Since it has been a while since my last post, I’m going to do something a little different and do a quick run down of ten Federal Circuit cases that have issued in the past few months and highlight some key takeaways. I’ll give a quick bullet point review of the cases, with slightly more detailed explanation below.
- Metacel Pharms. LLC v. Rubicon Research Private Ltd., No. 2023-2386, 2025 WL 1178384 (Fed. Cir. Apr. 23, 2025) – ANDA label suggesting storage at room temperature but permitting refrigerated storage did not induce infringement of claims to refrigerated storage
- Azurity Pharms., Inc. v. Alkem Lab’ys Ltd., No. 2023-1977, 2025 WL 1036994 (Fed. Cir. Apr. 8, 2025) – Brand’s disclaimer of propylene glycol by addition of “consisting of” preamble in parent patent precluded finding of infringement
- Janssen Pharms., Inc. v. Mylan Lab’ys Ltd., No. 23-2042, 2025 WL 946390 (Fed. Cir. Mar. 28, 2025) – Label describing reinitiation protocol for schizophrenia drug induced infringement of claims to reinitiation process
- Actavis Lab’ys FL, Inc. v. United States, 131 F.4th 1345, 1353 (Fed. Cir. 2025) – Generic Hatch-Waxman litigation expenses deductible as ordinary business expenses
- Regeneron Pharms., Inc. v. Mylan Pharms. Inc., 130 F.4th 1372 (Fed. Cir. 2025) – Claims reciting separate elements VEGF antagonist and buffer not likely to be infringed by product with self-buffering VEGF antagonist because of Becton presumption that separate elements are distinct components and intrinsic record supported, rather than rebutted, that presumption
- Merck Sharp & Dohme B.V. v. Aurobindo Pharma USA, Inc., 130 F.4th 1363 (Fed. Cir. 2025) – PTE statute applies to reissued patent that retained original claims based on regulatory delays following issue date of original patent
- In re Xencor, Inc., 130 F.4th 1350 (Fed. Cir. 2025) – Claims to method of treating a patient with an antibody not patentable for lack of written description support for treating a patient portion of preamble
- In re Strongbridge Dublin Ltd., No. 2023-2302, 2025 WL 751116 (Fed. Cir. Mar. 10, 2025) – Claims to a method of administering a drug without concomitant administration of other drugs did not require actively discouraging concomitant usage, but finding mere silence in prior art clinical report not substantial evidence of lack of concomitant use and remanding for further consideration
- CQV Co. v. Merck Pat. GmbH, 130 F.4th 1344 (Fed. Cir. 2025) – Remanding where Board did not fully address evidence of prior art’s public availability
- ImmunoGen, Inc. v. Stewart, 130 F.4th 1328 (Fed. Cir. 2025) – Affirming obviousness finding for claims to dosing regimen based on Adjusted Ideal Body Weight because ocular toxicity was a known problem for similar drugs, AIBW dosing regimens were known for avoiding ocular toxicity, and prior art taught dosing in similar amounts based on Total Body Weight
In Metacel Pharms. LLC v. Rubicon Research Private Ltd., No. 2023-2386, 2025 WL 1178384 (Fed. Cir. Apr. 23, 2025) the Federal Circuit affirmed a grant of summary judgment of non-infringement where the claims required storage at “at from about 2 to about 8°C” and the accused ANDA label instructed storage at room temperature and stated only that the product can be refrigerated and “can also be stored at 2°C to 8°C.”
The Federal Circuit described this type of label as an “if/then” statement, i.e., “if a downstream user decides to refrigerate the product, despite instructions to store the product at room temperature (which is noninfringing), then it should store the product at temperatures from 2°C to 8°C” and held “[t]hat is not inducement.” The panel then held that “where, as here, a label is unambiguous, circumstantial evidence cannot override its plain language.” It nevertheless considered the circumstantial evidence but found it too did not create a genuine dispute of material fact. First, it held “Rubicon’s FDA submissions are not available to downstream users and therefore cannot cause inducement.” It then considered expert testimony that a doctor would turn to the brand label which would lead to infringement, but rejected that argument “because it relies on downstream users turning to Metacel’s own Ozobax label as evidence of Rubicon’s specific intent to induce” which was “not persuasive.” The case reaffirmed that the induced infringement standard for ANDAs focuses first and foremost on the label and that the mere probability that users might infringe is not sufficient, absent evidence of specific intent from the ANDA holder.
