1. Introduction
The US federal government grants time-limited rights for brand-name pharmaceutical firms to sell prescription drugs free from direct generic competition, and generic firms can then enter the market once patent protection ends. 1 Such competition is, by far, the most important tool for lowering prescription drug prices in the US. Although approximately 80% of prescriptions are filled for generic drugs, these prescriptions account for just 20% of total spending. 2 As each new generic competitor enters the US market, prices tend to drop and eventually approach the costs of production.Reference Dave 3 This system is designed to promote innovation while ensuring that patients and payers ultimately benefit from access to low-cost therapy.
Key to the function of our prescription drug market is timely generic competition. The government permits monopoly pricing for a period before allowing the free market to operate. Unfortunately, many brand-name firms undermine this system by engaging in strategies designed to extend periods of market exclusivity and delay generic competition. These strategies not only harm patients, but they also raise costs for the US health care system by contributing to higher insurance premiums and taxes that fund federal health programs like Medicare and Medicaid.
This paper analyzes two key strategies that manufacturers employ to delay generic competition — patent thickets and product hops — and considers how several recently proposed legislative reforms aim to blunt these strategies. I will begin by first providing a brief background on patent thickets and product hops before turning to a particular type of pharmaceutical product that has proven especially vulnerable to gamesmanship: drug-device combinations, which contain active pharmaceutical compounds sold together with their delivery devices. 4 I then consider patenting practices on other types of pharmaceuticals before reviewing a set of policy reforms aimed at facilitating generic entry. My focus throughout will be on small-molecule drugs rather than biologics, as the two are governed by different regulatory structures (although biologic manufacturers employ similar strategies). I will argue that recent legislative proposals offer a valuable step forward in curbing patent gamesmanship on small-molecule drugs but that further reforms are needed.
2. Strategies that serve to delay generic competition
a. Patent thickets
One key tactic that brand-name manufacturers employ to delay generic competition is erecting large thickets of patents.Reference Feldman 5 Many of these patents do not cover active pharmaceutical compounds (otherwise known as primary patents) but cover peripheral features such as alternative formulations and methods of use (secondary patents) and delivery devices (tertiary patents).Reference Beall and Kesselheim 6 They may also be continuations of earlier patents that disclose no new substantive changes to the underlying inventions; indeed, a majority of pharmaceutical patents now listed with the FDA are continuations rather than original patents.Reference Tu 7 Brand-name firms augment large patent portfolios by adding numerous patents after FDA approval.Reference Horrow 8 Some of these strategies increase the density of a product’s patent portfolio (the number of patents) while others increase the duration of the product’s expected market exclusivity (the time from approval until expiration of the last-to-expire patent).
For small-molecule drugs, companies list their key patents in the FDA’s Approved Drug Products with Therapeutic Equivalence Evaluations (Orange Book). Maximizing both the density and duration of patent portfolios on brand-name drugs can be effective for pharmaceutical companies, because the FDA cannot approve generic drugs for marketing until patents in the Orange Book on these drugs expire or are successfully challenged. 9 The FDA does not review what gets listed in the Orange Book but instead serves in a “purely ministerial” role, listing patents that manufacturers choose for submission. 10 When generic firms challenge FDA-listed patents (via what is known as paragraph IV certifications), brand-name firms that sue for infringement within 45 days can receive an automatic 30-month stay while litigation resolves. The large patent thickets created by brand-name pharmaceutical companies often tie up generic firms in costly legal battles and deter generic firms from challenging patents in the first place.
