DNA Patent Eligibility Under Spotlight in Sarepta’s En Banc Bid

Sarepta Therapeutics Inc.'s bid for the full Federal Circuit’s review of a revived gene-therapy patent dispute with RegenxBio Inc. is drawing new attention to the question of when altered DNA becomes different enough from nature to be patented.

Sarepta requested in late March that the US Court of Appeals for the Federal Circuit review the case en banc after a three-judge appeals panel reversed a January 2024 ruling invalidating US Patent No. 10,526,617 in February.

The ‘617 patent covers lab-grown host cells containing recombinant DNA that combines adeno-associated virus, or AAV, sequences with other genetic material.

The Federal Circuit said those engineered cells don’t occur in nature—an observation central to its patent eligibility ruling. The panel sent the case back to the US District Court for the District of Delaware, which it said erred in finding the patent covered natural phenomena.

The ruling is being read less as a shift in patent law than as a clearer application of existing precedent, legal observers said. But it also gives litigants a road map for defending engineered biological materials and surviving early challenges, they said, while highlighting limits in how courts distinguish engineered DNA from natural material.

The decision corrected a district court ruling that “went off the rails,” said Kevin E. Noonan, a partner at McDonnell Boehnen Hulbert & Berghoff LLP who writes widely on biotechnology patent law.

The claimed cells, he added, “are not naturally occurring and have different properties"—features that satisfy established tests for patent eligibility.

Sarepta’s petition sharpens that dispute, arguing the panel went too far by letting routine manipulation of natural DNA serve as the basis for patent eligibility. The argument tees up debate on how far Section 101 of the US Patent Act extends to engineered biological materials.

Patent Dispute

RegenxBio and the University of Pennsylvania, which owns the ‘617 patent, sued Sarepta in September 2020 alleging the company’s use of cultured host cells with an AAV-based system to manufacture SRP-9001, its gene therapy for Duchenne muscular dystrophy, infringes the patent.

RegenxBio said in a May 2023 discovery conference it would seek more than $900 million in damages if it ultimately proves infringement.

Saurabh Vishnubhakat, a professor at the Cardozo School of Law, said the Federal Circuit opinion gives courts and litigants “a clearer template for evaluating engineered biological materials” under Section 101.

“For patentees, the strategy will be to claim the laboratory-made construct, insist on whole-claim analysis, and resist any effort to strip out recombinant or host-cell limitations as merely conventional,” Vishnubhakat said.

“Sarepta’s rehearing petition is an attempt to roll back this strategy,” he said, “as its view is that the panel effectively let routine manipulation of natural DNA do the work of eligibility and moved the inquiry toward a simple human-made-versus-natural distinction.”

The Federal Circuit panel said the claimed cells are patent-eligible because they contain recombinant DNA that “does not and cannot exist in nature” and must be created through human-directed genetic splicing. Eligibility must be assessed based on the claimed composition as a whole, the appeals court said, rejecting arguments that courts should disregard conventional laboratory steps when analyzing the claims.

Nod to Precedent

The decision reflects a straightforward application of Supreme Court precedent, according to David B. Gornish, a former patent litigator who now focuses primarily on patent preparation and prosecution at Eckert Seamans.

Gornish said RegenxBio’s asserted patent claims don’t explicitly describe a feature that makes the engineered cells different from anything in nature, unlike the patent claims in Diamond v. Chakrabarty, the 1980 Supreme Court decision that led to allowing patents on genetically engineered microorganisms.

The Federal Circuit acknowledged that distinction but treated those differences as implicit in the claimed invention itself—an approach that Gornish said could shape how patents are written and challenged going forward.

One takeaway for patent prosecutors, Gornish said, is that explicitly reciting a property that makes an invention “markedly different” from its natural counterpart, as in Chakrabarty, “would have blunted one of the primary attacks that Sarepta asserted.”

“At the very least,” he said, “patent prosecutors should consider reciting such a property in a dependent claim as fallback.”

Impact on Litigation

The decision could also shift how disputes are fought, Vishnubhakat said, by making patent-ineligibility arguments easier to survive at the pleading stage, potentially reducing the effectiveness of early dismissal efforts in district court.

The ruling may push disputes toward later fights over issues like novelty and obviousness, he said, because such challenges are “by far the most commonly asserted” grounds for unpatentability at that early stage.

The “practical tension” in Sarepta’s en banc bid, Vishnubhakat said, is that the company argues the claims require only a host cell containing the recombinant molecule, not one that performs any specific function, while the panel’s opinion “leaned on the claimed compositions’ potential utility for gene delivery—even though structural non-naturality is doing most of the analytical work.”

“That’s the fault line,” he said, “that the en banc court or else courts in future cases will have to address in disputes over engineered cells, vectors, and other biologic constructs.”

The case is RegenxBio Inc. v. Sarepta Therapeutics Inc., Fed. Cir., No. 24-1408.