The ongoing coronavirus disease 2019 (COVID-19) pandemic has magnified the value of real-world data (RWD) in oncology. The FDA is presently spearheading several initiatives aimed at refining the role of RWD in cancer care to guide clinical trial development, procure answers to pressing clinical questions, and support regulatory decisions for in vitro diagnostics, according to presentations shared during a major symposium at the AACR Virtual Annual Meeting II.
“We now find ourselves in a grand experiment,” said Paul G. Kluetz, MD, deputy director of the Oncology Center of Excellence at the FDA. “By necessity, COVID-19 has pushed the cancer clinical trial community out of its comfort zone to rapidly deploy remote assessments and digital health technology, [prompting] unprecedented collaboration to explore the use of real-world data. We cannot lose this opportunity to learn and advance our knowledge of the strengths and limitations of decentralized trials and real-world data to move our cancer research enterprise to a more efficient and more patient-friendly model.”
Long before COVID-19, calls to learn from RWD and “decentralize” standard components of cancer clinical trials for favor of more “patient-friendly” alternatives such as digital health technology to facilitate remote assessments were heard in oncology, Kluetz explained. However, only now has the oncology field so broadly begun to adopt different operational tactics and draw upon RWD to inform management approaches for both COVID-19 and cancer. “We’ve witnessed a great interest in taking advantage of real-world data to rapidly expand our knowledge of the natural history of disease and treatment,” Kluetz said.
With widespread interest in RWD in place, many clinical questions now require answers, including several that are particularly germane: for example, can we leverage data collected from cancer trials conducted during COVID-19 to understand the effects of remote assessments and decentralized trial procedures on safety and efficacy outcomes?; and how can we use RWD to characterize the natural history of COVID-19 and its effects on patients with cancer?1
The FDA is currently working with data holders, data analysts, and research institutions around the world to develop initiatives that not only address these inquiries but also probe beyond them, to clarify when and how specifically RWD can be most optimally used to improve current methodologies in oncology and inform new ones in the age of COVID-19.1
The COVID-19 Evidence Accelerator
The COVID-19 Evidence Accelerator, a collaborative, RWD-driven cooperative, unites partners at the FDA with fellow federal health organizations such as the National Institutes of Health and Patient-Centered Outcomes Research Institute, with the accelerator’s managers, the Reagan-Udell Foundation for the FDA/Friends of Cancer Research.1
Participants in the COVID-19 Evidence Accelerator developed a common set of core data elements and an initial set of priority questions regarding the therapeutic interventions, treatment settings, and associated outcomes that would guide early COVID-19 focused data aggregation efforts. The general epidemiology of COVID-19, the predictors of patients who would be at risk of developing severe COVID-19, and the patterns of general outcomes for people with COVID-19, including death and time to disease resolution, represent just 3 of the urgent clinical inquiries that multiple organizations and teams can simultaneously address. The questions have been crowdsourced, curated, and prioritized, so teams across the country can identify the highest priority items, according to Amy P. Abernethy, MD, PhD, principal deputy commissioner and acting chief information officer of the FDA.1
Beyond this general list of COVID-19 specific inquiries, questions regarding the management of COVID-19 in patients with cancer have also been crowdsourced, in an effort to define how RWD efforts can help inform the Oncology Center of Excellence’s COVID-19 related research endeavors. Questions included the safety and efficacy of immunotherapy in patients with COVID-19 and lung, melanoma, bladder, and other malignancies, as well as the rates and severity of COVID-19 mortality in patients with cancer.1
“Key to the accelerator is [our] focus on data holders, [such as] technology companies where large datasets are already well-curated and analytic teams are ready to analyze datasets,” Abernethy said. “We ask them to now start working on this prioritized list of questions in order to have pragmatic, real-time work directed at the COVID-19 pandemic.”
