Purdue Pharma L.P. v. Accord Healthcare Inc.

Case Name: Purdue Pharma L.P. v. Accord Healthcare Inc., No. 20-1362-RGA, 2023 WL 2894939 (D. Del. Apr. 11, 2023) (Andrews, J.) 

Drug Product and Patent(s)-in-Suit: OxyContin® (oxycodone HCl); U.S. Patents Nos. 9,763,933 (“the ’63933 patent”), 9,775,808 (“the ’808 patent”), and 9,763,886 (“the ’886 patent”), collectively “the Abuse-Deterrent patents,” 9,073,933 (“the ’73933 patent”), and 9,522,919 (“the ’919 patent”), together “the Low ABUK patents”

Nature of the Case and Issue(s) Presented: Purdue holds an NDA for OxyContin, an extended-release analgesic. The Abuse-Deterrent Patents claim a reformulation that makes the tablets hard enough to resist crushing and viscous enough to deter intravenous users. The Low-ABUK patents claim oxycodone containing 8α-14-dihydroxy-7,8-dihydrocodeinone (“8α”) and having particularly low levels of the impurity 14-hydroxycodeinone (“14-hydroxy”), an alpha beta unsaturated ketone, or “ABUK” that is a class of compounds thought to be genotoxic. Accord filed an ANDA seeking approval to market generic oxycodone HCl. Purdue filed suit and Accord alleged that the patents-in-suit were invalid as obvious. The court agreed.

Why Accord prevailed: The court first addressed the Abuse-Deterrent patents. The Abuse-Deterrent Patents departed from the prior art in two ways: (i) the prior art that taught sequential compression and heating used polymers other than polyethylene oxide (“PEO”), a significant claimed limitation; and (ii) no prior art used the same combinations of curing time and temperature ranges as those disclosed in the Abuse-Deterrent Patents.

In response to the first issue, the court found prior art referencing PEO tablets could be combined with heating techniques in other references provided there was a motivation to combine the two sets of prior-art references coupled with a reasonable expectation of success. Crediting Accord’s expert’s testimony over that of the inventors, the court found that a POSA would start with the PEO references and the knowledge that abuse by crushing was a real problem. That POSA would be motivated to apply heat to the tablets using an oven, a common and readily accessible tool for heating and curing tablets, which would lead to the claimed invention. Next, the court held that a POSA would have had a reasonable expectation of success after combining the prior art references. Specifically, the court found that “a POSA would need to balance opposing considerations in arriving at the claimed invention, ensuring that the PEO would harden, the active ingredient would not degrade, and the method would be practical” and “would reasonably expect there to be an optimal time and temperature that balances these considerations, discoverable through routine experimentation.” Responding to the second issue, once again crediting Accord’s expert testimony, the court held that the times and temperatures in the Abuse-Deterrent patents are the product of routine experimentation. “[Accord’s expert] testimony aligns with common sense.

Next, the court addressed secondary considerations associated with the Abuse-Deterrent patents. Purdue offered evidence of unexpected results, commercial success, skepticism, and failure of others. Although not claimed, Purdue argued that the decrease in tablet density, resulting in enhanced abuse deterrence, was unexpected. Accord argued that the decrease in density does not contribute to abuse deterrence and is therefore irrelevant to obviousness. The court found that Purdue established evidence of the existence of unexpected results, but that the nature of those results were not sufficient to overcome evidence of obviousness. “The testimony on the impact of a decrease in tablet density was too speculative for me to credit [Purdue’s expert] opinions above [Accord’s expert]—at most, they seem to be in equipoise.” On commercial success, the court found that reformulated OxyContin’s sales are the product of Purdue’s existing monopoly, not the claimed limitations. The court then dismissed Purdue’s evidence of industry skepticism, concluding that the inventor “does not serve as a stand-in for a POSA, or for the industry.” Finally, the court found that Purdue did not prove failure of others.

The court went on to address the Low-ABUK patents. The ’73933 patent claims a composition of at least 95% oxycodone HCl, 8α, and levels below 10 ppm of 14-hydroxy. It further claims that the final product have less than 25 ppm of 14-hydroxy and be made by a particular process that includes removing 8α from an oxycodone base before salt formation and employing HCl in the salt formation step. The ’919 patent claims that the ratio of 8α to oxycodone HCl be 0.04% or less, but allows up 15 ppm 14-hydroxy. After finding that Purdue was not able to antedate its invention date, the court framed the issues concerning the Low ABUK patents: (i) the obviousness of low levels of 14-hydroxy; and (ii) the obviousness of the inventors’ discovery of 8α.

A POSA would have been motivated to modify the prior art to lower the levels of 14-hydroxy as of September 12, 2002, based on a communication from the FDA sent to multiple opioid manufacturers about future ABUK restrictions. “Even if the requirements [to go below 10 ppm] were in the future, I think a POSA would be motivated to get ahead of the requirements.” The court then found that a POSA seeking to reduce 14-hydroxy levels would have a finite, small, and easily identified set of options, and, therefore, a reasonable expectation of success in achieving that goal. Purdue’s expert never testified, and Plaintiffs never argued, that a POSA would have had other options for reducing 14-hydroxy or that a POSA would not have known where to start.

Turning to the issue of whether the disclosure of 8α, either as an independent claim limitation not disclosed in the prior art, or as a necessary step to reaching the 14-hydroxy limitations, renders the patent nonobvious, the court held that the discovery of 8α itself would have been routine for a POSA and therefore does not render the Low ABUK claims patentable. In short, 8α was inherently present in prior art oxycodone compositions. Claim limitations requiring the presence of 8α are obvious because Purdue does not dispute that 8α was present in prior art compositions and identifying it is routine. Claim limitations requiring a certain ratio of 8α to oxycodone HCl were obvious because equally low levels were inherently disclosed in the prior art Lin reference. Finally, claim limitations requiring removing 8α are obvious because its removal in the invention itself is still the result of applying routine techniques to what a highly skilled POSA would have seen as a simple problem.