Purdue Pharma L.P. v. Amneal Pharms., LLC

Case Name: Purdue Pharma L.P. v. Amneal Pharms., LLC, Civ. Nos. 04-Md-1603 (SHS); 13-cv-3372 (SHS), 2015 U.S. Dist. LEXIS 45967 (S.D.N.Y. Apr. 8, 2015) (Stein, J.) 

Drug Product and Patent(s)-in-Suit: OxyContin^® (oxycodone); U.S. Pat. No. 8,337,888 ("the '888 patent") 

Nature of the Case and Issue(s) Presented: This case concerns the validity of the '888 patent. The '888 patent resulted from Purdue’s efforts to re-design its OxyContin product to prevent widespread abuse. OxyContin abusers were known to crush up tablets, dissolve them in solution and inject the solution directly into their bloodstreams. Purdue incorporated a gelling agent that caused the tablets to form a thick gel when dissolved in a liquid which was extremely difficult to pull into a syringe. Purdue submitted its NDA to the FDA in November of 2007 proposing the reformulated version, which incorporated the gelling agent. The FDA rejected the application initially, but granted it in April of 2010 after Purdue performed several additional studies. Purdue immediately ceased sales of its original product and began selling the reformulated version. After determining that the new version of OxyContin was effective at combating abuse, the FDA discontinued the original version of OxyContin and announced that it would no longer accept ANDAs for generic versions of the drug.

In July 2011, Amneal filed an ANDA seeking approval to market various dosage strengths of generic, reformulated, OxyContin. Purdue then sued Amneal, alleging infringement of the '888 patent. A five-day bench trial was held in July of 2014 to address Purdue’s allegations of infringement. The court found that Amneal infringed claims 5, 7, 23 and 24 of the '888 patent, but that those claims were all invalid as obvious, and that claim 7 was also invalid as indefinite. The court denied Amneal’s request for attorneys’ fees because Amneal did not establish that the case was exceptional in nature.

Why Amneal Prevailed: The court first addressed two claim-construction issues that arose at trial. The parties disagreed on the method a person of ordinary skill in the art (“POSA”) would use to assess the viscosity limitation in the asserted claims. The parties also disagreed on the substance that was required to impart the viscosity. Amneal argued that it was polyethylene oxide alone (“PEO”). Purdue argued that it was the gelling agent, which could include PEO in addition to other substances. The court found that the patent specification permitted a range of testing temperatures, but did not provide a precise boundary for those temperatures. The court also agreed with Purdue that the gelling agent as a whole, and not just PEO, could impart the viscosity to the solution.

In view of the new claim constructions, after reviewing the evidence before it, the court concluded that Amneal’s ANDA product infringed asserted claims 1, 5, 7, 23, and 24. The asserted claims addressed the viscosity of the tablets when dissolved in solutions at specific temperatures.

Next, the court addressed Amneal’s argument that U.S. Patent No. 6,488,963 (“the ‘963 patent”) anticipated the ’888 patent. The ‘963 patent covered a process for manufacturing a type of sustained-release tablets that incorporated PEO. The court found that the reference did not anticipate the ‘888 patent because it failed to disclose a dose between 2.5 to 320 milligrams of oxycodone. Amneal also argued that a separate WIPO application anticipated the asserted claims. That reference disclosed extended-release dosage forms that delivered an active ingredient at a linear rate of release. The court rejected Amneal’s argument because, although the WIPO application disclosed a tablet containing PEO and 100 milligrams of oxycodone, it did not disclose a 12-hour therapeutic effect. The court also found that the tablet disclosed in the WIPO application would not meet the claimed viscosity limitations described in the ‘888 patent.

With respect to the issue of obviousness, and in particular concerning the issue of incorporating gelling agents, the court noted that the prior art did teach that such agents could be used to combat potential drug abuse. The court specifically cited four prior-art references that taught the incorporation of gelling agents into drug tablets. The gelling agents served to either prevent intravenous injection, or the extraction of the active ingredient from a liquefied solution. The prior art also specifically identified PEO as a compound that could serve as a both a rate-controlling agent and a gelling agent. There were, however, some slight differences between the prior art and the ‘888 patent. For example, the court noted that the prior art did not explicitly teach that gelling agents should be used in conjunction with oxycodone to prevent abuse, but the court did not find this departure to be especially significant. Additionally, the court also found that the prior art did not disclose the quantitative level of viscosity that the gelling agent must produce. Nonetheless, the court found that the OxyContin abuse crisis, which was widely known by early 2001, would have provided the motivation to create an abuse-deterrent formulation. Methods for preventing dissolution and extraction of oxycodone were widely known. It was also well known in the prior art that PEO could provide the dual functions of acting as both a gelling agent and a controlled-release agent. The court also found that the patent’s quantitative-viscosity and tampering limitations were obvious. It would have required very little effort for a POSA to determine the viscosity at which syringing and injection would have become unreasonably difficult. The court also determined that Purdue’s evidence relating to secondary considerations of nonobviousness were not strong enough to rebut this conclusion.

Finally, the court turned to the issue of indefiniteness. Amneal argued that the ‘888 patent’s lack of specific guidance on shear rate, tampering and testing temperature, and extent of dissolution, rendered the claimed viscosity test indefinite. The court rejected this argument with respect to all asserted claims except for claim 7, because the shear rate directly impacted the results of the viscosity test, and therefore the determination of infringement. But the ’888 patent specification failed to provide any guidance that would have allowed a POSA to select the appropriate shear rate.