Federal Circuit Reverses Trade Secret and Patent Invalidity Rulings in Penile Implant Dispute: Prior Art Patents Destroy Secrecy, Inventor Contribution Standard Clarified

The U.S. Court of Appeals for the Federal Circuit issued a precedential ruling on April 18, 2026, substantially reversing a jury verdict in International Medical Devices, Inc. v. Cornell (Case No. 25-1580), a dispute centered on cosmetic penile implant technology. Writing for a panel that included Judges Reyna and Taranto, Judge Dyk authored an opinion that carries significant implications for both trade secret law and patent inventorship doctrine.

The plaintiffs — International Medical Devices, Inc. (IMD), its affiliate Menova International, Inc., and surgeon-inventor James Elist, M.D. — alleged that defendant Dr. Robert Cornell and his associates misappropriated confidential surgical techniques disclosed during a Penuma implant training session held in March 2018. Dr. Cornell had signed a nondisclosure agreement (NDA) before attending the session. The plaintiffs further alleged that Cornell used the disclosed information to develop a competing implant called the Augmenta and filed patent applications without naming Elist as a co-inventor.

A jury sided with the plaintiffs across the board, finding misappropriation of four trade secrets, breach of the NDA, trademark counterfeiting, and patent invalidity for failure to name the correct inventor. The jury’s damages award totaled over $18 million — $5.77 million in reasonable royalties, $11.54 million in exemplary damages, and $1 million in statutory damages for counterfeiting — along with a five-year permanent injunction. The Federal Circuit affirmed only the counterfeiting finding and reversed the rest.

The court’s trade secret analysis applied California’s Uniform Trade Secrets Act (UTSA) and turned on a straightforward question: were the disclosed techniques actually secret? The first three claimed trade secrets were (1) incorporating internal cavities within the silicone body to improve flexibility, (2) placing mesh tabs near the implant’s distal tip to encourage tissue ingrowth, and (3) using absorbable sutures to secure the mesh tabs. The Federal Circuit found that all three concepts were already disclosed in prior art patents — specifically the Subrini patent (1992) and the Finney patent (1980).

Applying the well-established California principle that “self-evident variants of the known art cannot be protected” as trade secrets, the court rejected the plaintiffs’ argument that those patents were irrelevant because they were never commercialized. The Federal Circuit was explicit: information disclosed in a patent enters the public domain upon publication, regardless of whether anyone ever builds the patented device. Transposing a known design solution from therapeutic implants to cosmetic implants does not generate a new trade secret when both contexts present the same technical problem addressed by the same solution.

The fourth alleged trade secret — a list of surgical instruments and materials — failed on different grounds. The list had been emailed to Dr. Cornell and to a third party, SCA Surgery, without any marking or indication of confidentiality. The court found that IMD had not taken “reasonable measures to maintain secrecy” as UTSA requires, and the trade secret claim therefore failed on its face. This finding carried a direct corollary for the NDA breach claim: because the disclosed information fell within the NDA’s carve-out for information “generally available to the public,” no breach occurred.

On the inventorship issue, the patents at stake were U.S. Patent No. 10,413,413 and U.S. Patent No. 10,980,639. The original patent applications had incorporated the three technical elements at issue. During prosecution, however, the examiner rejected the claims as unpatentable over prior art. The applicants responded by amending the independent claims to include a new feature: a “measured property of hardness difference between two locations on the implant.” That amendment was what made the patents allowable.

The Federal Circuit applied its longstanding rule that a person must contribute to the conception of the claimed invention to qualify as an inventor — and that contributing only information already known in the prior art does not constitute a conceptual contribution. Because Elist’s contributions were limited to the pre-amendment features, all of which were already in the prior art, and because there was no evidence that Elist contributed the hardness-difference feature that ultimately defined the allowed claims, the court reversed the jury’s finding that the patents were invalid for failure to name him as a co-inventor.

The counterfeiting claim survived appellate scrutiny. Cornell argued that the Penuma trademark was registered for goods, not services, and that he had offered only surgical services. The Federal Circuit found otherwise: Dr. Cornell’s own trial testimony indicated that his website advertising Penuma procedures was intended to promote the implant as a product, not merely a service. That factual foundation was sufficient to sustain the counterfeiting verdict.

Two practical takeaways emerge from this decision. First, companies and individuals who wish to protect technical information as trade secrets must audit that information against the full landscape of prior art patents — including obscure or uncommercialised ones — before disclosing it to third parties under NDA. A patent publication alone is sufficient to destroy secrecy, regardless of whether the patented technology ever reached a product. Second, in co-invention and technology collaboration disputes, the inventorship analysis focuses on the claims as finally allowed, not the original filing. Where prosecution amendments add new limiting features, parties must be able to show a specific contribution to those features; contributions limited to concepts already in the prior art will not qualify under Federal Circuit doctrine.