EpiPen’s last patent expired on September 11, 2025. Epinephrine has been off patent since the 1980s. For close to two decades, a spring-loaded injector mechanism, not the drug inside it, kept generic competitors out of a market worth hundreds of millions of dollars a year [1]. Three months earlier, a different kind of device patent had the opposite fate. In December 2024, the Federal Circuit ordered Teva to strip five patents covering the dose counter and canister of its ProAir HFA inhaler from the FDA’s Orange Book, ruling that patents on hardware alone were never eligible for listing in the first place [2].
Those two outcomes describe the same asset class producing opposite results within the same twelve months. That is the story of device patents in pharma right now. They can be the single most durable piece of a franchise, or the first thing a court, the FTC, or a well-funded generic challenger tears down. The difference comes down to how the patent was drafted, where it was listed, and which regulatory pathway governs the underlying drug. This piece works through the case law, the FTC’s three-year campaign against “junk” device listings, and what it means for how you build and defend a device patent portfolio.
What Is a Device Patent, and Why Do Pharma Companies Rely on Them?
A device patent, in the pharma context, claims the mechanical or physical components of a drug delivery system, an inhaler’s dose counter, an autoinjector’s spring mechanism, a pen’s drive train, rather than the drug substance itself. Companies rely on them to extend exclusivity after the underlying compound patent expires, because a device patent can run 15 to 20 years from a much later filing date than the original molecule patent.
Device Patents vs. Composition-of-Matter Patents: What’s the Difference?
A composition-of-matter patent covers the chemical structure of the active ingredient itself. It is the strongest patent a drug company can hold, because a generic maker cannot design around a molecule and still sell the same drug. A device patent covers something adjacent: the housing, the dosing mechanism, the needle-shield, the counter, the cartridge, or the drive spring that delivers the drug into the patient. The two patent types expire on different schedules and get treated very differently by the FDA, the FTC, and the courts, which is the entire subject of this article.
Why Drug-Device Combination Products Exist in the First Place
Insulin, GLP-1 agonists, biologics, and inhaled respiratory drugs are almost never delivered as a bare vial and syringe anymore. Patients need a device that meters an exact dose, protects a fragile molecule, and works reliably for someone who may be visually impaired, arthritic, or in anaphylactic shock. That clinical need is real. It is also a commercial opportunity: engineering improvements to the delivery device generate a fresh, later-expiring patent even after the drug substance itself is old news.
FDA’s Definition of a Combination Product
FDA classifies a drug-device combination product as a single approved product with two regulated parts, a drug constituent and a device constituent, that are physically combined or co-packaged. Combination products are approved under a single application, usually a New Drug Application or Biologics License Application, even though the device component is separately patentable. That single-application structure is exactly what created the Orange Book listing dispute this article spends the next section on.
The Orange Book Rule That’s Quietly Gutting Device Patents
Short answer: a patent can only be listed in the FDA’s Orange Book if it claims the drug substance, the drug product’s formulation, or a method of using the drug. Patents that claim only the delivery device, with no reference to the active ingredient, do not meet that bar, according to a string of Federal Circuit and First Circuit rulings running from 2020 through 2024.
What the Orange Book Actually Requires a Patent to Claim
Under 21 U.S.C. § 355(b)(1)(A)(viii), a patent submitted for Orange Book listing has to claim the drug for which the underlying application was approved [3]. Courts have read that language literally. If a patent’s claims never mention the active pharmaceutical ingredient, and only describe a canister, a counter, or a spring, it does not claim “the drug” no matter how central that hardware is to how the product actually works.
Teva v. Amneal: The Case That Changed Everything for Inhaler Patents
Amneal filed an abbreviated new drug application for a generic version of Teva’s ProAir HFA inhaler and gave notice of its Paragraph IV certification on August 24, 2023. Teva sued for infringement, which automatically triggered a 30-month stay of FDA approval, and Amneal countersued seeking to delist the patents entirely. On June 10, 2024, the U.S. District Court for the District of New Jersey ordered Teva to delist five of the patents, finding they claimed only inhaler hardware and never mentioned albuterol sulfate [4]. Teva appealed, and the Federal Circuit briefly stayed the order before affirming it on December 20, 2024, in Teva Branded Pharm. Prods. R&D, Inc. v. Amneal Pharms. of N.Y., LLC [2][4].
The appellate court’s reasoning is the part worth remembering. It held that the fact ProAir HFA was approved as a single drug-device combination does not make the device components a drug in their own right. As the court put it, the touchstone of the distinction between a drug and a device is that a device excludes an active ingredient [5]. Approval as a combination product does not launder a hardware patent into a drug patent.
