March 23, 2021
Republished with permission from BioCentury. This article originally appeared on March 19, 2021: https://www.biocentury.com/article/635095/pandemic-response-ushers-in-new-era-of-biopharma-data-sharing-says-covid-r-d-alliance
By Liz Roberts, UCB; Jared Christensen, Pfizer; Jules Desmond, Amgen; Annetta Beauregard, Janssen; Peter Mesenbrink, Novartis; Aaron Mann, Genentech And Roche; Matthew Bryant, Gilead; Anne Heatherington, Takeda
Co-authored by Anne Heatherington, Senior Vice President and Head of Data Sciences Institute at Takeda
Years before the novel coronavirus began cutting its lethal course across nearly every continent on earth, scientists and senior executives at leading biopharmaceutical companies were discussing ways of accelerating the design, efficiency and pace of drug development to deliver innovative medicines to patients. What the pandemic has broadened awareness of is the need to encourage and enable greater sharing of both summary-level and patient-level clinical trial data.
Every year, biopharmaceutical companies collect data of incomparable quality at nearly unimaginable scale. The data are amassed with rigor and depth, using detailed, frequent assessments and tracking endpoints of interest, as well as subject measures that are not available in real-world environments. However, once the research is complete, the drug developed or the study stopped, the data often do not live beyond to realize their full value.
At least, that was the case until recent years. COVID-19 reminded us that accelerating the sharing of trial data offers continuing value to scientists searching for treatments for this and future pandemics, as well as other diseases with great medical need.
Multiple initiatives by industry, academia and government over the past 15 years have reconsidered the R&D model and the ways that shared data could enhance and expedite the process. The challenge has been in how to increase transparency and data access while protecting intellectual property and patient privacy.
Considerable progress has been made in the last year. Still, more is needed with an expansion of data sharing across the pharmaceutical industry. In the future, we can envision data sharing occurring simultaneously with the release of data as part of a presentation or publication.
BioCentury & Getty Images
Past initiatives offered a solid foundation to build on when the true magnitude of the COVID-19 pandemic became apparent last spring. Soaring infection rates and alarming death tolls as the pandemic spread around the world made it imperative that biopharmaceutical companies shift the research and development paradigm. To respond to the greatest health challenge in a generation, organizations were ready to tackle barriers, use all available tools and technologies, and collaborate with unprecedented focus, energy and speed to address the urgent need. The COVID R&D Alliancewas born at this critical juncture.
The Alliance comprises 21 leading biopharmaceutical and healthcare organizations. We came together, reiterated the benefits of responsible data sharing, and made a joint commitment to share more and share faster. We focused on behaviors and technologies to enable sharing, combined our efforts with TransCelerate Biopharma Inc., and with various philanthropic organizations including the Bill and Melinda Gates Foundation, and mapped out a plan.
Alliance members committed to share data and information that could speed up the study of vaccines, immunomodulators, antibodies, anti-viral therapies, and other treatments in late-stage development or marketed for other conditions that offered potential benefits to COVID-19 patients. For therapeutics, we aspired to share both summary-level and patient-level data at a faster pace than the industry had ever seen before.
Achieving such collaboration meant standing on the shoulders of many organizations that came before. In 2005, for example, the International Committee of Medical Journal Editors (ICMJE) boosted the importance of data sharing when it required that authors who sought to publish in their journals record all studies in public registries. The purpose was simple: prevent selective reporting of outcomes and duplication of efforts. Since then, more than 350,000 studieshave been registered on ClinicalTrials.gov. More recently, ICMJE member journals have started asking that data-sharing plans be disclosed as part of each trial's registrationand that data sharing statements be included in submitted publications.
Researchers are now required to disclose high-level summary information from most clinical trials, a transparency effort that was championed by organizations including the PhRMA/EFPIA and IFPMA, as well as prominent groups such as the WHO and the National Academy of Medicine. One of the earliest examples of such disclosure was through ClinicalTrials.gov. These early steps enabled greater scrutiny of trial design and statistical methodology and furthered novel statistical analyses.
Initiatives that have encouraged expansion of data sharing to patient-level data include The Yale University Open Data Access (YODA) Project, the Clinical Data Study Request (CDSR) consortium, Vivli and TransCelerate. Many Alliance members already were among the 20 members of the nine-year-old TransCelerate, which collaborates across the global biopharmaceutical R&D community to drive the efficient, effective, and high-quality delivery of new medicines. TransCelerate’s ongoing initiatives aim to further data sharing across research and development.
Building on this history, COVID R&D Alliance members understood which obstacles needed to be removed for greater cross-industry collaboration and faster response to the pandemic. The standard timeline for sharing clinical trial data was at odds with the urgency presented by a global pandemic. The Alliance envisaged a different world of sharing summary-level and patient-level data more rapidly, within weeks or months of trial completion, while also respecting data privacy and informed consent considerations.
The Alliance began by analyzing the best and fastest approach to sharing both summary-level and patient-level data. Summary-level data from a clinical trial readout enables direct and indirect comparison of different treatment options, allows estimation of efficacy and safety at a population level, and assists prescribers. Patient-level data describes individual patient behavior rather than overall study results and is of the greatest interest to those planning future studies because it allows granular analysis of the course of disease and correlation to specific comorbidities, demographics, concomitant medication and more.
