2025-1236

Appeal from the United States District Court for the District of Delaware in Nos. 1:22-cv-00228-RGA, 1:22-cv00945-RGA, Judge Richard G. Andrews.

Decided: August 31, 2026

THOMAS SAUNDERS, Wilmer Cutler Pickering Hale and Dorr LLP, Washington, DC, argued for plaintiff-appellee. Also represented by GERARD ANDREW SALVATORE, BELLA M. WALKER, AMY K. WIGMORE; MADELEINE C. LAUPHEIMER, LISA JON PIROZZOLO, KEVIN SCOTT PRUSSIA, Boston, MA.

CHARLES B. KLEIN, Winston Taylor LLP, Washington, DC, argued for defendants-appellants. Also represented by KEVIN BOYLE, BRYCE COOPER, KURT A. MATHAS, Chicago, IL; EIMERIC REIG-PLESSIS, San Francisco, CA.

Page 2Before MOORE, Chief Judge , STOLL, Circuit Judge , and MOORE, District Judge . 1

STOLL, Circuit Judge .

MSN Laboratories Private Limited and MSN Pharmaceuticals, Inc. appeal the decision of the United States District Court for the District of Delaware holding that the asserted claims of United States Patent Nos. 11,091,439, 11,091,440, 11,098,015, and 11,298,349, which are owned by Exelixis, Inc., are not invalid. For the reasons discussed below, we affirm the district court's finding that the asserted claims of the ’439, ’440, and ’015 patents have adequate written description pursuant to 35 U.S.C. § 112(a) and dismiss MSN’s appeal as to the asserted claim of the ’349 patent.

Exelixis holds the New Drug Application for Cabometyx®, a tablet containing the active pharmaceutical ingredient (API) cabozantinib (L)-malate, which is indicated to treat kidney, liver, and differentiated thyroid cancer. In the early 2000s, Exelixis first evaluated freebase cabozantinib, filing a patent application claiming cabozantinib or a pharmaceutically acceptable salt thereof. Exelixis then partnered with a research organization to perform a salt screen and identify salts of cabozantinib for commercial development. Salts can exist as either crystalline or amorphous material, with both of these forms having the same chemical name and formula for a specific salt.

Page 3In an amorphous salt, the molecules are randomly arranged; in contrast, a crystalline salt has a regular repeating array of molecules that give rise to the crystal structure. Crystalline salts may exist in multiple different crystalline forms, called polymorphs, where different polymorphs of the same compound have a different repeating arrangement of molecules.

Exelixis reported in a 2015 NDA submission to the Food and Drug Administration that “[c]abozantinib ([L])malate was found to exist in two neat, closely related, crystalline solid forms (N-1 and N-2) that have similar properties' and '[n]o other [crystalline] forms were identified.' J.A. 3053. Exelixis filed patent applications with claims to crystalline cabozantinib (L)-malate forms N-2 and N-1, which issued as United States Patent Nos. 8,877,776 and 9,809,549, respectively.

Exelixis later filed for the ’439, ’440, and ’015 patents (collectively, the 'Malate Salt Patents'), which share a specification with each other and with the ’776 and ’549 patents. The patents are titled 'Malate Salt of N-(4-{[6,7Bis(Methyloxy)Quinolin-4-Yl]Oxy}Phenyl)-N'-(4-Fluorophenyl)Cyclopropane-1,1-Dicarboxamide, and Crystalline Forms therof [sic] for the Treatment of Cancer.” U.S. Patent No. 11,091,439 Title.² The '439 patent claims crystalline cabozantinib (L)-malate salts; the ’440 patent claims pharmaceutical formulations of the salts; and the ’015 patent claims methods of treating cancer with the salts. Each patent is subject to a terminal disclaimer limiting the patent’s term to that of the ’776 patent, which expires in 2030.

