by Dennis Crouch
The Federal Circuit recently affirmed the invalidity of several Purdue Pharma patents related to abuse-deterrent and low-impurity formulations of OxyContin. Purdue Pharma L.P. v. Accord Healthcare, Inc., No. 2023-1953 (Fed. Cir. Dec. 30, 2024). Although non-precedential, the appeal addresses three significant legal questions in obviousness jurisprudence: (1) The consideration given to discovering the source of a previously known problem under Eibel Process Co. v. Minnesota & Ontario Paper Co. (1923); and (2) how inherent properties of prior art compositions factor into obviousness analysis when combining multiple references similar to the Cytiva decision from earlier in 2024.
The appeal involved two sets of patents: (1) “Abuse-Deterrent Patents” (U.S. Patent Nos. 9,763,933, 9,775,808, and 9,763,886) claiming crush-resistant formulations using polyethylene oxide (PEO), and (2) “Low-ABUK Patents” (U.S. Patent Nos. 9,073,933 and 9,522,919) directed to reducing potentially genotoxic impurities in oxycodone.
In August 2020, Accord filed an Abbreviated New Drug Application (ANDA) seeking FDA approval to market a generic version of OxyContin, and including a paragraph IV certification claiming that Orange Book listed patents were invalid. Under the Hatch-Waxman Act framework (35 U.S.C. § 271(e)(2)(A)), the mere act of filing an ANDA with a paragraph IV certification constitutes an artificial act of infringement, allowing the patent holder to sue before actual generic sales begin. This gives courts a chance to resolve patent disputes in a declaratory manner before generic launch. This sort of declaratory judgment litigation is typically seen as aligned with traditional courts of equity, and so the Seventh Amendment right to a jury trial does not apply. Following this framework, Purdue sued Accord in Delaware district court, and after a three day bench trial Judge Andrews sided with the patent challengers — finding the claims obvious. Purdue Pharma L.P. v. Accord Healthcare, Inc., 669 F. Supp. 3d 286 (D. Del. 2023).
Obviousness Analysis of the Abuse-Deterrent Patents
As I’m sure almost all Americans are aware, oxycodone, the key ingredient of OxyContin, has been subject to widespread abuse. The original formulation tablets could be easily crushed and then snorted or injected in ways that produced a much more immediate and stronger high. The abuse-deterrent patents use polyethylene oxide (PEO) polymer that make the tablets physically harder to crush while also making them form a gel when combined with liquid – properties that help prevent these common methods of abuse.
Abuse-Deterrent Patents, a key issue was whether it would have been obvious to modify prior art processes (which used simultaneous compression and heating) to instead use sequential compression and heating steps. The court found that a person of ordinary skill would have been motivated to modify the prior art processes for large-scale production, with ovens being an obvious choice for heating tablets. The product-by-process patents claim a specific sequential manufacturing process:
- First, the tablet is “compression shaped”
- Then it is “air cured by heated air, without compression”
- The heating must be done for “about 10 minutes to about 10 hours”
- The heating must occur above PEO’s softening temperature (at about 70-85°C or 65-90°C)
Accord identified prior art that seemed to teach all of the elements, but not all together or in the particular sequence. Purdue argued that no reference had ever taught curing PEO tablets using heated air without simultaneous compression at the specific times and temperatures claimed. While the two key references (Bartholomaus and McGinity) taught using PEO, they used simultaneous compression and heating. And while a third (Oven Art) reference taught sequential processing (compression then heating), they dealt with different polymers, not PEO.
Purdue identified what it called a “surprising benefit” of their sequential process – it led to decreased tablet density that promoted faster gelling both (1) in the nasal cavity if someone tried to snort the crushed tablet; and (2) in other liquid if someone tried to inject it.
The courts were ultimately unpersuaded by these arguments. For the sequential processing, the court noted that Bartholomaus actually contemplated “subsequent exposure to heat” as one option, even if it wasn’t the focus. The court also found that scaling up production would have naturally led someone to consider sequential processing using heated air ovens that were already in use in the pill-making business.
As for the surprising benefit, while the court acknowledged this was an unexpected result, it found this secondary consideration insufficient to overcome the strong evidence of obviousness from the combination of references.
Inherency in Obviousness Analysis
The Low-ABUK Patents presented interesting questions about inherency in obviousness analysis. Following PAR Pharm., Inc. v. TWI Pharms., Inc., 773 F.3d 1186 (Fed. Cir. 2014), the court reaffirmed that inherency can supply missing claim limitations in an obviousness analysis. Some of the claims here required the presence of 8α – a chemical impurity that forms during the process of making oxycodone. The prior art relied upon did not expressly identify the presence of 8α, but the court found the limitation inherently present in prior art compositions – something not disputed by Purdue.
