Vertex Pharmaceuticals v. Lupin — Generic Kalydeco Formulation Does Not Infringe Narrow 80% Ivacaftor Claims

Background

Vertex developed Kalydeco, an ivacaftor treatment for cystic fibrosis. Because ivacaftor dissolves poorly, Vertex formulated it as an amorphous solid dispersion. Four asserted patents claimed formulations containing either exactly 80% ivacaftor by weight or “about 80%.” The patent history mattered: Vertex had earlier published a broad range, then persuaded the Patent Office that an 80% formulation produced unexpectedly strong solubility. When Vertex later sought a range of roughly 72% to 88%, the examiner rejected it and allowed the narrower “about 80%” language.

Lupin filed an abbreviated new drug application for a generic ivacaftor product. Its formulation contains approximately 74.257% ivacaftor, uses some different inactive ingredients, and is produced through wet granulation rather than Vertex’s dry process. Vertex sued under the Hatch-Waxman Act, alleging literal infringement of the “about 80%” claims and infringement of all four patents under the doctrine of equivalents.

The Court’s Holding

Judge Stephanos Bibas held after a bench trial that Lupin infringed none of the asserted claims. Literal infringement failed because a skilled pharmaceutical formulator would not understand “about 80%” to reach Lupin’s roughly 74% concentration. The court credited Lupin’s expert, who treated “about” as a modest variance, and rejected Vertex’s proposed 70%–90% “high drug load” range as unsupported by the patents, prosecution record, or a scientific definition.

The prosecution history reinforced the narrow reading. During examination, “about 80%” had been illustrated by 79.9%, while Vertex’s attempt to claim a materially broader range was rejected. The court therefore would not let Vertex use litigation to obtain scope that it had failed to secure from the Patent Office.

The doctrine of equivalents did not change the result. The numerical limitations were central to patentability, and expanding them to cover 74.257% would erase meaningful claim boundaries and risk recapturing prior art. Bioequivalence also was insufficient: generic drugs are designed to achieve comparable therapeutic performance, but that regulatory similarity does not by itself establish patent infringement. Lupin’s differing concentration, ingredients, and manufacturing process were substantial differences.

Key Takeaways

• Precise numerical claim limits can carry decisive weight when the prosecution record shows that a broader range was rejected.
• A patentee generally cannot use the doctrine of equivalents to recapture subject matter surrendered or left unproved during prosecution.
• Bioequivalence supports approval of a generic drug, but it does not automatically prove that the generic satisfies a patent claim.
• Expert testimony needs a concrete scientific and intrinsic-record basis; a broad label such as “high drug load” cannot substitute for the patent’s percentages.

Why It Matters

The decision underscores a practical limit on pharmaceutical patent scope. When a company wins claims by emphasizing a particular concentration and its unexpected properties, those representations can later prevent the company from stretching the claims across nearby generic formulations. For generic manufacturers, careful changes to concentration, ingredients, and process may matter even when the resulting medicine remains therapeutically equivalent.

More broadly, the ruling is a reminder that patent drafting and prosecution create a lasting record. Narrow language may secure allowance, but it can also define the boundary that a court will enforce years later.

Surfaced via Law360 IP.