Mylan Pharmaceuticals Inc. appeals the U.S. Patent and Trademark Office Patent Trial and Appeal Board's decision, which upheld the validity of claims 1–4, 17, 19, and 21–23 of U.S. Patent 7,326,708 (the '708 patent) owned by Merck Sharp & Dohme Corp. The Board determined Mylan did not demonstrate that these claims were anticipated or obvious based on prior art. The '708 patent pertains to sitagliptin dihydrogenphosphate (sitagliptin DHP), a dipeptidyl peptidase-IV (DP-IV) inhibitor used for treating Type 2 diabetes. Independent claim 1 specifies a sitagliptin DHP salt with a 1:1 stoichiometry, while dependent claims detail specific configurations and forms of the compound. Mylan's inter partes review petition contended that the claims were anticipated by Merck's International Patent Publication WO 2003/004498 and U.S. Patent 6,699,871, which describe a range of DP-IV inhibitors, including sitagliptin, and mention the formation of pharmaceutically acceptable salts from specific acids, including phosphoric acid. The court affirmed the Board's decision.
Mylan asserted that claims 1–4, 17, 19, and 21–23 were obvious based on Edmondson and publications by Brittain and Bastin. Brittain highlighted the prevalence of crystalline hydrates, noting that about one third of active pharmaceutical ingredients could form hydrates, with half being monohydrates, and discussed challenges in manufacturing hydrates, including lower solubility and chemical instability. Bastin focused on salt selection and optimization, recommending a range of salts for comparison and identifying disadvantages of certain salts like hydrochloric acid.
The Board found no express disclosure of the 1:1 sitagliptin DHP salt in Edmondson, rejecting Mylan's argument that a skilled artisan could fill in the gaps. It determined that Mylan failed to prove inherent disclosure of the salt, noting evidence showed it does not form in every reaction of sitagliptin and DHP. Consequently, claims 1–3, 17, 19, and 21–23 were neither expressly nor inherently anticipated by Edmondson.
The Board further ruled that claims 1–4, 17, 19, and 21–23 were not obvious in light of the cited references. It considered whether Merck could antedate Edmondson with evidence of prior reduction to practice, concluding that Merck had indeed reduced to practice more of the claimed subject matter than reflected in Edmondson. Thus, Edmondson was not a 35 U.S.C. 102(a) reference, only a 35 U.S.C. 102(e) reference due to common ownership or assignment obligations.
Merck did not claim prior reduction to practice for claims 3 and 4. The Board assessed the obviousness of claim 3, relating to (S)-sitagliptin DHP, and claim 4, concerning the crystalline monohydrate form of (R)-sitagliptin, finding that neither Edmondson nor Bastin disclosed relevant information about (S)-sitagliptin or any racemic mixture, leading to the conclusion that Mylan did not demonstrate the obviousness of claim 3.
Mylan failed to demonstrate that a person of ordinary skill would have been motivated to create the claimed crystalline monohydrate form of 1:1 sitagliptin DHP, nor did it prove that such a skilled artisan would have had a reasonable expectation of success in doing so. Consequently, the Board concluded that claim 4 was not obvious at the time of the invention. Mylan's claims 1–4, 17, 19, and 21–23 were also found not to be anticipated or obvious. Mylan subsequently appealed, raising three main challenges: (1) the Board's determination that a 1:1 stoichiometry of sitagliptin DHP was not anticipated by Edmondson; (2) the conclusion that the '708 patent antedated Edmondson; and (3) the finding that Mylan did not prove claims 3 and 4 of the '708 patent were obvious over prior art references (Edmondson, Brittain, and Bastin). The legal determinations by the Board are reviewed de novo, while factual findings are reviewed for substantial evidence. Mylan argues that Edmondson inherently discloses sitagliptin DHP through its mention of sitagliptin and pharmaceutically acceptable salts, specifically claiming that a skilled artisan would envision a 1:1 stoichiometry based on Edmondson's disclosures and supporting experimental data. Mylan contends that the Board erred in ruling against these assertions.
Merck argues that the Board's conclusion that the claims are not anticipated by Edmondson is well-supported by substantial evidence. They assert that a skilled artisan would not be able to “at once envisage” all members of the genus of DP-IV-inhibitor salts disclosed in Edmondson. Merck elaborates that combining the 33 compounds and eight preferred salts could yield 957 possible salts, many of which may not form experimentally, thus failing to satisfy the “at once envisage” standard. Merck criticizes Mylan for improperly narrowing the focus to a specific compound, 1:1 sitagliptin DHP, rather than considering the entire disclosed class. Mylan's expert acknowledged that Edmondson does not specifically direct a skilled artisan to sitagliptin among the 33 DP-IV inhibitors or disclose any phosphate salt. The Board's decision is supported by this expert testimony, confirming no express disclosure of the 1:1 sitagliptin DHP salt in Edmondson.
