IP Law: In Defense Of Polymorphs

By James T. Masters & Eric J. Fues

Crystalline forms of compounds and salts are frequently claimed in chemical patents, including in pharmaceuticals and agrochemicals. Different crystalline forms of the same substance — that is, different polymorphs — can have distinct properties, with one polymorph often being the most desirable for commercialization. As a result, patent claims to a particular polymorph can be a critical component of an innovator’s patent portfolio.

The Federal Circuit’s recent decision in Salix v. Norwich attracted attention for its conclusion that claims to a polymorph would have been obvious. But Salix dealt with a unique fact pattern, one where there was “no dispute” that the claimed polymorph could have been readily produced from crystallization conditions disclosed in the prior art. 98 F. 4th 1056, 1066 (Fed. Cir. 2024). In contrast, considerable Federal Circuit caselaw supports the non-obviousness of claims to polymorphs, even in instances where the prior art describes allegedly relevant compounds.

Does Salix Represent A Change In The Law?

At issue in Salix was whether claims to the b polymorph of rifaximin would have been obvious in view of prior art disclosure of several preparations of rifaximin. Salix, 98 F. 4th at 1059, 1064. The Salix Court credited expert testimony that those preparations would have yielded the b polymorph, and it also found it relevant that methods for characterizing the crystalline rifaximin resulting from the prior art preparations were “well known and readily available.” Id. at 1066. Against this backdrop, the court reasoned that the difference between the claimed subject matter and the prior art was “nothing more than the performance of routine characterization to identify” the material resulting from the known processes. Id.

However, the Salix Court did not articulate a fundamental change in the law; to the contrary, the court stated it did not hold there is always a reasonable expectation of success in accessing or characterizing polymorphs. Id. Rather, the court reiterated that the obviousness determination “is dependent on the facts of each case.” Id. at 1065 (citations omitted). The Salix Court emphasized the factual content of the prior art when finding no clear error in the district court’s obviousness conclusion, stating that Salix had only combined “known elements” to “verify readily accessible information.” Id. at 1066.

Salix Is Consistent With Other Polymorph Cases Supporting Non-Obviousness

Although the Salix Court’s determination that polymorph claims would have been obvious was unique, its emphasis on the facts underlying individual cases was not. Indeed, Salix’s close factual assessment is consistent with the analyses undertaken in prior Federal Circuit decisions finding polymorph claims non-obvious, including Grünenthal v. Alkem, Pharmacyclics v. Alvogen, and Mylan v. Merck.

In Grünenthal, the Federal Circuit affirmed a judgment that claims to crystalline Form A of tapentadol hydrochloride would not have been considered obvious over prior disclosure of a different form of that salt and general guidance on crystallization. 919 F.3d 1333, 1336-37, 1344 (Fed. Cir. 2019). One question before the court was whether a person having ordinary skill in the art (POSA) would have had a reasonable expectation of producing crystalline Form A by applying general experimental methods from a technical paper, Byrn, to known crystalline Form B. Id. at 1341-43. The court decided not.

The court acknowledged that Byrn identified variables that could be adjusted to determine if a compound exhibits polymorphism, meaning the ability to exist in multiple crystalline forms. Id. at 1337, n.5, 1341. But the court credited evidence on unpredictability, including evidence that a person having ordinary skill in the art “could alter any number of variables and still fail to find a polymorph of a particular compound.” Id. at 1342. In concluding that a POSA would not have reasonably expected polymorph screening to yield Form A, the court emphasized, among other things, the lack of guidance in the factual record regarding what conditions were likely to produce Form A. Id. at 1343.

In its non-precedential Pharmacyclics decision, the Federal Circuit affirmed a determination that a claim to crystalline Form A of ibrutinib was not obvious over references providing the chemical structure of ibrutinib and references that generally address polymorphism. 2022 WL 16943006 at *1, *5-6, *10-11 (Nov. 15, 2022). The prior art disclosed ibrutinib among other compounds, provided a method for synthesizing ibrutinib, and stated the disclosed compounds “may be in various forms,” but it did not assert that any crystalline form(s) of ibrutinib actually existed. Id. at *5. The Pharmacyclics district court concluded that a POSA would have been motivated to develop “a crystalline form of ibrutinib,” but it reasoned there would not have been motivation or a reasonable expectation of success in developing ibrutinib Form A specifically. Id. at *6 (emphasis in original).