The Malate Salt Patents’ specification describes cabozantinib malate and includes six examples describingPage 4 how to prepare both crystalline and amorphous cabozantinib malate. See id. at col. 18 l. 59-col. 24 l. 47. The specification discloses that “[t]he malate salts of [cabozantinib], and particularly Compound (1) [sic], have a preferred combination of pharmaceutical properties for development' and “exhibit beneficial properties over the free base and the other salts of [cabozantinib],” including stability at various temperatures and humidities, reversible water uptake, solubility, crystallinity, and moisture sensitivity. Id. at col. 7 ll. 10–13, ll. 32–36, Tbl. 1. The specification explains that '[a]nother aspect of this disclosure relates to crystalline forms” of cabozantinib (L)-malate and that “[e]ach . . . form of Compound (I) is a separate aspect of the disclosure.' Id. at col. 8 ll. 26–30. The specification goes on to describe two crystalline polymorphs of cabozantinib (L)-malate, N-1 and N-2. Id. at col. 9 l. 63-col. 11 l. 56.

Illustrative for the purposes of this appeal is dependent claim 4 of the ’439 patent:

1. 1. N-(4-{[6,7-bis(methyloxy)quinolin-4-yl]oxy}phenyl)-N'-(4-fluorophenyl) cyclopropane-1,1-dicarboxamide, malate salt, wherein said salt is crystalline.
2. 3. The N-(4-{[6,7-bis(methyloxy)quinolin-4-yl]oxy} phenyl)-N'-(4-fluorophenyl) cyclopropane-1,1-dicarboxamide, malate salt according to claim 1, wherein said salt is the (L)-malate salt or (D)-malate salt.
3. 4. The N-(4-{[6,7-bis(methyloxy)quinolin-4-yl]oxy} phenyl)-N'-(4-fluorophenyl) cyclopropane-1,1-dicarboxamide, malate salt according to claim 3, wherein said salt is the (L)-malate salt .

Id. at col. 32 ll. 22-24, ll. 29-36 (emphasis added).

Page 5The ’349 patent is titled “Processes for Preparing Quinoline Compounds and Pharmaceutical Compositions Containing Such Compounds” and discloses cabozantinib (L)malate compositions that are essentially free of certain impurities. U.S. Patent No. 11,298,349 Title, col.³ ll. 29–34. Synthesizing cabozantinib (L)-malate can result in a product with variable levels of the impurity 6,7-dimethoxyquinoline-4-ol (the '1-1 impurity'). The 1-1 impurity is genotoxic, meaning it can damage DNA and cause cancer.

Independent claim 3 of the ’349 patent was asserted in this case:

3. A pharmaceutical composition for oral administration comprising Compound IB;

IB

one or more fillers; one or more disintegrants; one or more glidants; and one or more lubricants, wherein the pharmaceutical composition is a tablet or capsule pharmaceutical composition; and

wherein the pharmaceutical composition is essentially free of 6,7-dimethoxy-quinoline-4-ol .

Id. at col. 34 ll. 30-51 (emphasis added). The specification defines “essentially free” as 200 ppm or less of the 1-1 impurity. Id. at col. 8 ll. 15-19.

Page 6In August 2019, MSN submitted Abbreviated New Drug Application No. 213878 seeking FDA approval for generic cabozantinib (L)-malate tablets. MSN used form S of cabozantinib (L)-malate, on which it received its own patent.

Exelixis brought two lawsuits, which were consolidated, alleging MSN infringed certain claims of the Malate Salt Patents and the ’349 patent. For the Malate Salt Patents, MSN conceded infringement but, as relevant to this appeal, argued the asserted claims were invalid for lack of written description under 35 U.S.C. § 112(a). For the ’349 patent, MSN contested both infringement and validity as to claim 3. The district court held a bench trial.

In its post-trial order, the district court found written description support for the asserted claims of the Malate Salt Patents. Starting with the legal framework, the court explained that a written description can adequately show that an inventor was in possession of a claimed genus in one of two ways: by disclosing either (1) a representative number of species falling within the scope of the genus, or (2) structural features common to the members of the genus so that one of skill in the art can visualize or recognize the members of the genus. Exelixis, Inc. v. MSN Lab'ys Priv. Ltd. , No. 22-228-RGA, 2024 WL 4491176, at *12 (D. Del. Oct. 15, 2024) (citing Ariad Pharms., Inc. v. Eli Lilly & Co. , 598 F.3d 1336, 1350 (Fed. Cir. 2010)). Applying this standard, the district court found that the specification disclosed structural features common to the members of the genus—crystalline cabozantinib (L)-malate salts—and analogized this case to GlaxoSmithKline LLC v. Banner Pharmacaps, Inc. ( GSK ), 744 F.3d 725 (Fed. Cir. 2014). The district court found that “the key feature of the genus is the chemical formula and structure of crystalline cabozantinib (L)-malate[, as a]ll crystalline cabozantinibPage 7 malate share the same chemical name and formula.” Exelixis , 2024 WL 4491176, at *13 (citation omitted). Furthermore, a skilled artisan 'would be able to identify whether the structure of a polymorph is crystalline” and 'could distinguish between crystalline and amorphous cabozantinib.' Id. (citation omitted). The district court also noted that the specification discloses processes used to make the invention. The district court found these facts sufficient to support adequate written description of crystalline cabozantinib (L)-malate salts.

