Teva v. Lilly: Conflicting Positions at PTAB and District Court Sink Patent Challenge

Need to know

The Federal Circuit has revived Teva’s headache-treatment claims and highlighted a recurring litigation risk: Arguments that win in one forum can backfire in another. 

In Teva Pharms. v. Eli Lilly, No. 2024-1094 (Fed. Cir. Apr. 16, 2026), the U.S. Court of Appeals for the Federal Circuit reversed the district court’s grant of judgment as a matter of law (JMOL) invalidating Teva’s method‑of‑treatment claims under 35 U.S.C. § 112 for lack of written description and enablement. The court emphasized that when the claimed invention is a new use of a well‑known antibody genus, disclosure of representative species — along with established techniques for making additional species — can be sufficient to satisfy written description and enablement.

The court’s holding underscores the importance of taking a long-term approach to litigation strategy. Lilly’s earlier § 103 obviousness arguments in related inter partes reviews (IPRs) proved difficult to reconcile with its later § 112 attack in district court. Teva capitalized on this tension, and the Federal Circuit took note. 

Background: IPRs plus parallel district court litigation  

In 2018, Lilly filed multiple IPR petitions challenging Teva patents covering (i) humanized antibodies that bind and inhibit a protein called CGRP (“antibody patents”) and (ii) methods of using these antibodies to treat headaches (“headache patents”). Anti-CGRP antibodies exist in mice, and converting non-human antibodies into human antibodies is called “humanization.” Before this appeal, the Federal Circuit had already reviewed the IPRs and affirmed the Patent Trial and Appeal Board (PTAB) findings that the antibody patents were headache patents survived, but affirmed that the antibody patents were unpatentable as obvious under § 103, but the headache patents were not.

In parallel with the IPR proceedings, Teva sought judgment in the District of Massachusetts that Lilly’s Emgality® product infringed the headache patents. The case turned on method claims that cover administering an “anti‑CGRP antagonist antibody” (including humanized antibodies) to treat headaches. Two representative claims are below:

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.

After the Federal Circuit decisions affirming the PTAB, the case proceeded to trial. A jury found willful infringement, and Lilly moved for JMOL based on a lack of written description and enablement. As to written description, Lilly argued that the claims are impermissibly broad because:

• Teva claims a broad genus of humanized anti-CGRP antibodies defined by their function (treating headaches) rather than structure, which could include countless species of antibodies.
• Teva only disclosed and was only in possession of a single representative species (“G1”) within this broad genus.
• Teva failed to disclose structural features common to the claimed genus of humanized anti-CGRP antibodies that would allow a person of skill in the art to recognize the antibodies of the genus.
• Teva had been unsuccessful in making additional species of humanized anti-CGRP antibodies and the disclosure of a single species constitutes a “research plan.”  

The district court granted JMOL and Teva appealed.  

On appeal: Teva focuses on its discovery of a new use for a well-known genus (and uses Lilly’s words against it)

On appeal, Teva’s central argument was two-fold: (1) The claimed invention is the method of treating headache, not the discovery of a new antibody genus, and (2) anti-CGRP antibodies and humanization techniques were already well-known, so Teva’s disclosure of a representative humanized anti-CGRP antibody (G1) plus multiple murine anti-CGRP antibodies was sufficient for § 112.

To bolster its argument, Teva pointed to Lilly’s statements in the prior IPRs (and related appeals) regarding the obviousness of Teva’s “antibody patents”:  

• “[A]nti-CGRP antagonist antibodies were well known in the art”; 
• The “prior art” was “replete with exemplary disclosures of anti-CGRP antagonist antibodies, including humanized antibodies, to treat human diseases and conditions”; 
• “[A]nti-CGRP antagonist antibodies that bound to and blocked the biological effect of CGRP were well known in the art”; and 
• “[T]he prior art already reported several monoclonal anti-CGRP antagonist antibodies, including those that bound to and blocked human CGRP.”

Relying on Lilly’s own statements, Teva argued that anti-CGRP antibodies were well-known as a class and that humanizing them would have been “routine and predictable,” so the specification’s disclosure was sufficient under §112.  

The Federal Circuit’s decision: Claims adequately supported by written description, given the state of the art  

The Federal Circuit framed the written description issue as “whether the specification disclosed a representative number of species of the asserted claims’ genus of humanized anti-CGRP antagonist antibodies.” Reviewing the JMOL de novo, the court reversed.

First, the court agreed that the relevant invention was the use of a well-known antibody genus to treat headache — not the antibodies themselves. The court placed the case in its line of precedent involving new uses for old genera. In those cases, disclosure of one or a few species of a well-known genus satisfied the written description requirement where the genus and tools for identifying additional working species were well-understood.

Second, the court repeatedly relied on Lilly’s own prior characterizations from the antibody IPRs calling humanized anti-CGRP antibodies “well-known,” “replete,” or “extensively described” in the prior art. And in holding that “[a] reasonable jury could have also found that humanization was a well-established and routine procedure by the priority date,” the court again quoted Lilly’s own statements in the antibody IPRs that humanization “was a well-established and routine procedure.”¹

Practical takeaways for multi-forum patent disputes

• Align § 112 and § 103 narratives early. It is common for patent challengers to attack a patent for obviousness before the PTAB while leaving § 112 arguments for a district court. Here, however, Lilly’s own obviousness arguments about the state of the art and knowledge of a skilled artisan in the antibody IPRs proved difficult to reconcile with its later § 112 arguments. 
• Beware of broad “well-known” language. Overbroad characterizations can become admissions. The Federal Circuit’s verbatim recitation of Lilly’s prior arguments to support what a reasonable jury could find illustrates the need for litigants to think strategically before taking positions that may hurt them in the long term.  
• Method claims may get a new look. Where the claimed invention is a new use of a known genus, courts may be more receptive to genus-level proof that “all such antibodies work” for the claimed purpose, supported by known screening/making techniques.

Bottom line: For many companies, particularly in the life sciences and pharmaceutical industries, it’s not uncommon for disputes to span multiple patents in different forums over several years. In these long-running disputes, positions taken in one proceeding (and related to one set of patents) rarely stay contained. Treat PTAB and district court briefing as a single, coordinated strategy across the patent family so that today’s “routine and well-known” argument doesn’t become tomorrow’s § 112 problem.