Purdue argued that inherent features should not be available for a prior art analysis that requires combining multiple references together. But, the Federal Circuit recently rejected this argument in Cytiva BioProcess R&D AB v. JSR Corp., 122 F.4th 876 (Fed. Cir. 2024). The black letter law here appears to be that an inherent feature cannot be used to support a motivation to combine. Honeywell Int’l Inc. v. Mexichem Amanco Holding S.A. de C.V., 865 F.3d 1348 (Fed. Cir. 2017) (“that which may be inherent is not necessarily known and that which is unknown cannot be obvious” internal quotation marks omitted). But, an inherent feature of a reference can be used for the obviousness analysis when references are combined for other reasons unrelated to the inherency. Cytiva.
Eibel: Discovery of Problem Source Does Not Guarantee Patentability
In its appeal to the Federal Circuit, Purdue leaned heavily on the Supreme Court’s century-old decision in Eibel Process Co. v. Minnesota & Ontario Paper Co., 261 U.S. 45 (1923), arguing that where an inventor discovers a previously unknown problem or source of a problem, the solution is almost automatically non-obvious. Eibel involved high-speed paper-making machines where paper quality deteriorated at higher speeds. Eibel discovered that the problem stemmed from unequal speeds between the paper stock and the wire mesh used to form the paper. His solution — tilting the wire at a higher angle to use gravity to accelerate the stock — led to significant improvements in paper production speed while maintaining quality.
The parallels to Purdue’s case seemed compelling at first glance. Like Eibel, Purdue claimed to have discovered an unknown source of a problem — specifically that 8α impurity was converting to 14-hydroxy during salt formation in oxycodone production. And like Eibel, Purdue developed a solution once the true nature of the problem was understood. The Federal Circuit acknowledged that “it is undisputed that ‘[t]he Low ABUK Patents were the first to report the presence of the molecule 8α in the synthesis of oxycodone.'”
The Federal Circuit, however, distinguished Eibel in two key ways. First, unlike in Eibel where the problem itself (paper quality degradation at high speeds) was previously unknown, here the underlying problem of 14-hydroxy content in oxycodone was well-known. The court noted that “testimony at trial . . . indicated that an understanding or suspicion that ABUKs were toxic existed even before September 2002.” Second, while Eibel’s case turned largely on the discovery of the problem’s source, the Federal Circuit emphasized that “even recognizing that Purdue may have discovered 8α, we disagree that ‘[t]hat should have ended the inquiry.'”
The Federal Circuit rejected Purdue’s sweeping argument that discovering “the source” of a problem automatically renders an invention nonobvious. While the court acknowledged Eibel as relevant precedent, it emphasized that “even Eibel Process demonstrates that obviousness is based upon underlying factual questions.” The court pointed to Eibel‘s detailed discussion of different factual scenarios that could lead to different patentability outcomes. The Federal Circuit then connected this to KSR’s warning against “rigid preventative rules that deny factfinders recourse to common sense.” The law is that discovery of the source of the problem will support non-obviousness of the solution, but is not determinative and does not automatically confer patentability. Instead, it becomes one factor in the overall factual analysis of whether the claimed invention would have been obvious to a person of ordinary skill in the art. Here though the court concluded that a person of ordinary skill would have “quickly postulated and easily confirmed” the existence of 8α through routine testing.
= = =
Objective Indicia of Non Obviousness: The Federal Circuit affirmed the district court’s analysis of the secondary considerations arguments. For commercial success, while Purdue pointed to OxyContin’s market performance after reformulation, the district court found no proven nexus since the new formulation simply replaced the old one without demonstrating increased success relative to other opioids. On industry skepticism, although Purdue emphasized FDA’s initial hesitance to apply an abuse-deterrent label, the court reasoned this reflected normal regulatory caution for a first-of-its-kind designation rather than technical skepticism about the invention itself. Regarding failure of others, Purdue highlighted two examples, but the court found those production failures actually supported the motivation to improve manufacturing processes. Finally, on unexpected results, while the district court acknowledged that the decreased tablet density was unexpected and could enhance abuse deterrence through faster gelling if crushed, it concluded this single secondary consideration was insufficient to overcome the strong evidence of obviousness from the combination of references.
Gregory Garre of Latham & Watkins argued for Purdue Pharma, with support from Alexander Siemers, Margaret Upshaw, Daniel Brown and David Kowalski (all from Latham). Additional attorneys on Purdue’s side included Gregory Castanias, Jennifer Swize, Gasper LaRosa, and John Normile from Jones Day, as well as Pablo Hendler from Potomac Law Group. Ben Mahon from McAndrews, Held & Malloy argued for defendant-appellee Accord Healthcare, with Bradley Loren and Alejandro Menchaca also on the briefs. The three-judge Federal Circuit panel consisted of Circuit Judges Prost, who authored the opinion, along with Judges Reyna and Taranto.