The Five ProAir HFA Patents the Federal Circuit Ordered Delisted
Teva had listed nine non-expired patents against ProAir HFA. Five were central to the fight, and all of them were directed at the metered-dose inhaler’s mechanical parts, chiefly an improved dose counter and canister design, with no claim reciting albuterol sulfate or any other active ingredient [6]. The court did not need to evaluate the underlying validity of the patents at all. It only had to decide that they were never listable in the first place, which made the case move faster and gave generics a cleaner path than a traditional invalidity fight.
In re Lantus: The First Circuit’s Warning Shot on Injector Pens
The Teva v. Amneal outcome did not appear out of nowhere. It followed a 2020 decision from the First Circuit involving Sanofi’s Lantus SoloStar insulin pen, a product with more than $7 billion in annual U.S. sales at the time [7]. Sanofi’s original insulin glargine compound patent expired in August 2014. In 2007, FDA approved a new disposable pen version, and in 2013 Sanofi listed a patent covering the pen’s drive mechanism in the Orange Book [8]. Direct purchasers sued, arguing the drive mechanism patent was improperly listed and that the resulting Paragraph IV litigation against would-be generics was a sham designed to trigger automatic 30-month stays.
A district court dismissed the case, citing ambiguity in FDA’s listing guidance. The First Circuit reversed in February 2020, holding that the drive mechanism patent did not claim insulin glargine or the SoloStar pen and therefore should never have been listed [8]. The court rejected Sanofi’s argument that any “integral component” of an approved combination product is automatically listable, reasoning that a patent on a car’s transmission does not become a patent on the whole car. That comparison, four years before Teva v. Amneal, effectively wrote the playbook that Amneal’s lawyers would later run.
How the FTC Turned Device Patents Into a Political Target
Since September 2023, the FTC has run three separate rounds of warning letters against what it calls improperly listed Orange Book patents, most of them covering inhaler and autoinjector hardware, and the campaign has survived a change in presidential administration and FTC leadership intact.
Timeline: FTC’s Three Waves of Orange Book Warning Letters (2023-2025)
The FTC opened with a policy statement in September 2023 announcing it would treat improper Orange Book listings as a potential unfair method of competition under Section 5 of the FTC Act [9]. Three concrete waves of enforcement followed.
November 2023: The First 100 Patents
On November 7, 2023, the FTC sent warning letters to ten manufacturers, AbbVie, AstraZeneca, GlaxoSmithKline, Mylan, Boehringer Ingelheim, Impax Laboratories, Kaleo, and subsidiaries of Teva and GSK, disputing more than 100 patents covering 17 branded products, mostly asthma inhalers, epinephrine autoinjectors, and Restasis multidose bottles [9][10].
April 2024: 300 More, Now Targeting GLP-1 Pens
On April 30, 2024, the agency expanded the campaign, challenging over 300 additional patents across 20 branded products and sending new letters to ten manufacturers [11]. This round pulled in device patents on injectable diabetes and weight-loss pens for the first time, not just respiratory inhalers, which is the moment GLP-1 delivery devices entered the FTC’s crosshairs [9].
May 2025: Chair Ferguson Picks Up Where Khan Left Off
Under new Republican FTC Chair Andrew Ferguson, the agency issued renewed warning letters on May 21, 2025, disputing more than 200 patents across 17 products, sending letters to Amphastar, Mylan, Covis Pharma, and three Teva entities [12]. The renewal came about five months after the Federal Circuit’s Teva v. Amneal decision, and the FTC explicitly credited the ruling with vindicating the legal theory behind its warning letters [13]. Bipartisan continuity on this issue is unusual, and it signals that device patent listing risk is not a policy fad tied to one administration.
Which Companies Delisted, and Which Companies Held Their Ground
The results after round one were mixed but real. Impax delisted both of its Adrenaclick patents. Kaleo delisted eight patents covering Auvi-Q. GSK withdrew three patents tied to Arnuity Ellipta, Advair HFA, Ventolin HFA, and Flovent HFA, while choosing to keep defending a fourth Arnuity patent it considered properly listed [10]. AstraZeneca, Boehringer Ingelheim, and GSK separately agreed to cap patient out-of-pocket inhaler costs at $35 a month rather than fight over the listings themselves [9].
Round two produced a very different response. According to FDA’s own database, none of the ten manufacturers targeted in April 2024 removed a single patent. Companies simply recertified that their listings complied with the statute, including several that had voluntarily delisted after round one [14]. That divergence is the practical lesson for any brand company weighing whether to fight or fold: a warning letter alone changes almost nothing. Litigated outcomes, like Teva v. Amneal, are what actually move the needle [15].
Why Generic Companies Target the Device, Not the Drug
Generic and biosimilar challengers go after device patents because they are procedurally easier to knock out than a composition-of-matter patent. A delisting argument does not require proving invalidity at all, it only requires showing the patent was never eligible for the Orange Book to begin with, which is a faster and cheaper fight than traditional Paragraph IV litigation.