The district court also addressed MSN's arguments that the disclosed N-1 and N-2 polymorphs have different crystal structures and physico-chemical properties than other polymorphs and thus cannot support written description for all crystalline cabozantinib (L)-malate salts. The district court reasoned that MSN did not explain why these differences meant the N-1 and N-2 polymorphs are materially different from the other polymorphs falling within the genus, thus distinguishing this case from AbbVie Deutschland GmbH & Co., KG v. Janssen Biotech, Inc. , 759 F.3d 1285 (Fed. Cir. 2014).

As to the '349 patent, the district court found that MSN's ANDA product did not infringe asserted claim 3. The district court also held that MSN failed to prove claim 3 was invalid. Relevant to the arguments made in this appeal, the district court determined that a prior art process disclosed in Brown 3 did not inherently disclose a cabozantinib (L)-malate API 'essentially free' of the 11 impurity, where the “essentially free” limitation was the only missing limitation from the prior art. Specifically, the district court determined that MSN did not meet its burden of proving inherency using either experimental results or expert testimony. The district court found that the threePage 8 batches of the cabozantinib (L)-malate API that MSN pointed to as experimental results showing the synthesis process formed less than 200 ppm of the 1-1 impurity did not clearly follow the prior art Brown process.

Accordingly, the district court entered final judgment against MSN as to the Malate Salt Patents, which the district court held infringed and not invalid, while it held the ’349 patent not infringed and not invalid. MSN filed a notice of appeal challenging the district court’s findings that claim 4 of the '439 patent, claim 3 of the '440 patent, claim 2 of the ’015 patent, and claim 3 of the ’349 patent are not invalid. See ECF No. 1 at 5. Exelixis initially filed a notice of cross-appeal indicating its intent to challenge the district court’s noninfringement finding as to claim 3 of the ’349 patent. See Notice of Cross-Appeal, Exelixis, Inc. v. MSN Lab'ys Priv. Ltd. , No. 25-1241 (Fed. Cir. Dec. 3, 2024), ECF No. 1 at 5. Before filing its first brief in this matter, however, Exelixis dismissed its cross-appeal, see ECF No. 26, rendering the noninfringement judgment as to claim 3 of the ’349 patent final. MSN nonetheless continued to maintain its appeal challenging the district court’s decision on the validity of claim 3 of the ’349 patent in its reply brief. See Appellants' Reply Br. 6 n.2. We issued an order instructing the parties to “come to [oral] argument prepared to discuss whether Appellants have Article III standing to appeal the final judgment of no invalidity for claim 3” of the ’349 patent. ECF No. 61 at 2. Following the order, MSN filed a motion to dismiss its appeal as moot and to vacate the underlying decision as to claim 3 of the ’349 patent. See ECF No. 63. Exelixis opposed the motion. See ECF No. 64.

On appeal, MSN asserts that the district court erred in finding written description support under 35 U.S.C. § 112(a) for claim 4 of the '439 patent, claim 3 of the '440 patent, and claim 2 of the '015 patent. While MSNPage 9 also initially challenged the district court’s finding of no inherency for claim 3 of the ’349 patent, it now urges us to dismiss this portion of its appeal as moot and vacate the underlying district court order as to claim 3. We address each issue in turn.

The written description requirement ensures that patentees adequately describe their inventions in exchange for the right to exclude others from practicing the claimed invention for the patent’s term. The test for adequate written description “requires an objective inquiry into the four corners of the specification from the perspective of a person of ordinary skill in the art.' Ariad , 598 F.3d at 1351. 'Based on that inquiry, the specification must . . . show that the inventor actually invented the invention claimed.” Id. '[T]he level of detail required to satisfy the written description requirement varies depending on the nature and scope of the claims and on the complexity and predictability of the relevant technology.” Id.

