Whereas nearly 2500 medicines have been approved by the FDA or marketed in the United States, a comparatively small number have been withdrawn from the market.
As part of its mission to provide transparency about the pharmaceutical industry, CRIB is initiating a program to catalog medicines withdrawn from the market and to assess the reason(s) for this action.
We are analyzing Congressional records and scientific sources to identify trends in drug withdrawals. We are always actively seeking collaborations with individuals interests in these questions, who can reach out to us at <this link>.
Below is a list of notable drugs withdrawn by the U.S. Food and Drug Administration (FDA): 1. Thalidomide (1957) - Withdrawn due to severe birth defects. 2. Phenylpropanolamine (2000) - Withdrawn due to increased risk of hemorrhagic stroke. 3. Rofecoxib (Vioxx) (2004) - Withdrawn due to increased risk of heart attacks and strokes. 4. Sibutramine (Meridia) (2010) - Withdrawn due to increased risk of heart attacks and strokes. 5. Rosiglitazone (Avandia) (2010) - Restricted use due to increased risk of heart attacks. 6. Propoxyphene (Darvon) (2010) - Withdrawn due to serious heart rhythm abnormalities. 7. Dextropropoxyphene (Darvocet) (2010) - Withdrawn due to serious heart rhythm abnormalities. 8. Lorcaserin (Belviq) (2020) - Withdrawn due to increased risk of cancer. 9. Ranitidine (Zantac) (2020) - Withdrawn due to contamination with a probable human carcinogen. 10. Hydroxychloroquine (2020) - Emergency use authorization revoked for COVID-19 treatment due to lack of efficacy and potential risks.
Moving forward, we will distinguish withdrawal decisions that are based upon toxicity versus withdrawals that reflect market inefficiencies (e.g., a product rendered obsolete by a superior competitor).
Our ability to develop future medicines is increasingly a sub- ject of considerable concern. This notion might seem absurd given that our knowledge of health and disease continues to grow exponentially. Yet an inability to understand causes of disease are not the limitations that threaten drug develop- ment. Rather, the challenge is our ability to span the divide between fundamental knowledge and its application. Medi- cines are amongst the most highly regulated products with development costs measured in billions of dollars and more than a decade of hard work [1, 2]. Moreover, our efficiency in developing new medicines has been declining persistently for at least 75 years [3].
This problem is well known and has acquired several monikers, including the “Valley of Death” and the “Great Divide.” Arguably the best descriptor is Eroom’s Law, a playful inversion of the Moore’s law of computing [4]. This designation contrasts the consistent improvements in the efficiency of computer processing speed, as popularized by Gilbert Moore (of Moore’s Law fame).
The inversion of the surname reflect the fact that whereas the efficiency of computing speed has consistently improved over time, the efficiency of drug discovery has persistently declined [5]. Indeed, the research that triggered the naming of Eroom’s Law revealed that the cost to develop a new medicine has been increasing at an exponential rate since at least the 1950s [3]. Analyses of the sources of pharmaceutical innovation suggest that the pharmaceutical industry began reacting to Eroom’s law in the mid-1970s.
These concerns com- pelled many attempts to circumvent the law, all of which have proven ineffective or counterproductive. The attempts include emphasis upon me-too drugs (fast followers), embracing high-margin niche markets (e.g., orphan indica- tions and oncology) and avoiding perceived low margin or demand products, (e.g., antibiotics) (Fig. 1a).
The example of antibiotics is particularly relevant given that market forces diminished the value of antibiotics just as microbial resistance became commonplace. The experience of the drug vancomycin reveals the challenge. Vancomycin was sufficiently innovative that the drug was literally and figuratively put on a shelf to limit its overuse. This was a highly rational public health decision given the abuses of antibiotics, which have compromised their long-term usefulness.
Consequently, the organization that developed van- comycin would not realize its value. By the time that the product would indeed become more widely deployed, its patent protection had expired, and the profits were realized by generic manufacturers, not by the originator. Unsurpris- ingly, the originator abandoned further antibiotic research, a decision that was mimicked by other companies in the years following.