Design-Around Strategy: How Generics Engineer Past a Patented Mechanism
Where delisting is not available, or the product is a biologic outside the Orange Book system entirely, challengers try to engineer a mechanically different device that performs the same clinical function without infringing the patented claims. This is harder than it sounds. Any injector that mimics EpiPen’s core spring-loaded mechanism risks infringement, and any injector that departs from it risks failing FDA’s usability and human-factors review, or worse, failing in the hands of a panicked parent during an actual anaphylactic emergency [16]. Sanofi’s own well-documented struggles bringing its Auvi-Q autoinjector to market safely illustrate how real that trade-off is [15].
The Cost Calculus: Litigation vs. Redesign
A generic manufacturer facing an improperly listed device patent has three options: fight the listing directly, as Amneal did, license the technology, or redesign the device and run a full human-factors validation study, which can take years and tens of millions of dollars. Delisting litigation has become the preferred route precisely because it sidesteps the redesign cost entirely. Win the listing fight, and the 30-month regulatory stay disappears without a single engineering change to the generic product.
Case Study: EpiPen’s Device Patent Outlasted Its Drug Patent by Decades
EpiPen shows the opposite failure mode from ProAir HFA. Epinephrine is a century-old, fully generic molecule, yet a chain of device patents on the auto-injector mechanism kept effective competition limited until the last patent expired on September 11, 2025 [1].
Why Epinephrine Being Generic Since the 1980s Didn’t Matter
EpiPen was first approved in 1987. By the time the pricing controversy exploded in 2016, the product was still shielded by four patents, three of which had been granted within the prior six years [17]. Mylan completed its acquisition of the EpiPen business, through Dey, on October 2, 2007. Just 23 days later, on October 25, 2007, the device’s original patent holder filed a new patent application covering the injector, which issued in 2010, extending the effective hardware monopoly well past the point where the underlying drug substance had any patent protection left at all [17].
The Spring-Loaded Mechanism Patent That Blocked Teva Until 2015
In 2008, the original patent holders sued Teva for infringing a 2005 patent covering the automatic spring-loaded injector mechanism, after Teva sought FDA approval for a rival device. The parties settled in 2012 on terms that barred Teva from launching its own product until mid-2015 [17]. Design-around attempts by competitors ran into the same wall repeatedly: any injector that used a comparable spring-and-needle mechanism risked infringing, and any injector different enough to avoid infringement risked failing FDA’s safety review for a product that has to work correctly in a life-threatening emergency without medical supervision [15].
What Changed When the Last EpiPen Patent Expired in September 2025
Once the final patent expired, medical device analysts expected a wave of new entrants able to build directly on EpiPen’s original mechanical design rather than engineering around it, adding features like Bluetooth dose tracking, smaller form factors, or child-specific ergonomics [1]. The pricing history behind that wait is well documented. A two-pack of EpiPens that cost roughly $100 in 2008 was selling for around $600 by the mid-2020s, and Viatris, formerly Mylan, settled antitrust claims for $73.5 million in January 2025 over allegations it conspired with Pfizer and Teva to delay generic entry and preserve that pricing [1][17].
Case Study: Advair Diskus Proved Patent Expiry Isn’t the Same as Generic Entry
Advair Diskus’s compound patent expired in 2010 and its Diskus inhaler device patent expired in 2016, yet the first substitutable generic, Mylan’s Wixela Inhub, did not launch until February 2019. The gap was regulatory, not patent-driven.
The Nine-Year Gap Between Diskus Patent Expiry and Wixela Inhub’s Launch
GSK’s fluticasone-salmeterol compound patent expired all the way back in 2010, and the Diskus device patent followed in 2016 [18]. On paper, that should have opened the door to cheap generics almost a decade before it actually did. Instead, FDA held generic dry-powder inhalers to an unusually demanding bioequivalence standard, requiring in vitro pharmaceutical equivalence, systemic pharmacokinetic bioequivalence, local lung deposition equivalence, and in-home device robustness testing, all at once [20]. Hikma’s generic was rejected by FDA in 2017 on a first attempt [19]. Mylan finally cleared the bar and launched Wixela Inhub in February 2019, priced at a wholesale cost 70 percent below Advair Diskus, and it was the first generic granted “substitutable” status, meaning pharmacists could swap it in without a new prescription [18][21].
