Opinion
Civil Action 22-913-WCB
09-09-2024
PURDUE PHARMA L.P. and PURDUE PHARMACEUTICALS L.P., Plaintiffs, v. ACCORD HEALTHCARE INC., Defendant.
FINDINGS OF FACT AND CONCLUSIONS OF LAW
WILLIAM C. BRYSON UNITED STATES CIRCUIT JUDGE
This patent case was tried to the court during the week of February 12, 2024. The parties have submitted post-trial briefs in the form of proposed findings of fact and conclusions of law. Dkt. Nos. 96, 97, 98, and 99. This opinion contains the court's findings of fact and conclusions of law. A judgment will be entered in accordance with these findings and conclusions.
I. Procedural Background
The technology at issue in this case is directed to tamper-resistant formulations for delivering drugs to patients. Certain pharmaceuticals, including opioids, are subject to abuse in the hands of persons other than patients needing them for pain relief. Opioid abusers often seek to access the drugs by crushing pills containing the opioids and extracting the opioids for illicit use. The industry has responded by seeking to devise extended-release pharmaceutical dosage formulations that are prepared in a way that makes the tablets containing the drugs highly crushresistant. The patent at issue in this case, U.S. Patent No. 11,304,908 (“the '908 patent”), which belongs to the plaintiffs Purdue Pharma L.P. and Purdue Pharmaceuticals L.P. (collectively, “Purdue”), is directed to one such formulation.
The '908 patent is listed in a U.S. Food and Drug Administration (“FDA”) publication entitled Approved Drug Products With Therapeutic Equivalence Evaluations, which is familiarly known as the “Orange Book.” Purdue asserts that the patent protects a product for which Purdue filed a New Drug Application with the FDA and which the FDA has approved for distribution in this country. See PTX-125. That product is a tamper-resistant extended-release pain relief medicine in matrix tablet form. The tablets contain the drug oxycodone hydrochloride (“oxycodone”), which is the active ingredient in each of Purdue's OxyContin products.
The plaintiffs' trial exhibits are denominated as “PTX.” The defendant's trial exhibits are denominated as “DTX.”
A matrix tablet contains an active drug dispersed in an inert material. The dispersion of the active drug within the tablet and the properties of the inert material, commonly a polymer, serve to regulate the rate at which the drug is diffused out of the tablet, and hence produces an “extended release” effect by gradually releasing the drug over time after the patient has ingested the tablet.
The defendant, Accord Healthcare Inc., submitted an Abbreviated New Drug Application (“ANDA”) to the FDA seeking permission to distribute a generic version of Purdue's OxyContin product. Purdue then filed this action in July 2022 alleging that Accord had infringed the '908 patent under 35 U.S.C. § 271(e)(2)(A) by the act of filing the ANDA for its generic product. Accord responded that the asserted claims are invalid and that Purdue's infringement action in this case is barred by collateral estoppel based on a prior judgment against Purdue on three related patents. See Purdue Pharma L.P. v. Accord Healthcare Inc., 669 F.Supp.3d 286 (D. Del. 2023) (“Accord I”), appeal docketed No. 23-1953 (Fed. Cir. May 30, 2023).
II. The Asserted Claims
Purdue has asserted claims 1, 10, 18, 23, 28, and 29 of the '908 patent in this action. Asserted claims 10, 18, 23, 28, and 29 are dependent claims. Asserted claim 10 depends from claims 1, 5, and 6. Asserted claim 18 depends from claims 1 and 17. And claims 23, 28, and 29 depend from claim 21. The asserted claims, along with the claims from which the dependent claims depend, are set out below:
1. A solid oral extended release pharmaceutical dosage form, comprising a shaped, convection heated, and cooled extended release matrix, said matrix comprising at least one polyethylene oxide (PEO) having, based on rheological measurements, an approximate molecular weight of at least 800,000, and at least one opioid analgesic, wherein (a) the shaped matrix is convection heated to an elevated temperature that is at least the softening temperature of said PEO for a time period of at least about 1 minute and thereafter cooled; and (b) a plurality of convection heated particles of PEO adhere to or fuse with each other within the matrix.
5. The dosage form of claim 1, wherein the total PEO content of the shaped matrix is at least about 15% (by weight) of said matrix.
6. The dosage form of claim 5, wherein said time period is at least about 5 minutes and said elevated temperature is at least about 55° C.
10. The dosage form of claim 6, wherein said elevated temperature is at least about 68° C.
17. The dosage form of claim 1, wherein the extended release matrix is shaped to form a tablet and heated in a coating pan; said time period is at least about 5 minutes; and said elevated temperature is at least about 60° C.
18. The dosage form of claim 17, wherein the opioid analgesic is oxycodone or a pharmaceutically acceptable salt thereof.
21. A solid oral extended release tablet, comprising a shaped, convection heated, and cooled extended release matrix, said matrix comprising at least one polyethylene oxide (PEO) having, based on rheological measurements, an approximate molecular weight of at least 800,000, and at least one opioid analgesic, wherein (a) the shaped matrix is convection heated to an elevated temperature of at least 55° C. for a time period of at least about 15 minutes and thereafter cooled at a temperature below 50° C.; and (b) a plurality of convection heated particles of PEO adhere to or fuse with each other within the matrix.
23. The table of claim 21, wherein the total PEO content of the matrix is at least about 50% (by weight) of the composition.
28. The tablet of claim 21, wherein said PEO has, based on rheological measurements, an approximate molecular weight of at least 4,000,000.
29. The tablet of claim 21, wherein the dosage form further comprises at least one of a cellulosic additive, magnesium stearate, talc, silica, fumed silica, colloidal silica dioxide,
calcium stearate, carnauba wax, stearic acid, stearyl alcohol, mineral oil, paraffin, glycerin, propylene glycol, polyethylene glycol, lactose, povidone, triacetin, and copolymers comprising methyl methacrylate.
Based on the language in claims 1 and 21 referring to “a shaped, convection heated, and cooled extended release matrix,” Purdue characterizes the asserted claims as being limited to matrices that are shaped (through the application of pressure), then heated (without the application of pressure), and then cooled. While the specification contains various descriptions and examples in which such a sequence is contemplated, see, e.g., JTX-1 at col. 21, ll. 41-56, the claim language itself does not necessarily foreclose a process in which the heating, like the shaping, is done while the matrix is compressed. However, Accord has not argued in favor of a broader claim construction that would permit such a process; in fact, Accord appears to have acquiesced in Purdue's construction of the claim language. See Accord's Post-Trial Reply Brief at 4 (“The claims here recite products-the same products already addressed in Accord I-i.e., PEO opioid formulations made by heating without compression.”). I will therefore apply what appears to be the agreed-upon construction, in which claims 1 and 21 (and their dependent claims) require that the matrices be under pressure during the shaping phase but not during the subsequent heating phase.
III. Findings of Fact
A. Background of the Invention
Oxycodone hydrochloride (or “oxycodone”), an opioid, is the active ingredient in OxyContin. It was developed in 1916 and has been used to treat pain since that time. Dkt. No. 86-1 ¶ 12. It is also subject to abuse, and if its use is not carefully controlled, it can lead to addiction and death. Id. ¶ 16; PTX-222.
In 1995, Purdue released OxyContin, which was an extended release oxycodone formulation. Tr. 207:4-11. It was known as “original OxyContin” or “OxyContin ER.” Because 4 OxyContin ER contained enough oxycodone in each tablet to provide 12 hours of pain relief, the tablets were very attractive to potential abusers. If an abuser could release and ingest the entire amount of oxycodone in the tablet at once, the drug would produce an intense euphoric effect. In order to obtain the full amount of oxycodone in each tablet for immediate release, abusers with access to OxyContin tablets would crush the tablets, extract the drug, and either inhale the drug or dissolve it to form an injectable solution. Tr. 209:3-215:14; PTX-311. The abuse of oxycodone became recognized as a serious problem in the years shortly after OxyContin ER was released. Tr. 207:12-18.
As the problem of oxycodone abuse became more widespread, the industry sought ways to make it more difficult for abusers to extract the oxycodone from the tablets. Two principal approaches were considered: (1) to add an “antagonist” to the oxycodone formulation, which would render the drug unusable for abuse; and (2) to make the tablets tamper-resistant so that abusers could not readily crush the tablets to release the drug inside. Tr. 215:18-222:18. Other possible strategies for abuse deterrence that Purdue and other companies considered during the period immediately before Purdue obtained the '908 patent were embedding the drug in a sticky polymer so that the drug could not be ground into a fine powder; use of a prodrug that would have to be metabolized by the liver to be converted into oxycodone; and adding an “aversive agent” to the formulation that would make the drug very unpleasant to ingest nasally or by chewing. Tr. 222:19-226:7; PTX-160; PTX-279.
In this context, an antagonist is a drug that blocks or dampens a biological response by binding to and blocking a receptor to which an active ingredient such as oxycodone would otherwise bind and produce a response. Ideally, the antagonist is not activated by ingestion of the extended release formulation, but is activated by attempts to extract the active ingredient from the formulation by crushing the tablet, solubilizing the drug, and injecting it. PTX-137.
On February 10, 2005, Johannes Bartholomaus and Heinrich Kugelmann published United States Patent Application Publication No. U.S. 2005/0031546 (“DTX-10”), which was directed to drug-containing tablet formulations made tamper-resistant by hardening the tablets. That application was assigned to Grunenthal GmbH, a German pharmaceutical company. DTX-10 (“Bartholomaus”). Among other anti-abuse techniques, Bartholomaus taught using an opioid drug formulation containing polyethylene oxide (“PEO”) having a molecular weight of at least 500,000 and up to 15,000,000, press-forming the formulation and heating it to the melting point of the PEO through preceding, simultaneous, or subsequent exposure to heat. Bartholomaus reported that the resulting tablets were extremely hard. They did not break, even when exposed to a force of 500 Newtons. In addition, the tablets could not be pulverized into pieces having an edge length smaller than 2 millimeters, and when the pieces were combined with water, a highly viscous gel was formed that could not easily be pressed through an injection needle. Id. at 12-13.