b. Product hops
A second type of strategy that manufacturers employ to limit generic competition is developing new products that are closely related to originators (reformulations) yet sufficiently different to receive new patents. For example, manufacturers may move an active ingredient from one delivery device (e.g., an inhaler) to another with extended patent protection even if the new device offers minimal added clinical benefit to patients — or none at all. Pharmaceutical firms may combine two existing drugs into a single pill with new patent protection or release new strengths, salts, or esters (e.g., metoprolol tartrate rather than metoprolol succinate), or single enantiomers of racemic mixtures (e.g., esomeprazole rather than omeprazole).Reference Darrow, Chong and Kesselheim 11 These reformulations are considered hard product hops when the original drug is discontinued altogether and soft product hops when both versions remain on the market but the company seeks to move patients to newer products to reduce generic uptake of older versions (e.g., by marketing newer brand-name versions to physicians even in the absence of improved clinical outcomes).Reference Gowda 12
Although some product hops result in formulations that are more convenient for patients, many offer little to no meaningful clinical advantage over existing products. For example, in the largest systematic review to date of single enantiomer vs. racemic mixture drugs, the majority of randomized controlled trials found no differences in efficacy or safety between the two.Reference Long 13 More importantly, the financial rewards of these small tweaks often far outstrip the benefits to patients, sometimes rivalling those bestowed on transformative new therapeutic breakthroughs.
Manufacturers rely on these strategies — increasing the density of patent portfolios and the duration of market exclusivity periods — to protect revenue streams. These strategies not only undermine the very rationale for time-limited, government-granted monopolies, but they divert resources away from investment in more meaningful therapeutic advances.
3. The challenges of drug-device combinations
Drug-device combinations, which include products like inhalers, insulin pens, glucagon-like peptide-1 (GLP-1) receptor agonists, buprenorphine transdermal patches, and epinephrine pens, are especially susceptible to patent thickets and product hops.Reference Sinha 14 They also represent a growing share of pharmaceutical spending. Of the 50 products with the highest 2022 gross Medicare Part D spending, 20 (40%) were drug-device combinations. 15 These include blockbuster drugs like Trelegy Ellipta (fluticasone-umeclidinium-vilanterol) and Symbicort (budesonide-formoterol) for asthma and chronic obstructive pulmonary disease (COPD) and Ozempic (semaglutide) and Lantus SoloStar (insulin glargine) for diabetes. Although manufacturers have engaged in harmful patenting behavior on a variety of products, I will focus on two of the largest classes of drug-device combinations — inhalers and GLP-1 receptor agonists — to illustrate key problems that threaten the function of our pharmaceutical system.
a. Inhalers for asthma and COPD
More than 27 million people in the US have asthma 16 and nearly 12 million have COPD. 17 Patients with both conditions rely on maintenance inhalers for daily use and rescue inhalers to manage acute symptoms. 18 List prices for these products can run more than $600 per month and, although payers negotiate rebates to lower costs, net prices are still substantially higher than prices in other countries. 19 Because out-of-pocket costs are often tied to list prices and because those without insurance do not pay insurer-negotiated rates, high list prices can also threaten affordability for patients. For many years, spending on inhalers represented approximately 3–5% of net spending on retail prescription drugs in Medicare.Reference Feldman, Gagne and Kesselheim 20
In research analyzing all 53 brand-name inhalers approved for asthma and COPD by the FDA from 1986 to 2020, my coauthors and I observed several worrisome features regarding the tactics employed by manufacturers to preserve market exclusivity.Reference Feldman 21 First, firms have not managed to find new therapeutic breakthroughs for inhaled medications to treat asthma and COPD. Instead, they have copied drugs in the same therapeutic classes. The last inhaler to reach the market with a new mechanism of action was Atrovent (ipratropium) in 1986. Since then, every inhaler for asthma or COPD to receive FDA approval has contained active ingredients with the same mechanisms of action already present in other inhaled products sold in the US. Second, manufacturers have released inhalers with an increasing number of patents on their products. The median number of patents at the time of approval rose from 2 in the first decade of our study (with a median of 0 device patents) to 11 in the final decade (with a median of 7 device patents). Third, inhaler manufacturers often file patents after FDA approval; manufacturers of 25 (47%) products employed this strategy, including 12 cases in which the new patents extended periods of market exclusivity (by a median of 6 years). Fourth, manufacturers file numerous device patents, which represented more than half of all patents listed on inhalers in the cohort. These device patents were on small plastic components of inhalers such as dose counters, filters, and nozzles. In some cases, the only patents listed on the product were on the delivery devices. Boehringer Ingelheim, for example, listed 25 patents on Combivent Respimat (albuterol-ipratropium) when the drug was approved in 2011, every single one of which was on the delivery device. 22 This product contains albuterol, first approved in 1981, and ipratropium, first approved in 1986, and yet still faces no generic competition.
