This article is provided to FT.com readers by Pharmawire—a news service focused on providing insight into the most price sensitive issues in the global pharmaceutical market. www.pharmawire.com
--------------------------------------------------------------------------------------------------------
A determination of whether Takeda made a fair and accurate representation of data in its patent application will be factored into the final judgement of Takeda v Teva Pharmaceutical, a number of lawyers have told Pharmawire. The two companies are awaiting conclusion from Delaware Judge Sue Robinson as to whether Teva has presented enough evidence to invalidate the patents on Takeda’s proton pump inhibitor Prevacid (lansoprazole). The two patents in question are the 4,628,098 structure and 5,045,321 formulation patents.
Calls to Teva were not returned.
One of the tactics Teva has employed is the assertion of inequitable conduct. This is an attempt to show that Takeda knowingly withheld information from the US Patent Office during its patent application for lansoprazole.
Most lawyers point out that inequitable conduct is a defence frequently raised but very rarely proved. The Federal Circuit often urges restraint in such cases because of the irreparable damage it can cause to individuals, one lawyer pointed out.
Yet Robinson is well known for determining in favour of an inequitable conduct charge asserted by Bayer in its lawsuit against Housey Pharmaceuticals. The interesting twist in this particular case was that Robinson’s initial determination in finding Housey incredible was rejected by the Court of Appeals for the Federal Circuit. However she responded to the rejection with a character assassination of the defence’s key witness, Dr Hausey, pointing to evidence that he fabricated data. “She basically accused him of lying,” a person familiar said.
Housey’s patents were subsequently invalidated because of Robinson’s credibility determination. The case was seminal: up until then a judge would not have been expected to go out of the way to determine that a person skilled in the art could intentionally fabricate data, especially after a CAFC remand.
But as one lawyer points out, “intent is an inherently fact-intensive inquiry that invariably depends on the totality of circumstances and frequently involves credibility determinations by the trial judge”.
In the case of Takeda, Teva has accused the Japanese company of withholding key data from the US patent office that showed lansoprazole as an obvious progression from prior art PPI compounds. Specifically, it argued that Takeda relied on a single test model in the prosecution of its patent to the PTO which showed lansoprazole to be 20 times superior to omeprazole, despite being in possession of multiple animal models with different levels of effectiveness. For example, it noted that the Japanese company had an animal model in dog that showed omeprazole to be superior to lansoprazole.
The point that Teva wanted to prove was not whether the withheld data would invalidate the patent because the Federal Circuit argues for “restraint in applying a defence that prevents enforcement of a valid patent” (Preemption Devices vs 3M, 1984), but rather whether the examiner would have wanted to consider the data in determining the patentability of lansoprazole.
In withholding this data, Teva argued that Takeda intended to mislead the PTO in bid to obfuscate the patentability of lansoprazole. In support of Teva’s argument, as a person familiar with the case pointed out, is the Federal Circuit, Cargill v Canbra Foods, as this decision is an important steer in determining inequitable conduct through data non-disclosure.
“[Cargill v Canbra] imposes a high burden on applicants to disclose data even if the data itself is distinguishable from the disclosed data and might not have affected patentability,” the source said.
In response, Takeda argued that the additional data is irrelevant and furthermore actually supports the inventors’ assertions that the 098 patent purports lansoprazole’s superior anti-ulcer effects.
As one lawyer pointed out, there is no US Patent Law that stipulates that every animal test conducted needs to be submitted in support of patentability. Furthermore as most clinicians would caution, animal models can never be a true proxy for humans.
Takeda pointed out in its post-trial brief that dogs cannot be used for anti-ulcer models because they would need to be sacrificed. Moreover, the Teva evidence had mistakenly focused on anti-secretory data and not the anti-ulcer IMAU model that supported the 098 application for lansoprazole, which was being claimed for its superior anti-ulcer effects.
The Takeda line of argument does have legal precedents that are counterpoints to Cargill since the failure to submit data does not in itself infer inequitable conduct. The Federal Circuit notes that information that is merely cumulative or amounts to just a simple omission from that already in possession by the PTO examiner can be regarded from a point of view of disclosure as a level of materiality that is not high (Litton Sys v Honeywell, 1996; Purdue v Endo Pharmaceuticals, 2006).
The final major argument that Teva proffered was that Takeda failed to disclose the IMAU model it used for its patent application when filing an IND with the FDA. But as Takeda pointed out that this was irrelevant to the PTO prosecution of the lansoprazole patent because in the case of the IND application, its aim was solely to obtain regulatory approval from the FDA by selecting “the models used by prior successful applications for anti-ulcer agents”. As the IMAU model was quite novel it had not been used in previous IND applications, it stated.
Perhaps the most important decision to the Takeda v Teva case with regards to intended non-disclosure is the one made by Judge Joseph Farnan in Pfizer v Ranbaxy Laboratories, 2005. Farnan stated that in the case of Lipitor he was not persuaded that Pfizer (fka Warner-Lambert) had manipulated or “cherry picked” data with an intent to achieve a deceitful result even though it failed to disclose certain in-vivo experiments cited by Ranbaxy that did not demonstrate significant difference in activity. “Pfizer has advanced reasonable and credible grounds for the non-production of certain data that weigh against a conclusion that Warner-Lambert scientists and employees were intentionally deceiving the PTO,” Farnan stated.
One legal source pointed out that while other aspects of Farnan’s decision to uphold the patent were rebutted by the Federal Circuit, this particular ruling of non-inequitable conduct has remained unchallenged. They also said that Farnan is well-respected by Judge Robinson
--------------------------------------------------------------------------------------------------------
For more information or to inquire about a trial please email sales@pharmawire.com or call Americas: +1 212-500-1384 or Europe: + 44 (0)20 7059 6251