In April 2005, Purdue employees, including Dr. Richard Mannion, one of the inventors on Purdue's '908 patent, reviewed the Bartholomaus application and visited the Grunenthal facilities to learn about the process described in the application. Tr. 227:5-228:22. Within months thereafter, Purdue had successfully produced a crush-resistant product by curing PEO tablets in a manner generally similar to that described in Bartholomaus. While the curing method described in the Bartholomaus application cured the tablets in a tablet press, a device that was not readily scalable for the purpose of producing commercial quantities of tablets, Purdue used a different heating mechanism to cure its tablets: a type of oven known as a coating pan, which could cure hundreds of kilograms of tablets at a time. Tr. 87:16-24. In August 2006, Purdue filed a provisional patent application claiming the formulation. That application ultimately matured into the '908 patent.
Purdue sought FDA approval for its abuse-resistant oxycodone formulation, noting that it was bioequivalent to Purdue's predecessor product, OxyContin ER, the extended-release formulation of oxycodone. Tr. 298:16-299:5; PTX-103, PTX-124. In 2010, the FDA approved the new formulation, and that formulation subsequently replaced OxyContin ER on the market. PTX-125. Three years later, the FDA approved Purdue's application to have the OxyContin ER label revised to reflect that the new tablets were abuse-deterrent. At that time, the FDA required Purdue to remove the Original OxyContin from the market so that it could be replaced by the abuse-deterrent formulation. PTX-229.
Accord thereafter filed an ANDA seeking the FDA's approval to market a generic version of Purdue's tamper-resistant OxyContin product. After Accord filed its ANDA, Purdue filed this action seeking a declaration that Accord's proposed tamper-resistant product would infringe the '908 patent. Accord responded by asserting that the '908 patent was invalid for obviousness over Bartholomaus and other prior art.
B. The Accord I Decision
Prior to this case, Purdue brought a separate action against Accord, asserting patents from the same family as the '908 patent. See Purdue Pharma L.P. v. Accord Healthcare Inc., 669 F.Supp.3d 286 (D. Del. 2023) (“Accord I”). That case, which was assigned to Judge Andrews, involved three patents directed to tamper-resistant tablet formulations, together with two other unrelated patents. The tamper-resistant patents at issue in that case were part of the same family of patents as the '908 patent at issue in this case. Like the '908 patent, two of the tamper-resistant patents at issue in Accord I claimed PEO matrix tablets. The third tamper-resistant patent at issue in Accord I claimed methods of making those tablets. Like this case, Accord I arose from ANDAs filed by Accord, and in that case, as in this one, Bartholomaus was central to the district court's analysis.
The court in Accord I held the asserted claims of all three tamper-resistant patents invalid. Judge Andrews' decision in Accord I deserves close attention both because the defendant is asserting collateral estoppel in this case based on Accord I, and because even if collateral estoppel does not apply, Accord I is sufficiently similar to this case that Judge Andrews' analysis in that case bears significantly on this case as a matter of stare decisis.
Based in part on a prior art patent to McGinity, U.S. Pat. No. 6,488,963, which was directed to tamper-resistant tablets for analgesics, Judge Andrews found that a person of ordinary skill in the art (“POSA”) would have understood that because the prior art tablets were made from melted PEO, they “had increased breaking strength that provided resistance to crushing.” 669 F.Supp.3d at 296.
McGinity, which used a hot melt extruder to heat the PEO formulation, was introduced into evidence in this case as DTX-009, but was not relied on by Accord's expert witness to support her invalidity analysis in this case.
In Accord I, Judge Andrews found that the equipment used to heat the PEO in Bartholomaus's examples was not scalable, but that a person of skill in the art would have understood that the purpose of heating the tablet was to melt the PEO. Id. Judge Andrews further found that Bartholomaus taught the steps of compressing the PEO formulations into tablets and heating them, either simultaneously or as separate steps. Id. He also found that Bartholomaus taught cooling and hardening of the PEO tablets. Id.
Analyzing the prior art, Judge Andrews found that although Bartholomaus and McGinity “broadly teach PEO matrix tablets formed with simultaneous compression and heating,” a POSA would look to Bartholomaus and McGinity “because abuse by crushing was a known problem,” and that a POSA would “then seek to modify Bartholomaus and McGinity because the processes disclosed in those references would not have been suitable for large-scale production.” Id. at 297, 299. In so doing, Judge Andrews found, a POSA “would have sought a process that could be scaled up,” and that “a desire to manufacture hardened tables at scale and with minimal switching costs is a motivation to modify the prior art.” Id. at 299. Judge Andrews concluded that Accord had presented clear and convincing evidence that a POSA would have been motivated to modify Bartholomaus and McGinity because “a POSA would seek a commercially viable process for producing hardened tablets.” Id. Noting that Accord's expert witness, Dr. Leah Appel, had testified that ovens were common and readily accessible tools for heating tablets, Judge Andrews found that it is “not much of a leap to infer that ovens would . . . be useful for applying heat to harden the matrix tablets disclosed in Bartholomaus” because “a POSA could reasonably expect similar success by simply changing the heating tool.” Id. at 300-01. At the very least, Judge Andrews found, “in the absence of testimony about other heating tools, employing a commonly available tool to apply heat to tablets is obvious to try.” Id. at 300. He added that “Bartholomaus and McGinity both teach hardening PEO by heating it to its melting point” and that “Dr. Appel credibly testified that a POSA combining these references would seek the same result in an oven.” Id.
Dr. Appel served as Accord's expert witness both in Accord I and in this case.
Judge Andrews also addressed Purdue's argument that a POSA would not have been motivated to modify Bartholomaus. Purdue argued that “adding an antagonist would have been a more obvious path” to abuse deterrence in light of the fact that the prior art taught using an antagonist for abuse deterrence. Id. Judge Andrews rejected that argument, noting that “a path does not need to be the most obvious or preferred path” in order to provide motivation for the invention. Id. (citing In re Fulton, 391 F.3d 1195, 1200 (Fed. Cir. 2004)). As Judge Andrews explained, the antagonist route may be a viable approach to abuse deterrence, “but it does not undermine the viability of the hardness route.” 669 F.Supp.3d at 300. He added that the plaintiffs did not argue that the prior art taught away from hardened opioid formulations for abuse deterrence; “they simply argue that the prior art taught an alternative.” Id. Judge Andrews therefore rejected Purdue's argument that the prior art discouraged physical abuse deterrence.
Judge Andrews next found that a POSA would have had a reasonable expectation of success in achieving the claimed invention. He was persuaded by Dr. Appel's testimony that a POSA “would understand that applying heat to melt PEO would cause it to harden,” and that “a POSA could reasonably expect similar success by simply changing the heating tool.” Id. at 301. In addition, Judge Andrews found that “Bartholomaus's discussion of subsequent exposure to heat [after compression] contributes to a reasonable expectation of success-a POSA would expect, upon reading Bartholomaus, to be able to achieve similar results with other heating methods.” Id.
Judge Andrews acknowledged that it is “possible that heating tablets without simultaneously compressing them could change their properties” and that it is “possible that applying heat with an oven rather than with the equipment used by Bartholomaus or McGinity would simply not work.” Id. at 302. However, Judge Andrews concluded that “[n]either of these possibilities seem likely given Dr. Appel's credible testimony. I appreciate that the prior art does not support a guarantee of success-but the law does not require a guarantee.” Id. In sum, Judge Andrews concluded, “I find there was clear and convincing evidence that a POSA would reasonably expect to produce hardened tablets by heating PEO tablets to their melting points in an oven.” Id.
Judge Andrews further agreed with Accord that in light of the guidance provided in the prior art references, optimizing the time and temperature used for curing PEO formulations would have been a matter of routine experimentation. As he explained, “I am persuaded by Dr. Appel's testimony that testing the curing procedure for various periods of time is simple and routine.... Plaintiffs have not offered any contrary evidence that arriving at a range of ten minutes to ten hours would be beyond the routine skill and creativity of a POSA.” Id. at 303. Based on those findings, Judge Andrews held that there was clear and convincing evidence that the claimed properties of the inventions would have been obvious. Id.
C. The Bartholomaus Patent Application Publication
The prior art Bartholomaus application discloses a number of the features that are recited in the asserted claims of the '908 patent. With respect to the limitations in claims 1 and 21 of the '908 patent, Bartholomaus discloses all of the following: (1) a “solid oral extended release pharmaceutical dosage form,” DTX-10 at ¶¶ 64-65, 82; (2) in which the dosage form can be a controlled release matrix and the active ingredient can be embedded in the controlled release matrix, id. at ¶ 101; (3) in which the active ingredient can be an opioid (and specifically oxycodone), id. at ¶¶ 15-16; (4) in which the dosage form comprises polyethylene oxide (PEO) having a molecular weight of between 1 million and 15 million, id. at ¶¶ 18, 116, 122, 126, 132, 137, claims 4-6; and (5) in which the dosage form is shaped and heated to an elevated temperature that is at least the softening temperature of said PEO and thereafter cooled, id. at ¶¶ 67, 101-06. In particular, Bartholomaus teaches that the heating needed to cure the PEO in the matrix tablets may be performed either simultaneously with, prior to, or after the application of pressure to generate the dosage form. Id. at ¶¶ 65 (“optionally after granulation, press-forming the resultant mixture to yield the dosage form with preceding, simultaneous, or subsequent exposure to heat”), 67 (“The resultant mixture is preferably formed directly by application of pressure to yield the dosage form according to the invention with preceding, simultaneous or subsequent exposure to heat”).