Many of these device patents were entirely disconnected from the drugs on which they were listed, making no mention of active pharmaceutical compounds, therapeutic classes, methods of use, or other features in their claims that might connect the patent to the drug. Companies are only permitted to list patents in the Orange Book that cover a drug (either a drug substance or drug product) or method of use. 23 A 2020 court case found that Sanofi had improperly listed a patent on the drive mechanism of an insulin injector pen because the patent made no mention (much less claimed) insulin or even the indication (diabetes). 24 In recent work, we applied a similar logic to analyze inhaler device patents and found that 77% made no mention of any active ingredients or molecular structures; 72% made no mention of any active ingredients, molecular structures, therapeutic classes, indications, device names, or the lungs. 25 Twenty-five inhalers had a device-only patent as the last-to-expire patent (with no mention of active ingredients), and these device-only patents extended periods of expected protection by a median of 8 years beyond other patents.
The FTC has recently identified hundreds of patents that have been impermissibly listed in the Orange Book. 26 The majority of these patents are device-only patents listed on inhalers for asthma and COPD. These patents increase the density of patent portfolios, which make challenges more difficult, and increase the duration of protection on the products they cover, which make challenges necessary in the first place. Indeed, a recent analysis of litigation following paragraph IV challenges on inhalers found that more than half of all litigated patents were on delivery devices.Reference Reddy 27 Brand-name firms are not just listing device patents, but they are building litigation strategies around these patents to keep generic competitors off the market.
Inhaler manufacturers have also engaged in extensive device hopping — or shifting active ingredients from one delivery device to another with new patent protection. Manufacturers of 15 different originator inhalers approved from 1986 to 2020 pursued this strategy, leading to 19 follow-ons. Some of these device hops were related to a ban on chlorofluorocarbons (CFC), which went into effect for pharmaceutical products from 2009 to 2013.Reference Wouters, Feldman and Tu 28 However, many of these device hops were unrelated to the ban. For example, GlaxoSmithKline released 5 different fluticasone products from 1996 to 2014: CFC-containing Flovent (1996) and then CFC-free Flovent Rotadisk (1988), Flovent Diskus (2000), Flovent HFA (2004), and Arnuity Ellipta (2014). Boehringer Ingelheim released two different CFC-free tiotropium inhalers: Spiriva HandiHaler (2004) and Spiriva Respimat (2014). By moving molecules to delivery devices with new patents, manufacturers can extend streams of revenue. The median duration of expected protection on inhalers with device hops was 28 years for the product lines (measured from approval of the first product to the last-to-expire patent listed for follow-ons). 29
These practices have been lucrative for inhaler manufacturers. From 2000 to 2021, manufacturers earned more than $178 billion on inhalers in the US, including $67 billion (38%) when patents on active ingredients were active, and $111 billion (62%) when these patents had expired.Reference Feldman 30 Some of the biggest blockbuster inhalers have earned staggering sums when only secondary and tertiary patents were active: Advair (fluticasone-salmeterol): $42 billion; Symbicort (budesonide-formoterol): $16 billion; Flovent (fluticasone): $12 billion; ProAir (albuterol): $8 billion; Spiriva (tiotropium): $5 billion. 31 In some cases, including AstraZeneca’s Symbicort and Teva’s ProAir, all manufacturer revenue accrued after primary patent expiration, as these patents had expired before product launch.
