The Federal Circuit recently reversed a district court finding that a Bayer patent covering formulations of the erectile dysfunction (“ED”) drug vardenafil hydrochloride trihydrate (Levitra®) was not obvious.  Bayer Pharma AG et al. v. Watson Laboratories, Inc., et al. 2016-2169 (Fed. Cir. Nov. 1, 2017).  In reaching this decision, the Court found that the district court judge erred in reaching his decision based primarily on a credibility assessment even though Plaintiff’s expert did not address important prior art evidence.

The claims at issue were directed to an orally disintegrating tablet (ODT) containing vardenafil and two sugar alcohols, mannitol and sorbitol, where the drug is released in the mouth without swallowing (immediate release).  One claim required a mixture of sorbitol and mannitol, while the other claim only required sorbitol.

Credibility Assessments.  At trial, Watson had introduced a number of prior art references disclosing the applicability of ODT formulations to ED drugs, some of which taught the use of vardenafil in an ODT.  The Court found that these prior art references directly contradicted the judge’s finding that ODTs were not considered particularly applicable to ED drugs.  Neither Bayer’s expert nor the judge addressed these references.  The Federal Circuit acknowledged that while it is within the judge’s discretion to credit one expert over another, it is error to “ignore the wealth of [prior art] evidence” based on a simple credibility assessment.

Lack of Commercial Availability.  The Court also concluded that the district court placed too much emphasis on the absence of an ODT formulation from the commercial market.  The Court explained that “[t]he motivation to combine inquiry is not limited to what products are forthcoming or currently available on the market,” particularly given the lengthy FDA approval process.

Teaching Away.  The district court also found that the prior art taught away from an ODT formulation because a POSA would have expected vardenafil to have a bitter taste and that an immediate release formulation would potentially increase bioavailability in elderly users.  The Court accepted the underlying factual findings—bitter taste and risk of increased bioavailability—but held that these potential concerns did not amount to a teaching away.  A reference teaches away when it suggests that a particular path is unlikely to provide a desired result, not when it merely identifies the existence of “a better alternative.”  Thus, even if the prior art taught a preference for a delayed-release formulation, that would not be a teaching away from an immediate release formulation.

The Federal Circuit’s Bayer decision serves as a strong reminder that cases do not often turn solely upon credibility determinations, but require careful consideration of the underlying prior art evidence.  It also lends legal support for generic defendants confronting arguments based on the lack of commercially available applications of the technology at issue—commercial availability is not a test for obviousness.  Finally, the Federal Circuit has reiterated that “teaching away” requires more than just a preference for one option over another, but a showing that the disfavored option likely will not work for its intended purpose.