Patient-level data also allows pooling across studies to answer research questions. For example, what is the effect of a medication on men aged 65-75 with asthma who also have elevated cholesterol? This might be impossible to determine from a single study, but patient-level data from multiple studies allows researchers to perform pooled and meta-analyses across comparable patient populations. The potential of such insights when applied to the heterogenous effects experienced by patients with COVID-19 would be powerful. While recognizing the value of patient-level data in this critical endeavor, Alliance members remained acutely aware of the need to ensure appropriate controls and legal commitments to protect patient privacy.
To further the sharing of summary-level data, the Alliance joined with a group sponsored by The Bill and Melinda Gates Foundation. For patient-level data sharing, the Alliance endorsed TransCelerate’s modular DataCelerate platform for patient-level clinical trial data that had recently added a new module dedicated to COVID-19 data sharing with a broad audience of qualified researchers.
With the Gates Foundation group, the focus was on enhancing an active treatment and control arm summary database. Work included developing a standardized way of sharing summary-level data, as well as a framework for rapid access to a summary-level data dictionary, allowing cross-study comparisons. Many research questions require access to data from multiple sources. This initiative enables researchers to discover metadata, run a data access request, and work with the data in a secure, audited workspace. Contributors decide how their data will be made available to researchers. Expert groups help ensure that statistical, scientific and ethical considerations meet international norms.
The goal was to ease the path for researchers, particularly those with less robust resources, to access and analyze summary-level data that may generate new insights. Qualified researchers working within the platform can perform a defined range of analyses. Such controlled access reduces liability for sponsors while allowing a broad group of researchers in Africa, Asia and elsewhere to probe for answers to questions related to their own COVID-19 research.
Responding to challenges associated with studying COVID-19, experts across life science companies recognized that collaborating with TransCelerate’s existing COVID-19 module would be the best way to enable rapid sharing of patient-level data. In May, a secure data sharing module was launched, enabling the input of patient-level data from COVID-19 studies. Within weeks, data sharing agreements were signed by 12 organizations with seven more joining over the summer.
Patient-level data sharing agreements allowed for participation by qualifying non-TransCelerate and non-Alliance members, at no cost, and sharing of both control and investigational product arms.
Data covenants and other mechanisms were put in place to address complex IP, competitive and regulatory hurdles. Sophisticated access controls were designed to create a virtual walled garden of trust where qualified researchers benefit from shared data while respecting commitments made to their fellow researchers and the public at large.
Throughout the past months, we have been encouraged by the willingness and enthusiasm of industry amid a global health crisis to allow greater data access and enhanced sharing of clinical trial data as soon as it is available.
We are in the early stages of understanding the full potential of how shared data enhances efficiency, reduces unnecessary costs, and accelerates delivery of treatments to patients. The pathway has been paved for the industry during the COVID-19 pandemic.
The vision is a landscape where shared control arm data can radically reduce the number of prospective control subjects required during clinical development. Organizations will use the data to model diseases more accurately, improve the search for biomarkers, and gain insights into how geography and demographics affect efficacy and safety. Leveraging shared data can help researchers interpret adverse events and other safety signals. It can assist researchers in better understanding patient characteristics and treatment patterns to help optimize inclusion and exclusion criteria in trials. These benefits become more vital when researchers are confronted with a novel disease in which time wasted equates to lives lost.
As an industry, the life sciences are on a journey from a model of increasingly costly R&D executed in closely guarded silos to a more collaborative system in which companies share data for the benefit of society and compete on the strength of their medicines. The COVID-19 pandemic experience may yet produce something positive if the pandemic motivates companies to make the practice of faster data sharing part of routine drug development for all diseases.
Data from COVID-19 trials are in early stages of release. Now is the time to see our established principles put into action with data from our clinical trials shared months before the usual timelines. This is especially important as the pandemic rages and new variants emerge.
Through the Alliance, we have shown that a common vision can be created across organizations and platforms. We have developed infrastructure to support this vision and demonstrated that data can be shared more quickly. We must now seek to challenge ourselves beyond COVID-19 to continue increasing data transparency and enabling access across scientific research in other diseases for the long-term betterment of all.
Liz Roberts is data policy and privacy lead at UCB S.A. Jared Christensen is VP of early clinical development & biostatistics at Pfizer Inc. Jules Desmond is strategic development director at Amgen Inc. Annetta Beauregard is head of global regulatory policy at the Janssen unit of Johnson & Johnson. Peter Mesenbrink is executive director statistician at Novartis AG. Aaron Mann is global program lead, industry collaborations data sharing initiatives at Genentech Inc. and Roche. Matthew Bryant is head of global technology & innovation for Gilead Sciences Inc. Anne Heatherington is SVP, head of data sciences institute at Takeda Pharmaceutical Co. Ltd.
Other companies in the COVID R&D Alliance are AbbVie Inc., Alnylam Pharmaceuticals Inc., Arbutus Biopharma Corp., AstraZeneca plc, Bayer AG, Bristol Myers Squibb Co., Eisai Co. Ltd., Evotec AG, GlaxoSmithKline plc, Eli Lilly and Co., Sanofi, Theravance Biopharma Inc., Versant Ventures, Vir Biotechnology Inc. and 5AM Ventures.