Composition or Method? How Amgen and Teva Are Reshaping Drug Patent Strategy

By Anthony D. Marino

Two recent decisions are quietly redrawing the boundaries of what drug and biologic patents can cover. The Supreme Court’s 2023 decision in Amgen Inc. v. Sanofi tightened the enablement screws on broad, functionally defined genus claims, which are the bread and butter for many life-science portfolios. On April 16, 2026, the Federal Circuit handed down Teva Pharmaceuticals International GmbH v. Eli Lilly & Co., which pointed the other way: when the asserted claims are methods of treatment using a well-known genus of compounds, § 112 can be satisfied even without disclosing every member of the genus.

Put together, Amgen and Teva create a practical dichotomy that every life-science innovator should be thinking about: composition claims are the straightforward path under § 101 but the most valuable claims face heightened § 112 risk after Amgen. Broad method-of-treatment claims may better withstand § 112 scrutiny under Teva, but they can attract § 101 eligibility challenges under the Mayo/Alice line of cases. Whether you are in-house counsel managing a mature portfolio or a founder deciding what to claim in your next filing, the strategic takeaway is the same: the choice between composition and method claims is no longer a drafting preference. It is a risk allocation decision with real downstream consequences.

Section 112(a) of the US patent laws requires patent specifications to do two things: describe the invention in a way that conveys the inventor had “possession” of it and enable a skilled artisan to make and use the invention without undue experimentation. For genus claims (i.e., claims covering a whole class of compounds), courts have long held that enablement must be commensurate with the scope of the claims. What has changed is how aggressively that principle is now being applied to broad, functionally defined claims in the life sciences, and how the claim type (composition versus method) shapes the analysis.

Amgen v. Sanofi (2023)

Amgen’s patents sought to claim an entire genus of antibodies that bind to specified residues on signaling proteins involved in cholesterol metabolism. The specification disclosed amino acid sequences for 26 antibodies and described two approaches for generating additional antibodies within the claimed class.

The Supreme Court affirmed invalidity for lack of enablement under § 112(a). The Court reiterated that “the specification must enable the full scope of the invention as defined by its claims,” and framed the principle categorically: the more one claims, the more one must enable. Amgen’s disclosures, the Court concluded, amounted to “little more than two research assignments,” leaving skilled artisans to engage in painstaking experimentation rather than enabling them to reproduce the invention.

Teva v. Lilly (Fed. Cir. April 16, 2026)

The asserted claims were directed to using antibodies targeting calcitonin gene-related peptide (CGRP) to treat headaches. A representative claim recited a method for reducing the incidence of or for treating a headache by administering a humanized “anti-CGRP antagonist antibody.” The specification disclosed amino acid sequences for one humanized antibody and multiple mouse-derived antibodies in addition to citing prior-art methods for humanizing antibodies. A jury found infringement and rejected invalidity. The district court then granted judgment as a matter of law that the claims were invalid for lack of written description and enablement under § 112(a).

The Federal Circuit reversed on both grounds and reinstated the jury’s verdict. The court’s reasoning turned on how it characterized the claimed invention. The claims, the panel emphasized, were not claims to the antibody genus itself. Instead, they were claims to a use of antibodies within a well-known class for a specific therapeutic purpose. Seen in that light, the court reasoned that evidence showing anti-CGRP antagonist antibodies and methods of making them were already well known, and that humanization was routine, was enough to support the jury’s written description finding. On enablement, the court distinguished Amgen directly: even if making every antibody in the genus would be burdensome, that work was not required to practice the claimed method and was “more akin to extra credit than a necessary research assignment.” The Federal Circuit also pointed to Lilly’s own positions in inter partes review (that anti-CGRP antagonist antibodies were well known, that the prior art was “replete” with such antibodies, and that humanization was routine) as support for the jury’s § 112 findings.

Read together, Amgen and Teva do not conflict; they draw a line between composition and method claims. When the claim is to the compounds themselves, Amgen‘s full-scope enablement rule controls and the specification must carry the weight of the entire claimed class. When the claim is to a method that uses compounds from a well-known class, the § 112 inquiry recenters on the claimed method and the input genus can contribute its background knowledge to the analysis.

