Thank you for your interest in CCO content. As a guest, please complete the following information fields. These data help ensure our continued delivery of impactful education.
Become a member (or login)? Member benefits include accreditation certificates, downloadable slides, and decision support tools.
To discuss biosimilars and their importance for oncology practitioners, it is important to define what a biosimilar is. The FDA defines a biosimilar as a biologic product that is highly similar to a US‑licensed reference biologic product for which there are no clinically meaningful differences in safety, purity, or potency compared with the reference biologic.
The development of biosimilars is one answer to the rising cost of healthcare, similar to the development of generic versions of small molecule pharmaceuticals. US spending on prescription drugs was estimated at $457 billion in 2015, accounting for 16.7% of overall personal healthcare spending. More than one third of spending on therapies is concentrated in the top 5 therapeutic cost areas: oncology, diabetes, mental health, respiratory diseases, and pain management.
Oncology spending is at the top of the list; it has been the highest for approximately the past 2 decades, and it is continuing to grow at a rapid rate. US spending on oncology-associated drugs increases at an average of 12% to 15% annually and is expected to reach $100 billion by 2022.
As we all know, healthcare costs in the United States have risen more rapidly than the US gross domestic product. The cost of medical care for cancer has risen even more precipitously, largely driven by the costs of delivering modern cancer therapies. US spending on the top therapeutic agents for multiple diseases reaches hundreds of millions of dollars per agent annually, with many oncology agents among the costliest.
The cost of cancer therapeutics is rising globally as well. In the past decade, costs of both oncology therapeutics and supportive care agents have risen sharply, with increases in the billions of dollars.
Another indication of the rapid increase in therapeutic cost is to consider the prices of these agents at the time of their FDA approval. Bach and colleagues analyzed the monthly and median costs of each cancer drug in the year it was approved. From 1965 through 1999, the cost of cancer drugs at approval increased at a fairly slow rate. Since 2000, that increase has risen very steeply. Modern cancer therapeutics now average well over $10,000 per month at the time of FDA approval. In most instances, the cost of these drugs continues to rise after approval when they are in the marketplace.
From the patient perspective, it is important to compare the costs of cancer care relative to household income. Median monthly household income in the United States has changed very little during the past 4 decades. Since the steep rise in launch price of new cancer drugs beginning at approximately Year 2000, median monthly cost of those agents has exceeded monthly median income and is now several‑fold higher than median household income. This poses an enormous challenge for patients and their families as they must weigh the opportunity to receive one of these new but very costly therapeutic agents against the ability to afford it, especially if they are underinsured or if, as frequently happens, the agents are not fully covered by their insurance plan. These challenges can inhibit objective conversations about the quality, costs, and value of oncology care.
The late‑phase oncology pipeline is experiencing rapid growth, heavily driven by targeted biologic agents, which have gained inroads into modern oncology treatment across a wide range of cancer diagnoses. Traditional cytotoxic agents have not changed much but now represent a smaller share of the pipeline than they did 15 years ago. Hormonal and radiotherapies remain a small piece of the treatment pipeline, although radiotherapeutics are slowly increasing in the United States.
Perhaps it is no surprise, then, that spending on biologic therapies has also increased precipitously, even faster than overall spending growth, accounting for 70% of drug-spending growth from 2010-2015. Global biologic sales are thought to represent approximately 20% of the global market value of therapeutics. These amounts continue to increase, raising concerns not only about the financial burden of these therapies to society but also whether high costs are limiting patient access to these effective but costly treatments.
Biosimilars are seen as a promising strategy to counter these rising healthcare costs, especially for biologic agents. We know that healthcare systems vary widely in the cost and quality of cancer care delivered, even within the United States, and we know that patients are bearing an ever‑increasing share of that cost: As copays go up, the caps on coverage get lower, shifting more of the cost of care to the patient and their family. This is particularly true for cancer care. These costs can affect whether a patient agrees to treatment or continues on treatment, affecting patient outcomes and survival.
It is important to understand that biologic agents are very different from small molecule drugs. Small molecules have a low molecular weight and a simple, well-defined structure that can be reproduced exactly in the laboratory through chemical synthesis. They are stable and most do not produce an immune response. Conversely, biologic therapies, such as monoclonal antibodies, are large, complex, high-molecular-weight molecules that can be very heterogeneous in structure. Because they are produced by living cells and organisms, they are impossible to replicate exactly through synthetic manufacturing. These molecules are often inherently unstable, are sensitive to external conditions, and have high potential for inducing an immune response because they are large, complex proteins.
