The US Court of Appeals for the Federal Circuit affirmed a Patent Trial & Appeal Board (Board) decision that prior art disclosing a class of 957 salts could not inherently anticipate claims to a salt within the class because a skilled artisan could not “at once envisage” every class member. Mylan Pharms. Inc. v. Merck Sharp & Dohme Corp., Case No. 21-2121 (Fed. Cir. Sept. 29, 2022) (Lourie, Reyna, Stoll, JJ.)

In the underlying inter partes review, Mylan alleged that Merck’s patent claims to sitagliptin dihydrogen phosphate (DHP) with 1:1 stoichiometry were anticipated by two similar Merck publications (collectively, Edmondson). Edmondson listed 33 enzyme inhibitors (including sitagliptin) and eight preferred acids for forming salts with the inhibitors. Mylan argued that the 1:1 stoichiometry between sitagliptin and DHP (which was required by the claims) was the only possible result when sitagliptin and phosphate were reacted.

In response, Merck experts declared that Edmondson did not expressly disclose any 1:1 sitagliptin DHP salts. They also declared that non-1:1 sitagliptin phosphate salts existed and had been created using conventional protocols, and that Edmondson encompassed approximately 957 predicted salts of DP-IV inhibitors.

The Board held that Edmondson did not expressly anticipate because it did not literally disclose the 1:1 sitagliptin DHP salt and Mylan could not attempt to fill in the missing claim limitation by arguing that a person of ordinary skill in the art (POSA) could “at once envisage” the “950+” salts. Merck’s evidence convinced the Board that non-1:1 sitagliptin phosphate salts “do exist and can form.”

Mylan tried to circumvent Merck’s antedation of Edmondson by asserting that it disclosed hydrates of 1:1 sitagliptin DHP, which Merck had not synthesized until months after Edmondson was published. The Board rejected this argument, noting that Edmondson only generically referred to hydrates. Since Mylan had not contested Merck’s common ownership of Edmondson’s subject matter, § 103(c)(1) applied and Edmondson became unavailable as an obviousness reference. The remaining claims to specific enantiomers and hydrates of sitagliptin DHP were deemed nonobvious because Mylan had not presented sufficient evidence to show motivation to make or reasonable expectation of success.

On appeal, the Federal Circuit found that substantial evidence supported the Board’s determinations with respect to explicit and inherent anticipation and obviousness. Mylan’s own expert had admitted that nothing in Edmondson directed a POSA to sitagliptin or to any phosphate salt. Edmondson’s disclosure of 957 potential salts was “a far cry” from the facts in the 1962 Court of Customs and Patents Appeals case In re Peterson, where a reference disclosing only 20 compounds was deemed inherently anticipatory. The Federal Circuit rejected Mylan’s antedation argument, noting that if Edmondson did not explicitly disclose 1:1 sitagliptin DHP, it could not disclose any hydrates of that compound either.

Finally, the Federal Circuit agreed with the Board that the claims to specific enantiomers or hydrates of sitagliptin DHP were nonobvious because Mylan had not shown any expected benefit to making the specific enantiomers claimed, the literature and experts for both sides reported many downsides to using hydrates, and experts for both sides had declared that salt formation was unpredictable.