This dividing line between composition and method claims has real consequences for patent portfolios. Broad composition claims offer easily enforced coverage and have a straightforward § 101 test under Myriad, but they now face a demanding § 112 gauntlet. Method-of-treatment claims can survive Amgen-style enablement attacks in view of Teva, but they expose the patentee to a harder § 101 test, especially when the claim can be recast as an application of a natural relationship. Patentees trying to avoid one obstacle too often steer straight into the other.

It is worth being clear about what these decisions do not do. Neither Amgen nor Teva rewrites § 112. They apply long-standing principles — possession, full-scope enablement, the Wands factors — to modern biologics claims. A method-of-treatment claim still has to clear every other hurdle that a method claim has always faced: novelty under § 102, non-obviousness under § 103, written description and enablement of the method itself, and definiteness. A composition claim that survives Amgen still has to withstand prior art. The headline news is the doctrinal line being drawn between claim types, not a reduction in what patentees must prove.

For In-House Counsel: New Filings, Continuations, and Litigation Posture

Portfolio Audits

A useful exercise for an in-house team right now is a focused audit of where method-of-treatment claims can strengthen the portfolio in light of Teva. For each asset, the questions are practical. Does the existing specification support a method-of-treatment claim tied to an input genus? Are there pending continuations or divisionals that could be repurposed to add method claims before the family closes? Teva suggests that method claims can clear § 112 when parallel composition claims cannot, and those opportunities should be identified and pursued deliberately rather than left on the table.

The audit should also flag assets where the portfolio leans too heavily on broad composition claims without a corresponding method-of-treatment. Post-Amgen, those composition claims can be fragile, and a method claim can often be the most efficient way to shore up coverage without starting from scratch. Finally, the audit should identify method-of-treatment claims (either existing or newly contemplated) for pressure-testing against § 101. This way, the cure for an Amgen problem does not quietly create a Mayo/Alice one.

New Filings and Continuations

For new filings in the unpredictable arts, the practical question is no longer “composition or method?” It is “how do we support both, and what does the specification need to disclose?” Invention harvesting should identify where method-of-treatment claims can provide complementary coverage consistent with Teva. New disclosures may benefit from a layered structure to the patent application: narrower composition claims or structurally defined subgenera paired with method-of-treatment claims, all supported by a specification drafted to carry both (when possible). In view of the give and take between obviousness and § 112(a) arguments, it may also be beneficial for US applicants to induce a restriction to gain the benefits of the safe harbor provisions of § 121 for both composition and method-of-treatment claims.

Litigation Posture

The lesson for litigation teams from Teva is uncomfortable but important: arguments made to defeat obviousness challenges can come back to support § 112 challenges. On the flip side, arguments made to defeat § 112 attacks can strengthen obviousness attacks. Both doctrines are measured against the same baseline of what skilled artisans knew and could do, and narratives must be coordinated across forums accordingly. Patent challengers may also benefit from shifting obviousness challenges to the newly invigorated ex parte reexamination process, where formal no estoppel is created.

For Business Owners and Entrepreneurs: Protecting Patent Value

What This Actually Means for Your Business

If your company’s value depends on patent protection for a drug, biologic, or other life-sciences asset, claim architecture is a business decision, not just a legal one. It shapes what competitors can design around, what licensees will pay for, and what defenses accused infringers will raise years from now. Amgen and Teva change the math on that decision, and they do it in ways that flow straight through to valuation and deal terms.

The simplest way to think about it: a pivot toward patenting methods of treatment instead of drug compositions may reduce exposure to one patent vulnerability, but it can also increase exposure to another. Either risk, if unmanaged, can undermine the strength of the patents you are relying on to protect your business. The companies that handle this well will treat patent strategy as something to be discussed at the same table as product and commercialization strategies, and not something handed off to counsel at filing time.

Key Steps to Consider

Amgen and Teva together are drawing a sharper doctrinal line between composition and method claims in drug patent law. Broad composition claims now bear the weight and risks of Amgen‘s full-scope enablement rule. Method-of-treatment claims may be evaluated through a different § 112 lens under Teva but also present their own unique § 101 risks.

Neither case is a silver bullet, and choosing composition claims does not foreclose method claims or vice versa. For life science innovators, counsel, and their business partners, the opportunity here is to build layered patent portfolios that calibrate and proactively manage risk across both claim types. The ventures that internalize this strategic choice early will likely be the ones holding the broadest and most defensible patents five years from now.