The Federal Circuit in Shire v. Amneal, a Hatch-Waxman case, reaffirmed that the burden rests with the defendants to prove motivation for a person of ordinary skill in the art to arrive at a claimed chemical compound. 1 The ANDA filer asserted that 18 claims of four Shire patents (U.S. Pat. Nos. 7,105,486; 7,655,630; 7,659,253; and 7,662,787) covering L-lysine-d-amphetamine (LDX) dimesylate (marketed as Vyvanse® to treat ADHD) were invalid due to obviousness. The court also considered whether an API supplier should be held liable for induced infringement for supplying the ANDA filers with the patented API.
The district court granted summary judgement in favor of Shire that the Shire patents were not invalid as anticipated or obvious, and the defendants infringed each of the Shire patents. The district court also ruled that the API supplier was liable for induced infringement.
On appeal, the defendants asserted that the Australian App. No. 54,168/65 (“AU ‘168”) and U.S. Pat. No. 3,843,796 (“Miller”) disclosed LDX as well as the mesylate salts of LDX, rendering the Shire patents obvious. Specifically, the defendants asserted that AU ‘168 disclosed amphetamine conjugated to amino acids, including lysine. AU ‘168 also suggests protected modifications of amphetamines, e.g., mesylated amphetamines. The defendants further asserted example 24 of AU ‘168 broadly taught N-tosyl-LDX, and claimed it was an obvious derivation of LDX.
In examining the references, the court emphasized reading the whole reference in context in determining whether a person of ordinary skill in the art would have arrived at the invention. Using this approach, the court found that defendants’ specific assertion that the disclosure of conjugating amphetamines to amino acids, including lysine, and a preference for L-amino acids and d-amphetamine in AU ‘168, did not provide sufficient motivation to envision LDX due to the broad disclosure of AU ‘168. AU ‘168 listed 18 amino acids “and the like, and state[d] they can belong to the D- or L- series” greatly increasing the potential amino acid groups to be utilized. 2The court further found that the disclosure and formulas in AU ‘168 did not disclose a finite and limited class including LDX or describe a preference that would lead a person of ordinary skill in the art to come to LDX.3
The court also found that example 24 of AU ‘168, which differs from LDX in that it contains a tosyl group, and a method of how to remove the tosyl group would not have rendered the Shire patents obvious because example 24 is a final product and not an intermediate synthesis product and the reference provides no motivation to perform such method. 4 Of the 30 specific examples, none were LDX. As such, the court noted that the only way for a person of ordinary skill in the art to reasonably select LDX would be through hindsight.
Inducement to Infringe of API Supplier
Johnson Matthey was the active pharmaceutical ingredient (API) supplier for the defendants. The district court had found that Johnson Matthey had induced infringement of the compound claims at issue.
Johnson Matthey argued that providing the ANDA defendants with the API so that the defendants could submit their ANDA was reasonably related to the submission of information under federal law, and thus within the safe harbor of § 271(e)(1). The Fed Circuit court agreed, ruling that Johnson Matthey had done nothing more than provide material for use by the ANDA defendants in obtaining FDA approval, and was thus protected from liability by the safe harbor of § 271(e)(1) because such activities were “reasonably related to the submission of an ANDA.” 5
2 Id. slip op. at 10.
3 Id. at 11-12 (“out of the thousands of possible compounds it discloses, AU ‘168 actually provides thirty specific examples, none of which is LDX”).
4 Id. at 11.
5 Id. at 15.