As such, the saga of vancomycin provided a valuable lesson taught in many a corporate board room for years to come.
As part of an ongoing effort to assess drug development activities in 2022, we contin- ued an evaluation of the pharmaceutical and biotechnology industries. We included information gathered from both the CDER and CBER. All underlying data, as well as much more information pertain- ing to these active pharmaceutical ingredients (APIs), including previously approved and clinical-stage drugs (never or not yet approved), can be accessed on the public website www.cdek.liu.edu and was curated consistent with our previous analyses.1–6 Given an increasing breadth in the types of medicine, which include human and nonhuman cells, we refer to these new entries as ‘entities’ rather than the more conventional ‘new molecular entities.’
We compiled and curated information about all US Food and Drug Administration (FDA)-approved entities by aggregating and cross-matching relevant information from multiple sources, including the Duke University-led initiative, clinicaltrials.gov (https://aact.ctti-clinicaltrials.org),7 Drugs@FDA,8 the National Center for Advancing Translational Sciences’ (NCATS) Inxight Drugs,9 as well as the National Library of Medicine’s PubChem and Unified Medical Language System (UMLS).10
For example, we captured all trial design and outcomes data conveyed on clinicaltrials.gov, a web-based resource for patients, healthcare providers, and researchers, and compared the results with searches of PubMed (a product of the National Library of Medicine) and Google Scholar (a product of Alphabet). As we have reported previ- ously,6 clinicaltrials.gov has often proven to be ambiguous, with 20% of data inter- pretation confounded by typographic errors, alternative spellings, and characters derived from non-English languages, as well as other challenges that caused ambiguity in determining the identity of APIs, spon- sors, and/or clinical indications.
All ambiguities were resolved and standardized using a blend of machine-based and manual cura- tion to create a synonym list of key words to unambiguously identify all APIs, spon- sors, and intended clinical indications, all of which can be found on a public website: cdek.liu.edu.5
For each API, we catalogued the identi- fier(s) for the API, the sponsoring institution(s), target(s), and mechanism(s), as well as the maximum clinical trial status (i.e., Phase I, II, or III) and overall regulatory status (approved or experimental). Furthermore, an assessment of public disclosures by or about all identified sponsors, includ- ing press releases, news coverage, and 10- Q and 10-K documents from the Securities & Exchange Commission, was conducted to capture additional information not identified by these approaches.
Again, all information is publicly available on cdek. liu.edu. To standardize the mechanistic basis and target recognized by APIs, a manual approach identified targets for each anti- body product and categorized these based upon a modified approach based upon an established database of bioactive mole- cules (www.ebi.ac.uk/chembl/). The modi- fication was necessary because ChEMBL primarily focuses only upon small- molecule therapeutics and it was essential to include mechanisms unique to other therapeutic entities, including gene or cell-based therapies, recombinant proteins (including monoclonal antibodies), and biologicals
The FDA approved 53 unique products in 2019, including new molecular entities (NMEs) given a green light from either the Center for Drug Evaluation and Research (CDER) or Center for Biologics Evaluation and Research (CBER). These medicines could be separated broadly into three categories: 11 biologics, 38 small molecules, and four biologicals (including gene and cell thera- pies) (Fig. 1).
To be clear, we define herein biologics reflecting products derived using re- combinant DNA technologies versus biologicals, which reflect products derived from primary animal or human sources. It is also important to point out that some products, such as Trifakta, contain multiple active ingredients.In such cases, we only consider an NME that was first approved in 2019. Although this rate of new approvals was lower than the 61 new molecules approved in 2018, the approval rate in 2019 exceeded the 5-year running average of 49.4 annual new medicine approvals.
The decline from 2018 reflects seven fewer biologics (18 in 2018 versus 11 in 2019), four fewer small molecules (42 in 2018 versus 38 in 2019), and one more biological (3 in 2018 versus 4 in 2019). The absolute and relative use of incentive regulatory mechanisms also declined. Twenty- one (or 40%) of medicines approved in 2019 utilized an orphan drug mechanism (Fig. 2), which is considerably lower than the highest- ever rate of 56% in 2018.
