What is Next for Enablement and Written Description of Antibody Claims?

by Dennis Crouch

I had been following the case of Teva v. Lilly for a few years.  Teva has traditionally been a generic manufacturer, but in this case sued Eli Lilly for infringing its patents covering methods of treating headache disorders like migraine using humanized antibodies that bind to and antagonize calcitonin gene-related peptide (CGRP), a protein associated with migraine pain. U.S. Patent Nos. 8,586,045, 9,884,907 and 9,884,908.  The patents cover Teva’s drug Ajovy, and allegedly cover Lilly’s Emgality. Both drugs were approved by the FDA in September 2018.  A Massachusetts jury sided with Teva and awarded $177 million in damages, including a controversial future-lost-profit award.

Although the jury sided with Teva, District Judge Allison Burroughs rejected the verdict and instead concluded that Teva’s patent claims were invalid as a matter of law for lacking both written description and enablement.  The appeal is now pending with Teva is asking the Federal Circuit to reinstate the verdict.  Lilly filed its own conditional cross appeal – arguing that the future lost profit award was not supported by the evidence or permitted by law.  The case has direct parallels to the Supreme Court’s 2023 decision in Amgen v. Sanofi, but also includes some distinguishing features.

• Teva v Lilly: DCT Judgment And Memo
• Teva v. Lilly: Opening Brief
• Teva v Lilly: Response Cross Appeal

District Court’s Invalidity Ruling

Although a jury sided with Teva, Judge Burroughs held that no reasonable jury could have done so. In particular, the court concluded that Teva’s patents failed to adequately describe the broad genus of antibodies claimed, because the specification discloses only a single antibody, “G1,” that falls within the asserted claims. The court emphasized the “very broad scope” of the claims, which “cover a method of treating headache using anti-CGRP antagonist antibodies,” while the specification does not “identify any narrower amino acid sequence or structure of antibodies possessing those functions.”  Dependent claim 30 is the key claim, it depends from claim 17:

17. A method for reducing incidence of or treating headache in a human, comprising

administering to the human an effective amount of an anti-CGRP antagonist antibody, wherein said anti-CGRP antagonist antibody is a human monoclonal antibody or a humanized monoclonal antibody.

30. The method of claim 17, wherein said anti CGRP antagonist antibody is a humanized monoclonal antibody.

Summarizing, what we have here is a 1-step claim requiring administering an effective amount of an anti-CGRP antagonist humanized monoclonal antibody.  If you recall from the Supreme Court decision in Amgen v. Sanofi.

Relying on Federal Circuit precedent that functionally-defined antibody genus claims require disclosure of either “a representative number of species falling within the scope of the genus or structural features common to the members of the genus,” AbbVie Deutschland GmbH & Co., KG v. Janssen Biotech, Inc., 759 F.3d 1285 (Fed. Cir. 2014), the court found Teva’s patents lacking on both fronts — only 1 species and no common structural features beyond those generally seen in humanized monoclonal antibodies. Although Teva’s patents did disclose several anti-CGRP antibodies and variants of G1, the court found these were not representative species because they fell outside the claim scope, which is limited to humanized antibodies. The court also found that Teva’s inventors had failed to make any other humanized anti-CGRP antagonist antibodies besides G1.

The district court also found it relevant to compare the differences between the disclosed G1 and Eli Lilly’s accused antibody “galcanezumab.” Those differences further supported the lack of representative species in the court’s view, because galcanezumab binds to a different region of CGRP and succeeded in treating cluster headache where G1 failed, demonstrating the structural and functional diversity of the claimed genus.  With regard to common structural features, the court rejected Teva’s arguments that all members of the genus share a “Y-shaped” antibody structure, complementarity to CGRP, and humanization, finding these are features of all antibodies and not specific structures that confer CGRP antagonism.

Additionally, the court held the patent claims were not enabled because the specification provides “nothing more than a ‘roadmap’ for a ‘trial and error’ process to identify and make antibodies within the scope of the Asserted Claims.” The court noted uncontroverted testimony that the number of candidate antibodies falling within the “extraordinarily broad” genus was “‘unknowable’ and thus not necessarily very large or small,” and that a skilled artisan would need to test each one to determine if it could antagonize CGRP, an iterative process that would take months and significant cost per antibody. See, Amgen Inc. v. Sanofi, 598 U.S. 594 (2023) (the roadmap amounts to little more than a research assignment).

