You may have heard recently in the news about the patent dispute between Ariad Pharmaceuticals and Lilly. Ariad had licensed a patent from Harvard, MIT, and Whitehead that covers use of NF-κB activity. This patent (No. 6,410,516), created by David Baltimore and colleagues, claims much about the NF-κB system — there are a total of 203 claims in the Baltimore patent.

While a patent can claim virtually anything, the patent office's examination should help rule out claims that will not be upheld in court. However, in general there is no guarantee that patent claims will be upheld unless they conform to certain rules. In particular, in the US patent system, patents must be:

• A patentable invention: patentable inventions are any new and useful process, machine, manufacture or composition of matter, or any new and useful improvement thereof.
• Novel: in contrast to the rest of the world, US patents are granted to the first-to-invent, not the first-to-file. The patent will be ruled invalid if the invention:
  • was invented previously by someone else,
  • was described in a US patent application that is later granted with an earlier priority date, or
  • was known or used by others in the USA or described in a printed publication in any part of the world.
• Non-obvious: One must convince the patent office that the invention would not be "obvious" to someone else with similar knowledge at their disposal.

(from Patents for Chemicals, Pharmaceuticals and Biotechnology, 4^th Ed., Philip W. Grubb)

The claimed methods as described in the Baltimore patent are to me of questionable specificity. The researchers discovered a process of nature, then patented many of the obvious theoretically possible therapeutic routes dealing with that natural process. While US patent law is sometimes a little vague about the novelty criterion, it is generally accepted that processes or phenomena of nature can not be inventions, but methods which harness natural processes are fair game. So is it obvious, after discovering NF-κB's natural function, that one could inhibit its activity for therapeutic benefit? That's perhaps a more nuanced question than it first seems, so I'll let you decide.

From what I understand, Ariad is suing Lilly because Ariad claims Lilly's Evista and Xigris drugs modulate the activity of NF-κB. In its response to the case, Lilly asserted that they had "discovered these drugs and disclosed their medicinal properties years before the patentees' scientists made their discovery." At first blush, I'm inclined to agree with Lilly, but I certainly haven't heard all the details of the case.

I suspect the patent will not be upheld on appeal. As a matter of commerce, if Ariad were to continue winning in court, most pharmaceutical research would come to a halt. This is really a test case, because if Baltimore, et al, can in effect patent the entire NF-κB signalling cascade, entire companies will stop small-molecule R&D out of fear of patent infringement.

However, before the courts is not a matter of commerce, but a matter of law. Although normally I think I would root for the scientists and academics, I think this patent just might be overly broad...

UPDATE: I almost forgot that about two years ago, a friend of mine had the pleasure of meeting David Baltimore when he was giving a talk down at Stanford. She got to drive him around campus on a little golf cart to take him to various labs, and said he told some great stories... Incidentally, Baltimore is now the president of Caltech, and Giao has his (actual) signature on her diploma. Sheesh it's a small world.

UPDATE: See also my more recent post: "Ariad v. Lilly, revisited"