What This Means for Your Own Combination Product Timeline
The financial impact once generic competition finally arrived was fast and severe. GSK’s Advair U.S. sales dropped 53 percent to £398 million within roughly a year of Wixela’s launch, and the company flagged an expected profit hit of around £800 million tied to the loss [18][19]. The strategic lesson runs in both directions. If you are the brand, do not assume patent expiry is your real deadline, the FDA equivalence bar for a complex device can buy years of extra runway even after your intellectual property is gone. If you are the generic challenger, budget for the device usability and bioequivalence program as the actual bottleneck, not the patent fight. FDA has continued monitoring adverse event reports tied to patients switching between Advair Diskus and Wixela Inhub, evaluating whether additional labeling reminders about device differences are warranted [22], a reminder that regulatory scrutiny of a combination product’s device half does not end at approval.
Case Study: Lantus SoloStar and the Antitrust Exposure of Over-Listing
Improperly listing a device patent is not just a losing legal argument, it can become an antitrust liability. In re Lantus let direct purchasers pursue Sherman Act claims against Sanofi over its SoloStar pen patent listing.
How a Drive Mechanism Patent Became a Sherman Act Problem
After the First Circuit found Sanofi’s SoloStar drive mechanism patent improperly listed, the case returned to the district court, where a class of direct purchasers pursued antitrust claims tied to the resulting Paragraph IV litigation and 30-month stays [8][24]. The theory: an improper listing that triggers automatic regulatory delay against a generic competitor is not just a patent-office paperwork problem, it can be exclusionary conduct that violates Section 2 of the Sherman Act if it delayed cheaper insulin from reaching purchasers who were forced to pay supracompetitive prices in the meantime [24]. The case eventually proceeded toward summary judgment findings and trial, with the court agreeing that Sanofi’s submission of the device patent for listing was improper as a matter of law [24]. For general counsel, the takeaway is blunt: an aggressive device patent listing strategy that would have looked like ordinary Orange Book housekeeping a decade ago can now become the factual predicate for a private antitrust class action.
Biologics Are Different: Why BPCIA Devices Don’t Get the Same Break
Everything above concerns small-molecule drugs governed by Hatch-Waxman and the Orange Book. Biologics run through the Biologics Price Competition and Innovation Act instead, which has no equivalent “improper listing” defense, and device and formulation patents on biologics have held up far better in court as a result.
Eylea’s Prefilled Syringe Patents and the Federal Circuit’s 2025 Injunctions
Regeneron’s Eylea, an aflibercept injection for retinal disease with $9.5 billion in global sales in 2024, is a useful contrast case [26]. Mylan, Samsung Bioepis, Formycon, Celltrion, and others each filed for FDA approval of Eylea biosimilars. Rather than a Hatch-Waxman ANDA fight, these cases ran through the BPCIA’s patent dance provisions. Regeneron sued Mylan in the Northern District of West Virginia in August 2022, later consolidating five more cases into a single multidistrict proceeding by April 2024 [29]. In June 2024, the district court granted preliminary injunctions against Samsung Bioepis, and on January 29, 2025, the Federal Circuit upheld those injunctions against both Mylan and Samsung Bioepis and Formycon, blocking their biosimilars from the U.S. market despite FDA approval [27]. Samsung Bioepis eventually settled and licensed its way back onto the market in a February 2026 agreement that lifted the injunction, leaving other biosimilar makers still tied up in litigation [28].
Why the “Improper Listing” Defense Doesn’t Exist Under the BPCIA
Biologics do not appear in the Orange Book, they appear in the Purple Book, and the Purple Book does not function as a listing-dispute battleground the way the Orange Book does. There is no statutory hook comparable to 21 U.S.C. § 355(b)(1)(A)(viii) forcing a biologic’s device or formulation patents to claim the active ingredient directly. Biosimilar patent fights instead run through direct infringement litigation under the BPCIA’s patent dance, meaning a prefilled syringe or formulation patent can be enforced on its own terms, and can support a preliminary injunction, without ever facing a delisting challenge in the first place. That structural gap is exactly why Regeneron’s device-adjacent patents held up in the U.S. even as international courts reached mixed results on the same family of syringe and formulation patents in the UK and at the European Patent Office [25].
What This Means for GLP-1 Pen Patents Going Forward
Semaglutide and liraglutide are peptides, but Novo Nordisk brought Ozempic, Wegovy, and Victoza to market through NDAs rather than BLAs, which keeps them inside the Orange Book system and exposed to the exact delisting risk described above [32]. The next generation of larger biologic GLP-1 and dual-agonist candidates entering development under BLAs, by contrast, would inherit Eylea’s more litigation-friendly posture for their pen and cartridge patents. Which regulatory pathway a molecule takes to market is turning into a genuine determinant of how defensible its eventual device patent estate will be.
GLP-1 Pens: The Next Battleground for Device Patent Fights
Novo Nordisk’s FlexTouch pen patents on Ozempic and Wegovy, and a pending lawsuit over how it listed device patents for the older drug Victoza, make the injectable weight-loss and diabetes category the most active front in device patent litigation right now.
