Federal Circuit Affirms Decision Preventing Generic Lyrica, Finding Sparse Evidence of Obviousness

A three judge panel of the Federal Circuit recently affirmed the District of Delaware’s judgment of infringement and validity related to U.S. Patent No. 6,197,819 (owned by Northwestern University and licensed to Pfizer) covering the drug Lyrica, which treats seizures and certain types of nerve pain.  The decision effectively prevents a generic version of Lyrica from entering the market until the ‘819 patent expires in 2018.

Teva, Actavis, Mylan, Sun and several other generic drug makers (the appellants) filed ANDAs seeking to market generic versions of Lyrica.  After a bench trial the district court held that the ‘819 patent was not obvious, and the generic drug makers infringed.  The only claim at issue on appeal was claim 2 of the ‘819 patent, the compound 3-isobutylGABA.

Appellants first argued that claim 2 should have been construed to mean only a particular stereochemical form of 3-isobutylGABA, and because their proposed products would not employ such a form, they did not infringe.  However, the Federal Circuit held that the plain language of the claim did not support such a narrow construction.  Furthermore, the fact that the specific stereochemical form appeared in a table in the specification showed that the patentee was capable of specifying particular stereochemical forms but chose not to in claim 2.

Next, appellants argued that claim 2 was not enabled because the specification did not teach how to prepare every conceivable form of 3-isobutylGABA.  The Federal Circuit rejected that argument, noting that the parent application disclosed the method of synthesizing the compound and specific forms could be prepared by methods already well known in the art.

Finally, the Federal Circuit rejected appellants’ argument that claim 2 was obvious, siding with the district court that appellants’ evidence was too sparse.  In part, appellants argued that the disclosure in three pieces of prior art of 3-isopropylGABA (as opposed to the claimed 3-isobutylGABA) would have directed artisans to modify other substituent groups at GABA’s 3-position.  The court, however, held that appellants failed to show why 3-isopropylGABA would have been selected for further research in the first place.  At the time, it was unclear which structures were significant for anticonvulsant activity.  In any event, a vague suggestion in the prior art pointing to a broad class of compounds, without identifying isobutyl among the millions of compounds, is not a teaching of specific molecular modification such that a lead compound could be modified with a reasonable likelihood of success.

This case shows once again that the mere identification of prior art disclosing a chemical structure related to a compound in suit is the easy part of an obviousness argument.  The difficult task for those attempting to invalidate a claimed compound is putting forward sufficient evidence teaching a specific modification.