Without legislative reform, these practices are likely to continue. The Senate Committee on Health, Education, Labor, and Pensions launched an investigation in January 2024 examining potentially anticompetitive practices by inhaler manufacturers over the last several decades.Reference Tirrell 32 This work builds on important action taken by the FTC to crack down on impermissibly listed device patents on inhalers. In response, 3 of the 4 major inhaler manufacturers announced that they would voluntarily cap out-of-pocket costs at $35.Reference Becker 33 While this is an important step for millions of Americans with asthma and COPD, it does not address the underlying problems that have limited generic competition, nor will it necessarily translate into less spending by payers as long as the same harmful patenting practices can continue.
b. GLP-1 receptor agonists for diabetes, weight loss, and cardiovascular disease
GLP-1 receptor agonists have shown remarkable promise for the treatment of diabetes, weight loss, and cardiovascular disease. But these drugs remain out of reach for many patients. List prices can run more than $1,000 per month and, even with large manufacturer rebates for payers,Reference Hernandez and Sullivan 34 net prices are still substantially higher than in other countries, 35 raising alarms about looming unmanageable budgetary impacts for US payers, including Medicare.
GLP-1 manufacturers appear to be relying on a similar playbook as inhaler manufacturers — establishing large thickets of patents, many of which are centered on delivery devices. This should not be surprising, because the two largest GLP-1 manufacturers, Novo Nordisk and Eli Lilly, used the very same strategies on insulin pens over the last 3 decades.Reference Olsen 36 Approximately half of all patents on insulin products approved from 1986 to 2019 were on the delivery devices of these products, and 85% of these patents made no mention of insulin in their claims. 35 The FTC has not targeted delivery device patents on insulin products because they are no longer listed in the Orange Book (they are now regulated as a biologics rather than small-molecule drugs).
Recent analyses of GLP-1 receptor agonists point to several problematic trends.Reference Olsen 36 Among 10 GLP-1 receptor agonists approved from 2005 to 2021, manufacturers listed a median of 20 patents per product, including a median of 17 before FDA approval and 2 added after FDA approval. The median period of expected protection on each product was more than 18 years. Because many of these products were recently approved, increasing numbers of patents are likely to be added in the coming years absent regulatory reform. Overall, manufacturers listed 188 patents on GLP-1 drug-device combinations, including 107 (57%) on the delivery devices. Not a single device patent on GLP-1 receptor agonists mentioned their active ingredients, molecular structures, therapeutic class, or device name in their claims. Only 1 patent (listed on 3 separate products) mentioned an indication (diabetes). In April 2024, the FTC announced a second round of improperly listed patents, including numerous device-only patents on GLP-1 receptor agonists. 37
Although many GLP-1 receptor agonists have only recently been approved, generic firms had filed at least 30 paragraph IV challenges on 5 different products through the end of 2023.Reference Alhiary 38 Brand-name firms sued for infringement in these cases on 70 patents, 40 of which (57%) were device-only patents that should likely not have been listed in the Orange Book in the first place. As with inhalers, device patents form a core part of litigation strategies that brand-name firms employ to delay generic competition.
c. Other pharmaceutical products
Although drug-device combinations are perhaps the most prominent exemplars of how pharmaceutical patenting practices have gone astray in the US health care system, brand-name firms employ similar strategies on other types of products. Research has found that the number of patents per new drug approval increased by more than two-thirds from 2000 to 2015. 39 Manufacturers obtain a wide variety of secondary patents to construct their patent portfolios, ranging from patents on formulations and inactive ingredients to methods of useReference Tu and Sarpatwari 40 and safety monitoring strategies.Reference Sarpatwari 41
Many of these secondary patents are continuations with terminal disclaimers (stipulations that these new patents will expire at the same time as original patents). Although such patents do not extend periods of market exclusivity, they increase the density of patent portfolios. Among brand-name small-molecule drugs approved from 2000 to 2015, 47% (1,857) were continuation patents. 42 While the number of original patents per drug listed in the Orange Book remained steady from 2000 to 2015 (increasing just 15% from 1.3 to 1.5), the number of continuation patents per drug increased by 200% from 0.6 for drugs approved in 2000 to 1.8 for those approved in 2015.
