Teva v. Eli Lilly: The Four Drafting Choices That Won $176.5M

The decision

Teva Pharmaceuticals International GmbH v. Eli Lilly & Co., No. 24-1094 (Fed. Cir. Apr. 16, 2026), reinstated Teva's $176.5 million jury verdict against Eli Lilly. The panel held that Amgen v. Sanofi, 598 U.S. 594 (2023), does not control method-of-treatment claims reciting a functionally defined antibody genus.

Teva sued Lilly for patent infringement in the District of Massachusetts in 2018. The asserted patents covered methods of treating headache by administering humanized anti-CGRP antagonist antibodies, the same therapeutic mechanism behind Teva's Ajovy® (fremanezumab) and Lilly's Emgality® (galcanezumab). At trial in 2022, the jury found willful infringement, rejected Lilly's §112 challenges, and awarded $176.5 million to Teva. 

The district court granted JMOL of invalidity, holding the specification did not support "the very broad scope of these asserted claims." On April 16, 2026, the Federal Circuit reversed, finding that the district court applied an overly stringent standard that ignored both background knowledge in the field and the nature of the claimed invention as a method of treatment.

The doctrinal move

Two categories of genus claim

The panel contrasted (i) claims directed to using a genus for the function that characterizes the genus, and (ii) claims directed to using a genus for a function different from the genus itself. The two categories are not subject to the same §112 analysis.

Teva's so-called “headache patents” claims fell into the second category. The genus, "humanized anti-CGRP antagonist antibodies," was defined by binding function. The claims used that genus to do something different: treat headache. Because the genus was not itself the invention, background knowledge could supply what the specification did not disclose. The line traces back to In re Herschler, 591 F.2d 693 (CCPA 1979), and Amgen did not disturb it.

The boundary of Amgen

The same patent family had already lost on the antibody composition claims. The PTAB found Teva's antibody patents unpatentable in IPRs and the Federal Circuit agreed. The headache patents, however, survived. The opinion "creates a distinction for written description and enablement under 35 U.S.C. §112(a) when the claims in dispute are method-of-use claims versus article claims." Amgen governs claims to the antibody genus itself. Teva governs claims to the use of that genus where the genus is not the inventive contribution.

The drafting choices that won $176.5M

Teva's specification disclosed one humanized anti-CGRP antibody (G1), several murine precursors, and a statement that humanizing murine antibodies was routine. Lilly did not dispute that statement. As Venable observes, "the patents stated that such antibodies were known in the art and could be humanized using routine procedures. Lilly did not dispute those statements." That combination was deemed to be enough.

Four early drafting decisions produced the result:

• Claim category. Method-of-treatment claims, not composition claims. The same disclosure supported the headache patents and sank the antibody patents.
• Functional anchoring. "Anti-CGRP antagonist antibody" tied the genus to a known biological mechanism within the POSA's knowledge.
• Reliance on routine art. Humanization was treated as a known technique. That shifted the §112 question from what Teva enabled to what a POSA could already do.
• One fully described embodiment. Antibody G1 supplied the proof of concept, and that was enough

Implications for life sciences portfolios

Existing genus claims

Method-of-treatment claims supported by only one or two embodiments may now be more defensible than corresponding composition claims in the same family. Validity, FTO opinions, and claim construction positions need to be reassessed against the Amgen / Teva distinction. Competitors and generics are running the same analysis to find weaknesses, not to defend them.

Pending applications

Where the inventive contribution is therapeutic use, the claim set should include method-of-treatment claims drafted against functional limitations and known background art, not just composition claims that carry Amgen's heavier disclosure burden. Dinsmore's takeaway is the right one: 

"Have a diversified claim strategy that includes article (antibody) claims, method claims, and method-of-use claims, especially when claiming a genus, because each class of claims may be evaluated differently."

How AI-assisted drafting and analysis changes the calculus

The drafting pressures Teva highlights share a feature: they require practitioners to hold a specific doctrinal frame across hundreds of specifications, dozens of attorneys, and a constant stream of office actions. The discipline is easy to articulate but hard to maintain at scale.

Solve Intelligence offers a patent copilot built for this work, with capabilities that map onto the challenges Teva v. Lilly creates.

Drafting with the Amgen / Teva distinction in mind

The four drafting decisions that won Teva (claim category, functional anchoring, reliance on routine art, fully described embodiment)  are the structural choices Solve's drafting copilot is built to surface during specification drafting. The supporting case law (Amgen, Teva, Baxalta, Idenix) shapes the spec as it's being written, not after the office action arrives.

Portfolio-scale §112, invalidity, and FTO analysis

Solve Intelligence's Charts product runs §112, invalidity, and FTO analysis work that would take attorneys weeks, with full citations and exposed AI reasoning in each cell. For teams reassessing portfolios against Teva, that means specifications, file histories, and prior art lined up in one workspace, with the §112 analysis traceable claim-by-claim and exposed for partner review.

Preserving scope during prosecution

Prosecution history affects how claims are construed when being asserted/enforced. Solve Intelligence's office action response tool addresses the core tension: overcoming examiner objections while maintaining the broadest defensible scope. For portfolios that may face Federal Circuit scrutiny years from now, clean prosecution histories at scale are a strategic advantage.

The road ahead

Teva v. Eli Lilly will be cited heavily on both sides of life sciences disputes for the foreseeable future. A higher-court development that recalibrates the line is possible but not imminent.

The practical message: drafting discipline is the variable, and it is a workflow problem as much as a doctrinal one. Maintaining drafting discipline across a modern life sciences practice (every spec, written by every attorney, every time) is the gap purpose-built tools are designed to close.

500+ IP teams use Solve Intelligence, with reported efficiency gains of 60–90%. If Teva v. Lilly has your team re-opening portfolios or rethinking drafting standards, request a demo.

FAQ

What did the Federal Circuit decide in Teva v. Eli Lilly?

The Federal Circuit reinstated Teva's $176.5 million jury verdict against Eli Lilly, reversing a district court grant of JMOL. The panel held that Teva's method-of-treatment claims, covering the use of humanized anti-CGRP antagonist antibodies to treat headache, satisfied 35 U.S.C. §112's written description and enablement requirements, even though the specification disclosed only one humanized antibody. It is the first precedential Federal Circuit ruling to mark a clear limit on Amgen v. Sanofi.

How does Teva v. Lilly limit Amgen v. Sanofi?

Amgen governs the disclosure required for claims to a functionally defined antibody genus. Teva holds that the same calculus does not extend to method-of-treatment claims that use such a genus where the genus is not itself the invention. Where the genus is well known in the art and routine techniques exist for producing additional members, background knowledge can supply what the specification does not disclose.

What drafting lessons does Teva v. Lilly carry for life sciences practitioners?

Four. Claim category matters: method-of-treatment claims and composition claims are evaluated differently under §112. Functional anchoring to a known biological mechanism positions the genus within the POSA's knowledge. Reliance on routine art shifts the §112 question from what the patentee enabled to what a POSA could already do. A single fully described embodiment can be enough where the surrounding background knowledge is robust.

How can Solve Intelligence help life sciences teams respond to Teva v. Lilly?

Solve Intelligence's drafting copilot surfaces the controlling §112 case law during specification drafting, in the attorney's voice. The Charts product runs §112, invalidity, and FTO analysis across thousands of documents, with full citations and exposed AI reasoning. Together they let teams hold Teva-aligned drafting discipline across hundreds of specifications and reassess existing portfolios against the Amgen / Teva line at scale.

‍