Panel at COVID-19 & Cancer Consortium retreat discusses goals for real-world data
During the COVID-19 & Cancer Consortium Scientific Retreat, former FDA Commissioner Andrew C. von Eschenbach, MD, asked representatives of data registries to reflect on the challenges and lessons of the past year.
He also asked them to look to the future.
“We want to spend most of our time looking at what is ahead, at the critical issues that we must address in 2021,” von Eschenbach said while moderating the panel discussion, titled “COVID-19 and Cancer Registries: Lessons Learned and Paths Forward.”
Von Eschenbach, who served as NCI’s 12th director prior to leading the FDA from 2006 to 2009, discussed the accomplishments of the COVID-19 and Cancer Consortium, which consists of 127 cancer centers that contribute data on their patients with solid tumors or hematologic malignancies who have been diagnosed with COVID-19. Using real-world evidence, the consortium has compiled and produced a wide array of scientific publications, abstracts, presentations and articles in the lay press.
“There’s an even bigger story here, though. On the one hand, we may view the [patient with cancer] as more vulnerable to the devastating impacts of COVID-19 infection but, ironically, it is the cancer enterprise that is most equipped and effective at discovering and unraveling the biologic mysteries of this pandemic,” von Eschenbach said. “We’re celebrating the 50th anniversary of the National Cancer Act, which empowered the NCI to create an equally impressive infrastructure of cancer centers, biorepositories, data registries and a host of resources that enable the rapid and extensive acquisition, aggregation and analysis of data. The COVID-19 and Cancer Consortium utilizes these resources to understand the impact of COVID-19 in the setting of a cancer diagnosis.”
‘Power of the grassroots movement’
Von Eschenbach began by asking each member of the panel to share a “topline sound-bite” that could summarize their insights regarding the COVID-19 pandemic over the past year.
“I would say the most important thing we learned [during the pandemic] is the power of the grassroots movement,” said Gilberto de Lima Lopes Jr., MD, MBA, FAMS, FASCO, interim chief of the division of medical oncology at Sylvester Comprehensive Cancer Center at University of Miami. “Starting with the medical colleges, we’ve learned about what we can do for our patients and how we can understand what the implications of COVID-19 are for those patients.”
Although she did not deal with COVID-19 in a clinical practice setting, Jennifer Whisenant, PhD, research assistant professor at Vanderbilt University Medical Center, said she learned a great deal through her work on the practice-based TERAVOLT registry, which emerged from a global effort to increase understanding of COVID-19’s impact on patients with thoracic cancers.
“Working with clinicians on this project, especially with TERAVOLT focusing on lung cancer malignancies, I saw that COVID-19 symptoms are very similar,” Whisenant said. “Distinguishing between them has been difficult, and I think TERAVOLT brought that to light, as well as the fact that these types of [patients with cancer] are at increased risk [for] mortality.”
Over time, however, the researchers said their registries have shown reductions in COVID-19-associated mortality among patients with cancer.
“The thing that really sticks out for me is the major changes we’ve seen in the prognosis of patients with cancer who have gotten COVID-19,” Elizabeth Garrett-Mayer, PhD, division director of biostatistics and research data governance at ASCO, said of her experience working on the ASCO COVID-19 registry. “Early in the pandemic, the risks seemed to be very great for various severe outcomes, including death. It’s shocking the way the cancer community was able to respond, share information and learn from the early cases. Now the prognosis has improved dramatically, from what we’ve seen in our data.”
Optimizing real-world data
In the second part of the discussion, panelists spoke about the future of using real-world registry data and evaluated the limits of its utility.
Garrett-Mayer referred to the statistical phenomenon of “informative missingness,” in which a piece of data is not accessible through electronic health records because it is part of the causal pathway of the outcome.
“A lot of the time, the reason we can’t get these data is because they’re related to the outcome we care about,” she said. “So, as a biostatistician, it’s sort of my job to be a ‘downer’ in these things. To be honest, I don’t know that real-world data are really going to get there, because until we get to a point that the way they are collected is more rigorous and less biased, we’ll never get them to be as good as a randomized trial.”
Whisenant agreed, adding that it might be useful to think of real-world data as the preliminary information needed to design a clinical trial. She cited a database in the U.K. that was similar to the TERAVOLT registry.
“It looked at longitudinal laboratory values, and at different time points they found certain lab values to be associated with COVID-19 severity and mortality,” she said. “Now, they’re planning a randomized clinical trial specifically looking at those lab values. If they hadn’t collected these real-world data, they might not have known what lab values to focus on. So, I think we could look at this as our preclinical data needed for a more targeted assessment in a rigorous clinical trial.”
Solange Peters, MD, PhD, head of medical oncology at Centre Hospitalier Universitaire Vaudois in Lausanne, Switzerland, and president of ESMO, discussed what she envisions as the possible future of real-world data registries and consortiums.
“In the short term, I think we could try to homogenize the definition we have of real-world data. What is the minimal set of criteria we consider with our data to build the hypothesis for a future randomized trial? That could be one usage,” she said. “The long-term idea is, to me, artificial intelligence; the implementation of a homogenized system where the data collection and mining will be inherent to the system you use, because we all could collect the same data at the same time formed with the definitions. I think that is probably the way of the future — to immediately have the data you need across centers.”
Peters emphasized that in the face of a public health crisis, the willingness to collaborate as researchers and research centers will always be paramount.
“I think the collaborative effort has been amazing,” she said. “We have been learning that when a major health problem occurs, collaboration goes above everything.”
Collapse
COVID-19 and Cancer Registries: Lessons Learned and Paths Forward. Presented at: COVID-19 & Cancer Consortium Scientific Retreat (virtual meeting). March 26, 2021.
Read more about