The patent system has been dragged into a decades-long debate over drug prices. The Biden administration, congressional leaders, policy activists, and academics have charged patents as a primary cause of “rising drug prices.”1 They allege that drug innovators acquire too many patents to create “patent thickets” that block competition from generic drug manufacturers or engage in “evergreening” by wrongly extending exclusive patent protection with these additional, improper patents.2 These accusations have prompted congressional leaders to propose legislation limiting drug patents, the U.S. Patent and Trademark Office (USPTO) to consider new regulatory restrictions on drug patents, and antitrust lawsuits against drug companies.3
A significant development in this patent-focused policy debate about drug pricing was the release of a long-awaited USPTO empirical study, Drug Patent and Exclusivity Study.4 The USPTO Studyis the first agency report to provide objective, empirical analysis of the number of patents covering drugs and the length of patent protections for the innovative drugs that have extended lifespans and increased the quality of life for people worldwide. Thus far, the policy debate about the connection between drug patents and drug prices has been based on information published by a policy organization known as the Initiative for Medicines, Access, and Knowledge (I-MAK). I-MAK has a dog in this policy fight, as it believes that a “root cause of the high cost of medicines” is “unjust patent monopolies.”5 Thus, many scholars have raised serious concerns about the veracity and reliability of its drug patent numbers and its related claims about “patent monopolies.”
The USPTO Study confirms the data on drug patents that I reported in a 2022 Hudson Institute essay, How Unreliable Data Have Infected the Policy Debates Over Drug Patents.6 The USPTO Study could not replicate the drug patent numbers published by I-MAK in splashy brochures in 2018 and 2019.7 The specific sources of these numbers and the specific methodology used by I-MAK to reach them remain shrouded in mystery, as I identified in my 2022 essay and despite a direct request by Senator Thom Tillis (R-NC) that I-MAK publicly disclose the specific data and methodology of analysis underlying its 2018 and 2019 brochures.
This policy memo summarizes the findings of the USPTO Study on the number of patents and the length of exclusivity periods for the drugs that I discussed in my 2022 essay. It first describes the USPTO’s findings concerning the specific number of patents covering the blockbuster drugs I reviewed in 2022 in a “spot check” of the I-MAK patent numbers. The USPTO’s data is consistent with the data on drug patents that I reported; notably, both my data and the data in the USPTO Study conflict by orders of magnitude with the large numbers of patents claimed to exist on these same drugs by I-MAK. Second, it describes the USPTO’s excellent analysis of the key failings in simple patent counting as a basis for proposals in innovation policy, even when these counting methods are done correctly and with reliable and replicable data. Third, it details the USPTO’s analysis of the average length of “market exclusivity”—the time period in which a drug patent owner can exclude competition from a generic drug company. The USPTO found an average period of market exclusivity of between 3–16 years, which is consistent with numerous rigorous studies by economists and other scholars. Again, this dramatically conflicts with I-MAK’s claims of many decades of market exclusivity for the “unjust patent monopolies” on new drugs.
In sum, the USPTO’s findings and analyses of drug patents not only fail to confirm I-MAK’s drug patent numbers and market exclusivity claims, but it also directly contradicts these claims. This is important because I-MAK’s patent numbers and market exclusivity claims have been “relied upon by those advocating for patent-based solutions to drug pricing” in Congress, among policymakers, and by academics.8 The USPTO Study should become the primary reference point for all data in the legal and policy debates concerning drug patents and drug pricing. It will ensure evidence-based policymaking in patent law, innovation policy, and healthcare policy.
