A Class of 957 Predicted Salts Is Insufficient To Meet the ‘At Once Envisage' Standard

October 26, 2022

Mylan Pharmaceuticals, Inc. (“Mylan”) appealed from the final written decision of the USPTO Patent Trial and Appeal Board (the “Board”) holding that Mylan failed to show that claims 1–4, 17, 19, and 21–23 of U.S. Patent 7,326,708 (the “’708 patent”) were anticipated or would have been obvious over the cited prior art.

Merck Sharp & Dohme Corp. (“Merck”) owns the ’708 patent, which describes sitagliptin dihydrogenphosphate (“sitagliptin DHP”). Sitagliptin DHP is a dihydrogenphosphate salt of 4-oxo-4-[3-(trifluoromethyl)-5,6-dihydro [1,2,4]triazolo[4,3-a]pyrazin-7(8H)-yl]-1-(2,4,5-trifluorophenyl)butan-2-amine. Sitagliptin DHP belongs to the class of dipeptidyl peptidase-IV inhibitors, which can be used for treating non-insulin-dependent (i.e., Type 2) diabetes. Independent claim 1 recites a sitagliptin DHP salt with a 1:1 stoichiometry.

Mylan petitioned for inter partes review of claims 1–4, 17, 19, and 21–23 of the ’708 patent and argued that the claims were anticipated by WO 2003/004498, a Merck-owned publication (equivalent to U.S. Patent 6,699,871 (collectively “Edmondson”)). Edmondson is directed to compounds which are inhibitors of the dipeptidyl peptidase-IV enzyme and are useful in the treatment of diseases in which the dipeptidyl peptidase-IV enzyme is involved, e.g., type 2 diabetes. Edmondson discloses a genus of DP-IV inhibitors and 33 species, one of which is sitagliptin and further discloses that pharmaceutically acceptable salts can be formed using one of eight particularly preferred acids. The salts may exist in crystalline forms, including as hydrates. Mylan also argued that the claims would have been obvious over Edmondson and two additional publications by Brittain and Bastin.

The Board determined that there was no express disclosure of all of the limitations of the 1:1 sitagliptin DHP salt in Edmondson, and that Mylan could not fill in the gaps by arguing that a skilled artisan would “at once envisage” what was missing. Also, Mylan had not proven an inherent disclosure of the 1:1 sitagliptin DHP salt in Edmondson, and experimental evidence and technical literature undeniably showed that 1:1 sitagliptin DHP did not form every time sitagliptin and DHP were reacted. The Board concluded that the claims were not expressly or inherently anticipated by Edmondson and would not have been obvious in view of Edmondson, Bastin, or Brittain.

The Board considered the issue whether Merck could antedate Edmondson with evidence that it had reduced to practice the subject matter of claims 1, 2, 17, 19, and 21–23 before Edmondson was published on January 16, 2003. The Board concluded that Merck had reduced to practice “at least as much, and in fact more, of the claimed subject matter than was shown in Edmondson. Thus, Merck could successfully antedate the subject matter of claims 1, 2, 17, 19, and 21–23, and thus Edmondson was not a 35 U.S.C. § 102(a) reference,” but was a 35 U.S.C. § 102(e) (pre-AIA) reference. Because it was undisputed that the inventions claimed in the ’708 patent and the subject matter of Edmondson were commonly owned by Merck at the time of the invention, the Board determined that the 35 U.S.C. § 103(c)(1) (pre-AIA) exception applied to the claims at issue.

The Board further considered whether claim 3, which recites the (S)-configuration of sitagliptin DHP, and claim 4, which recites the crystalline monohydrate form of (R)-sitagliptin, would have been obvious in view of Edmondson, Bastin, and Brittain and found that the references did not disclose “anything related to (S)-sitagliptin or even a racemic mixture of any sitagliptin salt.” The Board concluded that Mylan did not show that claim 3 would have been obvious. The Board also found that Mylan provided no rationale to explain why a person of ordinary skill would have been motivated to make the claimed crystalline monohydrate form of 1:1 sitagliptin DHP of claim 4 and failed to show that a skilled artisan would have had a reasonable expectation of success in making the crystalline monohydrate form of the 1:1 sitagliptin DHP salt. Thus, the Board held that Mylan had not demonstrated that claims 1–4, 17, 19, and 21–23 were anticipated or would have been obvious. Mylan appealed.

Mylan raised three challenges on appeal at the Court of Appeals for the Federal Circuit (“the Court”). First, Mylan contended that the Board erred in determining that a 1:1 stoichiometry of sitagliptin DHP was not anticipated by Edmondson. Second, Mylan contended that the Board erred in determining that the ’708 patent antedated Edmondson. Third, Mylan contended that the Board erred in determining that Mylan failed to prove obviousness of claims 3 and 4.

On appeal, Mylan argued that Edmondson anticipated the claims because it disclosed sitagliptin in a list of 33 compounds and asserted that Edmondson disclosed acids forming “pharmaceutically acceptable salts,” including phosphoric acid in a list of eight “particularly preferred” acids. Mylan asserted that sitagliptin DHP was effectively disclosed in Edmondson, and Edmondson thus anticipated the challenged claims. Mylan further asserted that a skilled artisan would “at once envisage” a 1:1 stoichiometry of the sitagliptin DHP salt for two reasons. First, Example 7 of Edmondson disclosed a sitagliptin hydrochloride salt having a 1:1 stoichiometry. Second, the experimental data presented by Mylan’s expert illustrated that under the conditions allegedly similar to those in Edmondson only a 1:1 sitagliptin DHP stoichiometry could be formed. Mylan contended that the Board thus erred in holding that a 1:1 stoichiometry was not anticipated by Edmondson.