“Adequacy of the written description is a question of fact.” GSK , 744 F.3d at 729 (citing Ariad , 598 F.3d at 1351). “After a bench trial, we review the district court’s findings of fact for clear error.' Id. (citing Pozen Inc. v. Par Pharm., Inc. , 696 F.3d 1151, 1166 (Fed. Cir. 2012)).

The district court used the legal framework set forth in Ariad and later applied in GSK to analyze the adequacy of the written description of the claimed genus. In Ariad , we explained that a “sufficient” written description for disclosure of a claimed genus “requires the disclosure of either a representative number of species falling within the scope of the genus or structural features common to the members of the genus so that one of skill in the art can ‘visualize or recognize' the members of the genus.' Ariad , 598 F.3d at 1350 (emphasis added) (citation omitted). We further “explained that an adequate written description requires a precise definition, such as by structure, formula, chemical Page 10name, physical properties, or other properties, of species falling within the genus sufficient to distinguish the genus from other materials.” Id. (citation omitted). MSN does not dispute that this is the correct legal framework for this case. See Appellants’ Br. 37–38, 44; Oral Arg. at 1:05–3:43, 30:30–31:17, https://www.cafc.uscourts.gov/oral-arguments/25-1236_06042026.mp3.

Rather, MSN contends that the district court legally erred by crediting allegedly cursory parts of the specification that fail to disclose structural features distinguishing the genus of crystalline cabozantinib (L)-malate salt such that a skilled artisan could visualize its members. We are not persuaded.

Specifically, we see no clear error in the district court’s finding that disclosing the chemical name and formula of cabozantinib (L)-malate salt, as well as that the structure of the salt is crystalline, is an identification of the structural features possessed by members of the genus. See Exelixis , 2024 WL 4491176, at *12-13; see also '439 patent Abstract, col.¹ ll. 26–39, col.² l. 58–col.³ l. 12, col.³ ll. 36–46, col. 5 l. 25–col. 6 l. 67, col. 8 ll. 26–28. The claims are no broader than the written description, as the claims require cabozantinib (L)-malate salt with a crystalline structure. Furthermore, while not dispositive of satisfaction of the written description requirement, the specification also discloses processes used to make the invention. Exelixis , 2024 WL 4491176, at *12; see also '439 patent col. 17 l. 10–col. 23 l. 60.

MSN does not dispute the accuracy of the disclosure relied on by the district court, merely its sufficiency. See Appellants’ Br. 39–44. But multiple of the factors laid out in Ariad are met here (i.e., structure, formula, chemical name), and the district court’s findings are not clearly erroneous. Moreover, the district court appropriately analogized this case to GSK , where we noted that '[d]escribing aPage 11 complex of dutasteride and solvent molecules is an identification of 'structural features commonly possessed by members of the genus that distinguish them from others,’ allowing one of skill in the art to ‘visualize or recognize the identity of the members of the genus.'' GSK , 744 F.3d at 730. And the claims here likewise have “no performance property (the claimed compound need not perform an identified function or produce an identified result) and hence raise[] no issue of insufficient structural, creation-process, or other descriptions to support such a property.' Id. at 729–30.

We are also unpersuaded by MSN’s reliance on the purported “different densities, melting points, solubilities, hygroscopicity, vapor pressure, and stability” of the disclosed N-1 and N-2 polymorphs. Appellants’ Br. 54 (citation omitted); see also Appellants' Br. 54-56. According to MSN, these properties shared by N-1 and N-2 might not be shared by other species of the claimed genus and thus a person of ordinary skill in the art would not think that the inventor invented other species. But MSN fails to explain, in the context of this particular invention, why these differences-which, by the way, are unclaimed and merely listed in general terms—show that the district court clearly erred in its fact finding. According to MSN, the properties of N-1 and N-2 polymorphs “cannot be used to predict the properties of a different form.” Appellants’ Br. 54 (citation omitted). Even were this so, the district court did not rely on the properties of the N-1 and N-2 polymorphs to identify other polymorphs. Rather, the district court reasonably relied on the specification’s disclosure of the chemical name and formula of cabozantinib (L)-malate salt and that the structure is crystalline as an identification of the structural features possessed by members of the genus.