Claims 1 and 21 of the '908 patent are quite similar. The only material difference between claim 1 and claim 21 is that claim 1 recites that the tablet is “heated to an elevated temperature that is at least the softening temperature of said PEO for a period of at least about 1 minute and thereafter cooled,” while claim 21 recites that the tablet is “heated to an elevated temperature of at least 55° C. for a time period of at least 15 minutes and thereafter cooled at a temperature below 50°.” Since the '908 patent identifies the softening temperature of high molecular weight PEO as being between 62° C. and 72° C., the only difference between claims 1 and 21 is that claim 1 recites that after heating, the subject tablets are “cooled,” while claim 21 recites that the subject tablets are “cooled at a temperature below 50° C.” Because it is understood that such tablets are routinely cooled to approximately room temperature (about 23° C.) after their processing, those limitations are not significantly different. Tr. 91:6-13.
Bartholomaus also discloses a number of the limitations found in the asserted dependent claims of the '908 patent. In particular, Bartholomaus teaches that the total PEO content of the shaped matrix is at least about 50% by weight of the matrix, which satisfies the requirements of claims 10 and 23. And Bartholomaus teaches that the temperature to which the matrix is heated is the softening temperature of the polymer (about 68° C.), which satisfies the various requirements in the asserted claims that the matrix be heated to “at least about 68° C.” (claim 10), “at least about 60° C.” (claim 18), and “at least 55° C.” (claims 23, 28, and 29). Id. at ¶¶ 16, 65, 67-68, 101-06, 116, 122, 126, 131, and 137. Finally, with respect to claim 29, which requires that the formulation contain at least one of a list of 19 excipients, Bartholomaus discloses the use of several of those excipients, which it refers to as “auxiliary substances,” including magnesium stearate, talcum, colloidal silica, stearic acid, and carnauba wax. See id. at ¶¶ 96-99.
The only limitations in the asserted claims of the'908 patent that are not expressly found in Bartholomaus are (1) that the formulation is “convection heated” (claims 1 and 21); (2) that the matrix is “shaped to form a tablet and heated in a coating pan” (claim 18, by way of claim 17); (3) that the heating of the PEO is “for a time period of at least about 1 minute” (claim 1) or about 5 minutes (claim 10 by way of claim 6), or about 15 minutes (claim 21); and (4) that “a plurality of convection heated particles of PEO adhere to or fuse with each other within the matrix” (claims 1 and 21).
D. The Claim Limitations Not Expressly Found in Bartholomaus
1. Convection Heating and Heating in a Coating Pan
Bartholomaus does not disclose the use of convection heating to cure its tablets, which is a limitation found in all the asserted claims of the '908 patent. In particular, Bartholomaus does not disclose the use of a coating pan, a type of a convection heating device, which is a required element of asserted claim 18 of the '908 patent. The device used to compress and heat the tablets in the examples described in Bartholomaus was a “tabletting tool” that was heated in a “heating cabinet.” See DTX-10 at ¶¶ 116-32. That device employed conductive heating, not convection heating.
Conductive heating is heating by direct contact between the heating source and the object to be heated. Convection heating is heating by the flow of heated air over the object to be heated. Conductive heating is considerably faster than convection heating. CITE FROM.
2. Time for Heating
Bartholomaus does not contain detailed information about the period for which the dosage forms are cured by heating. Bartholomaus's specification prescribes that the formulation, which includes the active pharmaceutical ingredient such as oxycodone and a polymer such as PEO is “briefly heated at least to the softening temperature” of the polymer and then cooled. Id. at ¶ 67. The examples set forth in Bartholomaus state that the mixture was compressed in a tool heated to 80° C. (and in one case to 90° C.) “for at least 15 seconds.” Id. at ¶ 117, 123, 127, 132. Thus, except to provide that the heating period had to be at least 15 seconds, Bartholomaus simply provided that the period of heating needed to be long enough for the mixture to reach the softening temperature of the polymer.
The asserted claims of the '908 patent are somewhat more specific. They provide that the claimed PEO matrix is to be heated to “at least the softening temperature of said PEO for a time period of at least about 1 minute and thereafter cooled” (claim 1); that it is to be heated for “at least about 5 minutes” and at a temperature of “at least about 68° C.” (claim 10, which depends from claim 6); and that it is to be heated to a temperature of “at least about 55° C. for a time period of at least about 15 minutes and thereafter cooled at a temperature below 50° C.” (claims 23, 28, and 29, which depend from claim 21).
3. PEO Particles Adhere to and Fuse With Each Other
The third limitation that Purdue argues is not found in Bartholomaus is the limitation in independent claims 1 and 21, and in each of the asserted dependent claims, providing that “a plurality of convection heated particles of PEO adhere to or fuse with each other within the matrix.” The only reference to that limitation in the '908 specification is a single sentence, which states: “Without wanting to be bound to any theory it is believed that the curing at a temperature that is at least as high as the softening temperature of the high molecular weight polyethylene oxide causes the polyethylene oxide particles to at least adhere to each other or even to fuse.” JTX-1 at col. 17, ll. 58-62. While there is very little discussion of this limitation in the '908 patent, the “adhere to or fuse” limitation is a key feature of Purdue's argument that the asserted claims of the '908 patent are non-obvious. In particular, Purdue contends that the curing process recited in the claims results in the adherence or fusion of the polymer particles in the matrix, which is what causes the matrix to harden and thus be tamper-resistant.
E. Accord's Evidence as to the Limitations Not Found in Bartholomaus
1. Convection Heating and Use of a Coating Pan Without Pressure
Purdue argues that Bartholomaus is distinguishable from the invention of the asserted claims both (1) because the claims require the use of convection heating (and in the case of claim 18 a coating pan) and (2) because the claims prescribe heating the matrix after the matrix has been shaped through pressure, while in Bartholomaus the pressure and heat are applied simultaneously.
Although Purdue argues that the use of convection heating (and in particular the use of a coating pan) distinguishes the asserted claims from Bartholomaus, the evidence does not suggest that any particular characteristics of convection-type heating in general, or the use of coating pans in particular, are uniquely responsible for causing the adherence and fusion of PEO particles and thus the hardness of the matrix after curing. Bartholomaus itself provides the best evidence to the contrary, as Bartholomaus was able to obtain very hard tablets with hardness values at least as high as the values obtained for the tablets produced in the examples set forth in the '908 patent, even when, in Bartholomaus's examples, heating and pressure were applied simultaneously. Compare DTX-10 at ¶¶ 124, 128, 133, 139 (Bartholomaus's tablets did not break when exposed to a force of 500 Newtons), with JTX-1 at col. 65, ll. 60-61 (tablets from Example 8 did not break when exposed to a force of 438 or 439 Newtons). See JTX-1 at col. 33, ll. 56-66; col. 55, ll. 59-63; col 72, ll. 61-67; and tables 5.1, 5.2, 5.3, 7.1.1, 7.2.1, 7.3.1, 13.1.2, 13.5.2.
Moreover, Dr. Appel testified that while the heated tablet press used in Bartholomaus was suitable for heating small numbers of tablets in a laboratory, such a heating system was not readily scalable. For purposes of heating large numbers of tablets simultaneously as part of a large-scale manufacturing process, she testified, a POSA would be motivated to use a convection heating device, such as an oven, a pan coater, or a fluid bed dryer that could cure large numbers of tablets at once. Tr. 82:5-83:12.
I find Dr. Appel's testimony on this point to be credible. In particular, she testified convincingly that convection heating devices, such as coating pans, ovens, or fluid bed dryers, are commonly found in facilities where pharmaceutical tablets are manufactured, and that there would be a strong motivation to substitute such devices in place of the laboratory equipment used in Bartholomaus when adapting the teachings of Bartholomaus to a large-scale manufacturing process. Id.
Dr. Appel also testified that “given the properties of PEO,” a POSA would have had “a good reason to expect that they may be able to use something like a pan coater.” Tr. 511:23512:3. She further pointed to the Encyclopedia of Polymer Science and Technology (DTX-21), which gives “specific information about polyethylene oxide” and “talks about the thermoplasticity of PEO,” as support for her conclusion that a POSA “would understand that they had a reasonable shot” when using convection heat. Tr. 512:20-513:14. I find Dr. Appel's testimony credible and unrebutted.
Purdue also argues that Bartholomaus taught only the simultaneous application of pressure and heat, and thus did not disclose the process claimed by Purdue, which involved the sequential application of pressure to form the tablets, followed by heating without pressure. I disagree. I find that Bartholomaus taught not only simultaneous application of pressure and heat, but the sequential application of pressure and heat, including using pressure to shape the tablets, followed by the application of heat without simultaneous heating.
Bartholomaus explicitly discloses “press-forming the resultant mixture to yield the dosage form with preceding, simultaneous, or subsequent exposure to heat.” DTX-10 at ¶ 65; see also id. at ¶ 67 (“The resultant mixture is preferably formed directly by application of pressure to yield the dosage form according to the invention with preceding, simultaneous or subsequent exposure to heat.”). In the case of direct tableting with preceding exposure to heat, “the material to be pressed is heated immediately prior to tableting at least to the softening temperature” of the polymer and then pressed into tablet form. Id.; see also id. at ¶ 68. In the case of direct tableting with simultaneous exposure to heat, Bartholomaus teaches that the tableting tool components “are briefly heated at least to the softening temperature of the polymer . . . and pressed together.” Id. at ¶ 67. And in direct tableting with subsequent exposure to heat, Bartholomaus explains, the tablets are first formed and then “are briefly heated at least to the softening temperature” of the polymer and then cooled. Id. at ¶ 67.
