In addition, many innovator biotherapies are (or soon will be) coming off patent protection. Just as companies entered the small-molecule drug world with generic versions of name-brand drugs, so companies are working to develop biosimilars or followon biologics that might be marketed once a name-brand biopharmaceutical comes off patent. Contract research organizations (CRO) are working with their biopharma clients to develop, optimize and validate methods that would be used in support of these products.
One of the most challenging areas, for sponsors and CROs alike, is in the functional analysis required to demonstrate the comparability of biosimilars to innovator products and to previous lots of the same product. The simple ANDA pathway to approval for small molecule generics will not work for complex biologics (For more on the concept of comparability, see Box)
Not only are biological products more complex than small-molecule drugs, they are also more prone to heterogeneity. Subtle differences may occur across product lots, resulting from the slightest differences in gene sequences used, or in conditions used in their production processes, including post-translational modification, cell passage and culture, production and purification.
Despite this inherent variability, global regulators, notably the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA), have the same strict expectations of quality for biosimilars that they have for small-molecule drugs. Any biosimilar or followon biologic submitted for regulatory approval must demonstrate comparable safety, purity, potency and stability to innovator product, and across product lots.
Comparability of potency and other critical parameters requires extensive characterization of the product’s physicochemical attributes, as well as assessment of its biological (functional) activity. Drug developers and their contract research organization (CRO) and analytical laboratory partners must evaluate and apply increasingly sophisticated tools and technologies to demonstrate biosimilar comparability.
CRO’s have the opportunity to support, or even lead the industry, in testing and applying new technologies.
Comparability Requirements Affect Analytical Method Development and Validation and Their Outsourcing
This article examines how these comparability requirements affect the outsourcing of analytical method development and validation, particularly for the functional analysis required, to demonstrate biosimilars and biopharmaceutical comparability. It highlights a bioassay method that PPD scientists In Middleton, WI, used to demonstrate the comparability of monoclonal antibodies from different sources.
The team decided to use a technology that capitalizes on the known cellular pathways that products target, using reporter genes to increase the robustness of several complex methods. Such assays have been applied toward the in vitro demonstration of MOA-directed potency for many biotherapeutics. Those that are sufficiently robust can be used to provide evidence of product consistency (lot release) and stability of products and comparability of biosimilar products to those of innovator products. PPD’s purpose was to demonstrate its application in comparing functional activities of anti-CD20 monoclonal antibody (mAb) products between manufacturers.
Therapeutic mAbs represent a growing group of biotherapeutics, and have been used successfully in the clinic to treat cancer and autoimmune diseases. Several innovator products are losing patent protection, giving rise to the development of biosimilars.
The known functional MOA’s of mAbs include binding to the target ligand through the variable antigen binding domains (Fab), other effector molecule binding functions through the constant (Fc) domain and triggering direct induction of apoptosis.
Effector functions that are triggered by Fc binding are antibody-dependent cell-mediated cytotoxicity (ADCC) and complement dependent cytotoxicity (CDC). This functional activity is part of the characterization performed to confirm consistency and reproducibility between lots, and is being applied to demonstrate comparability between innovators and follow-on products intended for licensure as biosimilars.
ADCC is a mechanism of cell-mediated immune defense in which an effector cell of the immune system, including Natural Killer (NK) cells, actively lyses a target cell. This results when a membrane-surface antigen on a target cell is bound by specific antibodies, and the Fc region of the antibody is bound by the Fc receptor on the effector cell. In vitro ADCC assays traditionally have used isolated primary peripheral blood mononuclear cells (PBMC) as the source for the NK cells. However, those bioassays generally show high variability, due primarily to both variable expression of Fc receptors and inherent genetic variability of these receptors between donors (mAbs 2012 4:p310).
Additional functional variation exists as a result of the differential affinities and cellular expression that occurs within the family of Fc receptors and their variants. Because patients receiving monoclonal antibodies possess genetic allotypes (VV, FF or FV) that are associated with different antibody binding affinities (and there are reports of this playing a role in the efficacy of at least one monoclonal antibody therapy used to treat neoplasms (Clin&Exp Immunol 2009, 157:p244)), it is important that data be collected to characterize the potential impact of these variants on biological product performance. Another source of variation results from differences in the type of glycosylation variation that can occur during production, post-translation (Blood 2008, 115:p2390).
Overcoming Variability in Cell Sources and Receptors
To overcome the variability resulting from both cell donor sources and Fc receptors, laboratories have developed modified cell lines that either enhance the NK ADCC response by down regulating NK cell inhibitory signals (J Immunol. 2008 180:6382) or transduce human Fc receptor and a luciferase reporter gene that allows modeling of the NFAT response elements of the IL-2 promoter to model PBMC derived ADCC (based upon the potency of antibody Fc effector function in ADCC pathway activation in effector cells, mAbs 2012 4:p310).