Furthermore, the Board’s conclusion that there is no inherent disclosure of the 1:1 sitagliptin DHP salt by Edmondson is upheld. The key distinction lies in the understanding of what constitutes a "limited class," as illustrated in In re Petering, which allows a skilled artisan to envisage members of a limited class without clear boundaries. Merck argues that the extensive list of 957 salts does not meet the “at once envisage” criteria established in Petering, especially compared to the narrower class of 20 compounds previously considered. Mylan’s expert also indicated that salt formation is an unpredictable process requiring trial and error. Ultimately, the Board did not err in concluding that the large class of potential salts is insufficient to satisfy the anticipation criteria. Additionally, Merck addresses Mylan’s claims regarding the obviousness of certain claims, indicating that Mylan has not proven that these claims would have been obvious to a skilled artisan at the time of invention.
Mylan challenges the Board's finding regarding antedation related to Merck's patent for 1:1 (R)-sitagliptin DHP salt and the applicability of the 35 U.S.C. 103(c)(1) exception. Under 35 U.S.C. 102(a) (pre-AIA), an invention can be patented unless it was known or used before the applicant's invention. A party may overcome this barrier by demonstrating that it conceived the invention prior to the reference's effective date, with diligence in reducing it to practice. Mylan does not contest that Merck reduced the 1:1 (R)-sitagliptin DHP salt to practice before the publication of Edmondson, nor that both were commonly owned by Merck. However, Mylan argues that Merck's reduction to practice does not antedate Edmondson because it addresses sitagliptin hydrates, which Merck purportedly did not create until March 2003, following Edmondson's January 2003 publication. Mylan also contends that the Board incorrectly determined that Edmondson does not disclose hydrates of sitagliptin phosphate.
Merck refutes these claims, asserting that it reduced to practice the subject matter of claims 1, 2, 17, 19, and 21–23 of the '708 patent before Edmondson's publication, rendering Edmondson a 35 U.S.C. 102(e) reference rather than prior art. Given that both inventions were commonly owned at the time, the 103(c)(1) exception applies, preventing Edmondson from being an obviousness reference. The Board's determination that Merck’s work antedated Edmondson is supported by substantial evidence, including Merck's development timeline of the 1:1 sitagliptin DHP salt. Mylan did not contest claim 4's anticipation by Edmondson, further weakening its position. The Board also found substantial evidence supporting that Edmondson does not disclose 1:1 sitagliptin DHP or a hydrate thereof, affirming that Merck’s reduction to practice exceeded what was presented in Edmondson.
Additionally, the Board addressed Mylan's assertion that claims 3 and 4 of the '708 patent would have been obvious to a skilled artisan. Mylan specifically argues against the Board's decision regarding claim 3, which pertains to the (S)-configuration of 1:1 sitagliptin DHP, asserting that a skilled artisan could have derived it from the combination of Edmondson and Bastin.
Mylan contends that Bastin, which highlights the drawbacks of hydrochloric acid in pharmaceutical formulations, would prompt a skilled artisan to substitute hydrochloric acid in Edmondson's Example 7. Mylan asserts that sitagliptin's single asymmetric carbon allows for a reasonable expectation of success in synthesizing both (R)-sitagliptin and (S)-sitagliptin. Mylan challenges the Board's ruling that it failed to prove the obviousness of claim 4, which pertains to the crystalline monohydrate form of (R)-sitagliptin, arguing that a skilled artisan would have had a reasonable expectation of success in creating this form based on Edmondson and Brittain. Mylan notes that Edmondson mentions multiple crystal structures, including hydrates, while Brittain provides motivation to investigate hydrates in drug development.
Merck counters that the Board correctly found claim 3 non-obvious and that its factual findings were backed by substantial evidence. The Board determined that Bastin lacked specific motivation, alongside Edmondson, to arrive at 1:1 sitagliptin DHP or the (S)-configuration. Merck also contends that Mylan failed to convincingly motivate the creation of the crystalline monohydrate form, citing evidence that artisans would typically avoid hydrates due to solubility and stability issues. Additionally, Merck argues that the monohydrate exhibits unexpectedly beneficial properties, contributing to its non-obviousness.
The Board's conclusions regarding claims 3 and 4 were supported by substantial evidence, as it found no motivation to combine the cited references and no reasonable expectation of success. Testimony from Dr. Chorghade emphasized the unpredictability of hydrate formation, while Dr. Myerson supported the notion that skilled artisans would avoid hydrates due to their inherent challenges. The Board did not err in its assessment of Mylan's claims about non-obviousness, ultimately affirming its decision as consistent with substantial evidence and law.