In addition to obtaining patents on peripheral features of pharmaceutical products, many of which are continuations, companies also develop patent thickets over time in a way that can create uncertainty for generic manufacturers and biosimilar firms. One such strategy is filing patents after FDA approval. In an analysis of the 10 top-selling prescription drugs in 2021, 72% (1,028 of 1,429) of patent applications were filed after a drug received FDA approval. For small molecule drugs, 58% (299 of 517) were filed after approval compared to 80% (729 of 912) for biologics. 43 Manufacturers can time these patents so that they are issued just as FDA-granted regulatory exclusivities are set to expire, forcing generic firms to further design around these patents or file new challenges. A key strategy for brand-name firms to control the timing of their patents is via the Track One prioritized review pathway, which allows manufacturers to receive expedited review at the USPTO (within one year of receiving this status). From 2011 to 2022, 27% (1,350 of 5,085) of all patents approved on small molecule drugs listed in the Orange Book had been granted Track One status, 86% (1,155 of 1,350) of which were continuations, and companies often obtained these patents just as regulatory exclusivities were expiring.Reference Tu and Feldman 44
4. Proposals for Reform
Addressing the problems of patent thickets and product hops across the pharmaceutical system will require a multipronged strategy that includes the FDA, USPTO, and FTC. The Senate Judiciary Committee moved three key bipartisan bills out of Committee during the last legislative session aimed at facilitating more timely competition. A modified version of the first bill passed the Senate by unanimous vote in July 2024 but was not ultimately taken up by the House during the legislative session; the other two were not brought for a floor vote in the Senate.Reference Silverman and Zhang 45 Although these bills — new versions of which could be taken up and passed by the current Congress — represent important advances, further reforms are needed to address the underlying problems of patent thickets and product hops in the US pharmaceutical system.
a. Three bipartisan bills
The Affordable Prescriptions for Patients Act of 2023 (S. 150, sponsored by Senator John Cornyn, R-TX), was first introduced in 2019 with more robust provisions than in later versions. In its original form, the bill applied to both small-molecule drugs and biologics and sought to tackle the problems of both patent thickets and product hops. The bill would have limited patent filings and reformulations during periods when generic or biosimilar competitors were about to launch and would have deterred some of the most problematic patenting strategies. In particular, the bill offered a broad definition of “patent thickets” — which included the common practice of filing patents on a drug after approval in the same family as existing patents — and considered patent thickets to represent an unfair method of competition absent compelling evidence to the contrary.
However, in the face of pharmaceutical industry pushback, lawmakers substantially watered down the bill before clearing the Senate Judiciary Committee and, again, before passing the full Senate. 46 In particular, recent versions of the bill exclude small-molecule drugs altogether, focusing instead only on biologics, and remove all provisions related to product hops. In addition, although the bill poses a cap on the number of patents that a brand-name biologic firm can seek to enforce against biosimilar competitors under federal code (allowing litigation on a maximum 20 patents), it grants exceptions that would allow manufacturers to file lawsuits relating to numerous patents above the cap. The Congressional Budget Office estimated that this bill will save taxpayers $1.8 billion over 10 years. 47 While a positive step forward, the bill in its current form would not address the underlying problems identified above.
The Interagency Patent Coordination and Improvement Act 2023 (S. 79, sponsored by Senator Richard Durbin, D-IL) would establish a formal task force to improve communication between the USPTO and FDA. Better communication could, among other things, help ensure that patentability standards are met at the USPTO and prevent invalid patents from being issued in the first place. Once issued, a patent is legally presumed to be valid and is therefore challenging to overturn in court, which underscores the importance of ensuring that the correct standards are applied in the first place. Key to the work of the task force would be identifying concrete, actionable strategies for collaboration to limit gaming of the patent system.