The Federal Circuit found no clear error in that determination. Id. at *10-11. In doing so, the court emphasized expert testimony that a POSA would not have been able to predict either whether a new compound would form polymorphs or what physical properties a crystalline form will have. Id. at *10. The Pharmacyclics Court also stressed the lack of teaching in the prior art regarding crystalline forms of ibrutinib as a fact supporting the conclusion that there would not have been a reasonable expectation of success in producing ibrutinib Form A.

The Salix decision addressed the “lack of clear error” determinations in both Grünenthal and Pharmacyclics. Salix, 98 F. 4th at 1065. Reiterating that an obviousness analysis “depen[ds] on the facts of each case,” the Salix Court contrasted the “distinct factual predicates” of Grünenthal and Pharmacyclics against those of Salix. Id. In the Salix Court’s view, the issue in both Grünenthal and Pharmacyclics was whether a POSA would have had a reasonable expectation of success “in producing a crystalline form,” whereas the dispute in Salix related to merely “characterizing [a] crystalline form” resulting from a known process. Id. (emphasis in original).

Lastly, the Federal Circuit’s Mylan decision also addressed whether a POSA would have had a reasonable expectation “in making [a] crystalline form.” 50 F.4th 147, 152 (Fed. Cir. 2022). Mylan appealed a Patent Trial and Appeal Board decision finding a claim to the crystalline monohydrate salt of (R)-sitagliptin to be non-obvious. Id. at 150. Prior art references disclosed sitagliptin, identified phosphoric acid as an acid from which a salt could be formed, disclosed that salts “may exist” in crystalline forms (including hydrates), and discussed the prevalence of crystalline hydrates. Id. at 150-151.

The Federal Circuit rejected Mylan’s obviousness argument on appeal. As in Grünenthal and Pharmacyclics, the Mylan Court stressed unpredictability, emphasizing expert testimony that a POSA “couldn’t predict” hydrate formation “with any degree of certainty.” Id. at 156. In addition, the court highlighted the “numerous downsides of hydrates reported in the literature” and expert testimony that “a skilled artisan would have sought to avoid hydrates.” Id.

Looking Forward, Patent Owners Should Stress The Traditional Factors Supporting Non-Obviousness

With Salix having reiterated the factual nature of the obviousness inquiry, patent owners facing obviousness challenges should emphasize the facts that were considered pertinent to a finding of non-obviousness in Grünenthal, Pharmacyclics, and Mylan. Those include the lack of knowledge that a compound is polymorphic, the lack of guidance regarding how to prepare a particular polymorph, the unpredictability associated with polymorph formation, and any disclosures suggesting the claimed polymorph would have been undesirable. By focusing on those well-established considerations, patent owners may be able to align with the “factual predicates” that the Salix Court recognized supported the non-obviousness of polymorph claims. Salix, 98 F. 4th at 1065.

About The Authors:

Eric Fues is a partner at Finnegan, Henderson, Farabow, Garrett & Dunner LLP’s Washington, D.C., office, and has significant experience with pharmaceutical patent litigation under the Hatch-Waxman Act, and in inter partes review proceedings before the Patent Trial and Appeal Board at the U.S. Patent and Trademark Office. While his cases involve a wide array of technologies, he focuses primarily on the fields of pharmaceuticals, agricultural chemicals, polymers, semiconductor devices and processing, and light-emitting diodes (LEDs). Eric has served as president of the ITC Trial Lawyers Association and as leader of Finnegan’s chemical and metallurgical practice group.

James Masters is an associate at Finnegan, Henderson, Farabow, Garrett & Dunner LLP’s Washington, D.C. office, and has an expansive intellectual property practice, covering patent prosecution, strategic client counseling, pre-litigation diligence, and patent litigation in the biotechnology, chemical, pharmaceutical, and medical device fields. James has considerable experience in a range of technological areas, including small molecule chemical compounds, polymorphs, antibodies and antibody-drug conjugates, polymer chemistry, medical devices, agrochemicals, and food and beverage products.