Despite the extensive clinical research needed to get new medicines approved, it can still be hard to answer the basic questions a patient might have about a drug. Is this the right medicine for me? Am I likely to get the side effects they warn about?
The scientists responsible for developing new therapies know that clinical trials alone can’t address every question. Elliott Levy, senior vice president for Global Development at Amgen, is frank in stating the limitations inherent in clinical research.
“Clinical trials gather data on a product’s performance patient by patient, data point by data point,” Levy said. “But the data from clinical studies should not be where our understanding stops. Data from real world use is critically important too so that we can learn more about a product’s performance once the medicine reaches the market.”
To paint a more complete picture of the real-world performance of medicines, Amgen and other companies are turning to real-world data. The trend has been fueled by rapid growth in the volume of medical information collected in digital formats and available for analysis. As of 2015, 87 percent of office-based physicians in the US and nearly all hospitals were using electronic health records (EHRs) to store patient data. With the right analytical tools, this ocean of data can yield real-world evidence—the conclusions that can be validly drawn from the data.
“The technology has evolved, so we can analyze these big data sets more efficiently,” said Cathy Critchlow, vice president and head of Amgen’s Center for Observational Research (CfOR). “Statistical methods have also evolved, which has given us greater confidence that we can derive accurate inferences from these data.”
The edge provided by real-world data is already large and poised to get much larger. “We’re on the threshold of a transformation in how we evaluate medicines,” said Levy. “Real-world data will dramatically improve our ability to acquire useful evidence about our products. It will allow us to get drugs to market more quickly, expand their usage to new indications more quickly, and answer critical safety questions to guide the safe and appropriate use of medicines.”
We’re on the threshold of a transformation in how we evaluate medicines. Real-world data will dramatically improve our ability to acquire useful evidence about our products.
A Head Start for Amgen
Interest in applying real-world data to drug development got a major boost with the passage of the 21st Century Cures Act in 2016. The law directs the FDA to develop guidelines on how real-world evidence can be used to support regulatory decisions, including requirements for post-approval safety studies and review of new indications for approved drugs.
“The new regulatory environment is spurring the rest of the industry to build up their capabilities in this area,” said Critchlow. “I’d say that Amgen is at the front of the pack because we’ve been integrating real-world data into our R&D programs for more than a decade now.”
“Some companies focus on just a few areas like safety, but our use of real-world data is very broad,” said Brian Bradbury, executive director and head of the Data & Analytic Center within CfOR. “Our platform enables us to answer a range of questions relating to disease, treatment patterns, and the standards of care available in many different geographies. We’ve built an infrastructure to support programs across all stages of the drug development cycle.”
Using real-world data in regulatory submissions
One of Amgen’s cancer therapies gained its first regulatory approval based on a single-armed Phase 2 study supported by real-world evidence from medical records. The company didn’t include a traditional standard-of-care comparator arm in the study because the patients enrolled had already failed to respond to standard therapies.
“In order to put the data from the active arm into context, we felt it was important to characterize treatment response rates and safety issues in a highly similar real-world population of patients who had not received our investigational therapy,” said Critchlow.
Expanding the indications for approved therapies
Real-world data has the potential to provide persuasive evidence for expanding the approved uses of a drug to new types of patients and new diseases. The evidence can be used to support investments in clinical studies to gain approval for new indications. In addition, the FDA is working on policies to define when such data may be incorporated in the approval process for new indications (see below).
The savings in time and money would be considerable. “Instead of doing a $10 million clinical trial that could take years to complete, we might be able to do a one-year, $1 million data study that would actually be more informative,” said Levy. “The FDA is inviting sponsors to come forward with proposals for using real-world evidence to support new indications, and Amgen is embracing this opportunity.”
Designing more effective and successful clinical trials
In designing trials, drug development teams must establish criteria they will use to include or exclude different patients. Real-world data can show the impact of different criteria on the potential pool of patients. “If your criteria are too limiting, the data might show that you’ve excluded most of the patients with the disease you’re studying,” said Bradbury. Using data to optimize the criteria can help accelerate patient recruitment and ensure that the results are more broadly relevant.
Data can also provide guidance on how large and/or long a study needs to be to allow a test drug to demonstrate an impact on disease outcomes. “When you’re thinking about how to power a study,” said Bradbury, “it’s very helpful to know the background rate of different outcomes in the patients you plan to include in the study.” Real-world data can give you that information and ensure a study has the right size and duration.
