Real-World Reluctance? Despite Guidance, Drug Developers Wary Of RWE-Based Trials

Regulators around the globe have released guidances to encourage biopharmaceutical firms to use real-world evidence strategies, but with hundreds of millions of dollars in research and development spending at stake for a single drug, companies are reluctant to walk away from traditional randomized, controlled clinical trials – even if substituting a typical control arm with an analysis of real-world data has the potential to save time and money.

Interest in real-world evidence (RWE)-based clinical trial strategies is growing, with biopharma companies considering such study designs more frequently now than they did a year ago, thanks to new guidance documents from regulators, Michelle Hoiseth, chief data officer at the contract research organization Parexel International Corp., told Scrip in an interview during the Drug Industry Association (DIA) annual meeting, held 23-27 June in San Diego, CA.

RWE was an inescapable topic during the DIA meeting, coming up in panel discussions on drug maker communications with payers, artificial intelligence in drug development and clinical outcomes assessments, as real-world data (RWD) becomes an increasingly important means of communicating a drug's safety and effectiveness to regulators, payers, doctors and patients.

Despite recently increased interest in real-world evidence strategies, Hoiseth said, "we're a tremendously risk-averse industry." While biopharma interest in incorporating RWE into clinical trials is "incredibly high" when companies' R&D teams begin to design studies, they back away from such strategies based on the uncertainty of whether regulators will sign off on innovative trial designs and RWD-derived results, she said.

Hoiseth also noted concerns about dealing with new and evolving sources of data and having to hire people with the expertise to identify and analyze sources of real-world evidence.

"I think the publications of the guidances over the past year have really, really helped people to know where to make their bets, but up until that point it had been hard to know that the data set that you created for analysis would be acceptable," she said.

RWE Framework Offers Guidance, Not Certainty

It's not surprising that companies are reluctant to adopt real-world evidence strategies, since regulatory guidance still is coming together in this area. For instance, while the US Food and Drug Administration released its RWE framework in December, it won't release the final guidance on the use of real-world evidence in trials of new indications for marketed drugs and post-marketing studies until December 2021. However, the draft guidance issued in December 2018 laid out the agency's thinking on the matter, including its requirements for high-quality data and study designs.

"The FDA framework that came out in December makes it clear, in fact, reinforces their interest in real-world evidence in a number of ways … but it's a long way from having clear guidance on what kind of evidence would be acceptable," Nancy Dreyer, chief scientific officer and global chief of scientific affairs for IQVIA Real-World & Analytic Solutions, told Scrip in an interview at the DIA meeting. (Also see "Real-World Evidence Takes Courage: A Conversation With IQVIA’s Nancy Dreyer" - Pink Sheet, 11 Feb, 2019.)

Dreyer noted that real-world data often looks different from data collected in randomized, controlled trials due to multiple variables, such as the age and ethnicity of patients and where they're treated.

The reliability of RWD is a big concern for both regulators and clinical trial sponsors. However, the time, money and other resources saved by conducting RWE analyses versus the expense, length and complexity of enrolling a comparator group of patients in a traditional randomized, controlled trial could be considerable.

Takeda Pharmaceutical Co. Ltd. president of research and development Andrew Plump told Scripearlier this year that as part of its investment in data science, the company is examining ways to use real-world evidence to support new safety and efficacy claims for its approved products. However, Plump said, Takeda knows the shift to such studies will not be an overnight process despite the draft guidance on the use of RWE in clinical trials. (Also see "Takeda's Investments In Data Science Seek Improved Trial Outcomes, R&D Efficiency" - Scrip, 14 Jan, 2019.)

Hoiseth also asserted that the shift to RWE-based studies will be gradual, but she sees a role for such strategies in earlier stages of development. The Parexel executive said the use of real-world data to show the value of approved drugs versus current standards of care is a well-established practice that biopharma companies are comfortable pursuing, but RWE has a place in R&D too.

By extracting electronic health record (EHR) data for a comparative analysis rather than enrolling patients in a control arm of a randomized trial, Hoiseth said, "that's the opportunity to speed the execution of studies and really drop the cost of development – not really today, but eventually," as data sources and as data extraction and analysis technologies and teams are developed.

Some Companies Reluctant, Others Are Embracing RWD

But despite anecdotal evidence of companies' reluctance to pursue clinical trials that rely on real-world evidence, drug developers are using natural history studies, health care insurance claims data and information gathered from electronic health records (EHRs) to provide some context for data generated in single-arm trials at increasing rates.

"In 2011 in Europe, there were less than five single-arm trials submitted to [health technology assessment (HTA)] bodies or payers in a year," Christina Mack, director of epidemiology and health outcomes at IQVIA, said in an interview at the DIA meeting. "In 2018, there were almost 60 single-arm trials and they're making major decisions on one small group of patients. Along with that, we've seen this appetite and hunger for what context can we put next to that, so that's where we see a big driver of the external comparator."

Regulators globally are racing to keep up with the industry with new RWE guidance documents, including Health Canada, which issued guidance in April in anticipation of increased filings based on real-world data. (Also see "Canada Develops RWE Guide Anticipating Increased Submissions" - Pink Sheet, 26 Apr, 2019.) 

The European Medicines Agency has been even more proactive, filling data gaps that companies chose not to pursue. The regulator revealed in May that it conducted 14 RWE-based studies of its own to aid its review of certain marketing authorization applications. The EMA also is working with doctors on the generation of RWE that eventually may support drug approval decisions. (Also see "EMA Enlists GPs’ Help To Generate More Real-World Evidence" - Pink Sheet, 11 Jun, 2019.)