While most of the examples in Bartholomaus involve the simultaneous application of heat and pressure, example 2 in Bartholomaus followed the procedure of heating the mixture “to 80°C and in [sic: “then”?] placed in the die of the tabletting tool. Pressing was then performed.” Id. at ¶ 121. And, as noted, Bartholomaus made clear in the specification that the process of pressforming the tablets to yield the dosage form could take place “with preceding, simultaneous, or subsequent exposure to heat.” Id. at ¶ 65. I therefore find that Bartholomaus teaches, as one means of producing the hardened tablet, applying pressure to form the PEO matrix tablets and then heating the tablets without pressure.
2. The Time Period That the Matrix is Heated
Because Bartholomaus teaches the use of a conductive heating mechanism to heat the PEO matrix tablets, it suggests the application of heat for shorter periods than the times set out in the '908 patent for convection heating. Dr. Appel explained that the time required to heat PEO to its softening or melting temperature (about 68° C.) would be much greater in the case of convection heating than conductive heating, the method used in Bartholomaus. Unlike in the case of conductive heating, she testified, the period necessary to heat PEO to its softening or melting point in a convection heating device “would be in the time frame of minutes . . . and probably tens of minutes or hours.” Tr. 73: 21-23; see also Tr. 123:4-16. A person of skill in the art, she explained, would understand that a convection heating source is “not going to heat up in less than a minute in an air environment like an oven or a pan coater or a fluid bed.” Tr. 82:1-3. Rather, she testified, it was well understood that convection heating typically requires minutes to hours to fully heat a tablet to the melting point of PEO, which is about 68° C. See Tr. 85:20-87:1. Determining the time necessary to produce a hardened tablet, she added, involves “routine kinds of work,” such as “maybe I pull a sample five minutes, ten minutes, all the way up to several hours, and test it and look for the desired property, in this case, tablet hardness.” Tr. 73:25-74:5.
Dr. Appel's testimony on this issue was convincing and unrebutted. Accordingly, I find that the time required for curing PEO tablets could readily be determined by routine testing and would depend on the type of heating device used and the surrounding circumstances, such as the volume of material being heated. Significantly, Judge Andrews reached the same conclusion in Accord I.
3. The Adhesion or Fusion of the PEO Particles
This limitation presents a more complex issue. With regard to the “adhering or fusing” limitation found in claim 1 (and thus by extension in dependent claims 10 and 18) and in claim 21 (and thus by extension in dependent claims 23, 28, and 29) of the '908 patent, Dr. Appel testified at trial that a person of skill in the art would understand that following the directions of Bartholomaus would lead to a formulation in which the PEO particles adhere to or are fused to one another. Tr. 79:3 to 80:3.
While Purdue disputes that assertion, Purdue's position in this case is contrary to the position it took in a previous case. After the Bartholomaus application issued as a patent, U.S. Pat. No. 8,114,383 (“the '383 patent”), Purdue took a license to the Bartholomaus patent and sued several pharmaceutical companies for infringing the patent. See In re Oxycontin Antitrust Litig., 994 F.Supp.2d 367 (S.D.N.Y. 2014). Although the '383 patent was ultimately held invalid, the court held that the defendants, who compressed and then heated tablets containing PEO, had infringed the '383 patent under the doctrine of equivalents.
In the course of the trial in In re OxyContin, Purdue's attorneys argued and presented evidence to show that the application of heat to the PEO tablets following their compression resulted in the particles of PEO adhering together and fusing, and that the adherence and fusing of the particles causes the tablets to become extremely hard. The opening statement of counsel for Purdue in the In re OxyContin case, which was read into the record in this case, was unequivocal on this point. Counsel said the following regarding the cause of the hardness of the tablets as recited in the Grunenthal patent (the specification of which is identical to the specification in the Bartholomaus published application):
In the Grunenthal patent, the function of thermoforming is to sufficiently fuse the polymer to infuse the structure that imparts the breaking strength. The way this is done is by applying pressure with simultaneous or preceding heat to soften the polyethylene oxide, and the result is the structural characteristic that the tablets have 500 units of breaking strength.Tr. 8:21-9:4.
Accord also read into the record in this case a portion of the testimony of Purdue's expert witness in the OxyContin case. The witness, Dr. Martyn Davies, explained that the function of the thermoformed dosage form recited in the Grunenthal patent was “to sufficiently fuse the polymer to achieve the structure that imparts the claimed breaking strength.” Tr. 15:17-19. Dr. Davies added that “[w]hat causes the polymer to fuse is the melting of the PEO above its melting point so that with the-sufficiently diffuse the polymer to give the structure to give the claimed breaking strength.” Tr. 15:21-25. That result is achieved, he testified, “through the application of the pressure with preceding or simultaneous heat to the component for a sufficient time at temperature to melt or soften the PEO,” which results in “a structural characteristic of the claim breaking strength of at least 500 newtons. It's what we're talking earlier about the way in which the PEO is able to interconnect and form this very strong skeletal structure, the matrix.” Tr. 16:3-14.
Based on that testimony, the court in the In re OxyContin case found that “the function of thermoforming is to fuse the PEO in order to create a scaffolding structure among the PEO,” resulting in a “500-N breaking strength of the tablet produced.” In re OxyContin, 994 F.Supp.2d at 420. The court also agreed with Purdue that heating subsequent to compression is equivalent to heating simultaneously with or prior to compression. Id. Purdue's evidence on that issue contributed to the court's finding of infringement in the In re OxyContin case.
Purdue's position in the In re OxyContin case is consistent with the position taken by the inventors in the specification of the '908 patent, where they stated that “it is believed that the curing at a temperature that is at least as high as the softening temperature of the high molecular weight polyethylene oxide causes the polyethylene oxide particles to at least adhere to each other or even to fuse.” JTX-1 at col. 17, ll. 58-62.
The evidence at trial in this case is to the same effect. Purdue's inventor and the experts for both parties all testified that, in their opinion, heating PEO particles in close proximity to one another is what causes them to adhere or fuse. See Tr. 77:24-80:10, 101:24-102:12 (Dr. Appel); Tr. 350:22-353:8 (Dr. Mannion); Tr. 465:14-17 (Dr. Constantinides).
Beyond that, the examples in Bartholomaus disclose that the process described in that application resulted in very hard tablets-at least as hard as the tablets shown in the examples of the '908 patent. In light of all this evidence, I find that the process disclosed in the '908 patent is the result of adherence or fusion of the PEO particles during curing and that the same is true of the process disclosed in Bartholomaus. See Accord's Post-Trial Proposed Findings of Fact and Conclusions of Law at 11-12. I therefore find that a POSA would have concluded that heating the PEO matrix tablets to the softening point of the PEO or above would cause the adherence and/or fusion of the particles of PEO, resulting in a substantial hardening of the tablets. See Tr. 8:21-9:4.
In sum, I find that Bartholomaus expressly discloses all the limitations found in the asserted claims of the '908 patent except for the limitations requiring that the formulation be cured by convection heating (all asserted claims), that the heating be performed in a coating pan (claim 18), and that the matrix be heated for at least about one minute (claims 1 and 18), five minutes (claim 10), or fifteen minutes (claims 23, 28, and 29). As discussed above, I find that each of those limitations were either necessarily present in Bartholomaus, such as the adhering and fusing limitation; were modifications of Bartholomaus that a POSA would have been motivated to make with a reasonable expectation of success, such as curing the tablets in a convection heating device such as a coating pan; or were measures that could have readily been developed as a matter of routine optimization, such as determining the ideal time for heating the tablets in a particular curing device.
F. The Reduced Density of the Dosage Form
The specification of the'908 patent notes that in certain embodiments of the invention, the curing step leads to a decrease in the density of the extended release matrix formulation as compared to the formulation prior to curing. JTX-1 at col. 23, ll. 41-46. The specification estimated the degree of the reduction in density (i.e., the increase in volume for a constant mass) as falling between 0.5% and 2.5%. Id. at col. 23, ll. 46-53. The inventors explained that “[w]ithout wanting to be bound by any theory, it is believed that the extended release matrix formulation, due to the absence of elevated pressure during the curing step (c) expands, resulting in a density decrease.” Id. at col. 23, ll. 54-57.
The reduction in density is not a limitation of any of the asserted claims of the '908 patent. However, Purdue's witnesses, Dr. Mannion and Dr. Constantinides, testified that the slight decrease in density that results from heating the matrix formulation without compression enhances the abuse-deterrent function of the claimed formulation. Purdue's theory is that the reduction in density increases the porosity of the formulation, thus allowing liquid to penetrate the formulation more quickly. The faster the influx of liquid, the witnesses said, the more quickly the PEO would go into a gelling state, making it more difficult for abusers to extract the drug from the formulation while it is in solution.
There are several problems with Purdue's argument regarding the asserted lower density of the formulation when made according to Purdue's protocol (i.e., first forming the tablets with pressure and then heating the formed tablets without pressure): First, as noted, Bartholomaus teaches the same process as an alternative to its principal focus on compressing and heating simultaneously. Therefore, if Purdue is correct that the reduction in the density of the tablets has a beneficial effect on the tablets' abuse-deterrence, the same would be true of tablets made in accordance with the alternative post-compression heating method taught by Bartholomaus.