Results from CDEK reveal that fewer than 10% of experimental medicines, those that have been tested in people, will eventually gain an FDA approval. CRIB is analyzing the rates of drug abandonment. We define an abandoned drug as one, where there is no objective evidence that a drug has advanced for at least five years.
Beyond the concerns that a low rate of acceptance implies about the efficiency of new product development, CRIB is beginning to analyze the diseases that are abandoned and the organizations that are most or least likely to abandon an experimental medicine.
We seek to assess whether decisions to abandon a product are linked with disease prevalence amongst different communities (e.g., based upon household income or other demographics of the afflicted population) and whether such decisions contribute to the inequities in healthcare that frequently capture the headlines.
This program includes:
[source: https://www.fda.gov/drugs/development-approval-process-drugs]
Using our CDEK database, CRIB has assembled a history of FDA drug Approvals to better clarify trends and distinctions year over year in pharmaceutical development.
Beyond creating discordance between market and public health needs, another response to Eroom’s Law was an emphasis upon industry consolidation [6]. Objective and measurable evidence reveals that pharmaceutical develop- ment has become ever more reliant upon mergers, acquisi- tions, and licensing. In the earliest stages of this approach, much of the deal making centered either on companies with either approved or late-stage drug development programs.
This strategy eliminated much of the risks that were associ- ated with early-stage products. Over the past half century, the efficiency and convenience of such acquisitions has evolved into a dependence. Presently, comparatively few “big pharma” companies have robust discovery or early development efforts. Most of those remaining are mere shad- ows of past efforts. At first glance, the interdependence between companies, large and small, established and start-up, early and late-stage development, does not itself seem to pose a particular cause for concern. After all, the logic goes, an acquired company is simply subsumed into the acquirer. Looking more closely, the problems come into focus.
The announcement of mergers is inevitably followed shortly thereafter by waves of layoffs and downsizing of both the acquired employees as well as staff from the acquirer (often to offset resources that were committed to the acquisition) [7]. Another concern is that an acquisition generally involves a company that has successfully and effi- ciently developed one or more promising products. During the transition or amidst layoffs that almost invariably follow mergers, such efficiencies are frequently lost as teams are disbanded or disrupted. Even were a laid off team to reform into a new venture, the activities needed for this new start-up to acquire funding would invariably negatively impact over- all productivity. Such outcomes thus ensure continued adher- ence to Eroom’s Law, which in part explains the persistence of this negative trend for more than seven decades. Another view of this phenomenon is to consider the mod- ern pharmaceutical industry as being akin to a terrestrial ecosystem.
Such networks range from top predators to other hunters, large and small, down to herbivores and the plants that they require for sustenance. The problem, our data suggest, likely lies at the bottom, which is often invisible from the highest peaks of the food chain [6]. A simple glance at the rates of corporate formation and dissolution suggests that the number of start-ups that are committed to the development of new pharmaceutical prod- ucts is not keeping pace with the rate of corporate acquisi- tions [8].
As one pharmaceutical executive has told me, “Our company, and all of our peers, are aware of every phase 3 and phase 2 clinical-stage asset. These are highly picked over. We increasingly find ourselves having to consider phase 1 and even preclinical assets.” Stated another way, we are eating and not replacing more and more of the proverbial seed corn. Such activities rely upon faith and presumptions that future harvests will spontaneously arise. The pharmaceutical industry has managed to get by with such a strategy.
The number of startups arising from aca- demic laboratories has been robust; sufficient to satiate the needs of a fast growing food chain in a mature industry [9]. However, the portents of a problem are easy to spot. For one thing, investors seeking a high multiple on their investment are increasingly cognizant of Eroom’s Law. This awareness of declining efficiency likely raised questions about com- mitting fortunes to endeavors that might require a decade of
After a run of 5 years with an annual rate of approval exceeding 50 new entities, 44 new entities were approved by FDA in 2022. Thirty-seven new entities entered CDER’s Orange Book, while CBER’s Purple Book gained an additional seven entities. This level reflects a drop in the absolute rate of approvals since 2016 but is aligned with the 5-year rates of approval seen during most of the past decade and is higher than the 5-year running average for most of the 20th century (Figure 1a). These newly approved entities could be broadly divided based upon their composi- tion. These categories included 20 small molecules, 17 biologics, and seven other therapies (which included cell- and gene- based therapies).