The Stop Significant and Time-wasting Abuse Limiting Legitimate Innovation of New Generics (Stop STALLING) Act (S. 148, sponsored by Senator Amy Klobuchar, D-MN) would fine firms that bring baseless citizen petitions up to $50,000 each day that the FDA spends reviewing the petition. Prior research has shown that brand-name firms often file citizen petitions to delay FDA review of generic drug applications, and the vast majority of these petitions are denied.Reference Carrier and Wander 48 By deterring frivolous citizen petitions, this bill could help the FDA approve generic drug applications in a more expeditious manner.
b. Other proposals
The Senate Judiciary Committee should be applauded for developing legislation to facilitate free market competition, but further action is needed. First, Congress should require manufacturers that submit patents for listing in the Orange Book to simultaneously submit these patents for reexamination by the USPTO. Given the stakes for the health care system of invalid patents, routine reexamination would help ensure that only truly patentable innovations can appear in the Orange Book.Reference Tu 49 Teams of seasoned patent examiners could reevaluate patents in light of new clinical data generated between patent issuance and FDA approval. Given that manufacturers may have incentive to downplay prior art to the USPTO (in seeking to demonstrate innovativeness) while potentially overstating prior art to the FDA (in seeking to demonstrate safety and effectiveness), enhanced transparency and data sharing could also help ensure more consistency between material provided to the USPTO and FDA. 50
Second, because not all patentable inventions are suitable for listing in the Orange Book (e.g., certain device-only patents), Congress should require the FDA to release more comprehensive guidance on the types of patents that can be listed and give the FDA the resources and authority to review all submitted patents to determine eligibility for listing in the Orange Book. The FDA has expressed reluctance to take on this role given limited patent expertise and few resources to conduct the work, 51 but an expanded role may be feasible with appropriate funding and close collaboration with the USPTO. The scope of such oversight would be limited since only a small number of patents appear in the Orange Book (with approximately 10,000 listed over the past 40 years compared to millions granted by the USPTO). 52 Improper patent listings in the Orange Book create unnecessary costs and delay generic competition; new enforcement responsibilities are needed to keep pharmaceutical firms from continuing to stretch the boundaries envisioned by the Hatch-Waxman Act.
Third, Congress should limit the number of patents that brand-name firms can assert when suing for infringement following patent challenges. The Affordable Prescriptions for Patients Act of 2023 sets a cap for biologic drugs (with numerous exceptions, as described above). A better approach, offered in recent legislation sponsored by Senator Peter Welch (D-VT), would also apply to small-molecule drugs and limit litigation to one patent per family among patents joined by terminal disclaimers.Reference Welch 53 Efforts like these, which serve to decrease the number of patents subject to litigation, could help reduce the costs and time for generic competitors to reach the market.
Fourth, to increase incentives for generic firms to file paragraph IV certifications on complex drugs, Congress should increase the 180-day exclusivity period awarded to first-to-file generic firms. Generic manufacturers of products like inhalers and other drug-device combinations not only must contend with large patent thickets when seeking FDA approval (including patents on the delivery devices), but they must also perform more extensive testing (including, in some cases, clinical trials) than is required for simpler drugs like oral tablets. Perhaps the rewards for generic firms that take on these risks should be higher for complex products. By the same token, to decrease the incentives for brand-name firms to file baseless litigation following paragraph IV certifications, Congress should reduce the length of automatic 30-month stays.
Finally, Congress should grant the FDA more authority to approve complex generic drugs like drug-device combinations that differ in slight ways from brand-name reference products containing the same active pharmaceutical compounds. This would enable generic firms to more easily design products that avoid infringing brand-name patents. Congress should couple such a law with a requirement that the FDA review post-marketing data on outcomes to further ensure clinical comparability between the generic and brand-name drugs.
5. Conclusions
Over the past several decades, brand-name pharmaceutical firms have developed a variety of strategies aimed at limiting generic and biosimilar competition to the detriment of patients. Chief among these strategies are patent thickets and product hops, which have proven particularly pernicious for drug-device combinations like inhalers and GLP-1 injector pens. Legislative measures are needed to help prevent these strategies and ensure that patients have timely access to low-cost therapies.
Acknowledgment
An earlier version of this paper was submitted as written testimony before the Senate Judiciary Committee in 2024 as part of a hearing on pharmaceutical patents. Dr. Feldman has received funding for his work in this area from the Commonwealth Fund, Arnold Ventures, the National Institutes of Health (K08HL163246) and the Food and Drug Administration.
Disclosures
Dr. Feldman has served as a consultant for Alosa Health and as an expert witness in litigation against inhaler manufacturers.