Second, Purdue has not pointed to any evidence that the slight decrease in density of the tablets prepared by the patented method had a significant effect on the tamper-resistance of Purdue's tablets. It is clear that the amount by which the density of the tablets is reduced is quite small; according to the specification of the '908 patent, the reduction in density of the tablet formulations following curing without pressure varied between 0.8 percent and 2.5 percent in the five examples given in the specification. See JTX-1 at col. 81, ll. 1-13. Purdue's expert witness suggested that the reduction in density could have increased the abuse-deterrence function of the invention by speeding up the process by which water entering the tablet would create a viscous gel. But Purdue did not introduce any empirical evidence at trial to show that such a phenomenon occurred or that it materially affected the abuse-deterrent capacity of the invention. Purdue could have done a comparison between a product heated under pressure and one not heated under pressure to determine whether heating with pressure would result, for example, in the drug being more syringeable, as Purdue's expert testified would be the case. But it did not conduct such a test, or at least did not offer evidence as to the results of any such test.
Third, it is not clear that the decrease in density resulting from heating the matrix tablet without compression was unexpected. Dr. Appel testified credibly that it was not surprising that when the PEO is heated to its softening temperature; “but [it is] not constrained, you have this mobility which allows for the polymer to relax and move. And so I don't think it's so surprising you might see some change in density.” Tr. 491:2-8.
For these reasons, I am not persuaded that the slightly reduced density in the tablets produced without simultaneous heating and pressure results in a material difference in the syringeability of the oxycodone once the tablets are exposed to liquid.
Judge Andrews addressed the same issue in Accord I. He treated the issue as going to whether the invention produced unexpected results, i.e., the unexpected result that heating the PEO matrix tablet would result in a decrease in density. His conclusion was that while the decrease in density was unexpected, he was not persuaded that it constituted a superior property or advantage. Rather, he found the testimony on the issue of the impact of the decrease in density to be too speculative to credit Purdue's expert over Accord's, and even if the decrease in density had some beneficial effect for abuse deterrence, “any increased gelling benefit the decrease in density might offer would not alone undermine the clear and convincing evidence that the invention's claimed properties are obvious.” 669 F.Supp.3d at 304.
Like Judge Andrews, I am not persuaded by the evidence that the minimal decrease in tablet density results in materially greater abuse deterrence. Although Dr. Mannion and Dr. Constantinides testified that there was a slight decrease in the density of the matrix formulation when it was cured without the simultaneous application of pressure, rather than being cured while under pressure, they did not suggest that the difference was material to the abuse-deterrent effect of the formulation.
Dr. Mannion pointed to examples in the '908 patent that showed that the tablet matrix prepared according to the '908 patent released much less of the drug than the original OxyContin capsules after they were crushed and exposed to a solvent. See Tr. 320:25-326:9. The slower release, according to Dr. Mannion, was the result of the faster swelling of the polymer, as it became gelatinous when exposed to a solvent. Tr. 321:22-322:10, 323:1-8, 325:24-326:9. He explained that “[i]f the polymer hydrated slowly and the drug went into the solution before the polymer hydrated, then the person abusing the dosage form could draw away the liquid.” Tr. 326:1-6. That testimony, however, is beside the point. It compared the reformulated OxyContin with the original OxyContin, which did not incorporate the Bartholomaus technology in any form.
For his part, Dr. Constantinides explained that a decrease in the density of tablet matrix would have a benefit for abuse deterrence because “[d]ecreased density will result in greater tablet porosity, which means faster water or fluid penetration into the matrix, faster gelation, and reducing syringeability.” Tr. 407:21-408:5. But Dr. Constantinides, like Dr. Mannion, did not testify as to the magnitude of the effect of the slight decrease in tablet density in tablets prepared without the simultaneous application of heat and pressure.
Dr. Appel did. Testifying for Accord in rebuttal, she responded to Dr. Constantinides' testimony about the effect of the reduced density of the tablets prepared according to the '908 patent by saying, “I don't think that the slight decrease in density that was observed matters in terms of abuse deterrence.” Tr. 489:14-15. She testified that when a tablet is heated, it gives the molecules more mobility, particularly at the softening point of the polymer. If the tablet is not under pressure at that point, she explained, the mobility “allows for the polymer to relax and move,” so she did not find it surprising that “you might see some change in density.” Tr. 490:11491:8. But she did not find the small change in density to be meaningful in this instance. Tr. 491:9-11.
Dr. Appel explained her opinion on that issue as follows: As soon as the polymer is exposed to liquid, the drug goes into solution and the polymer starts to gel. “The drug goes from a solid state to a liquid state. And then comes out.” Tr. 493:6-9. But at the same time, she explained, the liquid would cause the polymer to gel, and because the liquid penetrates the less dense tablet more quickly, the “gelling might happen a little more quickly.” Tr. 493:16-17. In substance, Dr. Appel explained that although the difference would be “trivial,” the greater porosity of the tablets prepared by the '908 patent protocol would release the oxycodone more quickly, but would also result in more rapid gelling with the more rapid influx of the solvent into the tablet. The net effect, she explained, would be that the '908 patent protocol “kind of goes the opposite of making [the tablet] more abuse-deterrent.” Tr. 493:22-494:2.
Purdue did not challenge Dr. Appel's analysis on cross-examination.
I find Dr. Appel's assessment of the effect of a lower density tablet to be more credible than that of Drs. Mannion and Constantinides. It makes sense that the slightly greater speed with which solvent enters the less dense tablet has two offsetting effects for present purposes: first, it releases the solid particles of oxycodone into solution more quickly; but second, it causes the gelling of the polymer to take place more quickly, thus trapping the oxycodone and making it difficult to extract. The speed with which the drug goes into solution is thus offset by the speed with which the polymer begins to gel. And even if the very minor increase in the porosity of the tablets produced in accordance with Purdue's protocol results in some incremental increase in the availability of the drug before the gelling occurs, I find Dr. Appel's explanation for why the difference would not be material to be credible.
In the course of its “reduction in density” argument, Purdue cites the Joint Statement of Undisputed Facts, at paragraphs 21 and 22. But those two paragraphs do not advance Purdue's argument, as they assert only that the compositions made pursuant to the '908 patent exhibit a lower density, and thus more porosity, than equivalent tablets that are heated prior to or with simultaneous compression. Dkt. No. 86-1 at 4. But those facts are undisputed and do not establish that the tablets are materially more abuse-deterrent.
In sum, I find as a matter of fact that although curing the matrix table without simultaneous pressure results in a slightly less dense tablet, the slight increase in the tablet's porosity does not increase its abuse-deterrence effect, or if it does, any such effect is not material.
G. Secondary Considerations
Purdue relies on various secondary considerations in support of its argument that Accord has failed to satisfy its burden of proof on the issue of obviousness. I do not find the secondary consideration evidence to be persuasive.
1. Long-Felt Unmet Need and Failure of Others
Purdue argues that Purdue met a long-felt unmet need for a solution to the problem of abuse of OxyContin products when Purdue used the invention claimed in the '908 patent to create its crush-resistant tablet and obtain FDA approval to add crush-resistance to its reformulated OxyContin's label.
It is true, as Purdue contends, that Purdue was the first to obtain FDA approval for an “extended-release opioid with an FDA-approved label stating that the medication was formulated with abuse-deterrent properties.” Plaintiffs' Opening Post-Trial Brief at 51. However, the solution to the problem that was approved by the FDA in 2013 was actually devised by Bartholomaus eight years earlier, in 2005.
Purdue argues that “numerous other companies in the industry undertook extensive efforts to try to address” the problem of opioid abuse, “but none had succeeded.” Id. That broad statement appears to be directed to all the various proposals for deterring the abuse of opioids, including educational programs, enhanced law enforcement efforts, and adding antagonists or aversive components to opioid tablets. Purdue does not contend that there was widespread failure on the part of the industry with regard to the specific anti-abuse mechanism of hardening the drugcontaining tablets.
In that respect, Purdue focuses mainly on one other party's asserted failure to achieve success with a tamper-resistant product similar to the reformulated OxyContin product. Purdue points to Opana ER, in which the active ingredient was oxymorphone, a stronger opioid than oxycodone. Tr. 476:13-15. Purdue characterizes Opana ER as incorporating “Grunenthal's commercialized hot melt extrusion version of Bartholomaus to increase tablet hardness to provide crush resistance.” Plaintiffs' Opening Post-Trial Brief at 51. Purdue argues that Opana ER “could not achieve the same unexpected decrease in tablet density as the claimed invention, and Opana ER, unlike reformulated OxyContin, was still susceptible to abuse by injection.” Id.
The FDA ultimately denied the manufacturer's request for abuse-deterrent labeling for Opana ER and directed that Opana ER be removed from the market because of the risk of abuse. Tr. 450:22-451:7; PTX-330. Comparing reformulated OxyContin with the reformulated version of Opana, the FDA's letter regarding Opana included the following observation: “When subjected to an aqueous environment [reformulated OxyContin] gradually forms a viscous hydrogel that resists passage through a needle such that it prevents oxycodone from being drawn into a syringe to any meaningful extent.... In contrast, [reformulated Opana] can be readily prepared for injection ....” PTX-330 at 8.
Based on the FDA's letter and what Purdue claims is a difference in the curing mechanism between reformulated Opana and reformulated OxyContin, Purdue argues that the difference in syringeability between the two products is attributable to the slight decrease in density in the reformulated OxyContin product resulting from the use of convection heating without pressure. Put differently, Purdue asks me to conclude that, if Opana-a different product with different ingredients-were made via Purdue's claimed process, its syringeability would have been reduced.