The Genesis of the USPTO Drug Patent and Exclusivity Study
Since the core allegation of “patent thickets” and “evergreening” is that drug innovators acquire too many patents—either to create an impenetrable thicket to stop competitors or to extend exclusive protections for a drug—the argument requires patent numbers to back it up. As I explained in my 2022 essay, I-MAK has been the primary source of the patent numbers behind these allegations—producing self-styled “reports” with eye-popping numbers of patents that cover blockbuster drugs.9 For example, I-MAK asserted in 2018 that Lyrica, Pfizer’s pathbreaking drug to treat nerve damage, had a whopping 68 patents covering this drug. I-MAK also claimed that these 68 patents artificially extended Pfizer’s “monopoly”—its market exclusivity—in the drug until 2038, approximately 20 years after the primary patent on Lyrica had expired.10
The problem was that the I-MAK patent numbers did not add up. There were massive discrepancies between I-MAK’s patent numbers and the patents listed in the Orange Book published by the Food and Drug Administration (FDA). The FDA’s Orange Book definitively identifies all relevant patents covering a drug for any company seeking to make a generic version of this drug.11 Instead of the 68 patents alleged by I-MAK to cover Lyrica, for example, the FDA identifies a total of three patents in the Orange Book.12 This was not a discrepancy that fell within any sense of a margin of error; it was a difference on the scale of orders of magnitude.
The inconsistencies between I-MAK’s patent numbers and the Orange Book’s drug patent listings raised concerns in Congress. The USPTO Study was undertaken at the behest of Senator Thom Tillis (R-NC), who requested that the USPTO and FDA investigate the claims by I-MAK following the release of my 2022 essay and some other reports raising similar questions about I-MAK’s patent numbers. Senator Tillis wrote to the USPTO and FDA on January 31, 2022, asking that the agencies undertake a study of the patents covering the drugs reported in I-MAK’s brochures.13 “In order to ensure an objective, measured, and appropriate approach to [the important goal of lowering drug prices],” wrote Senator Tillis, “it is fundamental that assumptions and premises be based on accurate facts and data from reliable, unbiased sources.”14
On the same date, Senator Tillis also wrote separately to Tahir Amin, co-founder and co-executive director of I-MAK.15 In this letter, Senator Tillis requested that Mr. Amin provide a “detailed explanation of your methodology for calculating the number of patents on a drug product that could be replicable by other researchers.”16 He also requested that I-MAK explain why its patent numbers covering drugs “differ so dramatically from public sources,” as well as explain why I-MAK claimed in its 2018 and 2019 brochures that some pharmaceutical companies will retain market exclusivity for many decades into the future when generic versions of these drugs were already being prescribed to patients in the healthcare market.17
In a lengthy response letter submitted more than a month later, Mr. Amin declined to answer Senator Tillis’ requests for disclosure of its specific data and counting methodology.18 Instead, Mr. Amin claimed that the Orange Book was not a valid source for the numbers of drug patents.19 Mr. Amin also decried “the hidden real-world workings of the industry when it comes to patents” and argued that the “system has deliberately been kept opaque by the pharmaceutical industry.”20 It is notable that, in this same letter, Mr. Amin chose to keep “opaque” and “hidden” the data and counting methodology used by I-MAK to reach its colossal patent numbers in its brochures.21
On April 1, 2022, Senator Tillis again wrote to the USPTO and FDA about conducting an impartial drug patent study given that this topic fell under the jurisdiction of both expert agencies.22 He reiterated that he was “specifically concerned about work from the Initiative for Medicines, Access & Knowledge (I-MAK),” and noted Mr. Amin’s refusal to disclose I-MAK’s data and counting methodology.23 He described further concerns about the unreliability of I-MAK’s patent numbers given indications that it “conflate[d] pending and abandoned patent applications with granted patents,” as well as similar concerns about the unreliability of the lengthy periods of market exclusivity alleged by I-MAK to exist given its claim of evergreening practices by drug innovators.24 Senator Tillis thus reiterated his request for the USPTO and FDA to “conduct an independent assessment and study” in order “to ensure that policymaking in this critical area is based on accurate, reliable, and replicable facts and evidence.”25
The Myth of “Patent Thickets” for Drugs
The USPTO Study states that its charge is to “conduct an independent assessment and analysis of certain sources and data that are being relied upon by those advocating for patent-based solutions to drug pricing.”26 These “certain sources and data” refer to I-MAK.27 Since it engaged in its own independent and objective assessment of the patents covering the top drugs, as requested by Senator Tillis, the USPTO acknowledged that it did not “specifically assess [I-MAK’s] data or analysis” in the “sources” identified by Senator Tillis.28 As the USPTO Study made clear, though, there was significant overlap in the drug patents independently studied by the USPTO with the assistance of the FDA and those identified in I-MAK’s annual brochures alleging that evergreening and patent thickets are causing higher drug prices.29
In his April 1, 2022, letter, Senator Tillis emphasized that the purpose of the USPTO Study is to ensure “accurate, reliable, and replicable facts” in the continuing policy debate over drug prices.30 In the sciences, the ability to replicate the findings of a study is one of the key indicators of accuracy and reliability. Thus, the primary question raised by the USPTO Study is its ability to replicate the drug patent numbers reported by I-MAK in its brochures.
