I recently served as the moderator for MassBio’s forum on “Closing the Real-World Evidence Gap: Pragmatic Clinical Trials & Observational Studies.”
This lively and informative discussion was led by a panel of industry leaders, which included Robert Califf, MD, MACC, Head of Medical Strategy and Policy, Verily Life Sciences and Google Health; Jane Liang White, ScD, Senior Director, Statistical Group Lead for Oncology/Hematology Franchise at Pfizer; and Rebecca Miksad, MD, MPH, Senior Medical Director of Flatiron Health.
A double-blind, placebo controlled design is the classic archetype of the randomized clinical trial (RCT). It is often considered the gold standard. There is increasing concern on the part of regulators that explanatory clinical trials, which confirm a clinical hypothesis, tend to overestimate the effectiveness and underestimate the safety of medicinal products and devices. Hence, results from these trials may not yield a true picture of how a drug performs in a real-world setting. Pragmatic RCTs, conversely, seek to choose the best treatment administered in the conditions of use via collection of data from day-to-day patient care. These trials offer more flexibility in treatment and wider recruitment of patients, due to fewer exclusion criteria, and simulate effects in those who will ultimately receive the new product.
Dr. Califf, a former FDA Commissioner, spoke to real-world evidence (RWE) in drug development and the evolving changes in evidence generation to assess benefits, risks and value of medical products. He suggested our current clinical research system is well-intentioned but flawed, stating, “a high percentage of decisions are not supported by evidence; health outcomes and disparities are not improving; and the current system is too slow, not reliable, and does not answer questions that matter most to patients. We are not generating the evidence we need to support the healthcare decisions that patients and their doctors must make every day.” He emphasized that the FDA needs to provide guidance based on lessons learned to inform the industry of the risks and benefits observed post-drug approval. As technology changes, he believes there is a need to change the way we do clinical trials and generate evidence; it needs to be done in technology-based systems that are not only focused on bringing drugs to market but also on how to use them in practice.
Dr. Liang discussed statistical and study design considerations for registrational studies with real-world data (RWD) as a synthetic control arm. She spoke to a recent trend in regulatory approvals. Amgen was among the first to use an historical RWD comparator to support accelerated approval of BLINCYTO® in March 2018. Her current company, Pfizer, received approval for IBRANCE® for male breast cancer, a label extension based on data from post-marketing reports and electronic health records (EHRs), in April 2019. In December 2018, FDA issued their Framework for FDA’s Real-World Evidence Program. The scope of this program is to evaluate potential RWE in support of the approval of a new indication for a drug already approved, or to satisfy post-approval study requirements. Dr. Liang addressed key points for future study design and analysis when seeking approval using RWD, which included: pre-specify the study design and analysis using RWD; have comparable patient populations (key selection criteria) in the trial and in the RWD; use propensity score methods to remove or reduce confounding effects; e.g., propensity score matching, inverse probability of treatment weighting (IPTW). Her key takeaway message was that sponsors should consult with the FDA in the early stages of trial planning, before study initiation, to get the agency’s buy-in on the study design and possible registrational path. The FDA is keen on collaboration and transparency.
Lastly, Dr. Miksad addressed sources of RWD, primarily EHRs, and the benefits of EHR data to pragmatic RCTs. EHR data facilitate pragmatic RCTs by allowing trials to be run in a more decentralized fashion. She spoke on how technology assists with data reliability and organizing EHR data - and monitors quality for efficient and reliable unstructured data processing (e.g., free-text notes, reports). Technology is also leveraged to transform structured data and map all data elements to a common data model. Dr. Miksad described RWD as being on a continuum with traditional clinical trial data. Determination of where and when to use RWD is based on the consideration of the research question and the appropriate study design, prior to study initiation. She noted that “fit for purpose” analytical methodology is an exciting area of growth in the field, meant to harness information we get from RWD. Dr. Miksad concluded that, the technology of RWD has evolved in the service of data quality and reliability which remain paramount.
There is no doubt that pragmatic RCT is an important tool in drug development especially for confirming effectiveness of new therapies. We must consider adopting RWD and associated technologies when they are appropriate for the research goals. At PROMETRIKA, we feel strongly about the importance of clinical innovations and have a dedicated staff focused on innovative trial design and execution including leveraging real world evidence as well as new technologies to help our clients deliver new therapies to patients faster and more efficiently.