Novo Nordisk’s FlexTouch Patent Stack on Ozempic and Wegovy
Novo Nordisk settled existing Ozempic patent litigation with Mylan, Dr. Reddy’s, Apotex, and Sun Pharmaceutical in the fall of 2024, on confidential terms that leave the market guessing at the real generic entry date [30]. Litigation over Wegovy and Rybelsus patents against Mylan, Sun, and Apotex remained pending as of early 2025 [30]. Beyond the semaglutide compound patents, expected to expire around 2031 to 2032 in the U.S., Novo has built a dense secondary patent estate covering the FlexTouch pen-injector device itself, the formulation’s specific concentration and pH, the weekly dose-titration schedule, and methods of treating obesity in defined patient populations [31]. Because semaglutide is a biologic, any true generic competitor has to clear the 351(k) biosimilar pathway rather than a simple ANDA, which realistically pushes a biosimilar semaglutide launch toward 2027 or 2028 even after the core compound patent falls [34].
The Victoza Lawsuit: A Preview of What’s Coming for Semaglutide
A pending antitrust suit against Novo Nordisk over its older GLP-1 drug Victoza alleges the company wrongfully listed device patents in the Orange Book to trigger Hatch-Waxman delays, and separately alleges that a 2019 settlement with Teva, the first generic Victoza challenger, included an unlawful reverse payment that pushed Teva’s launch out to June 24, 2024 in exchange for 180 days of exclusive generic sales [32]. Plaintiffs claim Novo also used the disputed listings to block other generic manufacturers from challenging Victoza until after Teva’s exclusivity window closed, delaying broader generic entry by at least 16 months [32]. Whatever the outcome, the theory of the case, that a GLP-1 pen device patent estate can be attacked using the exact same Orange Book listing arguments that worked in Teva v. Amneal and In re Lantus, is the clearest signal yet of where Ozempic and Wegovy patent challenges are headed next.
How to Audit Your Own Device Patent Portfolio for FTC and Delisting Risk
Pull every device or combination-product patent your company has listed in the Orange Book and check whether its claims recite the active ingredient by name. If they do not, that patent is a candidate for an FTC warning letter, a delisting counterclaim, or both.
Five Questions to Ask About Every Device Patent You’ve Listed
Does at least one claim explicitly recite the active ingredient or drug substance, not just a “drug delivery device” in the abstract?
Was the patent filed and listed years after the original NDA, in a pattern that resembles evergreening rather than genuine follow-on innovation?
Would the patent read the same way if the device were sold empty, with no drug inside it at all?
Has a similar patent from a competitor already been challenged, delisted, or voluntarily withdrawn by the FTC’s warning letter campaign?
Is the product regulated under an NDA and therefore Orange Book eligible, or under a BLA and therefore governed by the more litigation-friendly BPCIA framework instead?
Does the Claim Recite the Active Ingredient?
This is the single test that decided both Teva v. Amneal and In re Lantus. Courts are not weighing how important the device is to the product’s function or safety. They are reading the claim language on its face and asking whether the active ingredient shows up anywhere in it [5][6]. A dose counter claim that never says “albuterol sulfate,” or a drive mechanism claim that never says “insulin glargine,” fails this test regardless of how central that hardware is to the product’s real-world performance.
Is the Patent Directed at a Component or the Whole Approved Product?
Sanofi tried to argue that any “integral component” of an approved combination product should qualify for listing. Both the First Circuit and the Federal Circuit rejected that theory outright, using nearly identical logic: a patent on a car’s transmission is not a patent on the car [4][8]. If your patent claims a sub-part of the device rather than the device as approved together with the drug, treat it as high risk for delisting under current case law.
What Happens If Your Device Patent Gets Delisted
Delisting removes your patent from the Orange Book and, with it, your ability to trigger an automatic 30-month stay against a Paragraph IV generic filer. It does not invalidate the patent itself, and you can still sue for infringement, but you lose the built-in regulatory delay that made Orange Book listing valuable in the first place.
Delisting vs. Invalidation: Why the Distinction Matters
These are two different legal outcomes with very different consequences. An invalidated patent is dead, it cannot be enforced against anyone. A delisted patent is alive and can still be asserted in ordinary infringement litigation, it simply loses its special status inside the Hatch-Waxman framework [14]. That distinction is why some manufacturers, when hit with an FTC warning letter, have chosen to voluntarily delist rather than fight, keeping the underlying patent intact for conventional litigation while giving up the automatic stay.