In Azurity Pharms., Inc. v. Alkem Lab’ys Ltd., No. 2023-1977, 2025 WL 1036994 (Fed. Cir. Apr. 8, 2025) the court affirmed a finding following a bench trial that Alkem’s generic product could not infringe the patents because Azurity had disclaimed any use of propylene glycol during the prosecution.
During prosecution of a parent patent, Azurity attempted to distinguish a prior art reference using negative limitations excluding propylene glycol as a carrier and eventually obtained claims using the transitional phrase “consisting of” and distinguishing the claims from that prior art reference arguing “the absence of propylene glycol in the claimed invention, in part, distinguish it from [the prior art]” which the Examiner cited as a basis for allowance. During the later application that became the asserted patent, and following receipt of Alkem’s Paragraph IV letter, the applicants stated “For the record, Applicant did not disclaim propylene glycol when submitting the arguments in U.S. 15/126059, and reserves the right to claim propylene glycol in the instant and future cases in this patent family.” However, the claims still used the “consisting of” preamble and did not list propylene glycol.
Azurity argued that propylene glycol could be the claimed “flavoring agent” and thus was not excluded from the claims due to the closed “consisting of” language. The panel was not convinced, noting that the claims of the asserted patent were nearly identical to the claims of the parent patent, which Azurity had repeatedly argued excluded propylene glycol, and giving little attention to Azurity’s attempt to walk back the disclaimer. The panel noted that while Azurity had tried to narrowly exclude propylene glycol as a carrier, it had relented and used the closed “consisting of” language. Thus, the panel held that even if in the prior art propylene glycol functioned only as a carrier, “what matters most is the broad language that Azurity used to distinguish Palepu” and “[j]ust as the echo matches the shout, Azurity’s repeated, sweeping statements—endorsed by the examiner—return an equally sweeping disclaimer.”
Finally, the panel agreed with the district court that a stipulated finding of fact that “Suitable flavoring agents include for use in the Asserted Claims include flavoring agents with or without propylene glycol.” The panel noted that the section began with a statement that Alkem “contends that it does not infringe the Asserted Claims due to the presence of propylene glycol in Alkem’s ANDA Products” and it would be nonsensical for Alkem to have framed the infringement dispute that way, but then stipulated to infringement. Among other factors, the panel agreed that in context, the “district court correctly concluded that the disputed stipulation did not preclude application of disclaimer in this case.”
The case is a good reminder to prosecutors that broad disclaimers in a parent application can limit later claims and that later attempts to undo a disclaimer are unlikely to be effective. And while Alkem prevailed, the case is also a good reminder to be careful with pretrial statements of undisputed facts.
In Janssen Pharms., Inc. v. Mylan Lab’ys Ltd., No. 23-2042, 2025 WL 946390 (Fed. Cir. Mar. 28, 2025), the court affirmed a DNJ decision finding Janssen had proven induced infringement of claims to treating patients with paliperidone palmitate (PP) who had missed a treatment and that Mylan had not demonstrated invalidity by clear and convincing evidence.
PP, used to treat schizophrenia, comes in two forms, a form that lasts one month (PP1M) and a form that lasts three (PP3M). Janssen markets Invega Trinza, a PP3M. The asserted patent covers methods of reintroducing PP3M to patients who have missed a dose, by first administering PP1M. The label specifically instructs that if a patient has missed a dose, i.e., had his or her last dose between 4 and 9 months ago, “do NOT administer the next dose…[i]nstead, use the re-initiation regimen” recited in the claims.
