Sebela Int’l Ltd. v. Actavis Labs. Florida Inc.

Case Name: Sebela Int’l Ltd. v. Actavis Labs. Florida Inc., No. 14:6414 (CCC) (MF), 2017 U.S. Dist. LEXIS (D.N.J. June 16, 2017) (Cecchi, J.)

Drug Product and Patent(s)-in-Suit: Brisdelle® (paroxetine mesylate); U.S. Patents Nos. 7,598,271 (“the ’271 patent”), 8,658,663 (“the ’663 patent”) and 8,946,251 (“the ’251 patent”)

Nature of the Case and Issue(s) Presented: Sebela owns the NDA for Brisdelle, comprising paroxetine mesylate and indicated for treatment of moderate to severe vasomotor symptoms associated with menopause, including, e.g., “hot flashes.” The ’271 patent claims crystalline paroxetine mesylate having an IR spectrum with 18 peaks, each identified by wavenumber. The other two patents-in-suit claim methods of treatment: the ’663 patent claims the use of paroxetine mesylate; and the ’251 patent claims the use of any pharmaceutically acceptable salt of paroxetine.

The defendants stipulated to infringement of the method patents. They challenged those patents’ validity, and argued that their ANDA products did not infringe the ’271 patent.

Why Defendants Prevailed: The court found the ’271 patent not infringed by defendants’ ANDA products, because even using the IR spectrophotometer resolution proposed by Sebela, plaintiff’s expert was unable to find one of the claimed 18 peaks in their spectra. Relying on Glaxo, Inc. v. Novopharm, Ltd., 110 F.3d 1562 (Fed Cir. 1997), the court declined to “read out” peaks from the claims. The court also noted that plaintiff’s proposed spectrophotometer resolution was not supported by credible testimony; using higher resolutions resulted in even more claimed peaks not being found in the defendants’ products. Further, the court was not persuaded by the overall similarities amongst the IR spectra or DSC onset temperatures, as these characteristics were not claimed.

Next, the court found the two method patents invalid as obvious. The asserted claim limitations were found in the prior art except for the dosage strength, 7.5 mg/day. A first prior-art reference taught treating vasomotor symptoms with 10 mg/day of paroxetine hydrochloride. A second prior-art reference taught that paroxetine mesylate is preferable to paroxetine hydrochloride. A third prior-art reference taught a dose range of 0.21-8.4 mg/day. Further, the court observed that the method patents did not teach that the claimed 7.5 mg/day dose was critical, as the specification taught a range of 0.1-10 mg/mg. In light of that record, the court concluded that the third reference taught the claimed 7.5 mg/day dosage strength.

Having concluding that the differences between the claims and the prior art were “minimal,” the court looked at four secondary considerations of non-obviousness. There was no unmet need because physicians had been prescribing Paxil (paroxetine hydrochloride) for treating vasomotor symptoms. Further, any need was not long-felt because until as recently as four years before the patents’ priority date physicians successfully treated vasomotor symptoms with prescription hormones. The court did find that the failure of two others to win FDA approval for non-hormonal treatments for vasomotor symptoms did weigh in favor non-obviousness. But the court found no more than a “muted response” within industry for Brisdelle. Further, it rejected Sebela’s argument that its 7.5 mg dose was efficacious but also avoided the side effects known to be associated with higher doses. The court rejected this argument, pointing both to prior art that taught a dose-response relationship for paroxetine’s side effects, as well as the general understanding that side effects decrease as a dose is lowered. In light of the weak secondary considerations, the court found the method claims obvious.