The recent decision on Teva Pharmaceutical Industries Ltd. v. Astrazeneca Pharmaceuticals LP raises the issue of ‘secret prior art’ and also raises speculation on how (if at all) the America Invents Act will change the patent landscape in this regard.

Patent law would be a much simpler affair if all inventions were comprehensively described in patents. However, sometimes the information that supports (or more commonly, disproves) a patent claim exists before the start of the patent process. When you keep this information completely private, it’s called a ‘trade secret’ and rarely, if ever, sways a patent judgment in your favor. When it’s public, it’s called ‘prior art’, and has been a major part of patent law.

Still, there’s a gray area, referred to as ‘secret prior art.' The “secret” label is misleading -- if something can be proven to be applicable prior art, but is not generally available to the public, this information may be taken into account. 35 U.S.C.102 describes two potential circumstances for ‘secret’ prior art, involving a patent application which was not published, and involving a separate invention described in the course of another invention’s patent.

The 1926 decision in the case of Milburn. v. Bournonville (Alexander Milburn Co. v. Davis Bournonville Co., 270 U.S. 390) was a tacit admission that the patent process is often subject to delays of varying amounts -- for example, a later filing can be processed more quickly than an earlier filing, creating the opportunity for the later patent holder to sue the original inventor. Obviously, first-to-file would seem to eliminate this possibility.

However, Milburn said even more; it established that an invention that was disclosed but not claimed in a prior filing qualified as prior art, even though the disclosure was, for the time being, not published and therefore not public in the manner of ‘normal’ prior art. As long as the information was not “abandoned, suppressed, or concealed” (as stated in 35 U.S.C. 102 (g)(2)), it is considered sufficiently ‘disclosed’.

Given the increasing frequency and tendency to overlap, the priority of nearly simultaneous pending patent applications is a major question in, for example, the tech and pharmaceutical sectors -- where the characterization of novel and non-obvious can change in a mere fraction of a year.

The Teva vs. AstraZeneca case was not about the existence of prior art (this was well-established), but whether the compound included in the prior patent was utilized in the way that the later patent specified. To put it (very) simply, it’s almost as if the first company patented a rowboat with sails, only to be sued by the later inventor of the sailboat. This seems silly in the context of common sense, but the precise definition of what constitutes an ‘invention’ formed the basis of Teva’s appeal -- as it has for the vast majority of cases resting on 102(g)(2).

The AIA’s revised section 102 (d) seems to preserve the status of secret prior art exclusively in the matter of pending patent applications, while eliminating the “invention” language of the existing 102 (g)(2). The sort of quibbling over the language of ‘invention’ may be somewhat assuaged by a shift in focus to the filing date, although analysts seem confident that the present decision in this case would stand.

The controversial In re Hilmer case (359 F.2d 859 (C.C.P.A. 1966)) makes specific distinction between filing dates in the US and elsewhere, leading to decisions that generally placed more weight upon secret prior art via pending applications in the US, versus simultaneous or even earlier filings abroad. As the AIA was designed in no small part to ‘harmonize’ US patent law with that of the rest of the world, we can certainly expect In re Hilmer to be cited with less frequency, if not ultimately abandoned altogether.

Still, one can only speculate how the AIA will be interpreted for such cases as Solvay S.A. v. Honeywell International, Inc. (No. 09-1161 (Fed. Cir. Oct. 13, 2010)), in which both the ‘foreign’ aspect and the ‘functional’ aspect were invoked. On the face of it, either argument might seem to be discarded along with the ‘old’ 102(g)(2) language. Still, there’s a crucial element of the AIA that can still come into play -- namely, the one-year grace period for disclosure.

Designed to curtail secrecy and encourage innovation, the AIA grace period encourages inventor or third-party discussion (typically interpreted to include publicity and marketing) of an impending patent filing. Of course, as we’ve noted, a year is an extremely long period of time in the case of certain sectors -- and critics and analysts alike are dubious about whether this will spur innovation or merely chill actual discussion. The exact terms of this grace period -- which, unlike much of the AIA, has little or no correlation to the patent laws of other ‘first to file’ nations -- are likely to be hotly disputed in the courts for some time.

However, the ‘first to invent’ approach will continue to apply in the US until March 16, 2013 -- and, due to the lifespan of a patent, this means that (among other things) we may be quibbling over the definition of “invention” until 2033.