The second of the two formats has been shown to correlate to classic ADCC activity, and Promega, Inc., has developed this into a commercial bioluminescent cell-based bioassay. In this assay (Figure 1), Jurkat cell lines, stably expressing NFAT-luciferase reporter and human FcƴRIIIa (both variants have been transduced; currently, only the V158 variant is commercially available), are used to replace primary NK cells isolated from PBMCs as effector cells, while WIL2-S cells expressing the CD-20 antigen specific to the test products are used as target cells.
The ADCC RGA methods were qualified for precision (V158 variant mean RP 97.3, 5.2% CV; F158 variant mean RP 111.4, 9.8% CV), accuracy and linearity (V158 data, Table 1, Figure 2; F158 variant data not shown) and applied to support physicochemical and biological characterization of three anti-CD-20 biosimilar mAb products.
The results from using the V158 variant RGA were reported at the Biopharmaceutical Emerging Best Practices Association in Basel (BEBPA), Switzerland, in September 2013. That data demonstrated the capability of this method to differentiate between the functional activities of the three anti-CD-20 biosimilars, two being the same product manufactured in the U.S. (source 1) and Europe (source 2), and source 3 being that of a proposed biosimilar product (Figure 3, a and b).
Testing was extended to include qualification of and comparisons between the mAb sources using the F158 variant RGA. That comparison testing demonstrated that, regardless of the variant used, the Promega ADCC RGA was capable of differentiating potencies of the source 3 mAb from source 2, and, although the relative potencies were greater between source 2 and the two products than the source 2 product tested against itself, the results confirmed the much closer response similarity between the source 1 and 2 mAbs (Table 2). All product responses were evaluated using parallel line analysis, and passed for parallelism using the equivalence test for the difference of the slopes (data not shown).
Method Can Demonstrate Differences in Biological Activity
These results indicate that the ADCC method is capable of demonstrating that a measurable difference exists in the biological activity of the mAbs from these manufacturers. This supports data derived from other analyses, the first being Ion Exchange (CEX) HPLC showing different charges observed in mAb from source 3 relative to sources 1 and 2 (Figure 4). The second method was a comparison of the binding of the mAbs from all three sources to antigen target cells measured by flow cytometry, with the response EC50’s also demonstrating increased relative binding response of source 2 and 3, compared to sources 1 and 2 (Figure 5).
The Promega ADCC RGA technology appears to be capable of:
1) demonstrating differences in potency response between biosimilar products sourced from different manufacturers, in a format that is robust and reproducible; and
2) supporting other functional (binding to target antigen + cells) and physicochemical analysis of charge. This extends beyond characterization of the standard FcƴRIIIa binding and effector cell cytotoxicity responses to include applications designed to test for differences across FcƴRIIIa allotypes that have been shown to play a role in mAb therapeutic efficacy.
Other organizations, including Janssen, Catalent and Charles River Labs (IBC Bioassay Annual Meeting, Seattle, May 2013; 6th Annual BEBPA Meeting, Basel, Switzerland, September 2013) are applying this technological improvement for functional activity characterization in support of mAb-based biotherapy product regulatory submissions.
Based on published reports, this technology appears to allow for rapid and consistent functional performance testing applications that do not require either significant effort to optimize and stabilize NK cell lines (that might show extensive performance variability between cell lines and laboratories), nor are dependent upon sourcing donors that demonstrate adequately reproducible ADCC performance in either whole blood PBMC, or NK selected cell preparations.
A faster, more reproducible method could allow mAb manufacturers to monitor more effectively the functional activity across production lots and between biosimilar and innovator products. This could allow for much greater consistency in monitoring of responses by manufacturers and, perhaps more importantly, provide regulatory agencies responsible for determining the comparability and consistency of mAb-based biotherapies a better standard to use in new product submission reviews.
Delana Butz, Ph.D. Associate Research Scientist, Labs, PPD, joined PPD’s cell-based bioassay laboratory in 2010. She is responsible for the development, validation and maintenance of cell-based potency assays.
Manuela Grassi, Ph.D., Senior Group Leader, Biopharmaceuticals, PPD, oversees the development, validation and transfer of complex analytical methodology for bioassays. She joined PPD in 2006 and has more than 17 years of laboratory experience.
Onesmo Mpanju, Ph.D., Director, CMC Consulting, PPD Consulting, has more than 20 years of experience in drug development and more than 13 years in regulatory affairs. Prior to joining PPD, Dr. Mpanju was a product reviewer and microbiologist at the U.S. Food and Drug Administration.