Like the EMA, the US FDA is conducting its own RWE studies that will use various types of real-world data to try and replicate results seen in Phase III and IV randomized clinical trials for approved products. (Also see "Real-World Evidence: Sponsors Will Need Prospective Process Early In Development" - Pink Sheet, 12 Jun, 2019.) The agency admits that some of these endeavors will fail to replicate prior results, but it will still apply insights from these studies to RWE trials proposed by drug developers going forward. (Also see "Real-World Evidence: Replication of Controlled Trials Expected To Fail Sometimes, US FDA Says" - Pink Sheet, 26 Jun, 2019.)

Bayer AG vice president Patrick Brady, head of regulatory policy and intelligence for the German company, said during a 24 June DIA session on artificial intelligence in drug development that there are promising signs that RWE-based studies may be able to replace randomized, controlled trials, but he wasn't sure about how long that will take.

"I'm encouraged by some of the progress that I see from some of the leading health authorities around the world that are becoming much more open to different approaches to drug development," Brady said. "I'm encouraged by what I see in terms of the state of the dialogue on the use of real-world evidence for regulatory decision-making, but I think we're still very much in the early days."

Data Debate Ongoing, But Mining Continues

The debate about which sources of data and what kind of study designs are acceptable to regulators continues.

Even so, some big pharmas and large biotechnology firms have invested in large data sources of their own with the intention of using the information to aid in drug development, mining the data for new therapeutic targets and for RWE. (Also see "Real-World Evidence: Observational Studies Deserve More Respect In US FDA Framework" - Pink Sheet, 30 Apr, 2019.)

Roche used data from its Flatiron Health Inc. subsidiary to generate data for an external comparator arm to complement its single-arm pivotal trial for the tyrosine kinase inhibitor entrectinib in the treatment of patients with solid tumors that have an NRTK mutation and ROS1-positive non-small cell lung cancer (NSCLC). New drug applications (NDAs) for both indications are pending with the US FDA and have 18 August user fee dates. (Also see "Roche Outlines Use Of Real-World Evidence In Entrectinib NDA" - Pink Sheet, 11 Jun, 2019.)

Amgen Inc. mined its acquired deCode genetics EHF database for safety data to assist the FDA in its review of the new osteoporosis drug Evenity (romosozumab). A few years earlier, Amgen used RWE to support the agency's review of its bispecific T-cell engager (BiTE) Blincyto (blinatumomab), which won an initial acute lymphoblastic leukemia (ALL) indication based on the comparison of a single-arm trial to RWE – results that eventually were replicated in a confirmatory trial. (Also see "Real-World Evidence At US FDA: Bavencio, Blincyto Approvals Point Way Toward Broader Use " - Pink Sheet, 7 Aug, 2018.)

The approval of Pfizer Inc.'s Ibrance (palbociclib) for men with breast cancer on 4 April is the most recent use of RWE in an approval. An indication for women with HR+, HER2- breast cancer was extended to include men, based on electronic health records and post-marketing reports of use in men from the IQVIA Insurance database, Flatiron Health’s breast cancer database and Pfizer’s global safety database. (Also see "Keeping Track: Industry Channels Inner Gottlieb As Commissioner Departs US FDA" - Pink Sheet, 5 Apr, 2019.)

While Dreyer said RWE strategies can be used across both small and large indications, she noted that clinical trials in rare disease and oncology are the most likely studies to incorporate real-world evidence.

Indeed, the US FDA released a draft guidance on rare disease drug development in January that emphasized the use of natural history data to support the design of clinical trials and analyze their results, and several companies had such efforts under way long before that guidance was issued. (Also see "Baby Steps To Real-World Evidence Of Efficacy: External Controls Gain Popularity In Rare Disease Trials" - Pink Sheet, 16 Apr, 2019.)

Patient Voices In The Real World

Telba Irony, deputy director of the Office of Biostatistics and Epidemiology within the US FDA's Center for Biologics Evaluation and Research (CBER), said during a 24 June DIA session on enhancing patient-focused outcome assessment that in addition to looking at patient preferences and patient-reported outcomes, CBER is evaluating a lot of real-world evidence, including natural history studies. Irony noted that many CBER-assessed products are advanced therapeutics, such as cell and gene therapies, and treatments for rare diseases, which often involve small, single-arm trials.

She said "it might be unethical and sometimes impractical to get a control group, so we are very much interested in how to extract information from natural history of diseases to help in our analysis of clinical trials. We are bridging into the real-world data also with the same objective." That RWD may include input from patients on their treatment regimens and data collected from patients via wearable devices or other digital health technology.

Chris Pashos, vice president of global evidence strategy at AbbVie Inc., noted during a 24 June DIA session on drug developer communications with payers that AbbVie has incorporated what doctors and patients want from new therapies into its clinical trials, which the company discusses early on with regulators and with payers to make sure the studies generate data of value to all the stakeholders.

"We needed to share with them that this was important to patients, therefore it was important to us as we developed a new medicine," Pashos said. "And, indeed, over the last 20 years, the FDA has realized this isn't just a world of randomized, controlled trials; we have now an explosion of real-world data, we have now an explosion in the patient voice, and those two changes … are now very much a part of how we are developing and commercializing new medicines."