September 2016
Intellectual Property & Technology Law Journal
By Jeffrey M. Jacobstein; Amanda K. Murphy, Ph.D.; Leslie A. McDonell
Authored by Jeffrey M. Jacobstein, Leslie A. McDonell, and Amanda K. Murphy, Ph.D.
Antibodies are proteins produced in nature by the immune system and used frequently in biologic therapies for their ability to bind and alter signaling pathways involved in various diseases such as cancer, infection, and inflammation. Antibodies recognize particular target agents, called "antigens," which are bound via variable domains at the end of the antibody. The specific portion of an antigen bound by an antibody is referred to as an "epitope." Two antibodies that bind the same or similar epitopes may compete for binding, meaning the presence of the first antibody bound to the antigen prevents the second antibody from accessing and binding its epitope on the same antigen.
When drafting an antibody patent application, a question often arises of whether to include claims that define the antibody by its epitope or competitive binding properties, in addition to claiming the antibody by its amino acid sequence. The answer, as with most questions in patent law, is that it depends. It depends, mainly, on the amount of data available at the time of filing, or the likelihood that sufficient data will be generated later that could support an epitope claim. If crystal structures, alanine scanning, deletion studies, binning experiments, or other work that sheds light on epitope and competitive binding are available, then including that information in the patent application may be useful, along with some associated claims. While it is still an open question of how well a claim to any antibody defined by its epitope or competitive binding properties will stand up to challenge, a claim supported by extensive testing might significantly improve the chances of validity and, at the same time, sizably expand the scope of protection around important commercial products.
A recent case in the District of Delaware, Amgen Inc. et al v. Sanofi et al.,1 underscores this point.
The plaintiff, Amgen, developed Repatha™ (evolocumab), a monoclonal antibody targeting PCSK9 (proprotein convertase subtilisin/kexin type 9) for the treatment of high cholesterol. PCSK9 is an enzyme found in the liver, among other tissues, and binds to receptors including those for low-density lipoprotein (LDLR). The FDA approved Repatha on August 27, 2015, as an adjunct to diet and statin therapy to reduce LDL in certain patient populations.
Amgen was not the only company working to develop PCSK9 inhibitors to treat high cholesterol. Among other competitors, Sanofi and Regeneron collaborated on an anti-PCSK9 antibody, Praluent™ (alirocumab), which the FDA approved on July 24, 2015, for the treatment of high LDL cholesterol in adult patients.
Amgen subsequently sued Sanofi and Regeneron for marketing their PCSK9 antibody. Although Praluent was not alleged to have the same amino acid sequence as Amgen's antibody, it was asserted that the antibody bound an overlapping epitope and blocked binding of PCSK9 to LDLR.
Amgen is the owner of two patents directed to anti-PCSK9 antibodies, U.S. 8,829,165 and U.S. 8,859,741. The patents include claims to monoclonal antibodies that bind to particular residues on PCSK9 and block LDLR signaling. Claim 1 of the '165 patent is representative:
1. An isolated monoclonal antibody, wherein, when bound to PCSK9, the monoclonal antibody binds to at least one of the following residues: S153, I154, P155, R194, D238, A239, I369, S372, D374, C375, T377, C378, F379,V380, or S381 of SEQ ID N0:3, and wherein the monoclonal antibody blocks binding of PCSK9 to LDLR.
Amgen's two patents share a common specification, which provides testing data on, inter alia, the epitope contact residues of two antibody clones that prevent LDLR signaling. Epitope contact residues were determined for the two antibodies using X-ray crystallography, and from that information a "sweet spot" on PCSK9 was allegedly identified that disrupts LDLR signaling after antibody binding. Amgen also ran "binning" experiments for a series of additional antibodies they generated to identify those that competed for binding. Antibodies that competed with the two sequenced clones were considered to have overlapping epitopes.
At trial, the defendants alleged the claims were invalid for lack of written description and enablement because they covered a large genus of antibodies without providing sufficient examples or structural details of antibodies targeting the correct epitope. The defendants also argued the claims were obvious over prior art antibodies directed to PCSK9. Despite these arguments, the jury ultimately decided on March 16, 2016, after three days of deliberation, that the asserted claims were valid (infringement was previously conceded).
In post-trial motions, the defendants renewed their challenges to the written description and enablement of the asserted patent claims. Under the written description test laid out by the U.S. Court of Appeals for the Federal Circuit, a patent must provide (1) "a representative number of species falling within the scope of the genus" or (2) "structural features common to the members of the genus" sufficient for a skilled artisan to "visualize or recognize the members of the genus."2
The defendants argued that Amgen's patents do not provide sufficient representative species because crystal structures were only obtained for two antibodies, and the identified epitope contact residues for those two antibodies did not span the complete "sweet spot" asserted by Amgen. While the patents tested several other antibodies that competed for binding, the defendants noted that the actual epitopes for those additional antibodies had not been identified. They also pointed out that the exemplified antibodies were not "representative" because the skilled artisan could not gain any information from them on how to make and use any other antibodies that bound the same epitope, disrupted LDLR signaling, and thus fell within the scope of the claims. The defendants further argued that the specification failed to provide any common structural features for the claimed genus of antibodies, as the exemplary antibodies and their identified epitope contact residues would not provide the skilled artisan with the necessary structural features to design any other antibodies binding the claimed epitope.3
The defendants also responded to an argument raised by the patentee that Federal Circuit case law and U.S. Patent and Trademark Office (USPTO) guidance implicitly sanction broad antibody claims when a new protein or epitope has been discovered. Relying on Centocor Ortho Biotech, Inc. v. Abbott Laboratories,4 and the USPTO's "Written Description Guidelines," the patentee had argued that discovery of a newly characterized antigen (or in this case, an epitope) entitled them to claim any antibody that would bind that antigen without needing to specify structural or functional features. The defendants dismissed that argument as relying on nothing more than dicta from distinguishable cases and nonbinding USPTO interpretations of the law. They also noted that Amgen had not identified a new antigen (PCSK9 itself was a known protein), but rather a "sweet spot" in the protein for disrupting LDLR signaling. Thus, the defendants appear to reject the possibility of claiming based on the discovery of effective epitopes without providing further structural guidance.
