The pivotal role of real-world data in a pandemic
Drug development for COVID-19 will depend on real-world data and industry is stepping up to implement it
Real-world data is taking on a critical role in the age of COVID-19 drug development, where hasty authorizations are being made on limited datasets and natural history data is being collected in real time.
Teams across the industry have stepped up to adapt their data analysis tools to the crisis with the goal of applying previous learnings to benefit future patients. The challenge now will be scaling their systems to manage the unprecedented volume of patient data, while keeping pace with the evolving pandemic.
Before the pandemic, real-world data (RWD) had been gaining momentum as a tool to create a broader picture of drug safety and efficacy post-market, and to get drugs to market faster under accelerated and conditional approval pathways.
In the context of COVID-19, RWD can help guide clinical trial design and clinical practice, and assist in regulatory decisions.
RWD from COVID-19 patients have already begun to inform adjustments to clinical trial designs, endpoints and patient populations, and to informally guide clinical practice.
“What we’re learning from real-world data is feeding into not just clinical practice but how we’re designing clinical trials.”
However, industry has been slower to act on RWD that could guide regulatory decisions.
FDA has authorized two COVID-19 therapies under accelerated pathways that rely on RWD to keep patients safe and get them the most effective treatments, but both have raised questions around safety, efficacy and proper patient subgroups.
Evidence from randomized and observational clinical studies is mounting against the controversial drug hydroxychloroquine, authorized on anecdotal evidence alone, but it hasn’t been enough to drive FDA to rescind the authorization (see “Hydroxychloroquine Testing Hahn’s, FDA’s Independence”).
And although the antiviral remdesivir from Gilead Sciences Inc. (NASDAQ:GILD) was authorized on randomized controlled data, there isn’t enough trial data to determine which patients benefit from the drug, which is already in limited supply (see “Remdesivir’s Challenge”).
Those questions can be addressed by RWD, but with the growing demand comes a growing list of challenges. For COVID-19, there’s a volume of patient data that analysts haven’t seen before, and that needs to be turned into actionable information faster than ever.
Data are also being gathered from diverse sources, further complicating collection and analysis.
The industry groups turning to RWD believe the technological progress that’s been made, plus the unprecedented willingness to share data during the pandemic, are creating the perfect conditions to incorporate the tools into clinical practice and regulatory decision making.
Feeding RWD into clinical trial design
The COVID R&D consortium, a collection of biopharma R&D leaders, identified RWD as a priority early on in the pandemic, and it is directing its activities to an area where the data are already starting to make an impact: clinical trial design.
That application is directly in line with the consortium’s overarching goal of accelerating drug development (see “COVID R&D”).
“We’re learning about this disease as we’re designing the trials. There was very little prior information, very little published data and certainly no clinical trials to replicate,” said Anne Heatherington, SVP and head of the Data Sciences Institute at Takeda Pharmaceutical Co. Ltd. (Tokyo:4502; NYSE:TAK). She also leads COVID R&D’s data sharing component. “What we’re learning from real-world data is feeding into not just clinical practice but how we’re designing clinical trials and picking which drugs to use.”
The at least 15 biopharmas in the group are sharing clinical trial and patient-level RWD.
Its real-world data work, headed by Kathleen Gondek, VP of global health economics outcomes research and epidemiology at Takeda, and Heatherington’s data science component are brought together to collect and interpret the RWD.
“We recognized quickly that data sharing is a really important piece, and we have to involve real-world data because there wasn’t anything else. I think this is the first time in history that we’ve been in this place,” said Gondek.
For clinical trial design, using real-world data to understand the natural history of the disease is critical, said Gondek, who noted that the growing body of COVID-19’s natural history has been leading to changes in trial endpoints, inclusion criteria and therapeutic selection since clinical trials started rolling out. For example, Gondek noted that many trials pivoted to shortened, 14-day endpoints after learning more about the disease course and understanding that 28 days was too long.
“The more we use real-world evidence, the more we’ll understand about the strengths and limitations.”
“Only through having patient level real-world data will we learn what the relevant outcome measures are, what time point assessments are meaningful, and which patient populations to enrich,” she added.
Within the consortium, real-world evidence has led to discussions around trial populations and potential concomitant medication use. The consortium is now using it to determine timing for trials and to ensure trial design and adaptation are aligned with analysis output.
COVID R&D is organizing a master protocol trial to initially evaluate repurposed therapeutics, but it hasn’t disclosed details.
Jeremy Rassen, co-founder, CEO and chairman of healthcare technology company Aetion Inc., added, “I think the data can really help us understand the natural history of the disease, how it affects different groups such as older patients and those with kidney or cardiovascular disease, and that in turn becomes really important when thinking about who to include in clinical trials.”
Building on past success
While RWD hasn’t yet become the go-to tool for validating safety and efficacy of authorized therapies, the Reagan-Udall Foundation for FDA and Friends of Cancer Research are heading up an initiative -- dubbed the COVID-19 Evidence Accelerator -- to help guide FDA decisions.
And in May, FDA signed a deal with Aetion to generate information from RWD that will guide FDA decision making. Aetion gathers data from a variety of different sources, then uses analysis tools to interpret the information, but the company hasn’t disclosed exactly how the real-world evidence it generates will assist FDA.
“The collaboration is just kicking off, so while we can’t speak to study specifics yet, we anticipate the early work will focus on supporting FDA’s understanding of the disease and its impact, including characterizing COVID-19 patients, measuring medication use among patients with COVID-19, measuring diagnostic use and identifying risk factors for COVID-19-related complications such as hospitalizations, ventilation and intubation,” said an Aetion spokesperson.
A combination of technical advances made before the pandemic and data sharing during the crisis is creating the right conditions to accelerate the incorporation of RWD into regulatory actions and clinical practice, even beyond COVID-19.
“We were really fortunate coming into this that we had laid the groundwork for data collection and data sharing,” said Heatherington. “The technology platforms have been built. We have the capabilities.”
Heatherington noted that rare diseases were the launching point for the data sharing component.
“Careful consideration around data sharing in rare diseases in particular has helped us understand what types of platforms, legal agreements and security measures for privacy are required,” she said, although she noted that the process has been very slow in other indications.
Industry’s momentum in the COVID fight is changing that now.
“It took months and months for people to share data before this, between the agreements and third party brokers. Trials were finished before the data could be shared. But that work has enabled us to move rapidly into the galvanized industry. We have the technology and capability and now we’re seeing modified behaviors to find out how quickly we can act in the face of the pandemic,” Heatherington added.
The urgency of the pandemic means RWD is being shared at a faster pace, and Aetion CEO Carolyn Magill thinks the increased use of RWD will help the practice gain traction while fine-tuning the processes at the same time.
“More people will become familiar with how valuable these data are, and how more rapid cycle analytics and insights can impact clinical practice,” she said. “The more we use real-world evidence, the more we’ll understand about the strengths and limitations. That will better equip us to learn about how to apply it moving forward.”
The next step will be finding ways to address the bottleneck when it comes to cleaning and interpreting the data, given the influx coming in from different sources ranging from electronic medical records and insurance claims.
“There’s a typical lag time of at least a quarter and up to six months for real-world data curation, but now we don’t have the luxury of waiting. Now we’re getting data every two weeks, and the challenge is it’s a lot more messy,” said Gondek.
While the same level of data sharing and rapid drug authorization isn’t likely to continue past the pandemic, Gondek hopes the progress made with real-world data will.