The USPTO’s answers to this question are as striking as the comparison between the I-MAK patent numbers and FDA Orange Book drug patents that I reported in my 2022 essay:
Table 1. Comparison of Patent Numbers
Source: I-MAK, USPTO, and author.
I-MAK alleged in 2018 that 68 patents covered Lyrica. In my 2022 essay, I identified three patents and one reissue patent covering Lyrica. The USPTO Study similarly found three patents and one reissue patent covering Lyrica.31 The USPTO Study emphasizes that “a reissue patent replaces the original patent and does not prolong the duration of the original patent grant,”32 a point I likewise made in my 2022 essay. Thus, the total number of patents covering Lyrica is three patents.
I-MAK similarly alleged large patent numbers covering Eliquis, a drug that reduces blood clots after surgery. In 2018, I-MAK claimed that 27 patents covered Eliquis and, in 2019, I-MAK increased the total number of patents covering Eliquis to 31 (with no explanation for the increase). In my 2022 essay, I identified three total patents covering Eliquis. Again, the USPTO Study confirms the same drug patent numbers: three patents cover Eliquis.33
Lastly, I-MAK alleged in 2018 that 30 patents covered Xarelto, a drug that also reduces blood clots. In 2019, I-MAK increased the total number of patents covering Xarelto to 32 (with no explanation for the increase). In my 2022 essay, I identified six patents covering Xarelto.34 In this case, there is a slight difference between my findings and the USPTO Study, which reports only five patents covering Xarelto,35 as opposed to the six patents I found. But this is not a difference in orders of magnitude of the sort between the I-MAK numbers and the USPTO Study numbers.
In sum, the USPTO Study does not replicate I-MAK’s drug patent numbers based on the USPTO’s analysis of the public, official sources of drug patents, such as the FDA’s Orange Book. Yet, the USPTO Study does confirm the drug patents numbers I reported in my 2022 essay based on my own “spot check” of the I-MAK drug patent numbers. The conclusion is definitive and clear: I-MAK’s drug patent numbers are not “accurate, reliable, and replicable facts.”36
The Unreliability of Patent Counts (Even When Done Correctly)
Even if I-MAK had produced its data on the specific patents underlying its total patent numbers in its 2018 and 2019 brochures or explained its specific methodology in how it counted these patents, the USPTO Study further explains why simple patent counting of this sort is not a valid metric—even if it were done correctly and could be confirmed through replication. This is unsurprising, as economists have long recognized that there are significant limitations in the use of patent counts as a measure of output.37 Conversely, the same limitations exist in the use of patent counts in measuring market exclusivity for particular consumer products or services. Thus, policymakers should not rely on patent counts, even if such counts are done accurately and reliably.
