In an effort to bring new drugs to cancer patients faster, the Friends of Cancer Research (FOCR) and Alexandria Summit collaborated in a meeting in Washington, D.C., to consider how real-world evidence can be used along with data collected from traditional randomized, controlled clinical trials in the approval of new drugs. The organizations presented a draft document called Blueprint for Breakthrough: Exploring the Utility of Real World Evidence.
“These are real-world patients who are suffering,” said FOCR Chair and Founder Ellen V. Sigal, PhD. Real-world evidence, as defined by FOCR, is evidence derived from the use, benefits, and risks of medicines that fall outside the bounds of the classic clinical trial settings, including use of data routinely collected in the daily practice of medicine, and thus reflects the heterogeneous patients seen in real-world practice settings. Real-world evidence collected from practicing physicians can contribute to the continuous learning environment envisioned in reports issued by the Institute of Medicine (recently renamed the Health and Medicine Division of the National Academies of Sciences, Engineering, and Medicine).
The proposal states real-world evidence may be especially useful in certain situations, such as for a therapy designated a breakthrough therapy. The collection and sharing of patient data outside of clinical trials has been made possible in large part by the growing use of electronic health records (EHRs). In 2012, the FDA created the breakthrough designation as a way to expedite the approval of new drugs for serious medical conditions with a large unmet medical need.
Although the FDA has expanded trial populations, they are still not representative of the population at large, said FDA Commissioner Robert Califf, MD, MACC, in a keynote address. Therefore, he said, real-world data can be useful in supplementing evidence from clinical trials. Califf noted that he grew up in an era when there were two universes: one for clinical trials and one for clinical practice. “Now we need to bring these two worlds together,” he said. Bringing the two together is especially important in an era when the rapid acceleration of scientific knowledge is being translated into better outcomes for patients, emphasized Califf.
An effort to use novel pipelines of data to merge the clinical trial and clinical practice environments is already underway at FDA, said Sean Khozin, MD, MPH, Senior Medical Officer in the FDA’s Office of Hematology and Oncology Products. Khozin, a thoracic oncologist, is the co-founder and project leader for Information Exchange and Data Transformation (INFORMED), an FDA oncology initiative to advance regulatory science by building an organizational infrastructure for data from EHRs and biometric sensor technologies, along with aggregated clinical trial datasets.
“Essentially what we’re trying to do is reduce uncertainty over time,” said Khozin, noting that using novel pipelines of high-quality data in regulatory decision-making can reduce that uncertainty. While INFORMED is “a work in progress,” said Khozin, in the future, real-world data and other novel evidence may be a part of the data submission package of a sponsor seeking approval for a new therapy.
“All the stories embedded in the electronic health record are individual stories,” said Amy Abernethy, MD, PhD, Chief Medical Officer and Senior Vice President for Oncology at Flatiron Health, whose goal is to organize the world’s cancer data and make it actionable for providers, patients, and researchers. With the morass of data coming into the EHR, “we need a set of rules that helps us structure all this information” so it can be used to improve patient care, she said. Abernethy stressed this framework needs to be replicable and portable across multiple settings.
The American Society of Clinical Oncology launched its CancerLinQ continuous learning system, which is designed to improve patients’ outcomes and quality of life based on EHR data. But, notes the draft document, “it remains unclear whether such data could be suitable for regulatory purposes.”
Evidence in Practice Settings
The blueprint draft document released notes that real-world data collected within practice settings have the potential to decrease drug costs and increase drug development timeliness, define novel outcomes, minimize the number of patients exposed to a less efficacious therapy, and supplement post-marketing data collection. Specifically, it proposes that real-world evidence could be used in the following areas:
- expanding the safety profiles of a therapeutic product;
- identifying populations with enhanced benefit/risk for an already approved therapy to inform clinical practice;
- piloting studies to determine the potential correlation between feasible real-world measures (e.g., time to treatment switching) and more traditional clinical trial endpoints (e.g., time to disease progression);
- building evidence for a supplemental package to expand the indication for a therapeutic product; and
- supporting efficacy results observed in the clinical trial setting, particularly in areas of unmet medical need, when a new drug shows substantial clinical benefit.