As such, we agree with the district court that MSN has not made clear what the relevancy of these properties are in the structural analysis done here, making this case distinguishable from cases like AbbVie . See Exelixis , 2024 WLPage 12 4491176, at *14; AbbVie , 759 F.3d at 1300 (explaining that the fact finder heard specific evidence that the patents only described one type of structurally similar antibodies in an entire genus, and the accused antibodies “only share a 50% sequence similarity with the [disclosed] antibodies, which is far lower than the 90% sequence similarity shared among [those] antibodies described in AbbVie’s patents”). Additionally, the district court found, and MSN does not challenge, that here “[t]he maximum potential size of any pure polymorph genus is fourteen forms.' See Exelixis , 2024 WL 4491176, at *11 (citation omitted). Accordingly, this case does not seem to implicate the same concerns that some genus-claim cases with potentially vast numbers of species do. Nor did the district court need to analyze whether there were a representative number of species here, see Ariad , 598 F.3d at 1350 (“We held that a sufficient description of a genus instead requires the disclosure of either a representative number of species falling within the scope of the genus or structural features common to the members of the genus so that one of skill in the art can ‘visualize or recognize’ the members of the genus.” (emphasis added) (citation omitted)), so there is no need for us to reach this inquiry or vacate the decision for the district court to consider this issue in the first instance.

MSN also contends that the district court erred by “assum[ing] there are ‘more rigorous requirements for written description in support of functional claim language,’ which the court believed 'require more disclosure to meet the written description requirement.’” Appellants’ Br. 44 (emphasis removed) (quoting Exelixis , 2024 WL 4491176, at *12, *14). However, reading the district court’s order in full makes clear that the court was not using a lower standard for the structural claims than it would for functional ones. Instead, the district court was merely recognizing, as we have also recognized, that when there is functional claiming, supplying adequate written description can be more challenging. See Ariad , 598 F.3d at 1349 (explainingPage 13 that, when “a generic claim . . . define[s] the boundaries of a vast genus of chemical compounds, . . . the question may still remain whether the specification[] . . . demonstrates that the applicant has invented species sufficient to support a claim to a genus' and this 'problem is especially acute with genus claims that use functional language to define the boundaries of a claimed genus.”). We see no error in the legal standard applied by the district court to the claims here.

The remaining cases that MSN relies on are distinguishable. MSN cites ICU Medical, Inc. v. Alaris Medical Systems, Inc. , 558 F.3d 1368 (Fed. Cir. 2009), Tronzo v. Biomet, Inc. , 156 F.3d 1154 (Fed. Cir. 1998), and Eli Lilly & Co. v. Teva Pharmaceuticals USA, Inc. , 619 F.3d 1329 (Fed. Cir. 2010), for the proposition that genus claims defined by structural limitations can be invalidated for lack of written description. See Appellants' Br. 52. But this point of law is not in dispute. Under the facts of this case, however, we have determined that the district court applied the correct legal standard for written description for the asserted structural claims and did not err in finding written description support for these claims. Moreover, these cases are not analogous to the claims and specification at issue in this case. In ICU Medical , the patentee claimed spikeless medical valves when the written description only disclosed valves with spikes, arguing the spikeless claims could cover both spiked and spikeless medical valves. See 558 F.3d at 1376-79. We affirmed the district court's summary judgment of invalidity based on lack of written description, explaining that the spikeless claims lacked written description support because “figures and descriptions that include[d] spikes” could not “demonstrate that the inventor possessed a medical valve that operated without a spike” and the patentee had “failed to point to any disclosure in the patent specification that describe[d] a spikeless valve.' Id. at 1378-79. In Tronzo , the claims Page 14were directed to artificial hip sockets that included cup implants, where “the specification specifically distinguishe[d] the prior art as inferior and tout[ed] the advantages of [a] conical shape of the [claimed] cup.” 156 F.3d at 1159. We held that “[s]uch statements make clear that the [asserted] patent disclose[d] only conical shaped cups and nothing broader.' Id. at 1159-60 (emphasis removed) (holding that the specification of the asserted patent failed to provide the written description support for claims that were generic as to the shape of the cup). As to Eli Lilly , there the claims 'covered particle sizes before and after formulation into tablets, but the specification addressed only pre-formulation size,” and thus there was no written description support for particle size after formulation. GSK , 744 F.3d at 731 (citing Eli Lilly , 619 F.3d at 1344–45). The absence of disclosures over the claimed subject matter in each of those cases is not analogous to this case, where what is claimed is no broader than what is in the written description—i.e., a crystalline cabozantinib (L)-malate salt.