Merck argued that the combined list of 33 compounds and eight preferred salts, taking into account various stoichiometric possibilities, would result in 957 salts, some of which might not even form under experimental conditions. “That, Merck asserts, does not meet the standard set by the ‘at once envisage’ theory. Merck argues that Mylan seeks to expand the theory inappropriately, improperly focusing on whether skilled artisans could have envisaged 1:1 sitagliptin DHP among the members of the class instead of envisaging each member of the disclosed class.” Merck asserted that Mylan used hindsight to single out one compound from the large class and further argued that Mylan’s own expert conceded that Edmondson did not direct a skilled artisan to sitagliptin from among the 33 DP-IVs, and it did not disclose a phosphate salt of any DP-IV inhibitor.

The Court agreed with Merck that the Board had not erred in determining that Edmondson did not expressly disclose a 1:1 sitagliptin DHP salt. The Court pointed out that the Board grounded its finding in the testimony from Mylan’s own expert stating that nothing in Edmondson directed a skilled artisan to sitagliptin from among the 33 listed DP-IV inhibitors. Further, nothing in Edmondson singled out phosphoric acid or any phosphate salt of any DP-IV inhibitor, and the list of “pharmaceutically preferred” salts included 44 pages earlier in the specification. The Board reasonably concluded that Edmondson did not expressly disclose the 1:1 sitagliptin DHP salt. The Court also agreed with Merck that the Board had not erred in determining that Edmondson did not inherently disclose a 1:1 sitagliptin DHP salt.

In re Petering stands for the proposition that a skilled artisan may “at once envisage each member of [a] limited class, even though the skilled person might not at once define in his mind the formal boundaries of the class.” 301 F.2d 678, 681 (emphasis added). The key term here is “limited.”  The list of 33 compounds, with no direction to select sitagliptin from among them, plus the eight “pharmaceutically preferred” acids and various stoichiometric possibilities, results in 957 salts, some of which may not exist. “That is a far cry from the 20 compounds ‘envisaged’ by the narrow genus in Petering.” The Court noted that Mylan’s own expert even stated that salt formation was an unpredictable art that required a “trial and error process.” A specific number defining a “limited class” depends on the “class.” “But we agree with Merck and hold that the Board did not err in finding that a class of 957 predicted salts that may result from the 33 disclosed compounds and eight preferred acids, some of which may not even form under experimental conditions, is insufficient to meet the ‘at once envisage’ standard set forth in Petering.”

Concerning Mylan’s challenge to the Board’s determination that Mylan failed to prove that claims 1–4, 17, 19, and 21–23 would have been obvious the obviousness, Mylan argued that the Board erred in finding that Merck’s reduction to practice of the 1:1 (R)-sitagliptin DHP salt antedated Edmondson, because Edmondson disclosed sitagliptin hydrates, and Merck had not made hydrates of 1:1 sitagliptin DHP until March 2003, about two months after the January 16, 2003 Edmondson publication date. Mylan also argued that the Board erred in finding that Edmondson did not disclose hydrates of sitagliptin phosphate.

Merck argued that Edmondson could not serve as an obviousness reference and the Court agreed with Merck. Merck showed that it developed a 1:1 sitagliptin DHP salt in December 2001 with experimental confirmation in early 2002. As Merck highlighted, Mylan did not argue that claim 4, directed to a crystalline monohydrate, was anticipated by Edmondson, “which it could have done had it believed that Edmondson disclosed a crystalline monohydrate.” The Board’s finding that Edmondson did not disclose 1:1 sitagliptin DHP and did not disclose a hydrate of that salt was supported by substantial evidence. The Court agree with the Board that Merck reduced to practice “more . . . than what is shown in [Edmondson] for the claimed subject matter.”

The Court also agreed with Merck that the Board’s decision that Mylan failed to show obviousness of claims 3 and 4 of the ’708 patent was supported by substantial evidence. With respect to claim 3, the Board found that there was no motivation to combine Edmondson and Bastin to make sitagliptin DHP and that the two references did not provide motivation to make (S)-sitagliptin. The Board further highlighted that Mylan advanced no expected or theoretical benefit to making the (S)-enantiomer of 1:1 sitagliptin DHP, and that the general disclosure on diastereomers in Edmondson encompassed millions of potential compounds and salts with no motivation to make the (S)-enantiomer with a reasonable expectation of success, particularly in an unpredictable activity like salt formation. With respect to claim 4, the Board found that there was no motivation to combine the references and that a skilled artisan would have had no reasonable expectation of success in doing so. The Board credited Mylan’s expert testimony that a skilled artisan “couldn’t predict with any degree of certainty” hydrate formation. The Board further addressed the numerous downsides of hydrates reported in the literature, including those stating that a skilled artisan would have several reasons for avoiding hydrates. The Board also credited Merck’s expert testimony who stated that a skilled artisan would have sought to avoid hydrates, and that forming crystalline salts, including hydrates, was highly unpredictable.

The Court agreed with Merck that the Board’s decision was supported by substantial evidence and affirmed the Board’s decision that the claims of the ’708 patent were not anticipated and also would not have been obvious over the cited prior art.

Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp., Case No. 2021-2122 (Fed. Cir. Sep. 29, 2022