However, the portents of a problem are easy to spot. For one thing, investors seeking a high multiple on their investment are increasingly cognizant of Eroom’s Law. This awareness of declining efficiency likely raised questions about committing fortunes to endeavors that might require a decade of clinical-stage investigation to yield a payoff.
A combination of high risks with impatience, they might reason, does not seem as daunting for other technologies that are less tightly regulated, have lower barriers to entry, and are quicker to generate profitability. Compounding the concern, the past decade and its historically low interest rates had incentiv- ized many investors to consider higher-risk endeavors to gain a higher return. This era appears to be nearing an end, to be replaced by an extended period of higher interest rates. Consequently, other financial vehicles might become more attractive than higher-risk biotech start-ups. The urgency of this growing problem is particularly chal- lenging for some commonsensical reasons.
First, the dec- ade-long lag that separates early development from product approval means that a slowdown in the global portfolio of promising new medicines may go unnoticed for years. Rather than simply awaiting a disruption in the discovery or development of new therapeutics, it is important now to consider both the fundamental problems that have decreased the attractiveness of drug development, as well as ways to incentivize these crucial activities. For this, we return to the example of antibiotics. One possible solution may arise from innovative approaches that originated in the world of criminal justice. In 2010, British Justice Secretary Jack Straw announced the creation of a “social impact bond” to finance a prisoner rehabilitation program [10].
This financial vehicle defines a problem to be solved and attaches a success fee to innovators that achieve the goal. The concept is analogous to an Old West bounty akin to “Wanted: Dead or Alive.” The key to this approach is that the bounty must be set at a level that is perceived to provide value to those, both who post and receive the reward. When this level is understood, a social impact bond can attract enough bounty hunters, investor-backed entrepreneurs, to address the challenge. If we substitute a new and improved antibiotic for a Wild West desperado, then we can begin to appreciate how and why this approach might prove useful. Indeed, the idea of a bounty has already been introduced into the sciences.
One example is the InnoCentive challenges from the Wazoku, a London-based software company. This organization sponsors competitions for inventive ideas, awarding thousands of dollars for successful concepts. Were such a model to be amplified manifold, it could be utilized to provide multi-million or billion dollar incentives for much- needed public health measures, such as antibiotics. Investor perceptions that antibiotic development is insuf- ficiently valuable could be offset by social impact bonds that encourage greater risk taking and higher payoffs. Were we to rewind history and apply this model to vancomycin, its inno- vator might have been sufficiently rewarded to remain in the hunt for newer and even better antibiotics.
Despite this comparatively slow pace, the 5-year average rate of approvals remains near its historic high- est level, suggesting continued robustness of drug development (Figure 1b). To identify the disease areas targeted by NMEs approved in 2022, we relied upon Medical Subject Headings (MeSH), a con- trolled vocabulary developed by the National Institutes of Health (NIH) for indexing articles for PubMed. 11 For indications not readily defined using MeSH headings, we used ICD-10 guidelines for coding and reporting, as compiled by the Centers for Medicare and Medicaid Ser- vices (CMS). 12
An evaluation of the clinical applica- tions addressed by these newly approved medicines revealed a continue predominance of oncology indications. Cancer applications captured nearly one-third of new approvals, extending a decade long trend. Oncology was followed by eight new entities approved for autoimmune or inflammatory diseases.
These autoimmune/inflammatory disease approvals were an increase relative to 2021, which witnessed no new entities approved for this subset of indications. Two of the 44 approvals were for diagnostic agents. A growth of interventions that modify gene expression or function was reflected by five new entities, which encompassed two gene therapies (nadofaragene firaden- ovec and etranacogene dezaparvovec), two gene-modified cell therapies (ciltacabtagene autoleucel and betibeglogene autotemcel), and a single small interfering (si)RNA- based therapeutic (vutrisiran).