The FDA has created a regulatory approval process for biosimilars that aims to encourage and facilitate the development and introduction of biosimilars to provide competition for costly biologic agents. We are familiar with the approval process for a reference biologic: molecular characterization and preclinical work, usually in animal studies, followed by phased clinical trials—phase I dose‑finding studies, phase II to evaluate efficacy and safety signals, and then definitive phase III pivotal clinical trials to compare the biologic with the standard of care in a given disease entity. For approved agents, postmarketing surveillance is often used to identify any delayed or unexpected adverse events that did not emerge in the clinical trials. Requiring the same level, extent, and depth of evidence for a biosimilar as was required for the original biologic could lead the cost of developing a biosimilar to be the same or even higher than the original biologic. Not only would this provide little motivation to develop biosimilars, but the resulting agents would be unlikely to cost any less than the original biologic therapy.
Instead, the FDA decided to reverse the relative contributions to clinical predictability of the steps in biosimilar approval, putting most of the focus on molecular characterization. Of importance, laboratory analytic capacity to characterize molecular structure and function has improved dramatically during the development of this process, and the FDA examines a wide range of molecular features to assure that the structure and potential function of the biosimilar molecule are highly similar to the original biologic molecule. A biosimilar must undergo preclinical, pharmacokinetic, pharmacodynamic, and immunogenicity testing and basic clinical studies. Comparative clinical studies may be performed to confirm that efficacy and safety are similar to the original biologic. Postmarketing surveillance is generally required for approved biosimilars.
The structural characterization and preclinical work form the primary basis for regulatory review and approval of biosimilars, with early clinical trials conducted if residual uncertainty remains. Ultimately, the FDA needs to be convinced that there are no clinically meaningful differences between the reference biologic and the new biosimilar product in terms of safety, purity, and potency.
This forms a stepwise building process starting with analytical studies. At each step, the FDA evaluates the totality of evidence to that point to determine whether there is a need for additional studies to eliminate residual uncertainty about the similarity of the biosimilar to the original product. Although clinical studies may be done to increase confidence in the similarity of these biosimilars, there is certainly no requirement for multiple large phase III randomized trials, which often drive much of the expense of developing new therapeutics in oncology.
One different aspect related to biosimilars is the concept of extrapolation. This means that the data used to support approval of a biosimilar for one indication can be used to support its approval for other indications for which the reference product is licensed, even in the absence of additional studies demonstrating efficacy in those other indications.
Extrapolation is intended to simplify the approval process so that the cost of developing biosimilars does not meet or exceed the cost of developing the original agent. However, extrapolation to other indications does not occur automatically upon approval of a biosimilar. Convincing evidence, such as preclinical or clinical data, must be provided to support the rationale for extrapolation to a reference biologic’s other approved indications.
Scientific justification is required for extrapolation to each additional indication, with attention given to 4 main aspects. First, is the mechanism of action for a new indication the same as for the original? For example, are the binding sites and molecular signaling likely to be the same? Does the target or receptor show similar expression as in the original indication? Second, is there any reason to think that the populations for the original indication and the extrapolated indication would demonstrate any differences in pharmacokinetic or pharmacodynamic measures? Would biodistribution likely be the same? Third, is there any reason to think that toxicities would be different for the original indication vs the extrapolated indication? This often depends on the extent of differences between the patient populations for the original and the extrapolated indications. And finally, are there other factors, such as comorbidities or concomitant medications, that are likely to be similar or different between the populations for the original and extrapolated indications? If the FDA finds sufficient evidence that the biosimilar is likely to behave similarly in the extrapolated indication, it can grant approval. To date, several biosimilars have received extrapolation approval to additional indications in the United States.
Another concern, and it is important to note here that this applies not only to biosimilars but to reference biologics as well, is the variability and drift of these molecules over time or between different production locations. These large, complex molecules may be produced at different sites and different factories; any change in the manufacturing processes or the components that go into production could lead to variability in the end product. Approval by the FDA or European Medicines Agency (EMA) is required for any changes in the manufacture of both reference biologics and biosimilars, and manufacturers themselves must monitor production with the assumption that any changes can result in clinically significant changes in the product.
There are many different potential sources of variability for biologic molecules, including both reference products and biosimilars. They are produced in living systems or organisms, with various fermentation conditions and protein purification processes. There are opportunities for variation to occur throughout purification, potency, and activity studies and packaging and sterility testing. Variation over time can lead to drift. It is important to evaluate whether the end product is the same as when the originator was approved and whether any variation affects the efficacy and safety of the end product.[15,16]
A recent study by Pivot and colleagues that I participated in demonstrates what this variability and drift can look like in practice. We conducted a 1:1 randomized, controlled clinical trial comparing trastuzumab and a recently developed trastuzumab biosimilar product, SB3, in 367 patients with early‑stage HER2-positive breast cancer.