That hypothesis is not a compelling one. First, it depends on the assumption that Opana ER and OxyContin ER are prepared by different methods. A journal article in evidence that compares the two products reports that the two are made by similar processes. See PTX-232. Purdue challenged the accuracy of that statement at trial, offering into evidence a letter authored by two Purdue employees, including Dr. Mannion, contending that the processes used to make the two products were different. See PTX-334 at 2. That assertion, however, lacks a citation, and thus it is unclear how and why the authors reached the conclusion they did. Id. That gap is troubling especially in light of the fact that Dr. Mannion testified at trial but did not testify about the letter. Further, the third party article to which the letter responds comments that the INTAC process used to create at least Opana ER is “proprietary in nature and thus not accessible in any detail.” PTX-232 at 4. Without any explanation for how Purdue's employees were privy to the details of the INTAC process, I find the article and letter to be a wash and am unable to conclude that the two processes are meaningfully different.
Dr. Mannion's testimony at trial about the INTAC process was excluded as hearsay. Tr. 243:14-246:18.
Second, even if Purdue is correct that the processes for curing the two products are different, that does not mean that the asserted difference in syringeability between the two products is attributable to that difference (i.e., to a reduction in the density of tablets produced by the process taught in the '908 patent). Nothing in the FDA's letter suggests a cause for the different syringeability. In fact, because of the confidential nature of the data, the FDA's letter discusses the Opana data in only “general terms.” PTX-330 at 1 n.1. Thus, the FDA's letter provides no data that attributes increase in syringeability to any decrease in density of the Purdue product. Likewise, the journal article that discusses the two products speculated that “there may have been some modifications made” in the proprietary processes used by the two manufacturers “that made one superior to the other.” PTX-232 at 4.
Third, Purdue has offered no data to show that the slight decrease in density of its products produced a significant increase in the syringeability of the oxycodone compared to a similar product made without decreased density. There is thus minimal record evidence that the failure of reformulated Opana to deter abuse is attributable to the assumed fact that Opana is not made by the process claimed in the '908 patent.
And finally, Opana contained oxymorphone, a stronger opioid than oxycodone. The evidence at trial suggested that potential abusers of Opana were especially motivated to obtain access to that drug and were willing to make greater efforts to overcome anti-deterrence measures. See PTX-232 at 5. This explanation for the greater abuse of Opana is plausible and is unrelated to the claimed invention. In sum, the evidence of the failure of Opana ER as an abuse-deterrent tablet is insufficient to overcome the strong prima facie case of obviousness presented by Accord.
2. Industry Skepticism and Subsequent Acclaim
Purdue attributes the FDA's lengthy delay in approving its abuse-deterrent label for the reformulated version of OxyContin to skepticism about Purdue's invention on the FDA's part. The FDA approved the reformulated OxyContin in 2010, see PTX-125, but the agency did not at that time approve Purdue's proposed abuse-deterrent label. Instead, it required Purdue to prove that the abuse-deterrent version of OxyContin resulted in a decrease in misuse and abuse. Dkt. No. 86-1 at ¶¶ 28-29. After Purdue completed the required epidemiological studies, the FDA approved the use of the abuse-deterrent label for the reformulated OxyContin in April 2013. Id. at ¶¶ 30-31; PTX-226; Tr. 308:14-310:25.
I do not find the FDA's actions to be indicative of skepticism in the industry; rather, the agency's action appears to be consistent with its normal approach of requiring empirical evidence from the pharmaceutical company before allowing assertions regarding the characteristics of a drug to be set forth in the drug's label. Judge Andrews made the same observation in Accord I. He explained that it “seems natural that the FDA, as a regulatory body, would require real world studies before being satisfied that a hard tablet was indeed abuse-deterrent.” 669 F.Supp.3d at 306. The FDA's caution, he explained, is not indicative of industry skepticism. Id.
Purdue's other evidence of industry skepticism is based largely on the testimony of Dr. Mannion and Dr. Constantinides, Purdue's expert witnesses. While Dr. Mannion is, of course, a competent witness, he is aligned with the plaintiff, and his testimony must be judged in light of his affiliation. As Judge Andrews said, making the same point, “Dr. Mannion's own testimony as a named inventor, would seem to carry limited weight. He does not serve as a stand-in for a POSA, or for the industry.” Id. Moreover, Dr. Mannion's testimony was principally about his own skepticism that the Bartholomaus process would work, rather than being directed to skepticism on the part of the industry at large. See Tr. 332:5-333:1; Tr. 307:18-308:1 (expressing skepticism at the test results that Grunenthal showed him and whether the results would be consistent with his equipment). As for Dr. Constantinides, that testimony mainly reported that the reformulated OxyContin significantly reduced the incidence of abuse of the drug, rather than recounting skepticism in the field regarding whether Purdue's solution would work. See Tr. 455:14-456:7; PTX-211 at 7; PTX-328, at 3; PTX-226, at 1; PTX-161 at 9. For that reason, I also find the factor directed to the skepticism before and acclaim after the release of the Purdue reformulated OxyContin product does not significantly undermine Accord's showing of obviousness.
3. Unexpected Results
Purdue argues that the reduced density of the polymer produced by Purdue's curing method was an unexpected result and therefore provides further support for Purdue's claim that secondary considerations support a finding of non-obviousness. As to whether the reduced density of Purdue's matrix tablet was unexpected, the evidence is mixed. Dr. Constantinides for Purdue testified that the reduced density of Purdue's matrix tablets was unexpected. Tr. 406:17-20, 450:7-8. Dr. Appel for Accord testified that she did not regard it as “so surprising you might see some change in density” upon heating the matrix tablet without compression. Tr. 491:2-8.
Even assuming that the reduced density of tablets made by Purdue's process would have been surprising to a POSA, the evidence does not show that the slight reduction in density had a meaningful effect in reducing the likelihood of abuse. Although Judge Andrews in Accord I found that the reduced density of the tablets was surprising, he found that the testimony on the impact of a decrease in tablet density was “too speculative” to support an inference that the reduction was beneficial to abuse-deterrence. See 669 F.Supp.3d at 304. I agree with Judge Andrews and find that the evidence does not support Purdue's contention that slightly lower density of the tablets was material; the fact that the reduction in density ranged between less than one percent and less than three percent strongly suggests that the effect of change in the polymer's density would be minimal, as Dr. Appel surmised.
IV. Conclusions of Law
Accord argues that the asserted claims of the '908 patent, viewed in light of the knowledge of persons of skill in the art, would have been obvious. Accord contends that a POSA would have had a motivation to adapt the Bartholomaus process for generating a tamper-resistant matrix tablet on a commercial scale by using a convection heating device such as a coating pan and by heating the tablets without simultaneous exposure to pressure. Following such a procedure, according to Accord, would have given rise to a reasonable expectation of success.
Purdue's legal and factual challenges to Accord's theory for why Purdue's asserted claims would have been obvious in light of Bartholomaus can be summarized as follows:
(1) A POSA faced with the problem of finding a way to combat the abuse of OxyContin would not have been motivated to begin with the method of Bartholomaus involving heating oxycodone tablets to make them sufficiently hard to be crush-resistant;
(2) Even if a POSA would have begun with Bartholomaus, the POSA would not have been motivated to use convection heating of tablets, and in particular would not have been motivated to use a coating pan for curing;
(3) A POSA would not have been motivated to apply heat after compressing the tablets, rather than compressing them and heating them simultaneously; and
(4) A POSA would not have expected the heated particles of PEO to adhere to or fuse with each other in the matrix.
The first of those issues presents what is in part a question of law and in part a factual question. The other three issues present factual questions. As indicated above, I have resolved each of those factual issues in favor of Accord. In addition, Judge Andrews' decision in Accord I has resolved some of those factual issues and is binding as a matter of collateral estoppel, as discussed below. Because collateral estoppel plays a large role in the analysis of the obviousness question in this case, I begin with a discussion of the principles of collateral estoppel and how they may apply here.
A. Collateral Estoppel
Collateral estoppel “precludes a plaintiff from relitigating identical issues by merely switching adversaries” and precludes a plaintiff “from asserting a claim that the plaintiff had previously litigated and lost against another defendant.” Parklane Hosiery Co. v. Shore, 439 U.S. 322, 327 n.4, 329 (1979). With respect to several of the issues in this case, that is what Purdue is attempting to do.
The Second Restatement of Judgments contains the classic articulation of the governing principles of collateral estoppel:
When an issue of fact or law is actually litigated and determined by a valid and final judgment, and the determination is essential to the judgment, the
determination is conclusive in a subsequent action between the parties, whether on the same or a different claim.
Restatement (Second) of Judgments § 27 (Am. L. Inst. 1982); see In re Dockteroff, 133 F.3d 210, 214 (3d Cir. 1997); Tex. Instruments Inc. v. Cypress Semiconductor Corp., 90 F.3d 1558, 1568 (Fed. Cir. 1996). Those requirements are all met with respect to the collateral estoppel issues in this case, and none of the exceptions set forth in section 28 of the Restatement (or in the applicable case law) apply to the particular issues as to which collateral estoppel has been invoked here.
Because the patents at issue in Accord I were different from the patent at issue in this case, the decision in Accord I does not have broad collateral estoppel effect sufficient to foreclose Purdue's claims altogether. However, some of the same issues presented in this case were litigated in Accord I, and the resolution of those issues was necessary to the result in that case. Judge Andrews' decision in Accord I therefore has collateral estoppel effect as to those issues. See Parklane Hosiery, 439 U.S. at 326 n.5; United Techs. Corp. v. Chromalloy Gas Turbine Corp., 189 F.3d 1338, 1343 (Fed. Cir. 1999).