Another notable new entity was the first new bacte- rial cell-based therapy for gut reconstitution following Clostridium difficile infection in the form of donor fecal microbiota.
Nearly two-thirds (62%) gained an approval utilizing a priority review, again lower than the record rate of 71% set in 2018. However, these lower rates were more closely aligned with the 5-year running average of 46% for approved for an orphan indication and 59% reviewed on a priority basis. The FDA offers other mechanisms to expedite the approval of new medicines. Accelerated approvals are offered for indications, when a more easily obtained surrogate marker can be used in place of a longer-term clinical endpoint.
Our analyses revealed that accelerated approvals have increased steadily over the past 6 years, with eight NMEs approved in 2019 using this mechanism. Another mechanism invoked by FDA is break-through designation, which identified therapeu- tic options where preliminary information suggests a substantial improvement in clinical outcome. Fourteen NMEs were approved in 2019 with a breakthrough designation. When viewed over time, the rate of NMEs approved utilizing a breakthrough designation has generally increased, although the 14 NMEs approved utilizing a breakthrough designation in 2019 was somewhat lower than the 5-year running average of 16.7 NMEs awarded a breakthrough designation at the time of initial approval (Fig. 3).
As has been the case for most of the ongoing century, oncology applications continued to capture the largest number of new approvals (data not shown). Eleven new medicines were approved for the treatment of an oncology indication and one additional product was introduced for cancer diagnosis. All of these medicines were approved on a priority basis and six were initially approved for an orphan indication.
Whereas autoimmune diseases have generally reflected the next-most popular indication, this subset of diseases lagged considerably in 2018, with only four NMEs designated for the treatment of an autoimmune or inflammatory application. Eight new medicines targeted nervous system disorders, with two additional drugs targeting migraine-based pain and four other medicines approved for psychiatric indications.
There were other notable observations that distinguished 2019 from previous years. Infectious diseases in general and antibacterial agents in particular fared well, with six medicines approved for these diseases, all but one for a bacterial disease. According to our findings, 2019 was the first year since 1993 that lacked an approval for a treatment of vaccine for a viral disease.
By contrast, five different medicines were approved for dis- ease classes that are relatively uncommon, including non-oncological hematology indications (two for sickle cell disease, with one each for beta-thalassemia, erythropoietic protoporphyria, and thrombotic thrombocytopenic purpura) and three approvals were intended to remedy sleeping disorders (two for narcolepsy and one for insomnia).
A more contemporary example may be the situation with COVID-19 vaccines. In the early days of the pandemic, the world craved a vaccine and much investment, private and public, was invested into these lifesaving preventatives. However, a rather dramatic decline in vaccine uptake and booster shots has translated into massive negative implications for vaccine manufactur- ers; with one example being that this shortfall was cited as the primary reason for Pfizer’s decision to furlough roughly one-quarter of its staff in 2023 [11].
Beyond the immediate impact on Pfizer and other manufacturers, these actions could have dire implications for the future of new vaccine discov- ery and development. A social impact bond could provide an opportunity to incentivize future public health endeavors. Nor would a social impact bond model be limited to govern- mental organizations. A similar mechanism might be comparably applicable to the goals of certain non-profits (e.g., Gates Founda- tion, Chan Zuckerberg Initiative, etc.) and patient advocates, who might contribute to the bounties for diseases of interest.
While we do not expect a social impact bond strategy to provide a panacea for all public health issues, the innovation needed to develop new generations of medicines may not be limited to the pharmaceutical industry. Considerable value might be gained by leveraging new ideas in both the science and incentivization schemas to realize future biomedical breakthroughs.
Funding
This work was supported by the Laura and John Arnold Foundation. Data availability All data can be accessed at cdek.liu.edu. Declarations Conflicts of Interests/Competing Interests The author has no conflict- ing or competing interests with the subject matter cited herein.
