By Kevin E. Noonan —
In 2004, Purdue Pharma L.P. and its related companies were found to have obtained patents (U.S. Patent Nos. 5,656,295; 5,508,042; and 5,549,912) for their OxyContin® opioid pain relief formulations through inequitable conduct. The basis for this determination was that Purdue had maintained before the Patent Office that its formulations were patentable because they were able to relieve pain in 90% of patients using a reduced dosage range compared with prior art opioid formulations. The basis for these assertions was not clinical data but an "insight" by one of its scientists — the District Court ruling that Purdue had misled the Patent Office into believing its assertions of patentability were based on clinical data.
The Federal Circuit initially affirmed, but on rehearing, vacated the judgment and remanded the case back to the District Court for reconsideration. The Federal Circuit also affirmed the District Court’s judgment that defendants Endo Pharmaceuticals, Inc., Teva Pharmaceuticals, Inc., Boehringer Ingelheim GmbH, and Impax Laboratories Inc. infringed Purdue’s patents, and these defendants ultimately settled with Purdue. In the meantime, several other generic drug manufacturers — including Mallinckrodt, Inc., KV Pharmaceutical Co, and Actavis Totowa L.L.C. — filed ANDA applications with the Food and Drug Administration, and Purdue brought suit pursuant to the provisions of the Hatch-Waxman Act (35 U.S.C. § 271(e)(2)).
The District Court decided to try the inequitable conduct case first, and last Monday issued its opinion. The Court found that, while not entirely blameless, Purdue’s conduct did not rise to the level of inequitable conduct. The defendants’ inequitable conduct assertions included the following eight bases; for each of which the Court assessed the materiality of the action and the evidence of an intent to deceive:
1. Purdue lacked empirical evidence for claiming a reduced dosage range.
This basis had previously been successfully argued at the earlier trial and before the Federal Circuit. Its basis was that Purdue "implied" to the patent Examiner that clinical studies supported the assertion that the claimed formulation was effective over a reduced dosage range, when in fact the formulation was based on an insight by one of the named inventors.
The Federal Circuit had agreed with the District Court that Purdue’s misrepresentations in this regard were material to patentability, but had determined that the degree of its materiality "was not especially high" but was in fact "relatively low," based in part on its being merely an omission and not an affirmative representation.
The District Court next considered the evidence of an intent to deceive. In the first trial, the Court had relied on statements and documents relating to Purdue’s labeling claims to the Food and Drug Administration. The Court also considered Purdue’s reduced dosage range claims in the face of its inability to titrate the dosage to indicate an intent to deceive. The Federal Circuit held this to be erroneous, in part because the "quantum" of proof required to satisfy patentability requirements and FDA regulatory requirements were different.
In view of the Federal Circuit’s analysis, the District Court required the defendants to come forward with additional evidence of an intent to deceive. The District Court found that the defendants failed to adduce such evidence, proffering "only vague allegations" that Purdue’s business interests provided an incentive to obtain a patent on OxyContin®. This, the Court said, was no more than "scant evidence" of an intent to deceive, and thus the Court refused to find inequitable conduct in this behavior.
2. There was inadequate support for asserting that most other opioids had an eight-fold dosage range.
The defendants also asserted that Purdue had inadequate evidence for its assertions that, while other opioid analgesic formulations exhibited an eight-fold dosage range for controlling pain in 90% of patients, OxyContin® had a four-fold dosage range (thus requiring lower doses of drug and avoiding significant side-effects). The defendants alleged that Purdue had no evidence for these assertions.
The District Court disagreed, finding that Purdue had adequate evidence for morphine, from a report prepared by its expert two years prior to filing the parent patent application. The Court also credited an American Pain Society report that suggested opioid analgesics were effective over an eight-fold dosage range, and thus found that any misrepresentation on this basis was "vanishingly small." In view of this conclusion, the Court did not even reach the question of evidence for an intent to deceive.
3. Purdue failed to disclose clinical results that contradicted the dosage and titration claims.
The defendants also cited three scientific studies alleged to contain clinical results contradicting Purdue’s assertions regarding its reduced dosage range and ease of titration. The District Court found none of these reports to be material: one described an experimental protocol that precluded titration, and the other two failed to test the dosage range asserted by Purdue in support of its patentability arguments (nor did either of these references study the ease of titrating other opioid analgesics). In addition, the latter two reports had been asserted in the earlier litigation and the Court had not found them to be material. Here, the Court said that the defendants had not provided "a sound basis to reconsider its prior determination." Again, the Court did not reach the question of an intent to deceive in view of this determination on materiality.