Real-world studies that are able to support the preliminary magnitude of efficacy in a larger cohort may be sufficient to serve as post-market confirmation of clinical benefit.While randomized controlled trials remain the gold standard, the problem with the clinical trial process is that it takes too long, said Janet Woodcock, MD, Director of FDA’s Center for Drug Evaluation and Research (CDER). Woodcock said she can envision a registry of data collected from real-world evidence; clinical trials designed from real-world data collected in EHRs; an off-label use for an approved drug identified by real-world data; a new use for a drug based on mutations that cut across tumors, as identified by real-world data; and other expanded uses of real-world evidence.
Cancer Community Collaboration
To make the vision of FDA Commissioner Califf a reality—the merging of the clinical trial world with the clinical practice world—it will be necessary to work closely with community-based oncologists, noted Woodcock. “I’m a great believer that we have to enlist the cancer community oncologists in this enterprise,” she said.
When it comes to using real-world evidence for new drug approvals, “it really is going to require industry’s buy-in to do this,” said Richard Pazdur, MD, Director of FDA’s Office of Hematology and Oncology Products within CDER. He cautioned that trying to “shoehorn our concepts from clinical trials,” such as the use of progression-free survival as an endpoint, into the use of real-world data “may not be appropriate.
”On the other hand, what may be especially useful, said Pazdur, is using real-world data to focus on patients’ symptoms and to build health-related quality-of-life data into real-world evidence. Pazdur noted toxicity issues may not always surface in a clinical trial, but they could surface if real-world evidence is collected. As an example, he cited osteonecrosis of the jaw, a toxic side effect of the use of bisphosphonates in some patients that might have surfaced earlier if real-world EHR data had been collected and looked at systematically.
“We are still very tied to the clinical trial,” said Jeff Helterbrand, PhD, Senior Vice President and Global Head of Biometrics at Roche. But, he said, “I think oncology is getting to the point where we can’t always do a clinical trial.” That, he noted, is because the standard of care in oncology is changing very rapidly, and—agreeing with Woodcock—because “clinical trials take too long.” Helterbrand emphasized that “we need to open up the window beyond the clinical trial.”
Helterbrand praised the Targeted Agent and Profiling Utilization Registry (TAPUR) study launched in March 2016 by ASCO as an example of leveraging real-world evidence from clinical practice. TAPUR is a nonrandomized pragmatic trial whose goal is to collect data on the safety and efficacy of approved therapies in other disease settings. TAPUR, like other pragmatic trials, is a prospective intervention that leverages the existing clinical infrastructure to test therapies in everyday clinical settings to maximize their therapeutic applicability and generalizability.
On the whole, payers are “very supportive” of the idea of using real-world data as another pipeline in evaluating a new therapy, said Roy Beveridge, MD, a medical oncologist who is Chief Medical Officer and Senior Vice President at Humana. Beveridge noted that most of the cancer patients who are Humana beneficiaries are of Medicare age and are taking a number of medications for multiple conditions. Therefore, “we struggle trying to find out what is cost-effective and safe” for these patients, he said. Real-world evidence could help in making treatment decisions for this population.
Real-World Evidence in Regulatory Decision-Making
- What data quality elements need to be considered and should they differ by data source (e.g., data from EHRs)?
- How should quality be reported and presented for review?
- What details should be captured with respect to cohort selection when generating real-world evidence?
- What additional analyses need to be done to generate real-world evidence?
- What aspects of efficacy need to be captured with real-world evidence in addition to addressing safety concerns?
- How should changing data characteristics and needs for real-world data be accommodated over time?
- What specific adjustments (if any) need to be made to EHR recordings to allow data transfer to the FDA?
- What drug candidates in oncology could serve as potential case studies?
- What potential pilots in oncology could be used for clinical evidence generation to support regulatory decisions?