Beyond the clinical indications themselves, we assessed trends in the approval mechanisms utilized by FDA. The predom- inance of orphan indications continued in 2022. The 23 new entities targeting an orphan indication reveal the continued utilization of the incentives offered by the 1983 Orphan Drug Act (ODA). 13 The relative fraction of new entities approved using a priority, accelerated, or fast-track approval process remained at roughly their 5-year historical levels (Figure 2), as did the nearly one-third of drugs approved with a breakthrough designation (data not shown).
Ten of the 14 oncology drugs were approved for an orphan indication. This large fraction is consistent with concerns about the use, and potential overuse, of the incentives. Specifically, the concerns are that an orphan indication is merely a loophole used to obtain the lucrative financial incentives conveyed upon orphan drugs. Such worries further pre- sume that many of these oncology drugs will have comparable utility against much larger, non-orphan, markets. Looking fur- ther, we note that these ten oncology approvals reflect fewer than half of the 23 approvals that utilized the ODA process. This is a promising trend, which could indicate that the actions of drug develop- ers are aligning with the original intentions of the ODA (Figure 3).
Consistent with our activities over more than a decade, we calculated the number of ‘successful’ companies, which we define herein as private sector organizations that have contributed to the develop- ment of at least one FDA-approved entity. A total of 393 different companies contributed to the research and development of at least one new therapeutic entity that was approved by the FDA.
Of these 393 organizations, 265 had been acquired, gone defunct, or halted research into unapproved experimental medicines as of December 31, 2021, leaving a total of 128 companies that were still actively engaged in drug discovery (Figure 4a). In 2022, three additional compa- nies were subject to mergers or acquisitions. We also analyzed the impact of industry consolidation and other busi- ness decisions and found that an additional three companies demonstrated no evidence of continued development of new entities for at least 5 years (remaining focused instead on the marketing or distri- bution of products previously approved).
Consequently, by the end of 2022, the total number of ‘successful’ companies still contributing to the development of new drugs stood at 122. To put this in perspec- tive, the number of companies still active in drug development last reached this level in 1983, when our records indicate that a total of 124 ‘successful’ companies were engaged in drug development.
Taking a wider view, we also evaluated companies that were participating in clinical-stage drug development (including those that have not yet, or ever, received an FDA approval).2 These data were compiled by analyzing historic information about clinical-stage entities and their sponsors, as listed in ClinicalTrials.gov and its international equivalents.
Our data analysis included both US and ex-US data from participating countries that make their data publicly available (which includes the European Union, Japan, China, Australia, and many other high and low–middle income nations). We further cross-referenced this information to identify those subject to mergers, acquisi- tions, and other business decisions that might impact their continuation of drug development. This review identified all companies that have ever contributed to the develop- ment of at least one clinical-stage entity.
This analysis was intended to identify organizations that were still devoted to the development of new drugs (as opposed to extending applications or reformula- tions of existing medicines). Our analyses identified 3319 private sector organizations with prior experience in drug devel- opment, of which 1779 were still active in drug development as of December 31, 2021.
Beyond evaluating the medicines themselves, we continued the practice of evaluating the organiza- tions responsible for the research and development activities necessary for FDA approval. This work updates our past studies to analyze the organizations responsible for the research and development of all types of medicine (and biologicals) between 1800 and the end of 2019 [1].
By the end of 2019, a total of 535 different private sector orga- nizations had contributed to the research or de- velopment of an FDA-approved or US-marketed medicine (the latter designation includes over-the- counter medicines) (Fig. 4a).
From this total set of private sector research companies, only 146, or just over a quarter, remained extant and committed to the research or development of a new medicine (Fig. 5a). Most of those lost had been subject to a merger or acquisition. In 2019 alone, 17 organizations that had contributed to the research or development of a new medicine were subject to a merger or acquisition.
Five additional organizations were deemed ‘defunct,’ meaning that there was no evidence that the company was still extant, and three others remained viable companies but did not convey any evidence of ongoing research and development towards new medicines. Looking further, the net loss of 25 organizations in 2019 was consistent with a larger trend. Whereas the net number of research and de- velopment companies had grown over the years, this peaked at 334 different institutions in 2004. In the intervening decade and a half, 188 of these companies are either no longer extant or not doing research.