Mylan presented three non-infringement arguments: (1) Mylan cannot induce infringement because its label discourages missing doses in the first place; (2) Janssen failed to show infringement would inevitably occur; and (3) there was a divided infringement problem. The panel did not find any argument persuasive. On the first issue, the panel agreed with the district court that “the fact that Mylan’s proposed ANDA labels ‘discourage missed doses’ does not mean that the labels ‘discourage or make optional the practice of the asserted claims (or any claimed steps) in the inevitable situation that doses are missed.” On the second issue the panel found sufficient evidence that some users would inevitably miss doses and be reinitiation, including the testimony of Mylan’s own expert. On the third issue, the panel agreed with the district court that “Mylan’s divided infringement theory was not disclosed in its contentions, and appeared improperly for the first time in Mylan's rebuttal expert report” and that in view of the “deferential review standard” the district court did not abuse its discretion in not considering it.
Regarding invalidity, the panel found no clear error in the district court’s findings that “nothing in the prior art motivated a skilled artisan to use PP1M after a patient has been advanced to PP3M” and that even considering the prior art teaching of starting on PPM1M, there would have been no apparent reason to transition within the claimed 23rd-27th day where the art taught to stabilize a patient on PP1M for 17 weeks.
What’s interesting to note with this decision is that it effectively allows Janssen to block an undisputed non-infringing use (regular treatment with PP3M) based on the reinitiation language of the label. Of course substantial non-infringing uses is relevant to contributory infringement, not induced infringement. It is unclear from the record whether Mylan could have carved out the reinitiation language, but it should be a warning to generics that induced infringement can be found even for instructions that do not go to the primary use.
In Actavis Lab’ys FL, Inc. v. United States, 131 F.4th 1345, 1353 (Fed. Cir. 2025) the Federal Circuit agreed “with the trial court’s conclusion that Actavis may deduct its Hatch-Waxman litigation expenses as ordinary and necessary business expenses.” This was hardly a surprise, given the Third Circuit’s similar holding in Mylan Inc. v. Comm’r of Internal Revenue, 76 F.4th 230, 233-38 (3d Cir. 2023), but it’s an interesting read with a concise background of the Hatch-Waxman process. While the nuances of tax law are probably not our reader’s primary interests, the decision was interesting in that it rejected an argument that ANDA litigation was an expense occurred in obtaining capital, holding “the origin of the claim in Hatch-Waxman litigation is a patent claim brought by the NDA holder, and not the pursuit of effective FDA approval of an ANDA sought by the ANDA filer” because “the issues in the litigation and the issues in the regulatory approval process are different, and they are resolved by different decision-makers.” It similarly noted the “intangible asset sought by the ANDA filer is final, effective approval of the ANDA itself – and acquisition of that asset is not facilitated by Hatch-Waxman litigation.”
In Regeneron Pharms., Inc. v. Mylan Pharms. Inc., 130 F.4th 1372 (Fed. Cir. 2025) the court affirmed a WDVA court’s denial of Regeneron’s motion for preliminary injunction, finding the trial court correctly held that the separate listing of “VEGF antagonist” and “buffer” required two separate components and could not read on a self-buffering antagonist.
I wrote about the district court’s decision back in October and was unsurprised to see an affirmance based on the Becton line of cases holding that “where a claim lists elements separately, the clear implication of the claim language is that those elements are distinct components of the patented invention.”
The panel first rejected Regeneron’s argument that Becton did not apply because the district court had previously construed “buffer” to cover proteins like aflibercept (the alleged VEGF antagonist), stating that Regeneron “conflates two independent claim construction inquiries” because “claim construction inquiry relevant here, under Becton, is directed to whether a formulation is claimed in a way that clearly implies it requires distinct components” not simply whether the scope of “buffer” and “VEGF antagonist” can overlap.
Applying the Becton presumption, the panel agreed with the district court that not only did the intrinsic evidence overcome the implication, but that the claims and specification “only reinforce that the claimed components are distinct.” It agreed that all of the claims treat VEGF antagonist as separate from the buffer component, for example reciting their concentrations in different units, and the “the specification describes a formulation containing a VEGF antagonist plus a distinct buffer component” including a separate buffer in all eight examples. The panel noted that “Regeneron’s avoidance of the specification’s disclosures, or lack thereof, is telling and is an apparent concession” the specification did not help. Instead, Regeneron primarily argued that the plain meaning of buffer was broad enough to cover a aflibercept and Regeneron had not disavowed the full scope. But the panel found this unconvincing in light of Becton and the fact that the patent nowhere suggested a VEGF antagonist buffer was a part of the invention.
