The U.S. Patent and Trademark Office (USPTO) is taking another crack at how it evaluates inventions derived from nature. The agency has been forced to tighten its eligibility rules in light of recent Supreme Court decisions—including a 2013 ruling that struck down patents on human genes—but its first pass at new guidelines for examiners raised a stink. University groups and industry representatives feared the rules would chill investment by rendering potential new therapies and diagnostics unpatentable. A new guidance, released this week, goes a long way to quell those fears, though many are still uncertain how it will be implemented.
The document unveiled this past spring was troubling to the biopatent community for several reasons. It went too far beyond the court’s ruling on DNA, critics claimed, by suggesting that other naturally derived products—including chemical compounds, vaccines, seeds, and antibodies—must be “significantly different” from anything found in nature or else be immediately tossed out as patent ineligible. Many also protested the idea that an applicant had to demonstrate structural differences between a product and its natural counterpart. Amid criticism at public meetings and in written letters, the agency vowed to incorporate stakeholder comments into a revised guidance.
But while it’s eager to satisfy the applicants that keep it in business, USPTO also has to respect often murky directions handed down from the courts, says Dan Burk, a law professor at the University of California, Irvine. Particularly influential were last year’s ruling in Association for Molecular Pathology v. Myriad Genetics Inc. that naturally occurring human genes cannot be patented and the 2012 Mayo v. Prometheus decision, which invalidated a patent on a method of adjusting drug dosage using measures of blood metabolites because it relied on a “law of nature.”
And since the March guidance, another relevant decision came down from the high court: In Alice Corp. v. CLS Bank International, the judges invalidated a software patent by ruling that a method of financial analysis is an unpatentable “abstract idea,” even when implemented on a computer. That June decision may seem worlds away from biotech, but it aimed to lay out a definitive two-step test for deciding patent eligibility, Burk says, and it forms the backbone of the new guidance: If a patent claim describes “a law of nature, a natural phenomenon, or an abstract idea” (step 1), then it must amount to “significantly more” than what’s found in nature—a phrase borrowed from the Mayo decision—in order to be patentable (step 2).
Many biotech inventions may now get a green light at the first step, says Hans Sauer, intellectual property counsel for the Biotechnology Industry Organization in Washington, D.C. The guidance lets examiners decide at first look when an application involves a natural product or process, but obviously isn’t trying to claim it in a way that would prevent others from using it. “In practice, that could make a big difference,” Sauer says, and will ease the way for “a lot of claims that people complained about very much.”
An examiner can also decide to let an application proceed because the product is sufficiently different from its natural counterpart based on “structure, function, and/or other properties,” which may be reassuring to academia, says Robert Hardy, director of contracts and intellectual property management for the Council on Governmental Relations (COGR), a Washington, D.C.–based association of research universities. COGR had submitted a critique of the March guidance to USPTO. Hardy suspects academic technology transfer offices will receive fewer USPTO rejections on grounds of eligibility and believes the agency did “kind of a heroic job of trying to reconcile a great many court decisions.”
Still, it’s not clear what kinds of differences from nature will be enough to confer eligibility in the eyes of the examiners, Sauer warns. For example, if you purify an antibiotic from an organism, you might be changing it so that it can be formulated as a medicine for the first time, he says, but you’re probably not changing its essential bacteria-killing abilities. Does that count as a functional difference? “It seems to me that under PTO guidance the answer is probably no,” he says. And demonstrating a functional difference gets even harder, he notes, if a molecule’s original “function” in nature is unknown.
Another persistent concern is how evaluators will deal with patents on diagnostic processes or technologies related to personalized medicine. It’s not clear, for example, if the process of analyzing gene expression from a tumor biopsy to decide on a course of treatment can be protected by a patent. “Personalized medicine companies [have been] trying to deal with this for going on 2 years and keep expressing a lot of frustration,” Sauer says. “There have been lots of rejections, and they’ve been difficult to overcome.”
USPTO has announced a meeting in January to gather more feedback and may yet tweak its guidelines further. Meanwhile, another decision from the nation’s specialized patent court, the U.S. Court of Appeals for the Federal Circuit in Washington, D.C., may further muddy the waters. The 17 December ruling upheld a lower court’s decision to deny an effort by Myriad Genetics to prevent another company from developing genetic tests that Myriad says are still protected by its patents—despite the 2013 Supreme Court decision that knocked out some elements of those patents. And the Federal Circuit seemed to emphasize the need for structural differences from naturally occurring products, Burk says, which might just undermine part of the shiny new guidance.