Protecting the safety of patients
Realworld data is widely used to augment clinical trial data in determining a medicine’s safety profile. The data can be especially useful in measuring the risk of very rare but potentially serious side effects. “With some medicines, the risks of most interest to regulators may occur in one out of every few thousand patients,” said Critchlow. “To gather meaningful data on these risks, you would need to study hundreds of thousands of patients, and you couldn’t do a clinical study that large.”
So instead, Amgen has been working with companies that design electronic medical records to incorporate safety-related questions into EMRs. Physicians are prompted to ask about potential adverse events when they meet with patients using certain Amgen medicines. “We’re actually embedding studies within the EMRs to replace more costly but less efficient pharmacovigilance programs,” said Bradbury. “We’re not just analyzing data that’s already out there—we’re ensuring that safety data gets collected.”
Making Real-World Data Widely Available
Amgen is also using real word data to identify populations of patients whose needs aren’t being met by current therapies; to optimize inclusion and exclusion criteria for trials to help speed recruitment of patients and ensure relevant results; and to show how well different medicines fare in helping people stay productive and healthy.
To provide the company’s drug development teams with ready access to useful data, Amgen launched a Real-World Data Portal three years ago. The portal taps into anonymized records from more than 150 million patients. Complex queries can be executed and answered in minutes or seconds. “You can get the age distribution for people with a given disease as well as the other diseases they have and the medicines they are using,” Bradbury said.
The portal has generated high demand inside the company and recognition externally. At the 2016 BIO IT World Conference and Expo, Amgen’s real-world data platform received the Best Practices Award in the Clinical IT & Precision Medicine category.
Some companies focus on just a few areas like safety, but Amgen’s use of real-world data is very broad. We’ve built an infrastructure to support programs across all stages of the drug development cycle.
More Regulatory Clarity Expected
For real-world data to reach its transformational potential, regulators, health care providers, and drug developers need to evolve how they operate. In the U.S., the FDA has made this evolution a top priority.
In a speech to the National Academy of Sciences in September 2017, FDA Commissioner Scott Gottlieb said that real-world evidence “can help make sure doctors and patients are better informed about the clinical use of new products, enabling them to make more effective, efficient medical choices. ... As the breadth and reliability of real-world evidence increases, so do the opportunities for FDA to also make use of this information.” While cautioning that data won’t replace traditional clinical trials in many cases, Gottlieb said there were more possibilities in both the pre- and post-market context “to use real-world evidence where doing so will improve medical product development.”
The FDA has already started to use real-world data to make approval decisions about new medical devices and new uses for existing devices. However, when it comes to decisions about the medicines the agency regulates, Gottlieb acknowledged a need for greater clarity and guidance. “I recognize that FDA isn’t always clear about our approach to real-world evidence in regulatory decisions,” he said. “We are taking steps to provide more clarity. ... We’re now working on policies to support the use of real-world evidence in the approval of new indications for already marketed drugs. This may be especially relevant in settings like rare diseases or other unmet medical needs, where it can be hard to enroll patients in clinical trials.”
In that same speech, Gottlieb called for health care systems to ensure that EMRs include more data that can inform medical and regulatory decisions. The current incentives “force data to be structured in ways that are geared to billing. Ideally, we’d like to have a system where providers have the right incentives to enter clinically relevant information into EMRs at the point of care.”
In the biopharma industry, the expertise and processes used to develop drugs are deeply rooted in randomized clinical trials. Companies need to evolve in ways that accommodate real-world data and combine it with clinical trials to evaluate therapies faster and more efficiently. Levy said that Amgen is working to advance the new paradigm as rapidly as possible. He sees it as a solution to one of the largest constraints on the industry’s productivity.
“We’re making tremendous progress on the research side of R&D,” Levy said. “Large genetic datasets are enabling us to identify high-quality therapeutic targets that impact disease risk in humans. We have an expanding toolkit of drug modalities to interdict these new targets, including cell-based therapies and hybrid proteins. Our biggest challenge is the very high cost of successful clinical development, because that cost precludes from pursuing a lot of these highly promising opportunities. Real-world evidence can play an important role in helping to solve this problem.”