The panel also agreed with the district court that it need not even address the extrinsic evidence, but that it in any case it did not err in considering and discounting that evidence finding that it was reasonable for the district court to conclude that a contemporaneous reference’s teaching of buffer-free formulations “actually supports Amgen's contention that self-buffering proteins were not well known” and that the contemporaneous art “advanced the art over the ’865 patent precisely by disclosing certain buffer-free formulations in which the therapeutic protein is itself capable of maintaining pH stability.”
While the panel (and district court) seemed unequivocal that infringement could not be found, it didn’t need to go that far, concluding “there is at least a substantial question of noninfringement” and thus Regeneron was not entitled to a preliminary injunction. I expect Amgen to prevail on motion for summary judgment on remand shortly.
In Merck Sharp & Dohme B.V. v. Aurobindo Pharma USA, Inc., 130 F.4th 1363 (Fed. Cir. 2025) the court affirmed a DNJ district court’s conclusion that a reissued patent was entitled to a five-year patent term extension and had not expired.
Merck’s 340 patent issued on December 30, 2003, directed to cyclodextrin derivatives. Four months later, Merck applied for approval of sugammadex, the active ingredient in BRIDION®. During the pendency of FDA review, Merck filed for a reissue, retaining the original claims but also adding narrower claims to cover sugammadex specifically. BRIDION® was approved in 2015, and Merck successfully filed for PTE seeking the five year maximum, extending expiration of the patent from 2021 to 2026.
The generic defendants argued that the plain language of the PTE statute (35 U.S.C. § 156) only allows for an extension of “the time equal to the regulatory review period for the approved product which period occurs after the date the patent is issued” and that “the patent is issued” should mean the reissue date of the patent, dramatically cutting the length of the extension.
The panel first found that the statute was ambiguous as to which issue date the statute refers to. It then turned to the context of the PTE statute which it stated was clear: “to compensate pharmaceutical companies for the effective truncation of their patent terms while waiting for regulatory approval of new drug applications.” It sided with Merck, holding that such a construction “compensates Merck for the period of exclusivity lost due to regulatory delay” while the defendants’ reading “denies Merck compensation for all but a small period of the delay” and “[t]here is no reason why the Hatch-Waxman Act’s purpose would be served by disabling extensions of the unexpired term solely based on a patent holder’s decision to seek reissue.” It will be interesting to see if future cases apply Merck as a per se rule, or confine it to the facts of this case where the reissue patent contained the same original claims. For example, would the same result hold if Merck had cancelled its broader original claims and replaced them with all new narrower claims? Perhaps we’ll find out soon.
In In re Xencor, Inc., 130 F.4th 1350 (Fed. Cir. 2025), the court affirmed a PTAB Appeals Review Panel decision refusing to grant a claim to a method of “treating a patient by administering an anti-C5 antibody” and a similar Jepson claim as lacking written description support.
The panel first construed the claims and agreed with the PTAB that the “treating a patient” language in the preamble was limiting, because (1) the second part of the preamble, “administering an anti-C5 antibody” was limiting and it would not “splice” the clause into limiting and non-limiting portions; (2) “the more reasonable reading is that both sections of the preamble… give color and meaning to the other”; and (3) “the language in the preamble provides a raison d'être for the claim” and the claim would otherwise be a mere “academic exercise.”
Then panel then found substantial evidence supported the PTAB’s determination that the specification lacked written description support for treating a patient because it “does not define the term ‘treating,’ and it does not describe or provide any data associated with treating any patient with any disease or condition with any anti-C5 antibody, including an anti-C5 antibody with the claimed Fc modifications.”
The panel also agreed that the preamble of a Jepson claim requires written description support, otherwise a patent could evade the written description requirement, using the hypothetical that a “patentee cannot obtain a Jepson claim with a preamble that says that a time machine is well-known in the art without describing a time machine, in sufficient detail to make clear to a person of ordinary skill in the art that the inventor is in possession of such a time machine.” It then found that substantial evidence supported the Board’s conclusion that anti-C5 antibodies were not well known in the art.