As a preliminary matter, there is nothing significant in identifying more than one patent covering any specific product or service. The existence of multiple patents covering a single consumer product is not unusual or unexpected in the innovation economy. Nor is it even something new in the modern era. The foundational court case establishing that damages for infringement of a patent covering only one component of a multi-component product must be apportioned to the value contributed by only that patented invention was decided by the U.S. Supreme Court in 1853.38
Today, innumerable patents—some reliable estimates suggest it is thousands of patents—cover our ubiquitous digital devices, such as laptop computers, smartphones, and tablets. Even age-old consumer products are now covered by thousands of patents. Modern automobiles, for example, have thousands of patents covering varying aspects of their design, including the brake pedal mechanism, engine components, as well as the digital hardware and software that now control the vehicle’s operation, especially the 5G mobile capabilities in our “connected cars.”39 The USPTO Study recognizes the basic truism of the innovation economy that “multiple patents associated with a single marketed product are not unique to the pharmaceutical industry and are a common practice in many innovative industries, especially for complex products.”40
Beyond the simple fact that it is very common for multiple patents to cover a single product, especially technically complex products such as biotech-engineered drugs or mobile devices, patents also vary in scope, content, and even in their formal structure. All of this affects the technological value of any specific patent and delimits its exclusive effects. Again, this has been long recognized by economists studying patents and innovation. A 2020 report published by the Hudson Institute, 5G Technological Leadership, critiqued a similar practice of counting patents as a measure of contributions of technological innovation to our digital cellular devices.41 The report states that “numerous empirical studies have shown that the value of patents varies enormously, i.e., a small percentage of patents cover technology of great importance and other patents cover technologies of more marginal value.”42 Simple patent counts do not properly measure these massive variations between the technological innovation, value, and even form of specific patents.
The USPTO Study rightly recognizes that these insights in measuring innovation—or conversely exclusive effects—apply with equal merit to the multiple patents covering drugs. The USPTO states that “simple counts of patents can be misleading when every patent is counted equally, because the number of patents does not provide a clear picture of the landscape without a review of the scope of the claims in each patent.”43 Even if one might believe that the five patents covering Xarelto that are listed in the FDA Orange Book are too many patents, the straightforward counting of these patents tells us nothing about what innovations they secure and how they secure them.
Unreliable Patent Counting in the Evergreening Critique
The inability to infer any definitive empirical insights about innovation or exclusivity from simple patent counts, even if done correctly with reliable and replicable results, leads to the logical conclusion that one cannot infer the scope or temporal length of exclusivity from the number of patents that cover different elements of a product, such as a drug.
The USPTO Study observes that:
Simply quantifying raw numbers of patents and exclusivities is an imprecise way to measure the intellectual property landscape of a drug product because not every patent or exclusivity has the same scope. For example, one patent could contain different sets of claims directed to: (1) a pharmaceutical product, (2) a method of using the product, and (3) a process for manufacturing the product.44
Thus, the USPTO Study states that “patent expiration dates, like the number of patents, may not be predictive of the timing of actual launch of competing products.”45 In sum, simply counting patents and adding up the length of these patent terms without considering the nature of the patents being counted produce unreliable results, at best, or inaccurate results, at worst. “In some instances,” as the USPTO Study points out, “a higher number of patents may not necessarily delay a generic launch,” such as in the case Revlimid (one of the drugs the USPTO studied).46
This is confirmed again in I-MAK’s incorrect claims about lengthy periods of market exclusivity based on its oversimplistic counting of patents and their term lengths. This of course assumes I-MAK’s patent numbers are accurate, which is now an unjustified assumption. In its independent review of the relevant drugs, the USPTO Study recognizes, “For several [drugs] studied, generics were launched before the expiration of all applicable Orange Book-listed patent and exclusivity time periods for that [drug].”47
In my 2022 essay, for example, I explained that I-MAK claimed in its 2018 brochure that Pfizer would retain market exclusivity in Lyrica until 2038—more than 20 years after the publication of I-MAK’s brochure and 11 years after the expiration of the final patent listed for the drug in the FDA Orange Book.48 However, the main patent on the active ingredient for Lyrica expired in 2018, and the FDA approved nine generic versions for Lyrica in 2019 with a highly publicized entry into the market by generic drug manufacturer Amneal Pharmaceuticals.49 All of this occurred in the years I-MAK was publishing its brochures in 2018 and 2019 in which I-MAK claimed there would be no competition at all for Lyrica in the healthcare market for at least another 20 years. This is a dramatic example of how simplistic patent counting is fundamentally unreliable, even if it is not infected with overwhelming questions of invalidity and the inability to replicate its result.
