The use of real-world evidence (RWE) is evolving quickly.
Having cut their teeth using RWE to validate safety post-approval, companies are increasingly interested in using evidence of real- world efficacy to win approval in new indications, leading to landmark regulatory decisions for drugs including Pfizer’s Ibrance. Regulators on both sides of the Atlantic want to facilitate RWE use but barriers including infrastructure limitations and data quality concerns remain.
Pfizer and the FDA shed light on the evolution of RWE in April when breast cancer drug Ibrance secured approval in a new indication on the strength of data from electronic health records (EHRs) and other real-world sources. Analyses of real-world tumour responses and safety data persuaded the FDA that Ibrance is safe and effective in men, leading it to expand the label.
The RWE-focused strategy enabled Pfizer to work round the rarity of male breast cancer and clinical trial enrolment challenges it creates. Pfizer brought Ibrance to market using data from randomised controlled trials (RCTs) in women, before supplementing that evidence with real- world data (RWD) in men to expand the label.
With the FDA backing the approach, Pfizer sees the approval as part of a broader move toward RWE.
“We appreciate that our partnership with the FDA has allowed us to take a significant step forward in the use of real-world data to bring medicines to patients who are most in need,” a spokesperson for Pfizer told PME. “Real-world data is playing an increasingly important role in expanding the use of already approved, innovative medicines.”
Regulatory positions
The growing role of RWE in the US is supported by the 21st Century Cures Act, which led the FDA to create a framework for incorporating RWD into regulatory decisions, commit to publish guidance on the specifics and create the MyStudies app to support collection of results in the real world.
Yet, the industry is pressing the FDA to go faster. In feedback on the framework, the American Society of Clinical Oncology raised concerns that the FDA is “overly conservative”, while others including the Biotechnology Innovation Organization pushed for the agency to clearly state that RWE can support new drug approvals when two or more RCTs are not necessary or feasible.
Ned Sharpless, Acting FDA Commissioner
Similar discussions are taking place across the Atlantic. The European Medicines Agency has shown a willingness to consider data from outside RCTs, noting in 2018 guidance on factor VIII products that changes to the patient population have made it hard to enrol clinical trials and led to a growing role for real-world registries.
The EMA expanded on its approach to RWE through a document detailing how it sees regulatory science evolving through to 2025. In the document, the EMA stated that it is fostering an ‘adaptive approach’ that uses real-world data to supplement clinical trial results. The EMA sees this approach being used in areas of high unmet need that are hard to study via traditional methods.
Muted interest
The EMA’s limited embrace of RWE is echoed in the muted level of industry demand to use data from the real world in European development and regulatory strategies.
An analysis of requests for EMA scientific advice from July 2016 to June 2017 found 4% of discussions involved RWE. Ten of the 19 RWE- related requests involved the use of historical controls. The EMA agreed or partially agreed to seven of the historical control proposals.
The EMA’s willingness to diverge from its preference for RCTs in some situations is evident in its drug approvals. For example, the EMA approved MolMed’s Zalmoxis on the strength of an analysis that compared patients who received the allogeneic T cell therapy to people in a real-world database.
Most members of the EMA’s Committee for Medicinal Products for Human Use supported approval on the strength of the analysis but there was some pushback. In a dissenting opinion, three members of the committee called the results “plausible and promising” but argued against approval because of “critical issues that put the reliability of the available data into question”.
UK strength
The level of trust placed in RWE depends largely on the source of the data. Results gathered in the real world will, for better and worse, be the product of less controlled environments than RCT data.
That is part of the appeal of RWE, as it may be more representative of how real patients respond, but the variability and heterogeneity of RWD is a problem. Recognising that, the United Kingdom has sought to position its relatively standardised, single payer healthcare system for use in the generation of RWE. AstraZeneca is among the companies to recognise the potential of the UK.
‘The UK is an ideal setting for population-based studies of diabetes because diabetes care is largely coordinated by the GP, and information about metabolic parameters, cardiovascular risk factors, diabetes co-morbidities and disease outcomes are collected electronically,’ AstraZeneca wrote in the protocol for an observational study.
AstraZeneca is putting its confidence in UK infrastructure to the test in a pragmatic trial designed to compare the effectiveness of diabetes drug Forxiga to the standard of care. The trial is using a Clinical Practice Research Datalink (CPRD) platform to directly populate electronic case report forms from electronic health records (EHRs), collect patient-reported outcomes and help GPs find eligible patients.
The UK has high hopes for the study, which the Medicines and Healthcare products Regulatory Agency (MHRA) thinks is “the first large commercial real-world pragmatic trial of its type using EHRs on a single electronic trials platform”. The MHRA hopes the trial will “attract global attention” and “be the first of many” such studies run on CPRD infrastructure.
Identifying barriers
While CPRD may lessen challenges associated with RWE, AstraZeneca acknowledges that “there may be inaccuracies” in the data and variability in the availability of medical records.
The concerns are part of broader worries about the limitations of RWE. The EMA articulated some of those concerns in its 2025 regulatory strategy, noting that the ‘often heterogeneous nature of the data sources’ means further work is required ‘on the analytical and epidemiological methodologies needed to deliver robust evidence’.
European regulatory officials also have concerns about data privacy and security, and think more work is needed to answer fundamental questions such as where in the assessment process RWE may add value.
Other organisations think infrastructure shortcomings are preventing the collection and analysis of the robust data.
‘A significant limitation in RWD activities is related to the lack of infrastructure to collect all relevant data,’ Johnson & Johnson wrote in its feedback to the FDA’s RWE framework. J&J cited missing data fields, non-interoperable EHR systems and the need to capture measurements beyond those typically gathered by healthcare providers as examples of the problem.
J&J acknowledged that the FDA lacks the power to drive some infrastructure changes but thinks the agency can, and should, do more through its framework.
Ibrance gained a new approval based on data from EHRs and other real-world sources
The path forward
The diversity of these problems means a single fix is unlikely. Novartis has argued that the FDA needs to enable companies to share information about their RWE activities in a ‘learning environment’ that adds to the knowledge of industry and regulators, alike.
Elsewhere, the EMA has compared the FDA’s real-world Sentinel safety system favourably to
the patchwork of European systems, noting that it gives its counterpart ‘the ultimate level of control’. Sentinel is expensive, though, and the EMA sees its challenge as being how to achieve the same level of reassurance without exerting as much control.
It is unclear how these proposals and other efforts to improve RWE will play out in the coming years. Yet, the ongoing digitisation of healthcare and pressures to cut time to market create some near certainties. The breadth and depth of RWD is set to rise. The question is what industry and regulators do with it.