The decision highlights the importance of language in the preamble and the need to not allow patentees to overreach in claiming. Interestingly, the decision focuses heavily on the “treating a patient” language, begging the question of whether the claim would have been allowed if it simply recited “administering anti-C5 antibodies” without any identified purpose of administration, let alone test results. Perhaps it would have led to a utility/enablement problem. See In re '318 Pat. Infringement Litig., 583 F.3d 1317, 1327 (Fed. Cir. 2009) (claims to treating Alzheimer’s lacked enablement where no test results of any kind because “at the end of the day, the specification, even read in the light of the knowledge of those skilled in the art, does no more than state a hypothesis and propose testing to determine the accuracy of that hypothesis. That is not sufficient”). It’s interesting to me the panel did even mention whether such claims might have been patentable.
In In re Strongbridge Dublin Ltd., No. 2023-2302, 2025 WL 751116 (Fed. Cir. Mar. 10, 2025) the court confirmed that a method of administering a drug without “concomitant” drugs did not require actively discouraging use of the concomitant drugs, but only that a single patient not be given the concomitant drugs, but remanded for the Board to determine whether a POSA would have understood a clinical study not mentioning the concomitant drugs to have meant one or more patients was not taking such drugs.
The panel first construed the term “avoiding concomitant administration” and held that it did not require actively encouraging patients to avoid such drugs, because the “plain and ordinary meaning of ‘avoid’ is [] consistent with claim coverage in situations in which patients are simply not taking famotidine or methotrexate while they are taking dichlorphenamide” noting that by “way of example, a person who drives one route home from work avoids or keeps away from a traffic jam that takes place on an alternative route home even if that person does not take the active step of deciding to avoid that traffic jam.” The panel further noted that the applicants had originally sought claims with more active language and thus was not attempting “to rewrite the claims to confine them to the very claims it cancelled.”
Interestingly, the panel then turned to the Board’s finding of anticipation by a reference (Sansone) that disclosed the treatment method claimed, but did not mention the concomitant drugs, but held that there was not substantial evidence that the reference taught the method without the use of those drugs. It held that the “correct anticipation inquiry in this case is whether the examiner established on a preponderance of evidence that only one patient in Sansone was not taking” the concomitant drugs.
It then held that the Board’s findings that (1) Sansone does not mention the concomitant drugs and (2) the patient classes in Sansone are unrelated to the patient classes who take the concomitant drugs “are relevant but not substantial evidence in and of themselves” because “[t]here must be something more to suggest that the silence is significant, or other reasons that establish by a preponderance of the evidence that it is likely that one or more of the patients in Sansone were not taking famotidine or methotrexate at some time during the nine week and/or fifty-two-week length of the study.” “This could include, for example, proof from the structure of the study, the results of the study, or statistical analysis as to the likelihood that at least one patient was not taking famotidine or methotrexate.” I’m a bit surprised by this holding, as it seems to require the Examiner/Board to obtain expert testimony analyzing the Sansone reference or performing statistical analysis. It seems clear to me at least, that there is sufficient evidence for one to infer that a study would have mentioned concomitant administration of other drugs and their absence is at least prima facie evidence, sufficient to shift the burden to patentee. Otherwise, this seems to make it extremely difficult for examiners to ever reject claims with such negative limitations. Perhaps the Examiner/Board could have evaded such a problem by issuing an obviousness rejection in the alternative. It will be interesting to see how it handles the case on remand.
In CQV Co. v. Merck Pat. GmbH, 130 F.4th 1344 (Fed. Cir. 2025) the court remanded a final written decision finding claims not unpatentable because the Board failed to address all of the evidence showing a sample batch was prior art.
The court first addressed Merck’s argument that CQV lacked standing to appeal, but held that “[g]iven at least one customer's purchase and use of Adamas® products in the United States, Merck’s communications with that customer [asserting infringement], and CQV’s indemnity agreement with that customer, CQV has established that it has standing to pursue this appeal.”