The “Evergreening” Myth
Given the deeply troubling issues with the inability to replicate I-MAK’s patent counts, as well as the fundamental problems with the unreliability of patent counting even when it is done correctly, it is unsurprising that the USPTO Study exposes the “evergreening” critique as a myth. I-MAK and other drug patent critics have long accused drug innovators of wrongly extending their “drug patent monopolies” (to use I-MAK’s language) or market exclusivities far beyond the 20-year term on the active ingredient in a drug.50
As noted above, for instance, I-MAK accused Pfizer of extending exclusive “monopoly” protections for Lyrica at least several decades past the end of its patent term on its active ingredient. But given that patents vary qualitatively—the nature of the invention, scope of invention, and the structure of the patent protection, among other variables—it is untrue that anyone can merely count patents and conclude that the last patent that expires in a set of patents listed in the Orange Book that covers a drug is the temporal length of full market exclusivity.
In its independent review of drug patent protections, the USPTO Study finds that market exclusivity ranges “from about 3 to about 16 years.”51 Notably, this is less than the 20-year term that is legally secured in every patent. The USPTO’s finding on average years of market exclusivity is consistent with numerous statistical studies of drug patents published over more than a decade finding average market exclusivity periods of approximately 12–13 years. These rigorous studies have obeyed scholarly norms in data collection and statistical analyses, such as disclosing their methods of analysis and making their data available for testing and replication by other scholars.
Last year, Professor Erika Lietzan and Dr. Kristina Acri published the results of a study of average market exclusivity periods for 225 new drug applications at the FDA, and they found the average length of time before generic entry to be only 11.3 years.52 An earlier study by Professor Lietzan and Dr. Acri of 227 drugs that had received patent term extensions under the patent laws between 1984 and 2018 had an average period of market exclusivity of only 12.62 years.53
These numbers comport with other studies. For instance, a scholar in 2021 examined 370 drugs and found an average period of market exclusivity of only 13.3 years before there was market entry of a competing generic drug.54 A study published in 2019 of 170 top-selling drugs found an average period of market exclusivity of only 13.75 years for drugs that had received patent term extensions under the patent laws, and only 10 years of market exclusivity for drugs that were ineligible for patent term extensions.55 Another study published in 2015 found an average period of market exclusivity of only 12.5 years for 175 new drugs.56 Lastly, a study by Professor Scott Hemphill and Dr. Bhaven Sampat found an average period of market exclusivity of 12.1 years before generic entry for a set of 119 drugs that were approved by the FDA between 2001 and 2010.57
In sum, there is no evidence of systemic manipulation of the patent system by drug innovators—they are not engaging in “evergreening” to wrongly extend periods of market exclusivity beyond the 20-year patent term. Average periods of market exclusivity for new drugs, even drugs covered by patents that have legitimately had their terms extended due to delays caused by the FDA safety and efficacy testing system, are far less than the standard patent term of 20 years. On average, around 50 percent or less.
The USPTO Study confirms what numerous rigorous studies by economists and scholars have consistently found: accusations of “evergreening” are rhetoric employed by activists like I-MAK to create a moral panic in Congress and agencies in order to prompt legal actions to restrict or eliminate drug patents. The “evergreening” charge is based on policy-based evidence-making, it is not evidence-based policymaking. The USPTO Study confirms this fact.
Conclusion
For many years, I-MAK has created massive patent numbers and market exclusivity periods running many decades longer than patent terms of 20 years. It has done this to reduce a complex set of legal, regulatory, and economic determinants of drug prices down to an overly simplistic, sound-bite boogeyman: the patent system. When I-MAK’s patent numbers and market exclusivity periods were objectively assessed by the USPTO with the assistance of the FDA, the USPTO neither replicated nor confirmed any of I-MAK’s patent numbers or exclusivity periods, such as those claimed in its 2018 and 2019 brochures. I-MAK’s patent numbers are now contradicted by an official agency report and by other public, official sources of data on drug patents. I-MAK’s drug patent numbers are not “accurate, reliable, and replicable facts.”58 To enact laws or regulatory actions based on I-MAK claims is not evidence-based policymaking. Congress, the USPTO, and other agencies like the FDA or Federal Trade Commission should not consider any new laws or policies based on the proverbial fruit of this poisonous tree.