The 30-Month Stay You Lose, and What Still Protects You
Listing a patent in the Orange Book means that if you sue a generic filer for infringement within 45 days of receiving its Paragraph IV notice, FDA approval of that generic is automatically frozen for up to 30 months or until a court rules against you, whichever comes first [10]. Once a patent is delisted, that automatic freeze disappears. You can still sue for infringement and seek a preliminary injunction, as Regeneron did successfully with its Eylea prefilled syringe patents, but you now have to actually win a motion for injunctive relief on the merits rather than relying on a statutory timer that runs regardless of how strong your case is [27].
Building a Device Patent Strategy That Survives Scrutiny
The device patents that survive current scrutiny are the ones drafted to claim the drug together with the device, backed by separate method-of-use and formulation patents, and tracked continuously against FTC and delisting activity rather than filed once and forgotten.
Layering Device Patents With Method-of-Use and Formulation Claims
The lesson from Teva v. Amneal and In re Lantus is not that device patents are worthless, it is that device-only claims are fragile. A claim that ties the mechanical innovation to a specific dosing method, a formulation characteristic, or the active ingredient’s stability inside the device stands on much firmer ground than a bare hardware claim, because it satisfies the statutory listing requirement on its face. Formulation and method-of-use patents around a device, rather than patents on the device in isolation, are proving to be the more durable layer of the stack.
Monitoring Delistings and Corrections in Real Time
The Orange Book is not a static document anymore. Between FTC warning letters, voluntary delistings, court-ordered removals, and routine patent corrections, the list of what actually protects a given product changes month to month, and the changes are the leading indicator of who is about to face generic entry.
Patent-cliff analysts tracking Orange Book activity put the coming loss-of-exclusivity wave at roughly $350 billion in brand revenue between 2025 and 2030, a period in which delisting and correction activity has become one of the more reliable early signals of which drug-device franchises are actually exposed [33].
Tools built specifically to track this kind of movement, such as DrugPatentWatch’s Orange Book monitoring, exist precisely because expiry dates alone no longer tell the full story. A patent with eight years left on its face can be functionally worthless the moment it gets delisted, and a portfolio review that only checks expiration dates will miss that entirely.
Why Patent Cliff Forecasting Now Requires Orange Book Monitoring, Not Just Expiry Dates
A forecasting model that treats every listed patent as equally durable until its stated expiration date will systematically overstate exclusivity for any product with a device-only patent in its stack. The more accurate model weights each patent by listing risk, whether its claims recite the active ingredient, whether it has already drawn an FTC warning letter, and whether comparable patents in the same product category have already fallen, alongside the calendar date. Getting that weighting right is now a genuine competitive intelligence function, not a compliance afterthought, and it applies to how you evaluate competitors’ portfolios just as much as your own.
Key Takeaways
A patent claiming only device hardware, with no reference to the active ingredient, is not properly listable in the Orange Book under current Federal Circuit and First Circuit precedent.
The FTC has run three rounds of warning letters against device patent listings since 2023, and the campaign has continued unchanged under new leadership in 2025, making this a durable bipartisan enforcement priority rather than a temporary policy.
Device patents can still be enormously durable when the underlying drug is itself off-patent, as EpiPen showed for nearly two decades, because design-around attempts run into FDA safety and usability review, not just patent law.
Patent expiration and generic market entry are not the same event. Advair Diskus’s compound and device patents expired years before a substitutable generic actually reached pharmacy shelves, because of FDA’s bioequivalence standard for complex devices.
Biologics regulated under the BPCIA, like Eylea, do not face the same Orange Book delisting exposure as small-molecule combination products, which makes device and formulation patents meaningfully more durable for biologics than for NDA-approved drugs.
GLP-1 pens are the next major front. Ozempic, Wegovy, and Victoza are all NDA products sitting inside the Orange Book system, and a pending Victoza lawsuit is testing the exact delisting theory that already worked against Teva’s inhaler patents.
Frequently Asked Questions
Can a company list a device patent in the Orange Book if the device is essential to how the drug works?
No. Courts have rejected the argument that a device’s importance to the drug’s function makes it listable. The only test that matters is whether the patent’s claims recite the active ingredient itself [5].
Does delisting a patent mean a company loses its patent rights entirely?
No. Delisting removes the patent from the Orange Book and eliminates the automatic 30-month stay against generic approval, but the patent remains enforceable through ordinary infringement litigation [14].
Why did EpiPen’s device patents last so much longer than ProAir HFA’s?
EpiPen’s patents were not challenged on listing grounds the way ProAir HFA’s were. Its barrier was practical rather than purely legal, competitors struggled to design a mechanically different injector that could also pass FDA’s human-factors safety review for an emergency-use product [15][16].
Are biosimilar device and formulation patents at the same delisting risk as small-molecule device patents?
