In the case of Allergan vs. Sandoz, the United States Court of Appeals for the Federal Circuit (CAFC) has invalidated a key Allergan patent covering the use of prostaglandin F (PGF) analogs for hair growth, handing a significant win to generic manufacturer Sandoz Inc. The Federal Circuit reversed a Colorado district court judgment and held that claim 30 of US9579270B2 is invalid for lack of adequate written description under 35 U.S.C. § 112(a).

Allergan vs. Sandoz – Basic Details:

• CAFC Appeal Number: 2024-1078
• Decided on: Nov. 18, 2025
• Patent at issue: US9579270B2 (US’270)
• Patent Title: Compositions and Methods for Treating Hair Loss Using Non‑Naturally Occurring Prostaglandins
• Appeal from: The United States District Court for the District of Colorado in No. 1:18-cv-00997-RM-KLM

The US’270 patent is owned by Duke University and Allergan Sales, LLC. It describes methods of treating hair loss by topically applying prostaglandin F (PGF) analogs. Prostaglandins are signaling molecules that bind to cell receptors and change cellular function. Within the PGF family, numerous PGF analogs can be created by modifying groups attached to a common hairpin backbone.

At the center of the appeal was claim 30, a method claims that, when read with claims 17, 24 & 25, is directed to:

• A method of growing hair by topically applying to mammalian skin
• A composition comprising a PGF analog with a specified structure

• With strict limitations at key positions:
  • R1 is an amide, C(O)NHR3 
  • R2 is hydrogen 
  • R3 is methyl, ethyl, or isopropyl 
  • X is chosen from several linker options (e.g., covalent bond, —C≡C—, —CH=CH—, —(CH2)n—, etc.) 
  • Y is sulfur, oxygen, or NH 
  • Z is phenyl 

In practice, the claim targets a subgenus of PGF analogs used for hair growth.

Allergan vs. Sandoz – Latisse® and the accused product:

Allergan markets Latisse®, an FDA‑approved 0.03% bimatoprost ophthalmic solution for treating eyelash hair loss. Bimatoprost is a PGF analog with:

• An ethyl amide at the C1 (R1) position 
• A phenyl group at the omega (Z) end

Sandoz manufactures a generic version of Latisse®. Allergan sued Sandoz in 2018 in the District of Colorado (Case Number: 1:18-cv-00997-RM-KLM), alleging that the generic product infringes claim 30 of the US’270 patent. Sandoz stipulated to infringement but challenged the validity of claim 30 on multiple grounds, including written description, enablement, and obviousness.

A jury found the claim valid and awarded Allergan USD 39 million in damages. The district court denied Sandoz’s post‑trial motions, and Sandoz appealed.

Allergan vs. Sandoz – The Legal Issue: Adequate Written Description for Chemical Genus Claims

On appeal, the Federal Circuit focused on a single issue: whether claim 30 is invalid for lack of adequate written description.

Under 35 U.S.C. § 112(a), the patent specification must contain a written description of the invention sufficient to show that the inventor possessed the full scope of the claimed invention at the time of filing. For genus claims to chemical compounds, the Federal Circuit’s longstanding rule is that a patentee must provide:

• Either a representative number of species within the claimed genus, or 
• Common structural features shared by the genus members, described with enough precision that a skilled artisan can visualize or recognize the members of the genus.

The court applied this framework, citing cases such as:

• Regents of the Univ. of Minnesota v. Gilead Sciences, Inc.
• Ariad Pharms., Inc. v. Eli Lilly & Co.
• In re Ruschig 
• Fujikawa v. Wattanasin 
• Purdue Pharma L.P. v. Faulding Inc. 

The core question: Does the US’270 patent’s specification adequately describe the narrower subgenus of PGF analogs claimed in claim 30, including bimatoprost, within a vastly larger universe of prostaglandin compounds?

Allergan vs. Sandoz – Federal Circuit’s Analysis:

1. Scope mismatch between specification and claim 30:

Expert testimony established that:

• The specification is drafted so broadly that it encompasses billions of potential compounds, because many positions on the PGF hairpin are open and can be populated by numerous substituents.
• Claim 30, by contrast, covers only about 1,620 to 4,230 compounds, depending on how the subgenus is counted.

The court held that for written description to be adequate, the specification must guide a person of ordinary skill in the art (POSA) to this subgenus not just describe the entire forest of prostaglandin possibilities.

Importantly, Allergan did not argue that the patent disclosed a representative number of species of claim 30’s subgenus. Instead, it relied solely on the common structural features/blaze marks approach.

2. The generic prostaglandin hairpin is not enough:

The patent discloses a characteristic PGF hairpin backbone. Both parties’ experts agreed that this backbone is generic to prostaglandins and shared by billions of compounds.

The Federal Circuit held that this generic backbone:

• Is not unique to the claimed subgenus, and
• Does not help a POSA visualize the specific thousands of compounds in claim 30 from among billions of possibilities.

