This ISPOR conference advanced the discussion of the role and value of RWE in making regulatory and reimbursement decisions, with an RWE-dedicated sub-programme running throughout the conference.
Recent FDA Guidance Documents on RWE for Regulatory Decision-making1
In the fourth quarter of 2021, the FDA released a batch of four guidance documents for industry on RWE, to help inform appropriate, robust study designs (from medical claims data, electronic health records and registries) and analysis methodologies for the purposes of regulatory decision-making:
Assessing Electronic Health Records and Medical Claims Data to Support Regulatory Decision-Making for Drug and Biological Products
Assessing Registries to Support Regulatory Decision-Making for Drug and Biological Products
Considerations for the Use of Real-World Data and Real-World Evidence to Support Regulatory Decision-Making for Drug and Biological Products
Data Standards for Drug and Biological Product Submissions Containing Real-World Data
John Concato (Associate Director for RWE Analysis, FDA), noted the many misconceptions that persist around real world data (RWD) being a new, “special” form of data, whereas the FDA considers RWD to be just another “type” of data, which requires similar rigorous standards as required of clinical trial data in order to be reliable for decision-making.2 A recent example was cited of the approval of tacrolimus in a new indication (prophylaxis of organ rejection in patients receiving lung transplants) based solely on an RWE study. This was a registry study on all lung cancer recipients in the US over a ~20 year period, with ‘hard’ outcomes (organ rejection and death) unlikely to be influenced by bias.3 The FDA determined that this study met its evidentiary requirements and therefore was suitable to support a label extension, without requiring a randomised controlled trial (RCT). With the release of clear FDA guidance documents on study design and data standards, we can expect to see industry pivot towards RWE studies instead of clinical trials for label extensions.
The findings from the joint FDA-Duke Margolis RCT-DUPLICATE demonstration project were alluded to in a number of sessions at the conference, with the full results presented at a separate webinar in early May. The project aimed to emulate 30 RCTs using claims data to understand whether similar conclusions would be made if an RWE study was used instead of an RCT. Overall, RWE studies resulted in similar conclusions to RCTs regarding treatment effect when the study designs could be closely matched (in terms of population, comparators, outcome reporting, standard of care, follow-up, etc.). However, only half of RCT studies could be closely emulated by RWE studies; where this was not possible the results were more divergent from the original RCT, given that essentially the RWE studies were answering different clinical questions. These results support the value of rigorous, robust RWE studies in providing regulatory-standard evidence for decision-making and further pave the way for the expansion of how RWE is used in the drug approval process.
Role of RWE in Reimbursement Decision-making4, 5
RWE also plays an important role in reimbursement decision-making, allowing payers to better understand the impact of a treatment in clinical practice. Seamus Kent (Senior Adviser in Data & Analytics, NICE) presented an example of where NICE conducted an RWE study as part of their updated guidance on the use of direct oral anticoagulants for stroke prevention. The study results were similar to those derived from indirect comparisons of clinical trial data, again demonstrating the value of robust RWE studies in providing evidence for decision-making. Jon Campbell (Senior Vice President for Health Economics, ICER) presented the outcome of a reassessment of prophylaxis therapy for hereditary angioedema, which used RWE to better estimate the baseline attack frequency (an outcome to which the original cost-effectiveness analysis was sensitive). The claims databases analysis showed the frequency of attacks in clinical practice to be lower than in the clinical trials, and therefore suggested further discounts of prophylaxis therapy cost in order to meet the willingness-to-pay threshold. In both cases the question remains as to where the responsibility for generating post-approval RWE lies and whether such reassessments should be common or conducted only in cases with uncertainty in the clinical trial evidence.
The increasing acceptability of RWE to payers also raises the practical question as to whether RWE studies from other countries will be accepted by payers, or if local RWE studies will be required for most major markets. In 2021 and 2022, a number of HTA agencies – NICE, HAS and IQWiG – published guidance on the use of international data, with the overriding message being that the generalisability of such data to the country needs to be explored and justified. The example of the NICE evaluation of tepotinib, where RWE from an international cohort was used as part of an indirect comparison with the single-arm clinical trial, highlights the challenges of using international data. The results were considered highly uncertain by the committee, predominantly due to the lack of generalisability (different standards of care, different disease severity) to the UK population. The company had to resubmit a different indirect comparison, based on clinical trials in a more UK-relevant patient population, which the committee ultimately used as the basis of its decision-making. When planning global clinical trials, industry already must anticipate the different requirements from regulators and payers in different countries; increasingly such considerations will be required when planning RWE studies for regulatory and reimbursement purposes, in order to avoid the pain of conducting RWE studies locally in each major market.
HARPER Harmonised Protocol Template for Comparative Effectiveness Studies and RWE Study Registry1
The output of an ISPOR-ISPE (International Society of Pharmacoepidemiology) joint task force to produce a harmonised protocol for hypothesis-evaluating RWE studies was presented at the conference. The template has been developed based on four other templates published by relevant organisations. The same group also developed the recently launched registry of RWE studies, for which the new protocol template will help to ensure more consistent reporting. Only 21 studies were registered as of June 2022, 6 months after the launch of the registry, suggesting that a substantial cultural change is required for pre-registration of RWE studies to become normalised.
Craig-Brooks-Rooney, Global Scientific Director and Head of US (LinkedIn)