Duu-Gong Wu, Ph.D., Director, CMC Consulting, PPD Consulting, has more than 20 years of regulatory and scientific experience at the U.S. Food & Drug Administration and in the pharmaceutical industry. In his current role, he has been assisting clients with the development of MAb and therapeutic protein biosimilars at various stages of development.
Peter Wunderli, Ph.D., Associate Director, cGMP Cell Lab, PPD, provides oversight and review of method development and validation and stability testing projects. He has more than 25 years of experience as a bench scientist and managing groups involved in supporting the development of a wide range of products and response testing methods and models.
Biological products include immune response modulators (both agonists and antagonists), anti-cancer treatments and hormone replacement or enhancement therapies. Each involves a very different mechanism of action (MOA), such as cellular proliferation or cytotoxicity, gene activation or suppression, activation or inhibition of cytokine and metabolite secretion, effector phosphorylation, and, for gene therapies, transfection directed cellular activation. Results can be measured either directly, through quantitation of mRNA expression, or indirectly, by determining the product’s activity on stimulating or suppressing targeted cellular activities.
Functional in vitro analytical methods are designed to model a product’s known or theoretical MOA. Using these methods correctly requires extensive knowledge of the molecule’s behavior on living cells. Drug development teams at sponsor companies and CROs must be experienced in applying cellular systems to measure a product’s specific activity, reproducibly and robustly, across a range of identified and potential future biological products.
Variability makes this a challenge--not only variability in the materials themselves, but in the analytical methods that are used to measure them, which can be influenced by variation across instruments, critical reagents, operators, and simple day-to-day and inter-lab differences. Most functional methods apply either primary cells collected from donors or cultured cell lines that are known to respond to the biotherapies being evaluated.
Whatever the cell source, the use of cells to measure response mechanisms results in variability over and above that observed in most other analytical assays, because cells are extremely sensitive, even to subtle changes in their environment. As a result, the compounded variation makes it almost impossible to generate a method that allows an absolute measure of potency to be determined.
Instead, most methods in use today demonstrate the relative potency of a reference standard to the test product lot. Such determinations are based on the premise that a product’s standard and test sample will exhibit similar potency characteristics, which, when tested using the same cells, reagents and materials, will result in response curves across a concentration gradient that will exhibit response parallelism (USP<1032>, 2010).
Over the past two decades, regulatory agencies such as FDA, and standards-setting organizations such as the U.S. Pharmacopeia (USP) and the International Council for Harmonization (ICH) have drafted or issued guidance documents to set expectations for generating and validating robust and appropriate methods. The latest recommendations were set four years ago, in USP 111, Chapters 1032, 1033, and 1034.
Interpretations still vary widely on how analytical methods are developed and validated. CROs, in general, allow clients to guide their interpretation.
Since 1984, under the U.S. Drug Price Competition and Patent Term Restoration Act (Hatch-Waxman), the FDA has permitted generic drug manufacturers to submit Abbreviated New Drug Applications (ANDA) to demonstrate that a proposed generic drug has equivalent quality characteristics and pharmacokinetic profiles relative to the original drug product without clinical efficacy data.
Such a regulatory pathway, however, does not apply to biological products, due to their molecular complexity and other safety concerns. In fact, until 1996, the FDA often required the conduct of pre-clinical and clinical studies following manufacturing changes to demonstrate comparability of the pre- and post-change biological product made by the same company.
The concept of comparability for biological products, based on a 1996 FDA document, permits a step-wise approach to assessing the quality, safety and efficacy of a highly purified, well-characterized biological product following manufacturing changes. Such a comparability exercise normally starts with the initial, extensive characterization to compare the quality attributes of the product before and after manufacturing changes.
Comparative pre-clinical, safety and efficacy studies only are performed when there are observed differences or uncertainties regarding the similarities in the quality attributes. FDA’s comparability concept was incorporated into the ICH Q5E guidance (2004).
The concept of comparability was further extended and applied to biological products and their respective biosimilars with the passage first in the EU of Article 10 of Directive 2001/83/EC and, more recently, the Biologics Price Competition and Innovation Act (2009) in the U.S., resulting in regulatory agencies applying a risk-based approach to reviewing submissions for licensure. Other jurisdictions around the world have enacted similar legislation.
EMA and draft FDA biosimilar product guidelines recommend (in addition to other requirements) that pre-clinical studies by manufacturers include in vivo or in vitro functional assays, particularly for products with known or hypothesized MOA.
Regulators depend on the review of quality comparability data packages for proposed biosimilars in order to determine the amount of residual uncertainty toward demonstrating a high degree of similarity with the reference innovator product. The level of residual uncertainty determines the amount of expensive and time-consuming pre-clinical and clinical in vivo studies needed to demonstrate similarity between the innovator and the proposed biosimilar.