Accord identifies five issues on which, in its view, collateral estoppel requires a decision in its favor: (1) Judge Andrews' finding that a person of ordinary skill in the art would have had a motivation to find a different mode of heating from the non-scalable heating system used in Bartholomaus, and in particular would have been motivated to use an oven for heating, see 669 F.Supp.3d at 299; (2) Judge Andrews' finding that the industry's interest in the use of antagonists as a possible means of combatting the abuse of opioid tablets did not teach away from the Bartholomaus heating protocol, see id. at 300; (3) Judge Andrews' finding that a POSA would have had a reasonable expectation of success in obtaining hard, tamper-resistant tablets using an oven to cure the tablets, see id. at 301-02; (4) Judge Andrews' finding that the evidence that Purdue's manufacturing process resulted in a less dense tablet did not constitute a “superior property or advantage,” see id. at 304 (quoting Forest Lab'ys, LLC v. Sigmapharm Lab'ys, LLC, 918 F.3d 928, 937 (Fed. Cir. 2019)); and (5) Judge Andrews' finding that the evidence regarding Purdue's “density argument”-i.e., Purdue's argument that its process for manufacturing its abuseresistant tablets produced a less dense tablet that was more difficult to abuse-did not undermine Accord's obviousness case, see id. at 304. In addition to those issues, there is a sixth issue on which collateral estoppel from the Accord I decision is applicable: Judge Andrews' finding that determining the time and temperature to use in the curing procedure could have been achieved through routine experimentation. See id. at 303.
The decision in Accord I is currently on appeal to the Federal Circuit. See Purdue Pharma L.P. v. Accord Healthcare, Inc., No. 23-1953 (Fed. Cir.), filed May 30, 2023. Collateral estoppel can be, and has been, applied based on a district court decision that is still pending on appeal. See Uniloc USA, Inc. v. Motorola Mobility LLC, 52 F.4th 1340, 1347 (Fed. Cir. 2022); Pharmacia & Upjohn Co. v. Mylan Pharms., Inc., 170 F.3d 1373, 1380-81 (Fed. Cir. 1999); United States v. 5 Unlabeled Boxes, 572 F.3d 169, 175 (3d Cir. 2009). If the judgment in that case is vacated or reversed, the collateral estoppel effect of Judge Andrews' various rulings in that case may be negated.
Because of the pendency of the appeal in Accord I, I have not relied exclusively on collateral estoppel with regard to any of the issues that are common to that case and this one, but instead have addressed the merits of each of those issues independently, as alternative grounds for decision. Nonetheless, if the judgment in Accord I is sustained, or if Judge Andrews' findings on any of the overlapping issues are upheld, the principles of collateral estoppel will be applicable, and those issues will properly be deemed conclusively resolved against Purdue.
I now turn to a discussion of Purdue's four legal and factual challenges to Accord's obviousness case.
B. Bartholomaus as the Principal Prior Art Reference
As to the first issue, the parties disagree about whether Bartholomaus can properly be regarded as the principal prior art reference in light of the fact that other possible solutions to the problem of oxycodone abuse were considered at the time. Purdue contends that Bartholomaus would not have been an obvious starting point for an anti-abuse strategy, because at the time of the invention the industry was more focused on the use of other strategies, such as the use of antagonists to discourage potential abusers. Accord, however, argues that the existence of other possible strategies for abuse deterrence would not negate the obviousness of modifying Bartholomaus. Accord's position is legally correct.
The Federal Circuit has made it clear that the fact that there are other possible solutions to a problem does not affect the obviousness analysis as to one of those possible solutions. See PAR Pharm., Inc. v. TWIPharms., Inc., 773 F.3d 1186, 1197-98 (Fed. Cir. 2014) (“Our precedent . . . does not require that the motivation be the best option, only that it be a suitable option from which the prior art did not teach away.”); see also Galderma Labs., L.P. v. Tolmar, Inc., 737 F.3d 731, 738 (Fed. Cir. 2013); Bayer Healthcare Pharm., Inc. v. Watson Pharm., Inc., 713 F.3d 1369, 1376 (Fed. Cir. 2013); In re Fulton, 391 F.3d 1195, 1201 (Fed. Cir. 2004) (“Our case law does not require that a particular combination must be the preferred, or the most desirable, combination described in the prior art in order to provide motivation for the current invention.”).
Nothing in the prior art criticized the approach taken by Bartholomaus. Thus, the fact that other approaches to the problem were considered contemporaneously did not teach away from the solution suggested by Bartholomaus. See Novartis Pharms. Corp. v. W.-Ward Pharms. Int'l Ltd., 923 F.3d 1051, 1059 (Fed. Cir. 2019) (It is “improper to require West-Ward to prove that a person of ordinary skill would have select everolimus over other prior art treatment methods.”); SightSound Techs., LLC v. Apple Inc., 809 F.3d 1307, 1320 (Fed. Cir. 2015) (“[T]he ‘mere disclosure of more than one alternative' does not amount to teaching away from one of the alternatives where the reference does not ‘criticize, discredit, or otherwise discourage the solution claimed.'”) (citation omitted).
The facts support Accord's position that Bartholomaus should be treated as the principal prior art reference for this case. Bartholomaus was in the public domain and was fully available to a person of skill in the art as of February 2005. While there were other approaches to addressing OxyContin abuse that were being explored as of the priority date for the '908 patent, those other approaches were not so clearly preferable, or farther along in their development, as to wholly eclipse the crush-resistant approach employed by Bartholomaus, and ultimately by the inventors of the '908 patent. Moreover, defining the field of prior art depends on defining the problem to be solved. If the problem is defined as discouraging abuse, the solutions could be almost infinitely varied, and any single solution would not necessarily be apparent to a person of ordinary skill in the art. On the other hand, if the problem is defined as how to harden the tablets sufficiently to make them crush-resistant, the solution would be much more specific. In that setting, a solution such as Bartholomaus's would assume a more prominent role.
The latter is the more logical view. Where a problem has multiple possible approaches and some of the prior art has focused on one approach, a party using that approach should not be able to obtain patent protection simply because there were other possible approaches to solve the overall problem that could have been explored. In this case, for example, alternatives to creating tamperresistant tablets for combatting abuse could be as diverse as aggressive educational programs; strong enforcement measures; regulatory insistence on strict accounting of tablets distributed to patients; restricting the number of physicians authorized to prescribe the drugs; use of a “sticky” polymer that would make it difficult to grind the dosage tablets into a fine powder; incorporating antagonist drugs in the tablets; adding aversive agents to the tablets that would discourage use by abusers; binding the drug to ion-exchange resins that would be so small as to be almost uncrushable; and developing prodrugs that would need to be ingested and metabolized by the liver in order to be converted into the opioid itself. See Tr. 218:14-227:4; PTX-279.
It would make little sense to say that an applicant can obtain patent protection for a solution that is obvious from the prior art simply because the field was crowded with other proposed solutions. A POSA is deemed to be aware of all the pertinent prior art in a field. In re GPAC, Inc., 57 F.3d 1573, 1579 (Fed. Cir. 1995); Custom Accessories, Inc. v. Jeffrey-Allan Indus., Inc., 807 F.2d 955, 962 (Fed. Cir. 1986). Purdue's invention therefore cannot be regarded as non-obvious because a person of ordinary skill in the field might have been tempted to follow one of the other possible ways of discouraging abuse rather than the one Bartholomaus highlighted.
C. Obviousness Based on Bartholomaus
Bartholomaus is a highly pertinent reference. Viewed as a whole, Bartholomaus teaches the following regarding a formulation with strong anti-abuse potential by making tablets crushresistant:
(1) Use of an opioid formulation containing PEO;
(2) Heating the opioid formulation to and above the softening temperature of the PEO;
(3) Doing so while applying pressure either before, during, or after heat-curing the formulation;
(4) Then cooling the cured tablet.
The '908 patent contains only three limitations that are arguably not found in Bartholomaus:
(1) The “heated particles of PEO adhere to or fuse with each other within the matrix”;
(2) Using convection heating (including, in claim 18, heating in a coating pan) to heat the tablet matrix; and
(3) Heating the tablet subsequent to compressing the tablet for at least about one minute (claim 1), at least about 5 minutes (claim 10), or at least about 15 minutes (claim 21).
As discussed below, each of these limitations is either found in Bartholomaus or would have been obvious in view of Bartholomaus.
1. “Adhere to or Fuse”
With regard to the “adhere to or fuse” limitation, Accord argues that the statements of Purdue's attorney and expert witness in the In re OxyContin case, addressed at pages 18-20, supra, give rise to judicial estoppel, which would have the effect of barring Purdue from denying that heating the PEO to its melting point under the conditions set forth in Bartholomaus results in the PEO particles adhering to one another or fusing. See New Hampshire v. Maine, 532 U.S. 742, 749 (2001) (“The doctrine of judicial estoppel prevents a party from asserting a claim in a legal proceeding that is inconsistent with a claim taken by that party in a previous proceeding.”) (quoting 18 Moore's Federal Practice § 134.30 at 134-62 (3d ed. 2000)); id. at 750 (“courts regularly inquire whether the party has succeeded in persuading a court to accept that party's earlier position, so that judicial acceptance of an inconsistent position in a later proceeding would create ‘the perception that either the first or the second court was misled.'”) (citing Edwards v. Aetna Life Ins. Co., 690 F.2d 595, 599 (6th Cir. 1982)). Because Purdue prevailed in the In re OxyContin case on the infringement issue by successfully arguing that the Bartholomaus process results in the adherence or fusion of the heated PEO particles, Accord argues that Purdue may not take a contrary position in this case.
