Ariad's trek through the legal system has not yet ended! This story has been running for years now - I think the original lawsuit was filed in 2002. Back in the spring, a decision by the Court of Appeals for the Federal Circuit reversed a Massachusetts District Court ruling in Ariad v. Eli Lilly. That decision invalidated a lot of Ariad's key patent claims regarding the Nf-kB signaling pathway, and some of us thought (well, I did) that this would be the end of the story.

But no, Ariad filed a petition in June (PDF) for a rehearing, and that has now been granted. So this fall, the decision will be revisited. It looks like this time, though, the question will not be decided so much on the science and history of Nf-kB, but on a question of patentability.

There are several requirements to get a patent, of course, novelty and utility being the first big ones. You also have to have a complete written description of the invention, and (if you want your claims to stand up) you're going to want to enable them - that is, actually show that you can do what you say, and prove that you have. For pharmaceuticals, that means you need to make real compounds, show physical data for them sufficient to prove that you've made them, and (if you're claiming their effects) show that they do what you're claiming they can do.

The Ariad v. Eli Lilly decision in April turned on written description. Basically, the court held that the company had not described any molecules that could do the vast numbers of things the claims staked out. There was a 1997 case (also, oddly enough, involving Lilly) that raised the standard in that area, and the famous University of Rochester v. G. D. Searle case (COX-2 inhibitors) was decided by applying the same standard. There's been a lot of controversy about the 1997 ruling, though, with many people complaining that the court sort of superglued a tougher written description requirement onto the existing patent law. Ariad has invited the CAFC to take this opportunity to clear things up. That's probably a good thing, since this issue was going to have to be resolved at some point, but it pains me to see Ariad's ridiculous patent case be the means for this.

Personally, I think that Ariad's claims could be tossed by considering the enablement requirement, rather than just written description. (If you think that they didn't do a sufficient job of describing what they wanted to claim, you should see how they reduced it to practice). Here's a post that agrees with that view, and goes into much more detail. It appears, though, that the courts haven't yet come up with a good way to use enablement to chop humungous patent claims down to size. Perhaps this will eventually happen, and the whole written-description era will come to seem like a detour.

I suppose we'll be returning to this issue something this coming winter. Until then, Ariad's patent walks the earth still.