In addition to addressing written description, the defendants also renewed their enablement challenges in the post-trial motions. In order to satisfy the enablement requirement, a patent must "teach those skilled in the art how to make and use the full scope of the claimed invention without undue experimentation."5 Although they provided less extensive arguments on this ground, the defendants again questioned whether two examples were sufficient to enable the skilled artisan to make and use any other antibodies that would bind the claimed epitope and disrupt LDLR signaling. The defendants supported this position by pointing to the expert testimony given at trial indicating that the skilled artisan could not extrapolate from the known epitope to the structural details of any other antibody targeting that epitope.
Instead, the defendants alleged, the skilled artisan would have to make additional antibodies arbitrarily and test for those that bound the correct epitope and disrupted LDLR signaling (e.g., by checking for competition with the two characterized antibodies). This process, they alleged, would be unguided by the disclosure in Amgen's specification, imposed undue burden on the skilled artisan, and thus the claims were a mere invitation to experiment.
While the outcome of the Amgen case could still change if the defendants prevail on their post-trial motions or on appeal, the initial jury decision highlights the potentially broad reach of epitope claims. The defendants in this case had developed their own antibody that provided a unique variable domain structure distinct from that in Amgen's commercial product. Traditional antibody claims directed to Amgen's binding domain regions (such as complementarity determining regions or full variable domains) likely would not have covered the defendants' antibody. Yet broader epitope and competitive binding claims helped expand the scope of protection around an innovative antibody-based therapy.
Moreover, such claims are not only possible to obtain from the patent office, as demonstrated by Amgen's granted epitope claims in this case, but they also have the apparent ability to withstand challenge. In part, this may depend on the amount of epitope testing in the application, including X-ray crystal data and competitive binding experiments.
Thus, the outcome of this case (at least at the trial court level) demonstrates the importance of including a solid evidentiary basis for epitope and competitive binding claims when drafting antibody applications. Patent practitioners should work with their inventors to obtain as much data as possible on epitopes and competitive binding for their newly-developed antibodies, along with data showing any other functional aspects of the constructs, and consider carefully whether to include that data in the application. If presented properly, the data may support broader protection for a commercial product.
With carefully designed studies and binding data in place, a defendant faced with epitope claims may have no alternative but to directly challenge the underlying case law basis for allowing any such claims. This appears to be the strategy employed by the defendants in the current case. It remains to be seen whether written description and enablement challenges will be successful in the long run when they rely on arguments about the amount of effort required to obtain the structure of an antibody from a known epitope. Particularly as high-throughput screens and other advanced techniques become more common in antibody drug development, this type of challenge based on the uncertainty and undue burden of rational antibody design may only become harder to maintain.
Endnotes
1 Amgen Inc. et al v. Sanofi et al., DED-1-14-cv-01317.
2 Ariad Pharm., Inc. v. Eli Lilly & Co., 598 E3d 1336, 1353 (Fed. Cir. 2010).
3 See, e.g., AbbVie Deutschland GmbH v. Janssen Biotech, Inc., 759 E3d 1285, 1301 (Fed. Cir. 2014).
4 Centocor Ortho Biotech, Inc. v. Abbott Labs., 636 F. 3d 1341 (Fed. Cir. 2011).
5 Genentech, Inc. v. Novo Nordisk A/S, 108 F.3d 1361, 1365 (Fed. Cir. 1997).
Originally printed in Intellectual Property & Technology Law Journal in September 2016. This article is for informational purposes, is not intended to constitute legal advice, and may be considered advertising under applicable state laws. This article is only the opinion of the authors and is not attributable to Finnegan, Henderson, Farabow, Garrett & Dunner, LLP, or the firm's clients.
Lecture
Patent Protection for Software-Related Inventions in Europe and the USA Training Course
June 5, 2024
Hybrid
Workshop
Life Sciences Workshop: Updates and Key Trends in Pharmaceutical and Biotechnology IP Law
May 2, 2024
Cambridge
INCONTESTABLE® Blog
April 19, 2024
Due to international data regulations, we’ve updated our privacy policy. Click here to read our privacy policy in full.
We use cookies on this website to provide you with the best user experience. By accepting cookies, you agree to our use of cookies. Please note that if you opt not to accept or if you disable cookies, the “Your Finnegan” feature on this website will be disabled as well. For more information on how we use cookies, please see our Privacy Policy.
Finnegan is thrilled to announce the launch of our new blog, Ad Law Buzz, devoted solely to breaking news, developments, trends, and analysis in advertising law.