Guest Column | April 24, 2018

How Should Biosimilar Companies Approach Real-World Evidence?

By Anita Burrell, principal of Anita Burrell Consulting LLC

How Should Biosimilar Companies Approach Real-World Evidence?

Real-world evidence (RWE) has become increasingly important for innovative manufacturers looking to convince reluctant payers that their new medicine has a place in patient care. Indeed, many claim the changing nature of the data landscape in healthcare will revolutionize the way we consider evidence, but does this apply to biosimilars? Is there a real need to look at data from actual clinical practice before or after market entry for these products?

How Is Real-World Evidence Different?

A recent framework from the Duke-Margolis Center for Health Policy (Berger M. D., 2017) defines RWE as “evidence derived from data relating to patient health status and/or delivery of health care routinely collected from a variety of sources.” It is now widely acknowledged that analysis of observational data on routine clinical practice is more pertinent and informative for coverage decisions than data collected during trials conducted for regulatory approval alone. Why is this?

The answer relates to the nature of the information collected and the purpose for which it is intended. Clinical trials conducted during the drug development process are designed to answer questions regarding cause and effect of the treatments studied, i.e., if the efficacy and safety noted in the trial is truly due to the arm of the trial to which the patient was assigned. To ensure the study’s outcomes can be attributed to the choice of therapy, these experiments are highly controlled to reduce confounding and bias. They use randomization and have strict inclusion and exclusion criteria to guarantee the populations for each arm are comparable. Similarly, the use of protocols to limit choice to the agents under inspection and dictate treatment observation and intervention reduce individual variation which could lead to bias in the outcomes. While these restrictions ensure the differences in the study results can be assigned to the treatment choices, they make the extension of trial results into medical practice difficult to hypothesize. Will the treatment work to the same extent, better, or worse when applied to a population outside of that studied? How will it stack up against a different comparator than the one in the clinical trial? For these reasons and many others, manufacturers and payers have analyzed RWE. But in recent years, the interest in the information that can be gleaned from this data has risen dramatically.

Why Is RWE Becoming A Hot Topic?

Four major trends are driving the increase in the importance of RWE in our healthcare systems, with the first and perhaps the most influential being the availability of electronic health records. This and other clinical and administrative data offer examination of information more broadly representative of clinical settings and patient populations than those found in traditional clinical trials. This data meets the second trend of increased tailoring of therapies to improve individual rather than population outcomes. In addition to this wealth of traditional data, the third major force is the advent of new technologies that enable patients to generate their own health data, providing opportunities to learn from patient experiences with treatments and their impact on day-to-day life. RWE is already used for multiple purposes in the U.S. For example, manufacturers include information on medical needs to inform targets for drug development; the safety monitoring process has relied upon the collection of post-approval information in real-world settings, e.g, SENTINEL (Office of Critical Path Programs, 2008); and the FDA has used RWE to inform regulatory decisions where a randomized controlled trial (RCT) has been impossible or impractical, e.g., rare diseases.

However, in the last few months the fourth trend of regulatory interest in RWE has intensified, in particular with respect to the eagerly awaited guidance from the FDA regarding the incorporation of RWE into regulatory decisions. The guidance is a requirement of the 21st Century Cures Act (Energy and Commerce Committee, 2016) and is also included in the recently agreed-upon Prescription Drug User Fee Act (PDUFA) IV (Center for Drug Evaluation and Research, 2017) with a deadline for a draft framework by the end of 2018 and a draft guidance by the end of 2021. The FDA must establish a program to evaluate the potential use of RWE to support the approval of new indications for an approved drug and/or to support or satisfy post-approval study requirements for the 21st Century Cures Act. It is widely anticipated that the guidance will be similar to the already-released guidance for medical devices and in line with the Duke-Margolis workshop and subsequent white paper on “A Framework for the Regulatory Use of Real World Evidence” released at the end of last year. The use of RWE across the spectrum — from new drug application filing to post-market commitments — is considered with a focus on the use of the evidence relating to confidence that the information is transparent and reproducible. Motivation to use RWE by more conservative stakeholders is not restricted to the U.S., as drug regulatory bodies such as the European Medicines Agency are also involved in projects such as the Innovative Medicines Initiative GetReal to explore the application of these data sources.

