While the IP law world is being rocked by new developments in biosimilars, patent-eligible subject matter and IPR’s, it is almost refreshing to see a decision upholding the validity of claims to a “small molecule” drug based on application of the principles of obviousness–particularly the no-no of hindsight reconstruction. Shire LLC v. Amneal Pharmaceuticals, LLC, Appeal no. 2014-1736 et al. (Fed. Cir., September 24, 2015). (A copy of this decision can be found at the end of this post.)
This was a straightforward Hatch-Waxman infringement suit, during which Shire had to defend the validity of four patents covering a derivative of “speed,” L-lysine-d-amphetamine (“LDX”) dimesylate, marketed as Vyvanse® to treat autism. Defendants needed to convince the court that it would be obvious to both make the L-lysine derivative of d-amphetamine and then to make the dimesylate salt.
To meet their burden, Defendants had a published Australian patent application (AU ‘168) and US Pat No. 3,843,796 (“Miller”). Without using the term “picking and choosing,” the court still upheld the district court’s grant of S.J. of validity. While AU ‘168 disclosed L-lysine and d-amphetamine, the court focused on the breadth of the ‘168 disclosure, which D- or L-amino acids, “and the like” as combinable with D- or L-amphetamine. The court found that defendant’s expert had conceded that the closest formula, “does not indicate any preference” among the different compounds usable in the formula.
Example 4 of AU ‘168 seemed pretty close, since it disclosed Nα-tosyl-L-lysine-D-amphetamine, but the court stated that the mere fact that the tosyl group could be removed to yield LDX did not establish motivation to actually remove it, since the compound in the Example was disclosed to be a final product. The Miller patent was dismissed as showing a “base compound” that was not amphetamine, but rather had two structural alterations. [Ed.’s note: Is this a return to the old “two changes” rules of structural unobviousness that I learned in the 80’s?]
The court found that the Defendants relied on hindsight reconstruction:
“[AU ‘168] provides ‘no direction as to which of many possible choices is likely to be successful.’ [citing Unigene v. Apotex] Thus AU ‘168 does not make LDX obvious to try….Defendants can only come to LDX by ‘retrac[ing] the path of the inventor with hindsight.'[citing Ortho-McNeil Pharm. v. Mylan] We therefore reject the hindsight claims of obviousness. [citing In re Cyclobenzaprine…Capsule Patent Litig.].”
In another interesting gloss on KSR, the court stated the “rule” as “[f]or a patent to be obvious, ‘some kind of motivation must be shown…so that a jury can understand why a [POSA] would have thought of either combining two or more references or modifying one to achieve the patented method.” [citing Immunogenetics v. Abbott]. That is a big “or” since it suggests that if it is necessary to both modify references and then to combine them, a prima facie case of obviousness cannot be established.
The safe harbor of s. 271(e)(1) was also widened a bit when the panel found that the activities of Johnson Matthey, the drug supplier to the Defendants, did not constitute induced infringement. J-M did not intend to market the drug, and the court found that its activities were “reasonably related to the submission of an ANDA.” Since J-M did not file an ANDA, they are not liable for infringement under s. 271(e)(2).