submitted work; grants from the EU-Horizon program, outside the submitted work; support for ECCMID congress 2021 from Pfizer, outside the submitted work; working as co-leader of the Belgian Guidelines on Therapeutics for COVID-19, outside the submitted work; acting as a president for the Belgian Society of Clinical Microbiology and Infectious Diseases, outside the submitted work. J. Poissy reports payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from Gilead, outside the submitted work; patents planned, issued or pending from Gilead, outside the submitted work. C. Andrejak reports grants from the FEDER, outside the submitted work; support for ERS Conference from Homeperf, outside the submitted work; acting as member of Covid group for the French High council of Public Health and acting as responsible of the Scientific council for the French Society of Respiratory Diseases, outside the submitted work. J.A. Paiva reports consulting fees from MSD, Jansen, Pfizer, Astra-Zeneca, outside the submitted work; payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from Gilead, Cepheid, AOP Orphan Pharmaceuticals, outside the submitted work. R. Greil reports consulting fees from Celgene, Novartis, Roche, BMS, Takeda, Abbvie, MSD, Janssen, Astra-Zeneca, Merck, Gilead, Daiichi Sankyo, Sanofi, outside the submitted work; payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from Celgene, Roche, BMS, Takeda, Abbvie, Sandoz, MSD, Novartis, Astra-Zeneca, Merck, Gilead, Daiichi Sankyo, Sanofi, Amgen, outside the submitted work; support for attending meeting and/or travel from Celgene, Roche, BMS, Abbvie, MSD, Novartis, Astra-Zeneca, Gilead, Daiichi Sankyo, Amgen, Janssen, outside the submitted work; participation on a Data Safety Monitoring Board or Advisory Board for Celgene, Novartis, Roche, BMS, Takeda, Abbvie, Astra-Zeneca, MSD, Janssen, Merck, Gilead, Daiichi Sankyo, Sanofi, outside the submitted work; research funding from Celgene, Roche, Merck, Takeda, Astra-Zeneca, Novartis, Amgen, BMS, MSD, Sandoz, Abbvie, Gilead, Daiichi Sankyo, outside the submitted work. D. Costagliola reports grants or contracts from any entity from Janssen for Inserm, outside the submitted work; lectures fees from Janssen and Gilead, outside the submitted work; payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from Gilead, outside the submitted work. All other authors declare no competing interests.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. The data supporting this publication will be made available to researchers whose purpose of use is approved by the DisCoVeRy Steering Committee. The request shall be addressed directly to the corresponding author ([email protected] or [email protected]) or to the sponsor's representative ([email protected]) to obtain a data access form.