Here's an idea. If your company is looking for a bit of extra income to support its research and clinical programs and you've gone to all your usual sources of financing and still come up short, why not trawl around to find a piece of intellectual property (IP) from three prominent academic centers that stakes a claim to virtually every therapeutic approach under the sun modulating a pivotal pathway in biology—say one central to inflammation, cancer and osteoporosis, for starters—and then license it exclusively. It doesn't really matter whether the IP is related to your company's R&D interests; the important thing is that the patent is broad enough to make tens (perhaps hundreds) of drug companies subservient to your license. Simply send out 50 or so letters to your (former) friends and counterparts at companies around the nation informing them that they are infringing your patent and ask for an appropriately exorbitant level of remuneration. Sit tight and wait for the money to roll in.

Though it may seem far-fetched, this situation is essentially what has transpired in the case of US Patent No. 6,410,516. This is one of a family of patents covering methods of treating human disease by regulating nuclear factor (NF)-κB, exclusively licensed in 1991 from the Massachusetts Institute of Technology, the Whitehead Institute for Biomedical Research and Harvard University by Cambridge, Massachusetts-based biotech Ariad Pharmaceuticals. Four years ago, to test its patent's mettle, Ariad filed suit in the US District Court in Massachusetts, accusing Indianapolis-based pharma giant Eli Lilly of patent infringement.

On May 4, a federal jury in Boston ruled Ariad's patent to be both valid and infringed by two Lilly drugs, small molecule Evista (raloxifene) and recombinant protein Xigris (drotrecogin alfa), currently marketed for use in osteoporosis and septic shock, respectively. The jury awarded Ariad ∼$65 million in back royalties and a healthy 2.3% royalty on future US sales of the two drugs until the patent's expiration in 2019. The sum should cover Ariad's operating losses in 2005 (see p. 625), if not its legal fees.

The decision has sent shock waves through the industry not least because Ariad's patent, and similar method-of-treatment patents like it, could fence off whole swaths of biology, preventing other innovators from developing medicines because they may trespass (however tangentially) upon the patented pathway(s). Ariad's patent is particularly worrisome because NF-κB regulates the expression of more than 175 other genes and is involved in the mechanism of >200 marketed compounds, including aspirin, antibiotics and such biotech drugs as Velcade (bortezomib), Enbrel (etanercept) and Kineret (anakinra). To make matters worse, NF-κB is involved in virtually every disease you can think of, including cancer, arthritis, chronic inflammation, asthma, neurodegenerative disease and heart disease. Almost no therapeutic indication is safe from its clutches.

Indeed, the IP lawyers appear to have gone to extraordinary lengths to ensure that anything that comes within so much as a whiff of NF-κB will be drawn into the '516' patent's black hole. As one industry insider puts, the patent's claims—an eye-popping 203 of them in all—are a “relentless paving machine, spreading hot asphalt on everything in sight and spraying lane markers for the toll booths.” No wonder the patent took 16 years of prosecution at the US Patent and Trademark Office (USPTO) before eventually issuing.

This is not the first instance of a patent describing therapeutic methods that modulate a natural pathway. Other examples include the University of Rochester's US patent awarded in 2000 on methods for inhibiting the cyclooxygenase 2 (COX-2) pathway to create anti-inflammatory drugs, Pfizer's patents on modulation of the phosphodiesterase 5 pathway in erectile dysfunction, and Cambridge, Massachusetts-based biotech Curis' patent on blocking Hedgehog signaling proteins in cancer. There are many others. Intriguingly, Lilly itself has also deployed the same IP strategy; in 1996, it filed an international patent application on Evista entitled—yes, you guessed it—“Methods of Modulating of NF-κB Pathway” (one wonders how well this went down with the Boston jury).

None of these patents, which have much narrower claims than the Ariad patent, has stood up to courtroom challenge. That is why the Boston court's decision to uphold Ariad's patent is so surprising. It goes against long-standing practice that patents should not be issued that would remove known materials from the public domain (in this case, Evista, Xigris and the 200 or so other marketed drugs brushing the NF-κB pathway). It ignores Lilly's original work on the discovery and identity of Evista and Xigris, which dates back to the 1980s, long before the 1991 priority date for Ariad's patent. And it disregards the patent's failure to describe methods that “enable any person skilled in the art” to make an NF-κB inhibitor “without undue experimentation.”

If this decision is upheld—and it's a big if, given the ongoing reexamination of the patent by the USPTO, Lilly's request for the Boston judge to overturn the jury's verdict, the likelihood of an appeal, a separate bench trial that will examine the patent's enforceability and yet another suit from Amgen in the US Court for the District of Delaware for a declaratory judgment of invalidity and noninfringement—the flood gates could open for 'method-of-treatment' patents. This would be bad news for innovative drug developers, bad news for patent examiners and bad news for the courts, which will have to untangle a burgeoning morass of methods and composition-of-matter patents with conflicting claims and priority dates.

Patents are supposed to encourage invention, commercialization, disclosure and societal benefit in return for a limited market monopoly. They were never intended as a means for a single company to hold the rest of the industry to ransom. The courts must now act swiftly to invalidate this patent. And, more importantly, they must provide clearer guidance to the industry and USPTO as to the proper scope of patentable subject matter. Let's hope patent '516' goes down in flames. The sooner, the better.