4. Whether Purdue’s finding regarding "peak" plasma levels and duration of pain relief were "surprising."
Purdue had asserted during patent prosecution that its finding that OxyContin® achieved peak plasma levels at 2-4 hours after administration, and provided 12 hours of pain relief was "surprising." The defendants argued that oxycodone was the fifth opioid to exhibit this behavior, and thus it could not be a "surprising" result since Purdue had devised formulations of morphine, dihydrocodeine, hydromorphone, and codeine that exhibited the same pattern. In response, Purdue asserted that since every drug is different, finding this relationship between peak plasma concentration and pain relief for OxyContin® was indeed surprising and was not a misrepresentation. These assertions were supported by expert testimony regarding the unpredictability of pharmacokinetics, and by the American Pain Society report discussed above, which stated that "different patients respond differently to different opioids."
The District Court said that the defendants had provided no scientific evidence contrary to Purdue’s assertions and evidence, but instead relied upon the "common sense proposition" that a result cannot be surprising the fifth time it is obtained. The Court characterized this as "facile reasoning" not supported by Purdue’s "unchallenged evidence," and held that the defendants had not sustained their burden of showing that Purdue made a material misrepresentation to the Patent Office when it characterized its OxyContin® results as being "surprising."
5. Purdue failed to disclose other controlled-release opioid formulations.
The defendants further maintained that Purdue committed inequitable conduct by not disclosing the existence of other controlled release opioid formulations to permit the Examiner to assess whether he agreed that the results asserted for OxyContin® were surprising. Purdue had in fact disclosed controlled-release morphine and hydromorphine, but not codeine and dihydrocodeine.
The District Court found that this failure to disclose the other controlled-release opioid formulations was a material omission, being inconsistent with Purdue’s representations that the results obtained with OxyContin® were surprising. While stating that the level of materiality was low, the Court also found that the evidence of the other, undisclosed controlled-release opioids was not cumulative, since they doubled — from 2 to 4 — the number of other formulations the Examiner could have considered for comparison.
However, the District Court did not find Purdue’s failure to disclose the existence of these additional controlled-release opioid formulations to be inequitable conduct, because it found no evidence of an intent to deceive. The Court found credible Purdue’s explanation for the failure to disclose: that the two opioids it did disclose were more relevant to OxyContin® because they, like OxyContin® were used to control moderate to severe pain, while the undisclosed opioids were used for controlling mild to moderate pain. The Court characterized Purdue’s behavior in deciding whether to disclose the different opioid formulations as a "good faith" conclusion that morphine and hydromorphine were relevant to patentability of OxyContin®, and codeine and hydrocodeine were less relevant (and cumulative). While disagreeing with this conclusion, the Court found that the defendants had provided no evidence of any intent to deceive.
6. Failure to disclose that their expert witness was not an independent expert.
This allegation of inequitable conduct was based on Purdue’s failure to expressly inform the Patent Office that its declarant, Dr. Kaiko, was not an independent expert but was an employee. The District Court’s determination that this was a material omission stemmed directly from the Federal Circuit’s decision in Ferring A.V. v. Barr Labs, Inc.
However, the District Court found "overwhelming" evidence that the omission was made in good faith, based on Dr. Kaiko’s status as a named inventor on each of the patents-in-suit. The Court reasoned that had Purdue wished to conceal Dr. Kaiko’s identity or interest, it would not have named him as an inventor. The Court also considered the substance of Dr. Kaiko’s declaration to indicate to the Examiner that he was involved in developing the invention, and distinguished this case from Nilssen v. Osram Sylvania Inc., where the patentee actively concealed the affiant’s affiliation from the Examiner.
7. Failure to disclose U.S. Patent No. 4,861,598.
The District Court agreed that Purdue failed to provide this patent reference to the Patent Office. However, the omission was not material because the Examiner uncovered the reference and used it to reject certain claims during prosecution. The Court noted that a reference cannot be found to have been withheld by the Examiner when the evidence is that the Examiner considered the reference, and thus that Purdue’s nondisclosure was not a material omission.
8. That Purdue’s conduct, taken as a whole, comprised a pattern that was evidence of an intent to deceive.
Having failed to persuade the District Court with clear and convincing evidence of inequitable conduct, the defendants asserted that, taken together, Purdue’s actions were sufficient to comprise a pattern that evinced an intent to deceive. The Court reviewed all of the defendants’ allegations, and found that at best the defendants had shown Purdue to have committed a single misrepresentation of low materiality, a single immaterial omission, and two omissions made in good faith. On this evidence, the Court declined the defendants’ invitation to find a pattern of behavior consistent with an intent to deceive.
Having had the integrity of its patents (and reputation) restored, Purdue is now in a position to assert these patents against the defendant ANDA filers. Provided that none of the defendants have modified the formulation in a non-infringing way, it is likely that Purdue will prevail, and consumers will need to wait for a successful challenge of these patents, or until their expiry, before generic OxyContin® will become available.
Patent Docs 
Leave a comment