Early results from this trial suggested a trend toward greater efficacy of the biosimilar compared with trastuzumab. This actually raised concerns because a biosimilar should be neither more nor less effective than the originator. Higher efficacy could require regulatory approval as a new originator.
To better understand this, the investigators sought to identify any change in the originator over time. They obtained batches of both US-produced and EU-produced trastuzumab from 2012 to 2019 and compared a range of quality parameters. The products are supposed to fall within a designated range for each parameter. As illustrated by one parameter, antibody‑directed cell‑mediated cytotoxicity (ADCC), which is thought to be a mechanism of action of anti‑HER2 effects of trastuzumab, all the batches up until about 2018 fell within the target boundaries.
Surprisingly, batches of trastuzumab from both the European Union and the United States dating to late 2018 and early 2019 fell below acceptable measures. This apparently went unnoticed—or at least unreported. More recently, ADCC appears to have come back up, at least in the US version, but for perhaps a year or longer, ADCC activity fell below what the FDA has established as acceptable levels.
In the 3‑year follow-up of these patients, researchers analyzed the subgroup of patients who received at least 1 dose of the “drifted” version of trastuzumab. In both event‑free survival and OS, the patients who always received undrifted trastuzumab had virtually the same outcome as patients who received the biosimilar SB3. However, patients who received 1 or more doses of the lower‑performing, drifted trastuzumab had a higher likelihood of recurrent disease (HR: 0.47; 95% CI: 0.26-0.87) or death.
This was a retrospective, post hoc subgroup analysis and should be confirmed a priori. However, it demonstrates that not only is drift of biologics real, but it can occur in either a new biosimilar or in an original biologic therapy.
I believe this is a wake-up call to all of us, including clinicians, manufacturers, and regulatory agencies, that we need to monitor the performance of these products, reference or biosimilar, over time.
Immunogenicity is an important consideration with biosimilars since, as mentioned, these are complex, large proteins that can produce an immune reaction. As with variability and drift, immunogenicity is a concern of all biologics, both original biotherapies and biosimilars. If an agent triggers production of neutralizing antibodies that affect the critical sites and the mechanism of action, it can lead to loss of efficacy and possibly a generalized immune reaction such as a rash or even anaphylaxis. The FDA requires monitoring immunogenicity at the time of development and over time to account for drift or a change in molecular characterization. The likelihood of an immune response to a biosimilar does not need to be zero, but in a head-to-head comparison, it should be no greater than the likelihood of immune response to the original biologic.
Although it has not yet gained much traction in cancer therapy, biosimilars are subject to interchangeability or substitution—that is, switching from an original biologic to a biosimilar or between 2 biosimilars with no change in result. The FDA has established a regulatory process to obtain a higher level of regulatory approval as an interchangeable biosimilar. It requires additional clinical data to show that switching back and forth between products has no meaningful impact on efficacy or safety. The EMA has not yet created any process to approve agents as interchangeable.
One component of interchangeability is that an interchangeable biosimilar could be substituted for an originator without notifying the provider or the patient of the change. Concerns from clinicians and state regulators have led nearly 40 states to pass state substitution laws that would allow substitution to occur if a product has obtained interchangeable biosimilar status but would require notification of the provider that the change has occurred. At the heart of this issue is transparency about substitution. The clinician and, I believe, the patient should be aware that the treatment has changed.
To date, no biosimilar has been granted interchangeable designation. It is unclear whether manufacturers will be likely to pursue this designation; of course, there is considerable additional cost for conducting the additional clinical trials and developing the evidence base required. In addition, observational studies and postmarketing regulatory studies tell us that switching is already occurring. Providers may have different preferred agents and patients may be given a different form of the biologic if they change practices or locations. To date, no evidence of a loss of efficacy or safety has emerged based on that switching. With little incentive for the industry to make the effort to obtain regulatory approval as an interchangeable biosimilar, it remains unclear whether this designation will be used in the future.
To summarize, an interchangeable biosimilar must demonstrate the same clinical result as the reference biologic and that changing from the originator to the biosimilar or vice versa produces no meaningful change in efficacy and safety. And again, this higher standard of regulatory approval has yet to be granted to any biosimilar in the United States.
The FDA established a naming process for biosimilars to aid pharmacovigilance and postmarketing surveillance and reporting of any unexpected complications. All FDA-approved biosimilars have a core, nonproprietary name of the agent, followed by a random 4‑letter, nonmeaningful, nonproprietary suffix. The only exception to this is the first biosimilar approved in the United States, filgrastim‑sndz, which was named before the FDA nomenclature was in place. Every new biosimilar will have a unique name so that it can be tracked over time in the postapproval process.