The court then addressed CQV’s arguments on the merits and found them persuasive. The Board considered three pieces of evidence that demonstrated that “Sample C” was publicly available: (1) “general statements about the availability of the Xirallic® product line that were not linked to Sample C”; (2) fact testimony that “CVQ purchased Sample C in about October 2011”; and (3) evidence that “Merck manufactured Sample C in 2007 and is incentivized to sell a batch as soon as possible so as not to waste shelf-life.” However, the Board did not address other evidence, such as testimony that after being released from quality control, a customer can purchase product. The panel stated that “CQV raised highly material and unrebutted evidence that Sample C would have been made available to the public within a few weeks of being placed into quality control, which the Board discarded without explanation.”
While noting that “failure to explicitly discuss every issue or every piece of evidence does not alone establish that the tribunal did not consider it,” (quoting Novartis AG v. Torrent Pharms. Ltd., 853 F.3d 1316, 1328 (Fed. Cir. 2017)) the panel held that the Board’s decision “goes beyond a failure to discuss a ‘cursory argument’” and the panel “cannot reasonably discern whether the Board followed a proper path.” Thus, it remanded the case for further fact finding. Interestingly, the panel did not address whether substantial evidence would have supported the Board’s determination.
I expect this decision to be cited often in appeals from IPR decisions whenever a party perceives that the Board did not fully address all arguments. It will be interesting to see how future panels strike a balance between the holding here and the holding in Novartis.
In ImmunoGen, Inc. v. Stewart, 130 F.4th 1328 (Fed. Cir. 2025), the court affirmed the district court’s agreement with the PTAB/examiner’s rejection of claims to dosing regimens of a drug for treating ovarian and peritoneal cancer.
The panel first addressed ImmunoGen’s arguments that because the claimed drug was not known to cause ocular toxicity, such concerns could not have motivated a POSA to modify the prior art, but rejecting it noting that while “[w]here a problem was not known in the art, the solution to that problem may not be obvious,” (quoting Forest Lab’ys, LLC v. Sigmapharm Lab’ys, LLC, 918 F.3d 928, 935 (Fed. Cir. 2019), “it does not follow that a claimed solution to an unknown problem is necessarily non-obvious.” The panel then found no clear error in the district court’s “findings that a person of ordinary skill in the art, despite not knowing of IMGN853’s ocular toxicity, would have nonetheless been motivated to monitor for those side effects when administering the drug to humans” because ocular toxicity was a “well-known adverse event in the administration” of related compounds.
It then addressed ImmunoGen’s arguments that a POSA would not have been motivated to try a “AIBW” dosing regimen, but again found no clear error in the court’s findings that such a regimen “would have been within the range of knowledge of a person of ordinary skill in the art when confronted with dosing-induced toxicities, and particularly when confronted with dosing-induced ocular toxicity” because “AIBW dosing was a well-known methodology that had been implemented on drugs both smaller and larger than IMGN853 and had been used to specifically reduce ocular toxicity.”
Then panel then addressed whether a POSA would have been motivated to select the claimed dose of 6 mg/kg AIBW (AIBW “refers to a size descriptor that accounts for sex, total body weight, and height” and is a modification of total body weight or TBW). The panel then noted that two prior art references taught dosing of the drug at 6 mg/kg of TBW and 5 mg/kg TBW, and agreed with the district court that a POSA “would start with doses of around 5 mg/kg or 6 mg/kg AIBW and then determine the precise dose based on routine optimization.” Indeed, the panel noted that evidence showed that “for patients who weigh exactly their ideal body weight, a dose of 6 mg/kg AIBW is identical to a dose of 6 mg/kg TBW,” and thus for some patients the prior art explicitly taught the claimed dosing amount.
Finally, the panel rejected ImmunoGen’s argument that the district court failed to find a reasonable expectation of success “that a 6 mg /kg AIBW dose would solve ocular toxicity” because the “claims are silent as to any ocular toxicity problem.”
The case is a good reminder of the flexibility of the obviousness inquiry and the difficulty on appeal of attacking motivation to combine factual findings.
Comments