So Can Biosimilars Benefit From RWE?

Biosimilar entry to markets has rarely resulted in the highly anticipated cost savings. This is due to a number of reasons, particularly reluctance from providers (and in some cases patients) to switching from originator products. At a recent press briefing at Pfizer, Richard Blackburn, global president for biosimilars, referred to the need for education to clear up misinformation about biosimilars. Here RWE can demonstrate the product performance and management of disease in practice do not alter significantly from originator to biosimilar.

A more innovative use of RWE, though, may approach the need for fair and equal market access, which Blackburn also referred to. In the U.S. especially, the entry of biosimilars has been blocked through contracts with originator manufacturers which offer short-term rebating for preferential placement or exclusive access to a payer’s formulary. John Kennedy, U.S. head of biosimilars at Pfizer, noted highly integrated systems such as the U.S. Department of Veterans Affairs accounted for about half of the share of Inflectra in the U.S. mainly because they “were able to prioritize sustainable cost savings over short-term rebates.” This means when we are able to link the spend at the pharmacy with medical outcomes and the rest of the healthcare system, the incentive for using higher-cost agents (and thus gaining greater rebates) is reduced.

RWE can link systems that are already integrated to demonstrate the savings to the system from biosimilars but also as a performance indicator for providers seeking quality patient outcomes while bending the cost curve. As of January 2018, the number of accountable care organizations (ACOs) participating in Medicare ACO shared savings initiatives grew to over 561 nationwide, with nearly 10.5 million members (Center for Mediare and Medicaid Services, 2018). Medicare ACOs are one target for accessing RWE on biosimilars, as they have the continuum-of-care data needed to show the outcomes of treatment in specific patients. They are also potential consumers of RWE produced by similar organizations since they would use this information to support access for biosimilars, especially if it contributed to shared savings measures. Similarly, education on the use of biosimilars as a way to enhance outcomes and reduce medical costs is increasingly important for other stakeholders such as integrated delivery networks and active employer groups such as the National Business Group on Health, as they can influence their pharmacy benefit managers or third-party administrators when designing benefit packages.

Even more interestingly, RWE can offer a way for biosimilars to enhance or hone product uptake either at the point of market entry or before. The new exploration of ways to incorporate this data extends across the drug development spectrum. For example, in the latter stages it can be used to optimize commercial spend effectiveness through targeted promotion to selected physician-patient segments at launch. An illustration would be understanding physicians who have high use of a certain agent but are not fast adopters of the biosimilar in order to provide evidence showing efficacy and safety in real-world use to allay concerns and enhance uptake. RWE can also be used to identify areas of greater need in current medical practice to guide market entry, perhaps related to a subgroup of patients with a specific medical history within the marketed indication. For biosimilars this could be one way to crack barriers through gaining experience with a payer in a relative niche in order to allay any mistrust and simultaneously gain demonstrative data on performance. Similarly, RWE can be used to gain insights into possible agreements with payers to deliver performance or bear some of the risk since it clearly indicates the current epidemiological set, costs, and alternatives available. However, this type of information is also patient-centric and can be used to understand the patient experience of treatment both in terms of making a market entry competitive through value-added services but also as a way to differentiate to improve the acceptance of biosimilars. One example of this is Samsung’s Benepali, which created an improved delivery device relative to the originator Enbrel in order to speed market entry with nurses and patients (Thakur, 2016).