To structure researchers’ investigative efforts, the Reagan-Udell Foundation for the FDA/Friends of Cancer Research has defined common data elements and translation tables between data models that researchers are encouraged to use. Common protocols for repeated analyses of priority research questions; meetings and forums for rapid cycle feedback and learning from the insights of participating data partners; and individual accelerator subcommunities dedicated to specific topics, such as therapeutics, diagnostics, and vaccines, have also been established.1
Data holders and aggregators participate in 1 or more focused work streams, which currently include Therapeutics, Diagnostics, and Vaccines. Individual research questions are generated and investigated within each work stream. Evidence Accelerator collaborators meet weekly to discuss findings from their respective weekly parallel analyses. In contrast with the Therapeutics, Diagnostics, and Vaccines work streams, the oncology work group, established in June 2020, does not limit or divide its investigative focus to just 1 of these areas, but rather evaluates each of them.1
Following the COVID-19 Evidence Accelerator’s launch with its first lab meeting on April 16, 2020, the pool of data holders and number of weekly lab meetings has increased, with a Diagnostics Evidence Accelerator established on May 28, 2020. The overlapping efforts of both Evidence Accelerators and the oncology workgroup within the COVID-19 Evidence Accelerator reflect the FDA’s commitment to “creating opportunity out of crisis,” according to Kluetz. Additionally, the efforts support the widely recognized importance of moving expeditiously to find answers to pressing clinical questions during this unprecedented time in medicine, Abernethy said.
“The work of the Evidence Accelerator sits in this larger global data community, where partners across the community understand 1 critical thing: we need to address the questions of COVID-19 with urgency, and we need to understand the natural history of disease and the effectiveness of potential treatments, and use the information to plan clinical trials,” Abernethy concluded.1
In addition to shaping future oncology clinical trials, RWD generated and collected during the COVID-19 pandemic and thereafter can also be used to evaluate investigational medical devices and extend marketing approvals where necessary. However, as with all clinical efforts that involve RWD, the data that supports these decisions must be fit for purpose, meaning that the data are “complete, consistent, accurate, and contain all critical data elements needed to evaluate a medical device and its claims,” said Wendy Rubenstein, MD, PhD.1
“A key problem in leveraging real-world diagnostics is the lack of information on the specific test used and the manufacturer,” added Rubenstein, director of Personalized Medicine at the FDA. The Systemic Harmonization and Interoperability Enhancement for Lab Data (SHIELD) project is the FDA’s response to this widespread problem.
“SHIELD is a cross-center and multi stakeholder initiative that supports efforts to harness nontraditional in vitro diagnostic data sources while reducing burdens to the health care ecosystem,” Rubenstein explained. At present, the SHIELD network encompasses the FDA, Centers for Disease Control and Prevention, National Institutes of Health, Office of the National Coordinator of Health Information Technology, Centers for Medicare and Medicaid Services, US Department of Veterans Affairs, in vitro diagnostic manufacturers, electronic health record (EHR) vendors, laboratories, College of American Pathologists, standards developers, Pew Charitable Trusts, National Evaluation System for health care Technology, and academia.2
Crucial to SHIELD is the development of code mapping manuals to allow for the consistent mapping of the Logical Observations Identifiers Names and Codes (LOINC) that correspond with a given in vitro diagnostic. LOINC is the most widely used coding system for clinical laboratories and EHRs. For in vitro diagnostics, LOINC codes are used to identify the type of diagnostic used to test a clinical measure.2 Although the code mapping process was instituted to streamline electronic searches for in vitro diagnostics and corresponding test results, across institutions, codes are often “inconsistently applied and this ambiguity erodes the usability of the data,” Rubenstein said. Non-uniform entry of the codes is not limited to institutions: in vitro diagnostic manufacturers and laboratories also frequently assign different and inaccurate codes for a given device.2 SHIELD will “provide a process for consistent definition and use of codes,” thereby supporting interoperability, Rubenstein added.
Leveraging RWD for public health initiatives involving both COVID-19 and cancer is an overarching goal of SHIELD, which aims to establish interoperability to standardize data sharing among in vitro diagnostic data sources and support regulatory decisions involving diagnostics. Despite the widely recognized reality that interoperability is necessary to facilitate fluid and complete data sharing among and between these data sources to enable the use RWD in this manner, full-scale interoperability has not been achieved, and has consequently impeded this specific application of RWD.1,2
However, the clinical shifts catalyzed by the COVID-19 pandemic have further illuminated the value of RWD in the oncology space and in medicine at large, and are providing the impetus for change, vis-á-vis SHIELD: “The urgency of the COVID-19 pandemic is supplying the activation energy we need, and the societal reaction is collaboration across all sectors to build the infrastructure to finally harness RWD. Once interoperability and data exchange are accomplished for COVID-19, success will be within reach to leverage RWD for public health more widely,” Rubenstein said.