Teva’s Arguments on Appeal

There is an important difference between Teva’s claims here and those in Amgen v. Sanofi. In Amgen, the claims were directed to a broadly claimed genus of antibodies. In Teva, however, the invention is a novel method of using antibodies that were (arguably) already well-known.  Teva explains:

This Court has therefore never deemed a method of treatment claim using an already well-known class of antibodies invalid on written description grounds due to an allegedly inadequate disclosure of the underlying antibodies. . . . [P]recedents from this Court and its predecessor demonstrate that method claims employing an already well-known genus should not be treated the same as claims involving a novel genus.

Teva relies on cases like Ajinomoto Co. v. International Trade Commission, 932 F.3d 1342 (Fed. Cir. 2019), and In re Herschler, 591 F.2d 693 (C.C.P.A. 1979), as more analogous than the novel antibody cases the district court invoked.  In Ajinomoto, the critical component of the invention was not the discovery of new promoters but the application of known promoters to increase the production of aromatic L-amino acids through genetically modified E. coli.  Thus, extensive disclosures of the promotors was not required to satisfy the 112(a) requirements.   Similarly, in Herschler, the CCPA concluded that the claimed method enhancing steroid penetration did not require extensive disclosure of the underlying steroids themselves.

Teva’s argument here is a bit of a stretch because, although murine (mouse) anti-CGRP antibodies were well known, the humanized versions were not. And, the asserted claims require the humanized antibodies. In its briefing, Teva brings this together by arguing that the court wrongly disregarded the murine anti-CGRP antibodies disclosed in the specification and prior art:

The district court treated humanized anti-CGRP antagonist antibodies as a novel genus of antibodies, of which antibody G1 was the only known species. . . . The court refused to consider the many murine anti-CGRP antagonist antibodies already known to a POSA, even though humanizing them would have been routine and a POSA would know that they all would treat headache, because they had not actually been humanized. . . . That was error.

Teva Brief.  Teva also points out that Eli Lilly, in its IPR petitions against some of the patents, successfully argued the prior art was “replete with exemplary disclosures of anti-CGRP antagonist antibodies” that were “well known,” and Eli Lilly’s own expert admitted humanization was “routine” as of the invention date. Given this evidence, Teva contends that a reasonable jury could find practicing the full scope of the claims had been both described and enabled without requiring undue experimentation.

Eli Lilly’s Response

On written description, Eli Lilly argues the claims “broadly cover methods of treating headache using any humanized anti-CGRP antagonist antibody” yet the patents “disclose[] only one humanized anti-CGRP antagonist antibody.”  Eli Lilly contends Teva cannot rely on the unclaimed murine antibodies as representative species to make up for this shortcoming – especially for the written description requirement –  because they “are not within the scope of the claims, and thus cannot be representative.”  Lilly goes on to argue that this requirement persists regardless of whether the claim is an antibody composition or method of treatment claim. And while Teva argues humanizing the murine antibodies would have been routine, Eli Lilly asserts “whether a skilled artisan could make and use the claimed subject matter is irrelevant to written description if the specification does not contain sufficient disclosure showing that the inventors possessed what they claimed.”  On enablement, Teva argues that even if humanization of antibodies is routine, this does not eliminate the need to synthesize and screen each candidate to determine if it binds and antagonizes CGRP and “effectively” treats headache as claimed.

Finally, Eli Lilly’s conditional cross-appeal argues that the future lost profits award was speculative and therefore should be rejected. The brief does not reject the potential that future lost profits could be awarded in an appropriate case, but notes that past lost profits already requires a high standard of proof that is made much more difficult without the 20-20 vision of hindsight.  Focusing on the evidence presented, Eli Lilly particularly notes how Teva’s damages expert’s analysis was forced to revise his initial forecast down by over 50% in one year due to changing market dynamics, and uncertainty remained as to whether his revised estimate was reliable.

Although it will depend upon the particular panel, my view is that the Federal Circuit is likely to affirm the written description and enablement holdings — making incrementally even harder to protect genus claims for biotech innovations.