The problem with Accord's position is that the Third Circuit has consistently taken a narrow view of the scope of judicial estoppel. In addition to requiring that a party's position be clearly contrary to a position previously taken in the same or another action, the Third Circuit has held that a district court may not invoke judicial estoppel unless the court finds that the party that has changed its position has done so in “bad faith.” See MD Mall Assocs., LLC v. CSX Transp., Inc., 715 F.3d 479, 486 (3d Cir. 2013); Chao v. Roy's Constr., Inc., 517 F.3d 180, 186 n.5 (3d Cir. 2008); Krystal Cadillac-Oldsmobile GMC Truck, Inc. v. Gen. Motors Corp., 337 F.3d 314, 31920 (3d Cir. 2001) While the Third Circuit has not defined the term “bad faith” with precision, it appears that the court regards a change in position, without more, to be inadequate to justify applying judicial estoppel. And in this case, Accord has not identified any aggravating factors pointing to bad faith on the part of Purdue.
Although any appeal in this case will go to the Federal Circuit, not the Third Circuit, the Federal Circuit has made clear that in an appeal raising an issue of judicial estoppel, the court will apply the law of the regional circuit in which the district court sat-in this case, the Third Circuit. See Egenera, Inc. v. Cisco Sys., Inc., 972 F.3d 1367, 1378 (Fed. Cir. 2020); Akamai Techs., Inc. v. Limelight Networks, Inc., 805 F.3d 1368, 1374 (Fed. Cir. 2019).
With that said, however, Purdue's volte-face on the issue of whether Bartholomaus's process results in adhering or fusing justifies other remedial measures. As explained above, the district court in the In re OxyContin case agreed with Purdue that “the function of thermoforming is to fuse the PEO in order to create a scaffolding structure among the PEO,” resulting in a “500 N breaking strength of the tablet produced.” In re OxyContin, 994 F.Supp.2d at 420. Even if judicial estoppel is not applicable, the doctrine of collateral estoppel bars Purdue from arguing to the contrary in this case, because the court's finding was essential to the court's conclusion that Purdue had successfully shown infringement of the Bartholomaus patent in that case.
The Bartholomaus patent, U.S. Pat. No. 8,114,383, to which Purdue took a license and which it asserted in the infringement action involved in the In re OxyContin case, has a specification that is essentially the same as the specification in the Bartholomaus Patent Application Publication.
Even in the absence of judicial estoppel or collateral estoppel, Purdue's statements in the In re OxyContin case-through its counsel and its expert witness-constitute evidentiary admissions. See Paice LLC v. Toyota Motor Corp., 504 F.3d 1293, 1312 (Fed. Cir. 2007); Diebold Inc. v. Continenetal Cas. Co., 430 Fed.Appx. 201, 205 (3d Cir. 2011). Although an evidentiary admission is not binding, and the fact-finder is free to weigh the admission along with other evidence in the case, see Paice, 504 F.3d at 1312, it is nonetheless record evidence that can be used against the party who has made the admission. In this case, the admission was clearly purposeful and was central to Purdue's position in the OxyContin case. It is therefore entitled to substantial weight.
Finally, the evidence in this case provides solid support for the proposition that Bartholomaus satisfies the “adhere to or fuse” limitation of the '908 patent. As discussed above, both parties' witnesses testified that in their opinions, heating PEO particles in close proximity to one another causes them to adhere to one another or fuse, which is what causes the tablets to become very hard. See Tr. 77:24-80:10, 101:24-102:12 (Dr. Appel); Tr. 350:22-353:8 (Dr. Mannion); Tr. 465:14-17 (Dr. Constantinides). And the examples in Bartholomaus showed that the heating process produced tablets that were at least as hard as the tablets shown in the examples of the '908 patent. Based on that evidence, I find that a POSA would have concluded that heating the PEO matrix tablets to the softening point of the PEO or above would cause the particles to adhere or fuse, thus satisfying the “adhere to or fuse” limitation of the '908 claims.
Purdue alleges that in her testimony, Dr. Appel “impermissibly relied on the inventors' disclosure in the 908 patent . . . rather than any prior art teaching.” Plaintiffs' Opening Post-Trial Brief at 32. I disagree. After noting that the '908 patent stated the inventors' belief that heating the matrix to the softening temperature of the high molecular weight PEO caused the particles to adhere to each other or fuse, Tr. 78:18-25, Dr. Appel stated that in her opinion, “that's what a POSA would think was happening in Bartholomaus as well.” Tr. 79:1-2. Contrary to Purdue's assertion, that comment does not reflect hindsight on Dr. Appel's part, but merely constitutes an observation that a POSA would reach the same conclusion that the '908 inventors reached: that heating high molecular weight PEO in the manner set forth in Bartholomaus would result in adherence or fusion.
2. Heating by Convection or in a Coating Pan
With regard to the heating method limitation, as discussed in my Findings of Fact, I find, based on Dr. Appel's testimony, that a POSA would have had a motivation to use convection heating as a way of scaling up the Bartholomaus process and an expectation of success of doing so. I also credit Dr. Appel's testimony that a coating pan is a common piece of equipment in a tablet manufacturing facility, and that a POSA would find it attractive to cure the matrix tablets in a coating pan because doing so would offer the advantage of being able to cure the tablets and coat them using the same equipment.
Moreover, Judge Andrews found in Accord I that a pharmaceutical company seeking to use the Bartholomaus abuse-deterrence invention for the large-scale production of opioid tablets would have had a motivation to find a different mode of curing Bartholomaus's PEO formulation because Bartholomaus's heating method was not scalable. That finding, which was contested in both Accord I and in this case, is entitled to collateral estoppel effect.
3. Compression Before Heating and Time of Heating
With regard to sequencing compression and heating, as discussed in my Findings of Fact, I find that, while Bartholomaus does teach compressing and heating the tablet simultaneously,. Bartholomaus also teaches heating the tablet without compression after the tablet has been compressed.
Further, the court in In re OxyContin accepted Purdue's argument that heating subsequent to compression is equivalent to heating simultaneously with or prior to compression. 994 F.Supp.2d 420. In this case, Purdue has reversed field and now claims that heating by convection, including heating in a coating pan, is entirely different from conduction heating of the sort used in the examples in Bartholomaus. Purdue's position now is contrary to the position it took in the In re OxyContin case, in which it prevailed on its claim of infringement under the doctrine of equivalents. Once again, Purdue's contrary position in the In re OxyContin case serves as an evidentiary admission on this issue, supporting my conclusion that Bartholomaus teaches this limitation.
Finally, as I found previously based on Dr. Appel's testimony, the time required to heat the PEO matrix tablets to the melting point of PEO would be considerably longer for heating by convection than for heating by conduction. I also credit her testimony that the time required for heating the PEO to its melting point could readily be determined by routine experimentation. As a legal matter, the difference between the time period estimated in Bartholomaus for heating by convection does not undermine Accord's showing of obviousness. Moreover, as noted above, Judge Andrews specifically found in Accord I that determining the time and temperature for heating the PEO in a convection device would be a matter of routine optimization, a finding that has collateral estoppel effect in this case.
D. Product-by-Process Claims and Written Description/Enablement
The parties agree that the asserted claims are “product-by-process” claims, i.e., claims in which a product is defined at least in part in terms of the method or process by which it is made. See Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 158 n.* (1989); Greenliant Sys., Inc. v. Xicor LLC, 692 F.3d 1261, 1268 (Fed. Cir. 2012); SmithKline Beecham Corp. v. Apotex Corp., 439 F.3d 1312, 1315 (Fed. Cir. 2006). Such claims cannot validly claim products that are already known in the art; “while the process set forth in the product-by-process claim may be new, that novelty can only be captured by obtaining a process claim.” Id. at 1318-19. As applied to this case, therefore, the asserted claims of the '908 patent are valid only if the products made by the claimed process (including convection heating and heating in a coating pan) are significantly different from the prior art products. See Kamstrup A/S v. Axioma Metering UAB, 43 F.4th 1374, 1381 (Fed. Cir. 2022).
Accord argues that the product produced by the process limitations in the asserted claims of the '908 patent is not significantly different from the product produced by the process described in Bartholomaus, and thus the asserted claims of the '908 patent are invalid. Purdue contends otherwise.
On the merits, the problem with Accord's argument is that in order to prevail on this issue, Accord must show that the product produced by the process claimed in the '908 patent is the same as the product produced by the process set forth in Bartholomaus. See Amgen Inc. v. F. Hoffman-LaRoche, Ltd., 580 F.3d 1340, 1370 (Fed. Cir. 2009). In particular, Accord must rebut Purdue's argument that the product produced by Purdue's process differs from the product produced by Bartholomaus in that Purdue's process produced a less dense product than was produced by the Bartholomaus process. Although the lower density of the tablet was not a claimed feature of the '908 invention, it is not necessary for the structural or functional distinction between the invention and the prior art to be claimed. See Amgen, 580 F.3d at 1370 (“Those structural and functional differences are not explicitly part of the claim, yet are relevant as evidence of no anticipation because of the source limitation.”).
Although there is force to Purdue's argument that the asserted claims of the '908 patent are not invalid based on their character as product-by-process claims, it is not necessary to resolve the parties' dispute over this issue in light of my findings and legal conclusions regarding the issue of obviousness, in which I have addressed all of the limitations of the asserted claims, including the process limitations.
In addition to its “product-by-process” argument, Accord has argued that the asserted claims of the '908 patent are invalid under 35 U.S.C. §112 for lack of an adequate written description and lack of enablement. Given the disposition of the obviousness issue, those issues also need not be addressed.
V. Conclusion
For the reasons set forth above, I conclude that Accord has proved by clear and convincing evidence that the asserted claims of Purdue's '908 patent would have been obvious and are therefore invalid.
IT IS SO ORDERED