In other words, simply pointing to a shared prostaglandin scaffold does not establish written description for a much narrower subgenus within that massive chemical space.

3. “Maze‑like” options at the C1 (R1) position:

A key limitation in claim 30 is that R1 must be C(O)NHR3 (an amide), with R3 being methyl, ethyl, or isopropyl. Allergan argued that the specification undisputedly discloses only 13 options for the C1 position, including this amide group.

The Federal Circuit rejected this characterization:

• Only four of the 13 options are single, fixed substituents. 
• The remaining nine are broad categories (e.g., esters, amines, sulfonamides, various carbon‑length chains) that themselves contain many sub‑options.
• Even once C(O)NHR3 is chosen, R3 is defined by 12 further categories, many of which are again highly variable.

The court described this as a “maze‑like path” of branching choices, citing Regents and Ruschig: listing broad functional groups and categories at open positions does not, by itself, “blaze a trail” to the specific combinations later claimed.

4. Blaze marks point away from the claimed invention:

The specification also identifies certain C1 options as “preferred” or “more preferred.” Crucially:

• C(O)NHR3, the amide required by claim 30, is not among those preferred groups. 
• Instead, the specification prefers other groups such as carboxylic acids and esters.

Thus, the only explicit blaze marks at C1 actually direct a skilled artisan away from the subgenus claimed in claim 30. This supports the conclusion that the inventors had not clearly signaled possession of amide‑containing PGF analogs with the specific R3 options recited in the claim.

5. Z (omega) position: Phenyl is buried in broad categories:

At the omega or Z position, claim 30 requires phenyl. The specification defines Z as selected from eight broad categories: carbocyclic, heterocyclic, aromatic, heteroaromatic, and their substituted counterparts.

Although the patent mentions phenyl as a preferred aromatic group, that preference only applies:

• After the artisan has already chosen “aromatic group” from among the eight starting categories, and
• In some cases, only after choosing a specific linker option at X (e.g., —C=C—).

Allergan vs. Sandoz – The Federal Circuit noted that:

• The specification’s 95 example compounds include at least 10 with phenyl at Z, but also 18 examples with fluoro‑benzene at Z, without any stated reason to favor phenyl over other options.
• Nothing meaningfully points a POSA to combine phenyl at Z with C(O)NHR3 at C1 and the other specific limitations that define claim 30’s subgenus.

Overall, the court characterized the disclosure as a “laundry list” of possible moieties at multiple positions, falling short of the required blaze marks for the claimed subgenus.

Allergan vs. Sandoz – Holding: Claim 30 Is Invalid for Lack of Written Description

Applying its precedents, the Federal Circuit concluded that no reasonable jury could have found that Sandoz failed to prove lack of written description by clear and convincing evidence. The court held that:

• The US’270 patent does not disclose a representative number of species within the claimed subgenus. 
• The patent does not identify structural features that sufficiently distinguish the narrow subgenus from the broader “universe” of billions of disclosed compounds.
• The specification’s guidance and preferences often steer a POSA away from, rather than toward, the specific combination of features required by claim 30.

As a result, the court reversed the district court’s judgment and held claim 30 of the US’270 patent invalid for lack of adequate written description under § 112(a).

Because invalidity on written description grounds was dispositive, the Federal Circuit did not reach Sandoz’s other arguments, including obviousness, enablement, collateral estoppel, and alleged jury instruction errors on teaching away.

Allergan vs. Sandoz – Key Takeaways:

For professionals, especially those working with chemical, pharmaceutical, and biologic genus claims, this decision reinforces several strategic points:

• Avoid pure laundry list drafting. Simply enumerating large sets of possible substituents at many positions, without clear guidance, invites § 112(a) challenges. 
• Use true blaze marks. If you later intend to rely on a specific subgenus (for example, amides at one position and phenyl at another), the specification should:
  • Highlight those combinations as preferred or especially useful. 
  • Provide embodiments actually falling within the later‑claimed subgenus. 
• Representative species matter. The absence of even a single disclosed species within claim 30’s subgenus was a major weakness for Allergan at the Federal Circuit. 
• Be mindful of scope mismatch. When a specification broadly covers billions of possibilities but the litigated claim is a narrow subset of a few thousand, courts will closely scrutinize whether the inventors truly “possessed” that subset at filing.

For generic challengers, Duke/Allergan v. Sandoz showcases an effective written‑description attack on broad pharma genus claims that sit on top of widely known chemical scaffolds.

Allergan vs. Sandoz – Conclusion:

The Federal Circuit’s Nov. 18, 2025 decision in Duke University, Allergan Sales, LLC v. Sandoz Inc. underscores the continuing vitality of the written description requirement as a tool to police overbroad chemical genus claims. By striking down claim 30 of the US’270 patent, the court limited Allergan’s exclusivity around Latisse®‑type PGF analogs and clarified once again that patentees must do more than disclose a forest of possibilities and later claim a specific tree.

Citation:

• Duke University, Allergan Sales, LLC v. Sandoz Inc – CAFC Opinion 2024-1078