By Jason Rantanen
Teva Pharmaceutical Industries Ltd. v. Astrazeneca Pharamceuticals LP (Fed. Cir. 2011) Download 11-1091
Panel: Rader, Linn (author), Dyk
"Secret" prior art is prior art that is not available to the public as of the date of invention for the relevant patent. The most well-known category of secret prior art is that of pre-publication patent applications. 35 U.S.C. 102(e). Less commonly asserted is prior invention by another under 102(g)(2). This case illustrates the strength of such prior art under the current novelty framework.
Teva, the owner of Patent No. RE39,502, sued AstraZeneca, asserting that an AstraZeneca drug (CRESTOR) infringed several claims of the '502 patent covering a pharmaceutical formulation stabilized exclusively by an amido-group containing polymeric compound ("AGCP compound"). The problem Teva faced was that AstraZeneca had made the first batches of its drug in the summer of 1999, prior to Teva's asserted invention date of December 1999. Given this prior invention by AstraZeneca, the district court granted summary judgment of anticipation under 102(g)(2). Teva appealed.
Teva's principal argument on appeal was that because AstraZeneca had failed to appreciate the stabilizing role the AGCP compound played in its drug product in 1999, it had not conceived of - and therefore not made - the invention as required by 102(g)(2). This argument rested on the undisputed fact that although AstraZeneca's drug contained an AGCP compound, it was designed to use a non-AGCP compound as the stabilizer; the AGCP compound was believed to be a disintegrant.
Note: Because both the product that Teva accused of infringing and the 1999 batches were the same formulation, AstraZeneca conceded infringement for purposes of its motion for summary judgment of invalidity. In other words, it was undisputed that the AGCP compound was acting as the exclusive stabilizer even if AstraZeneca was unaware of that fact in 1999.
Conception requires an appreciation of what is made, not why it works: Rejecting Teva's argument, the panel concluded that AstraZeneca had met the requirement for conception of its prior invention by appreciating "that the compound it created was stable and what the components of the formulation were." Slip Op. at 12. AstraZeneca did not need to appreciate which component was responsible for the stabilization or why the invention worked.
The "invention" as distinct from the claims for purposes of 102(g): In arriving at its ultimate conclusion of anticipation, the court drew a distinction between the "invention" and the claims, at least for purposes of 102(g). Traditionally, an anticipation analysis involves first establishing that something is prior art and then demonstrating that this prior art contains all the elements of the claimed invention either expressly or inherently. Under a conventional approach to anticipation, then, Teva's claims would be invalid once the prior art status of AstraZeneca's 1999 batches was established since there was no dispute that the batches met all the limitations of Teva's claims.
Section 102(g) operates somewhat differently, however, due to its origins in the interference context. It requires that another inventor have made "the invention" that is at issue. To address this issue, the court in Teva invoked a broad view of what constitutes the "invention" that allowed it to conclude that AstraZeneca had conceived of the same invention as Teva despite AstraZeneca's failure to think of it in the claimed terms. Quoting from Dow Chemical Co. v. Astrao-Valcour, Inc., 267 F.3d 1334 (Fed. Cir. 2001), the court applied a concept of the "invention" as distinct from the claims. "The invention is not the language of the [claim] but the subject matter thereby defined." Slip Op. at 12, quoting Dow (quoting Silvestri v. Grant, 496 F.2d 593, 597, 599 (CCPA 1974)). “The language of the count is but one way to define the new form and certainly not a unique definition [. . . and a]ny claim they might have written, based on this or other information specific to [the new form] would still define the same subject matter as the count even though in different terms.” Slip Op. at 9, quoting Silvestri at 601 (CCPA 1974)). “[T]hat which we call a rose [b]y any other name would smell as sweet.” Slip Op. at 9, quoting William Shakespeare, Romeo and Juliette act 2, sc. 2. Based on this view of the "invention" as a metaphysical entity existing independent of any defining claim language, the court held that AstraZeneca had conceived of Teva's invention, despite doing so in different terms than Teva chose to use in the patent claims.
The Impact of the AIA: With the change to a first-to-file-or-first-to-disclose system, the America Invents Act eliminates 102(g)(2) prior art. The impact of this change is significant in cases such as this one: under the future 102, AstraZeneca would not have been able to bring this validity challenge. Nor is it probable that AstraZeneca could seek protection under the framework of prior user rights: even if its batches met the requirement of being a composition of matter used in a commercial process (there's a strong argument they would not), AstraZeneca's manufacture began less than one year before the filing date of Teva's provisional patent application.
Correction: The first sentence of the original version of this post referred to secret art as art that is unavailable to the public as of the patent's priority date. That is incorrect. Secret prior art is prior art that is unavailable to the public as of the invention date for the patent (the relevant date for purposes of 102(e), (f), and (g)).
Addendum: Several of the comments have pointed to the sentence in the court's opinion that says "By late summer 1999, AstraZeneca had disclosed the ingredients and quantities for its rosuvastatin formulation matching those of all commercial drug dosage strengths," and pointed out that this implies that AstraZeneca's formulation was publicly disclosed prior to Teva's invention date. The court's use of "disclosure," however, appears to just be an unfortunate word choice. According to AstraZeneca's brief, the only relevant events that took place during that time period were the manufacture of the batches and an internal company presentation on the finalized sales formulation of the tablets. Any form of public disclosure took place much later, such as through its patents and applications, the first of which was filed in Great Britain on January 26, 2000.