Accelerated Approval: Experts weigh in on the role of RWE in confirmatory trials

Regulatory NewsRegulatory News | 08 August 2022 |  By 

As Congress considers how to reform the US Food and Drug Administration’s (FDA) controversial pathway for providing accelerated approval to drugs that treat serious or life-threatening diseases, there may be an expanded role for real-world evidence (RWE).
Both the House and Senate versions of the FDA’s user fee reauthorization legislation – pending in the Senate – have a section with changes to the accelerated approval program, including language that requires drug sponsors to conduct an appropriate postapproval study, “which may be augmented or supported by real world evidence.”
This language has drawn criticism from some health services researchers who say that real-world data (RWD) sources are not robust enough to provide postapproval confirmatory evidence of clinical benefit in the accelerated approval setting. At the same time, companies working in the RWE area, say that these sources, when used appropriately, could help fill evidence gaps when confirmatory clinical trials are not feasible.
“Our concern really is that Congress has opened the door for the possibility of real-world evidence to be used in place of clinical trials for confirming clinical benefit when there’s not really sufficient evidence that the sources that we have right now – that were not designed for that purpose – could be used to do so,” said Reshma Ramachandran, MD, professor of medicine at Yale University and one of the directors of Yale’s Collaboration for Research Integrity and Transparency. “Based on current evidence, this would be like letting the horse out of the barn a little too quickly.”
Over the last year, FDA has published four draft guidances looking at the use of RWD & RWE in regulatory decision making, tackling the use of RWD from sources such as electronic health records, medical claims, and patient registries. In the future, the agency plans to publish additional guidance documents examining the use of RWD in external control arms, as well as study designs for RWD.
The agency defines RWD as data related to a patient’s health status or routine health care, collected from a variety of sources including EHRs, claims, registries, patient-generated data, and mobile devices. RWE, on the other hand, is the clinical evidence on the use, and potential benefits and risks, of a medical product that are derived from analysis of the RWD.
The FDA routinely uses RWD and RWE in the postmarket setting to examine safety and adverse events. More recently, in 2021, the FDA used RWE, in conjunction with historical clinical trial data, as the basis of approving a supplemental indication for Prograf (tacrolimus) for preventing organ rejection in adult and pediatric patients receiving lung transplantation.
Emulating trial data
The use of RWE as a complement to clinical trial data makes sense, Ramachandran said, but it is not ready to be used as a substitute, especially not in the setting of confirming clinical benefit in accelerated approval. The concern is that if RWE is used to replace a confirmatory trial, it would have the potential to overstate the benefit of the treatment due to confounding factors, she said.
She pointed to research, published in JAMA Network Open in 2021, that examined the feasibility of using RWD to emulate the postapproval confirmatory trials for new drugs that received accelerated approval between 2009 and 2018. Across the 50 postapproval confirmatory trials during that time period, the researchers concluded that none could have been emulated using available claims and/or structured EHR data (RELATED: Study: RWD not ready for postapproval prime time, Regulatory Focus 11 November 2021).
Joseph S. Ross, MD, a professor of medicine and public health at Yale University and one of the authors on the JAMA Network Open study, said using structured data to confirm clinical benefit is especially challenging because some common trial endpoints – from symptom measures to imaging changes – are not routinely collected in clinical practice as structured data. Additionally, the complex decision making around treatment selection in the real-world clinical setting is vastly different from the rigid inclusion/exclusion criteria seen in clinical trials.
“I continue to worry that people oversell the progress of real-world evidence without realizing the limitations,” said Ross, who is also a member of the Center for Outcomes Research and Evaluation at Yale.
Not a one-size-fits-all solution
But Irene Nunes, vice president and head of regulatory affairs and policy at Flatiron Health, said that RWE should be considered on a case-by-case basis, depending on if the RWD are relevant and reliable and if the study design is appropriate, with adequate methods and analytical tools that can be used to generate evidence to answer clinical questions.
“It’s not black and white,” Nunes said. “Considering the use of real-world data includes clearly defining what evidence is needed and what type of evidence can be generated.”
One circumstance where RWE can come into play is when there is not an opportunity to conduct a post-approval trial. In the post-approval setting, in particular, it can be difficult to conduct a confirmatory trial because patients may be unwilling to be randomized to a control arm treatment regimen, such as chemotherapy, when there is another approved option. “The feasibility of conducting trials and representativeness of patients from the real-world setting in a trial are important elements when considering post-approval studies,” she said.
It is also critical to think about integration of RWD sources, Nunes said. For example, a single source of RWD may not provide the total picture of a patient’s medical care but linking medical claims and EHR information, for instance, may reveal a more complete picture of the patient’s journey through the health care system. “It’s not just about the data, it’s also about understanding the contribution of each source of real-world data, the methodologies that can stitch everything together, and then analyzing the data holistically,” she said.
RWE in the future
In the future, Ross said he could see an important role for RWD as part of a randomized clinical trial. For example, it might be possible to use digital technologies to pull together individual medical records and then randomize patients, passively embedding the trial within the health care infrastructure. Another long-term strategy to improve RWD sources would be to make refinements in what information health systems collect. Though it would not be possible to do this on a national scale, the FDA could work with select health systems to collect information that would shed light on clinical benefit, he said.
Ramachandran suggested that future studies using RWD could either be integrated into the registry or have their own study registry, allowing FDA officials and the public to see the study design, including the endpoints selected and inform them early in the process.


© 2023 Regulatory Affairs Professionals Society.

Discover more of what matters to you