No. Biologics are governed by the BPCIA rather than Hatch-Waxman, and the Purple Book has no equivalent improper-listing dispute mechanism, so biosimilar device and formulation patents are typically enforced through direct infringement suits and preliminary injunctions instead [27].
What is the practical difference between a Paragraph IV challenge and an Orange Book delisting challenge?
A Paragraph IV challenge argues the listed patent is invalid or not infringed. A delisting challenge argues the patent should never have been in the Orange Book at all, which sidesteps validity and infringement analysis entirely and has proven faster for generic challengers [4].
Did the FTC’s warning letter campaign actually remove many patents from the Orange Book?
Results were mixed. The first round in November 2023 led several companies to voluntarily delist patents, but none of the manufacturers targeted in the April 2024 round removed a single patent, choosing instead to recertify their listings [10][14].
Why did it take almost nine years after Advair’s key patents expired for a substitutable generic to launch?
FDA required generic dry-powder inhalers to clear a demanding, multi-part bioequivalence standard covering in vitro performance, pharmacokinetics, lung deposition, and in-home device robustness, a scientific and regulatory bar that had nothing to do with patent status [20].
Is Ozempic’s FlexTouch pen protected by its own separate patents from the semaglutide compound patent?
Yes. Novo Nordisk has listed secondary patents covering the FlexTouch pen device, the formulation’s concentration and pH, and the dose-titration schedule, in addition to the underlying semaglutide compound patents, which expire separately and generally later [31].
Can a company be sued for antitrust violations over an improperly listed device patent?
Yes. In re Lantus allowed direct purchasers to pursue Sherman Act claims against Sanofi over its SoloStar pen patent listing, on the theory that the resulting Paragraph IV stays constituted anticompetitive conduct [8][24].
How can a company track Orange Book delistings and patent corrections across a competitor’s portfolio?
Dedicated patent intelligence platforms such as DrugPatentWatch track Orange Book delistings, corrections, and supplements as they happen, which is a more reliable early signal of generic entry risk than expiration dates alone [33].
References
Medical Device Network. (2025, August 8). EpiPen patent expiry opens new era for innovation. https://www.medicaldevice-network.com/news/epipen-patent-expiry-opens-new-era-for-innovation/
Federal Trade Commission. (2024, December 20). FTC statement on appellate court decision ordering delisting of Teva inhaler patents. https://www.ftc.gov/news-events/news/press-releases/2024/12/ftc-statement-appellate-court-decision-ordering-delisting-teva-inhaler-patents
Patently-O. (2024, December 20). Orange Book device patent listings: Understanding Teva v. Amneal. https://patentlyo.com/patent/2024/12/orange-listings-understanding.html
American Bar Association, Antitrust Law Section. (2025, May 12). Federal Circuit affirms device patent delisting in Teva v. Amneal. https://www.americanbar.org/groups/antitrust_law/resources/newsletters/federal-circuit-affirms-teva-amneal/
Cooley LLP. (2025, January 2). Teva v. Amneal ruling interprets Orange Book listing statute, affirms delisting of device patents. https://www.cooley.com/news/insight/2025/2025-01-02-teva-v-amneal-ruling-interprets-orange-book-listing-statute-affirms-delisting-of-device-patents
Duane Morris LLP. (2025, January). Federal Circuit affirms delisting patents claiming device components of drug product without active ingredient. https://www.duanemorris.com/alerts/federal_circuit_affirms_delisting_patents_claiming_device_components_drug_product_without_0125.html
The FDA Law Blog. (2020, March 16). If FDA won’t regulate, maybe the courts will: First Circuit opines on listing device patents in the Orange Book. https://www.thefdalawblog.com/2020/03/if-fda-wont-regulate-maybe-the-courts-will-first-circuit-opines-on-listing-device-patents-in-the-orange-book/
Cozen O’Connor. (2020). First Circuit permits antitrust claims for improperly listing a device patent on the FDA’s Orange Book to move forward. https://www.cozen.com/news-resources/publications/2020/first-circuit-permits-antitrust-claims-for-improperly-listing-a-device-patent-on-the-fda-s-orange-book-to-move-forward
Congressional Research Service. (2026, January 21). Patent listing in FDA’s Orange Book (IF12644). Congress.gov. https://www.congress.gov/crs-product/IF12644
