Last week, the US Patent and Trademarks Office ruled on the most-watched patent proceeding of the 21st century: the fight for Crispr-Cas9. The decision was supposed to declare ownership of the rights to the revolutionary gene editing technique. But instead, the patent judge granted sorta-victories to each of the rival parties---a team from UC Berkeley and another with members from both MIT and Harvard University’s Broad Institute. That’s great for those groups (and their spin-off, for-profit gene editing companies with exclusive licenses). But it leaves things a bit murkier for anyone else who wants to turn a buck with gene editing.
The Crispr discoverers now have some authority over who gets to use Crispr, and for what. And while exclusive licenses aren’t rare in biotech, the scope of these do stand out: They cover all the 20,000-plus genes in the human genome. So this week, legal experts are sending a formal request to the Department of Health and Human Services. They want the federal government to step in and bring Crispr back to the people.
Crispr is new, but patent laws governing genetic engineering date back decades. In 1980, shortly after the Supreme Court ruled that genetically engineered microbes were patentable, Congress passed something called the Bayh-Doyle Act. The law gives permission for universities to patent---and license---anything their researchers invented with public funds, making it easier to put those inventions back in the hands of citizens.
The law’s original intent was to patent mature discoveries, things like a genetically modified crop, or biofuel-farting yeast. Over the years, however, universities started filing patents further upstream---on everything from protein structures to bits of DNA. This frenzy of molecule-grabbing can actually work counter to Bayh-Doyle, locking up promising discoveries that don’t need help getting commercialized. “Crispr totally epitomizes that,” says Michael Eisen, a Berkeley biophysicist and long-time advocate of open science. “Everybody in the universe is chomping at the bit to use it. But patents are an obstacle to that happening right now.”
In the 20 years after Bayh-Doyle, as scientists developed new genetic techniques, the government came up with ever-more convoluted ways to make sure publicly-funded science got to the public. In 1999, the NIH recommended that patent holders nonexclusively license research tools developed with federal funds, so more entrepreneurs could commercialize them.
Berkeley and the Broad are following that recommendation---sort of. They’re not doing anything to stop scientists who want to use Crispr for science’s sake; both have granted nonexclusive licenses to researchers at universities and nonprofit institutions. But the buck stops the moment any of those license-holders try to take a Crispr-ed product to market. At that point, the researcher needs to buy the appropriate sublicense from whichever company---Editas or Caribou or Crispr Therapeutics---holds it. These biotechs are surrogates for patent holders like Berkeley and the Broad, taking over the role of patent owner (plus the majority of profits). And that extra licensing step has the potential to stop innovative applications of Crispr.
Here’s the problem: Crispr---which can function as either a tool or a treatment---is such a powerful technique that the 90’s-era definition of ‘research tool’ can’t contain it. And three companies can’t possibly develop all of its possible applications. Jorge Contreras, a law professor at the University of Utah, says the NIH should update its guidelines to treat technologies like Crispr more like Wi-Fi or the internet, and license it on a “fair, reasonable and non-discriminatory” basis. No favorites.
He’d also like the NIH to get tougher on enforcing their recommendation. “There aren’t any real penalties for noncompliance,” he says. But something like disbarment from seeking federal grant funding for a few years? “That would turn some heads in the academic research community!”
Bayh-Doyle caused the rampant university patent problem. But it might also provide the solution. If any company or public interest group believes a patent’s license agreement is too restrictive, the law says they can petition the agency that funded the patent’s foundational research. (In Crispr’s case, that’s the HHS, of which the NIH is a part.) That agency can then compel the license-holders to loosen up.
Today, James Love is spending the afternoon putting final touches on a letter to the HHS, asking the agency to intervene in the Crispr patent situation. As the director of the nonprofit Knowledge Ecology International, he’s been building a case that the federal government should ensure an open and non-discriminatory licensing paradigm for Crispr. Specifically, he’s hoping that the NIH will exercise its so-called march-in rights. “That’s the part of the act that says the research has to be made available to the public on reasonable terms,” says Love. “But it’s very rare for parties to agree on what those reasonable terms are.”
Love has filed these kinds of petitions before. They’re usually pretty anticlimactic: Since 1980, the NIH has only held march-in hearings four times, and they’ve never moved into a full march-in proceeding. But that’s fine---because threats work, too. Back in 1997 a company called CellPro petitioned the Clinton administration for a compulsory license to four patents held by Johns Hopkins University, under Bayh-Dole. The NIH held a hearing on whether to march-in or not, and while they decided against it, Hopkins eased its agreement to allow in a competitor. “March-in has never fully been used, but it has definitely created some shadow effects,” says law scholar Arti Rai, a professor at Duke University.
A director at the NIH declined to comment on Love's petition. But the agency does have an interest in not exercising its legal leverage. If it were to start granting marching-in requests, NIH grants would look less appealing to researchers. Taking federal money would mean accepting weaker patent rights. And that could impact the agency's ability to draw high-profile collaborators.
With the patent fight still smoldering and Berkeley contemplating an appeal, it’s not clear yet if the licensing agreements really are a barrier to development. But with Editas, Caribou, and Crispr Therapeutics all expecting to start clinical trials by this time next year, the field may soon find fresh incentive to advocate free Crispr for all.
