Case Name: Depomed, Inc. v. Actavis Elizabeth LLC, No. 12-1358-JAP, 2014 U.S. Dist. LEXIS 118096 (D.N.J. Aug. 18, 2014) (Pisano, J.)
Drug Product and Patent(s)-in-Suit: Gralise® (gabapentin); U.S. Patents Nos. 6,635,280 (“the ’280 patent”), 6,488,962 (“the ’962 patent”), 7,438,927 (“the 927 patent”), 7,731,989 (“the ’989 patent”), 8,192,756 (“the ’756 patent”), 8,252,332 (“the ’332 patent”), and 8,333,992 (“the ’992 patent”)
Nature of the Case and Issue(s) Presented: Depomed sued Actavis for infringing seven patents related to a dosage form capable of being retained in the stomach, allowing for the delayed-release of gabapentin in the small intestine. Gabapentin is unique in that it is only absorbed in the upper portion of the small intestine and can only be absorbed in small amounts at a time. For a once-daily gabapentin tablet to be effective, it must remain in the stomach for a long time, slowly releasing gabapentin into the small intestine. The patents-in-suit disclose a solution that alters the size and shape of the tablet to swell in the presence of gastric fluid. When food is present in the stomach, the pylorus contracts to prevent large food particles from leaving the stomach and entering the small intestine. This state is known as the “fed mode.” Because the tablet swells in the stomach, it cannot pass through the pylorus in the fed mode, allowing the slow release of gabapentin over time.
Depomed alleged that Actavis’ ANDA product directly infringed the ’927, ’989,’756, ’280, and ’962 patents and indirectly infringed the ’927, ’756, ’332, and ’992 patents. Actavis responded that the patents-in-suit were invalid as obvious and that the ’280 patent was indefinite. The court found the patents-in-suit valid, definite, and infringed.
Why Depomed Prevailed: The ’927, ’989, and ’756 patents claim that the tablet “swells…to increase its size to promote gastric retention.” The court found that Actavis’ ANDA product infringed because it contained a similar amount of hydrophilic polymers, which swell in gastric fluid, to Depomed’s Gralise product. In addition, the proposed ANDA labelling noted that the tablets swell in the stomach to gradually release gabapentin. The parties also stipulated that Actavis’ ANDA products release gabapentin by diffusion, controlled through the use of high molecular weight hydrophilic polymers that swell in water. Studies, corroborated by expert testimony, demonstrated that the generic tablets swell and gain mass in aqueous solutions simulating gastric fluid or in water. Thus, the generic tablets function according to the same mechanisms as Gralise, and infringed the asserted patents.
The ’280 patent claims that a tablet, when swollen, “is of a size exceeding the pyloric diameter in the fed mode to promote retention in the stomach during said fed mode.” Actavis argued that Depomed did not prove the generic tablet swells and, even if swelling did occur, presented no evidence that the swollen tablet’s size exceeded that of the pyloric diameter during the fed mode. Alternatively, Actavis alleged that its 600 mg tablet, which already exceeded the size of the contracted pylorus, could not infringe because the swelling did not cause the tablet to become larger than the pyloric diameter. Testing of the generic tablets, along with expert testimony, demonstrated that the size of the swollen tablets was greater than the contracted pylorus. Plaintiffs had argued that Actavis’ 600 mg tablet could infringe because the claim required only that the swollen tablet be larger than the pylorus during the fed mode, regardless of its initial size.
Actavis further argued that its 600 mg tablet could not infringe the ’962 patent because the tablet had a modified capsule shape, not an oval, as required by the claims. Actavis’ contention was based on a definition in tableting literature, and was different from the stipulated construction agreed to by the parties. The court rejected Actavis’ argument, relying on expert testimony that a skilled artisan would understand that the tablet’s shape was an oval. Further, Actavis referred to an “oval shape” when it purchased dies used to manufacture the tablet. Actavis also failed to provide credible evidence to support deviating from the stipulated construction of the term.
Depomed also alleged that Actavis was liable for contributory and induced infringement of the ’927, ’332, and ’992 patents. The court found that physicians and patients directly infringed the patents when using the generic tablets as directed. Actavis knew about the patents because they were a part of Actavis’ Paragraph IV certification to the FDA. Further, as labelled, the generic tablets are used to treat only postherpetic neuralgia. Because an ANDA product cannot be marketed for an off-label use, significant non-infringing uses did not exist. Actavis stipulated that its product treats the same disease as Gralise, and thus is a material part of the patented invention. Additionally, the court recognized that the generic product’s label instructs the user to take a tablet orally, once each day, with a meal. Actavis knew that doctors and patients would follow these directions to use the generic tablet. Thus, Actavis knowingly and intentionally induced the infringement of the asserted patents.
Next, the court found that the patents-in-suit were not obvious. Specifically, the claims included five limitations not present in the prior art. Further, the court found that a skilled artisan would have been dissuaded from creating a delayed-release gabapentin tablet. At the time of filing, it was believed that gabapentin’s bioavailability quickly reached a saturation point, and did not significantly improve when the drug was administered over an extended time. In addition, it was believed that the stomach’s environment caused the degradation of gabapentin to lactum, which is toxic to humans. A skilled artisan would not have thought to pair a gabapentin tablet with food because the protein content in the food changed the amount gabapentin absorbed, making it difficult to formulate an effective dosage of the drug. Gabapentin also has different pharmacokinetic properties from those of other compounds disclosed in the prior art, and therefore substituting gabapentin would require more than mere routine optimization.
Additional considerations also supported the court’s finding of non-obviousness. There was a long felt need for an extended-release gabapentin drug to treat neuropathic pain. Gralise, the embodiment of the asserted patents, fulfilled that need upon release. And although not wildly successful, Gralise performed moderately well commercially in a mature and competitive marketplace.
Actavis presented a different validity argument for ’962 patent, which discloses the use of an oval-shape tablet to prolong the length of time the tablet remains in the stomach. The prior art taught pills that would be retained in the stomach, swell, and release a drug. But the prior art did not disclose that an oval-shaped tablet would lead to greater retention times. Although oval-shaped tablets existed in the prior art, they were meant to dissolve, not be retained in the stomach. Additionally, the prior art taught tablets capable of swelling, but did not disclose limitations on how long it took the swelling to occur. Thus, the court also found the ’962 patent was valid.
Finally, the court held that the ’280 patent was not indefinite. Claim 1 of the ’280 patent refers to the “pyloric diameter in the fed mode.” During the fed mode, the pylorus, while generally shut, periodically opens to allow some food particles to pass into the small intestine. Actavis argued that the claim was indefinite because it was impossible to tell if the “pyloric diameter” referred to the open or closed state. The court disagreed, and ruled that the context of the claim made it clear that the claim language referred to the pylorus in the open state. Further, the specification of the patent and expert testimony detailed sizes of particles that could pass through the pylorus during the fed mode corresponding to the open diameter, not the closed.