BioSpace. (2024, October 3). Three companies relent to FTC demands, delist patents from FDA’s Orange Book. https://www.biospace.com/three-companies-relent-to-ftc-demands-delist-patents-from-fda-s-orange-book
Federal Trade Commission. (2024, April 30). FTC expands patent listing challenges, targeting more than 300 junk listings for diabetes, weight loss, asthma and COPD drugs. https://www.ftc.gov/news-events/news/press-releases/2024/04/ftc-expands-patent-listing-challenges-targeting-more-300-junk-listings-diabetes-weight-loss-asthma
Hogan Lovells. (2025, May 23). Republican FTC renews challenges to Orange Book patent listings. https://www.hoganlovells.com/en/publications/republican-ftc-renews-challenges-to-orange-book-patent-listings
IPWatchdog. (2025, May 22). FTC renews campaign to remove ‘junk’ inhaler and device patents from Orange Book listings. https://ipwatchdog.com/2025/05/22/ftc-renews-campaign-remove-junk-inhaler-device-patents-orange-book-listings/
White & Case LLP. (2024, August 20). The current status of FTC’s Orange Book listings challenge: A mixed bag. https://www.whitecase.com/insight-our-thinking/current-status-ftcs-orange-book-listings-challenge-mixed-bag
Sherkow, J. S., & Zettler, P. J. (2021). EpiPen, patents, and life and death. NYU Law Review. https://nyulawreview.org/wp-content/uploads/2021/08/SherkowZettler-fin-1.pdf
Seven Pillars Institute. (2017, September 14). Mylan’s EpiPen pricing scandal. https://mail.sevenpillarsinstitute.org/mylans-epipen-pricing-scandal/
St. Louis Post-Dispatch. (2024, April 12). Behind the EpiPen controversy are questions about patents granted to drugmaker. https://www.stltoday.com/business/local/behind-the-epipen-controversy-are-questions-about-patents-granted-to/article_54bbc38b-57b1-5a41-bbe4-a92aa9f85751.html
pharmaphorum. (2019, February 13). Mylan launches Advair generic at 70% discount in US. https://pharmaphorum.com/news/mylan-launches-advair-generic-at-70-discount-in-us
pharmaphorum. Chasing Mylan, Hikma tries again with rejected generic Advair in US. https://pharmaphorum.com/news/chasing-mylan-hikma-tries-again-with-rejected-generic-advair-in-us
pharmaphorum. FDA finally approves Mylan’s Advair generic. https://pharmaphorum.com/news/fda-finally-approves-mylans-advair-generic
Citeline. (2024, September 28). Adverse events for generic of GSK’s Advair Diskus on US FDA’s radar. https://insights.citeline.com/PS141609/Adverse-Events-For-Generic-Of-GSKs-Advair-Diskus-On-US-FDAs-Radar/
PMC (National Center for Biotechnology Information). Usability and robustness of the Wixela Inhub dry powder inhaler. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8060712/
Hagens Berman. Lantus (insulin glargine) direct purchaser antitrust lawsuit. https://www.hbsslaw.com/cases/lantus-insulin-patent-antitrust
Axinn, Veltrop & Harkrider LLP. (2025, June 2). FTC peeling back the layers of the Orange Book. https://www.axinn.com/en/insights/axinn-viewpoints/ftc-peeling-back-the-layers-of-the-orange-book
JUVE Patent. (2025, October 22). Setback in UK after victory in Netherlands for Regeneron in Eylea case. https://www.juve-patent.com/cases/setback-in-uk-after-victory-in-netherlands-for-regeneron-in-eylea-case/
Patently-O. (2025, January 29). Biologics patent enforcement: Regeneron patents block biosimilar entry for EYLEA. https://patentlyo.com/patent/2025/01/biologics-enforcement-biosimilar.html
National Law Review. Federal Circuit upholds injunction in Regeneron v. Mylan. https://natlawreview.com/article/eye-catching-biosimilars-injunction-prevails
Markman Advisors. (2025, February 7). What is the patent landscape for Novo Nordisk’s semaglutide products, Ozempic, Wegovy and Rybelsus? https://www.markmanadvisors.com/blog/2025/2/7/what-is-the-patent-landscape-for-novo-nordisks-semaglutide-products-ozempic-wegovy-and-rybelsus
HealthRX. (2026, May 25). Wegovy patent portfolio and generic timeline: When will generic semaglutide 2.4 mg be available? https://healthrx.com/wegovy/patent-generic-timeline
Finviz. Novo Nordisk sued over alleged scheme to delay generic Victoza in favor of Ozempic. https://finviz.com/news/293529/novo-nordisk-sued-over-alleged-scheme-to-delay-generic-victoza-in-favor-of-ozempic
DrugPatentWatch. (2026, February 28). Delistings, patent corrections, and supplements: Monitor changes and predict the patent cliff in the FDA Orange Book. https://www.drugpatentwatch.com/blog/delistings-patent-corrections-and-supplements-monitor-changes-and-predict-the-patent-cliff-in-the-fda-orange-book/
Telehealth Ally. (2026, April 29). Generic Ozempic 2026: Semaglutide patent expiry, biosimilar timeline and cheaper alternatives. https://telehealthally.com/guides/generic-ozempic-semaglutide-timeline-guide