Finally, with the investigation of the ways in which RWE can be included in new indications for existing products or even the granting of an initial approval, there is reason to think strategically about the type of evidence being produced and the ways it may permit increased acceptance and/or a chance to break through for biosimilar products. Physicians have requested additional information on all of the indications the originator product may have in order to feel comfortable with the issues of safety, immunogenicity, and efficacy. While most of the information has come from clinical trials in the past, there is a chance to use the results of real-world observations to indicate populations of special interest that are either not currently treated with the originator product or have enhanced outcomes when treated with the biosimilar outside of the initial patient population studied. The question currently being posed is what this evidence must look like. Do we need to have some level of control or blinding to consider RWE fit for purpose in regulatory decisions?

Barriers To Use Of Real-World Evidence

Despite the clear potential for RWE to influence healthcare in the U.S., there are major barriers to making full use of the data. In particular these relate to the potential for selection bias and confounding in observational data sets. Indeed, this is one reason the PDUFA IV requirements include the need for the FDA to hold a public workshop with key stakeholders (e.g., patients, industry, academia) by the end of 2018 and initiate (or fund) activities (e.g., pilot studies or methodology development projects) aimed at addressing key concerns and considerations in the use of RWE by the end of 2019 (Center for Drug Evaluation and Research, 2017).

While this is a concern, there are statistical techniques which adjust for biases and potential confounding, and several initiatives are underway to improve the confidence in RWE analyses. These include the International Society for Pharmacoeconomics and Outcomes Research and the International Society for Pharmacoepidemiology joint task force (Berger M. S., 2017) for good procedural practice, web resources such as the GetReal RWE Navigator (, as well as the Center for Medical Technology Policy’s RWE Decoder (, which seek to educate and provide resources for users of RWE studies.

One other issue in conducting RWE analyses relates to the dispersion of evidence across different data sets held for reasons specific to the data owner. Consider bringing the data on prescription claims together with the information from primary care calls, employment history, and hospital and clinic visits for the same patient across all healthcare systems they interact with, and you start to see the problem. But the unspoken issue is that even if you have all this data, to gain insights it needs to be harmonized (combined). Different data sources use different vocabularies, and interoperability (getting them talking to each other) can cost time and (lots of) money. Luckily, advances in Big Data analytics allow the integration of different data sets into a translated image which can be interrogated. This is where machine-learning solutions provide the most value; with artificial intelligence solutions, larger and more complex data sets can be connected and analyzed more rapidly.

While these techniques will provide one solution, we will also need to ensure they are employed in a responsible manner so we are not just data mining for information that will support our cause but truly representing the actual situation and implications of change. This last barrier will be probably the hardest to break down since we all seek to support our own arguments. However, it stands to reason that the same level of scrutiny employed for prospective RCTs will be engaged for retrospective (and prospective) observational data considered by both regulators and payers. Indeed, efforts such as the Duke-Margolis White Paper, the recent publication of a joint white paper between the Institute for Clinical and Economic Review and the Office of Health Economics on standards (Pearson, 2018), and the anticipated FDA framework should be clear steps in establishing the credibility of RWE research.


RWE has huge potential to improve the use of biosimilars both in placement and tailoring of therapy to individuals’ needs. Although challenges still exist in tapping the potential of RWE for manufacturers and other stakeholders in the healthcare market, the opportunity to do so is increasingly important. Biosimilars have to establish value over and above price, and the use of RWE to inform decisions and look for competitive differentiation makes this an area of research that should see increased investment in the coming years.


About The Author:

Anita Burrell, principal of Anita Burrell Consulting LLC, has 23 years of leadership experience in the pharmaceutical industry. She is a founding member of the Aurora Project, which seeks to explore and promote patient centricity in the industry, and is an adjunct professor at the Ernest Mario School of Pharmacy at Rutgers University. As a consultant, she has helped companies to: understand market dynamics and payer behavior, design big data platforms, develop predictive models to inform health technology assessment submissions, and understand the implications of digital health and possibilities to use behavioral economics. She holds a BA (Hons) in economics from the University of Stirling, an MA in economics from Dalhousie University, and an MBA from Kingston University.