We turn next to MSN’s initial challenge to the district court’s finding of no inherent obviousness for claim 3 of the ’349 patent and its subsequent motion to dismiss its appeal as moot and vacate the district court’s judgment as to this claim. We agree with MSN that this portion of its appeal is now moot.

Under the Supreme Court's decision in Cardinal Chemical Co. v. Morton International, Inc. , a declaratory judgment action challenging validity does not automatically become moot upon a determination of noninfringement. 508 U.S. 83, 95-98 (1993). Rather, 'courts are entitled to presume, absent further information, that jurisdiction continues.” Id. at 98. Nonetheless, if “either party ha[s] advised [the court] of a material change in circumstances that entirely terminated the party’s controversy, it would [be] proper either to dismiss the appeal or to vacate Page 15the entire judgment of the District Court.' Id. That is the situation here.

MSN concedes that the case has become moot. See ECF No. 63 at 8–9 (“MSN’s appeal of the district court’s final judgment of no invalidity for claim 3 of the ’349 patent is moot . . . [and] should be dismissed because there is no case or controversy sufficient to support jurisdiction under Article III.”). Exelixis, on the other hand, argues that MSN has a continued redressable injury not based on the existence of the ’349 patent itself, but rather by asserting that our holding on the inherency issue could potentially have collateral consequences in another case between Exelixis, MSN, and other defendants, which involves a patent related to the '349 patent.⁴ See ECF No. 64 at 8-10. We are not convinced. These purported collateral consequences are too speculative and hypothetical to confer standing. See Best Med. , 46 F.4th at 1353 (“[T]he potential for collateral consequences is insufficient, on its own, to confer standing.” (quotation marks and citation omitted)).

Having determined that MSN’s appeal as to claim 3 is moot, the appropriate resolution in this case is vacatur of the district court's judgment on this claim. See United States v. Munsingwear, Inc. , 340 U.S. 36, 39-40 (1950); U.S. Bancorp Mortg. Co. v. Bonner Mall P’ship , 513 U.S. 18, 24 (1994). The Supreme Court has explained that “[v]acatur is in order when mootness occurs through happenstance—circumstances not attributable to the parties—Page 16 or, relevant here, the 'unilateral action of the party who prevailed in the lower court.'' Arizonans for Off. Eng. v. Arizona , 520 U.S. 43, 71-72 (1997) (quoting Bancorp , 513 U.S. at 23). “A party who seeks review of the merits of an adverse ruling[] but is frustrated by the vagaries of circumstance” or the “unilateral action” of the appellee “ought not in fairness be forced to acquiesce in the judgment.' Bancorp , 513 U.S. at 25. Vacatur avoids this unfairness. It “‘clears the path for future relitigation’ by eliminating a judgment the loser was stopped from opposing on direct review.” Arizonans , 520 U.S. at 71 (quoting Munsingwear , 340 U.S. at 40). Accordingly, because Exelixis’s unilateral decision to drop its cross-appeal frustrated MSN’s attempt to seek review on the merits of the district court’s no inherency finding, we vacate the judgment as to claim 3. We thus grant MSN’s motion to dismiss its appeal as to claim 3 of the ’349 patent and vacate the district court’s judgment of nonobviousness of claim 3.

We have considered the parties’ remaining arguments and find them unpersuasive. We affirm the district court’s decision finding written description support pursuant to 35 U.S.C. § 112(a) for the asserted claims of the ’439, ’440, and ’015 patents. We also grant MSN’s motion to dismiss its appeal challenging the no inherency finding for claim 3 of the ’349 patent and vacate the district court’s judgment of nonobviousness of claim 3.

COSTS

No costs.