We then analyzed the impact of indus- try consolidation and other business deci- sions on the number of companies still participating in pharmaceutical research and development. Unsurprisingly, we did not identify any companies that were formed in 2022 that subsequently filed an investigative new drug application (IND) in that same year. Yet, 44 companies were lost as a result of mergers and acquisi- tions that year. From the list of the remaining companies, we sought out evidence of whether the organization was still extant, based upon searches of business records and websites.
This action identified an additional 15 companies that were lost in the year 2022. In addition, we identified 62 extant organizations (i.e., companies that could still be verified using business records and active websites), which were removed from the list of drug developers because they had not participated in a clinical trial based upon a review of informa- tion compiled from clinicaltrials.gov and other relevant data sources. We then subtracted the number of organizations that had been subsumed by a merger or acquisition, had been deemed defunct, or with no evidence of continued development of novel drugs.
As of December 31, 2022, 1666 companies had demonstrated evidence of clinical-stage drug development in support of at least one new entity over the past 5 years. It must be appreciated that these assess- ments of companies contributing to drug development are lagging indicators of the overall health of the drug development enterprise because they do not reflect organizations that will eventually, but have not yet, contributed to the development of at least one clinical-stage novel entity.
In general, our data set suggests that an average of 5 years elapses between the foundation of a company and the initia- tion of clinical-stage investigation (data not shown). Consequently, the trends identified in Figure 4b are dynamic.
2019 was rather unremarkable in many ways, The year witnessed a reversion to the mean from 2018, which had set records in terms of new drugs approved and a net gain in the overall number of companies that contributed to re- search and development. The maintenance of a level of approvals at the 5-year average is no mean feat given that the exceedingly high rate set in 2018 factors into this average. Nonetheless, issues raised by expert analysts and por- trayed broadly in the media suggest that the numbers posted in 2019 were somewhat enhanced.
Specifically, the FDA issued a spate of approvals late in the year, which in some cases came ahead of the scheduled deadline. Although viewed by some as a positive sign of regulatory efficiency, this trend could be troubling in light of a recent report that suggested that such ‘desk-clearing’ behaviors by regulators tend to increase the likelihood of approvals for medicines with increased toxicities[2]. The past year also gave rise to other trends or. more accurately, breaks in trends, that might ultimately prove interesting. For one thing, the year witnessed a rare downturn in the steady growth in the use of the provisions of the 1983 Orphan Drug Act [3]. Whereas 56% of medicines approved in 2018 took advantage of these incentives, far fewer, both in absolute and relative terms, did so in 2019 [4]. This might reflect a statistical hiccup but will be worth watching further.
A flurry of approvals late in the year induced conversation as to whether the FDA might be approving drugs too quickly, with questions being raised about declining stan- dards in assessing safety or efficacy [5,6]. Were such scrutiny to continue or a high-profile toxicity ensue, future years might see a more cautious regulator and the end to the record shattering rates seen in recent years.
Another significant change is the acceleration in mergers and acquisitions. Over the past 2 years, we had noted an easing in the decline associated with industry mergers and acquisitions. 2019 bucked this trend, witnessed a record number of companies subject to mergers and acquisitions. Importantly, the companies tracked herein reflect a privileged few organizations that have successfully contributed to the research or development of at least one approved medicine. Given the recent emphasis on new start-ups, this population of relatively ‘aged’ companies might not portend imminent disruptions in pharmaceutical pipelines.
Nonetheless, the reversion back to accelerating consolidation should be monitored further.
The past decade has witnessed historically high levels of FDA approvals, exceeding 50 new approvals in 6 of the past 10 years.
The 44 approvals in 2022 might simply reflect a reversion to the mean. Two other continuing trends bear watching. The first is the continued and outsized emphasis upon orphan indica- tions, which again exceeded the 50% level in 2022. This response to the incentives conveyed by the ODA has continued to be a source of controversy and has inspired calls for legislative action to redress what is seen by many as an abuse of a law that has been in place for 40 years. 14–16
The year 2022 witnessed a relative decline in orphan drugs for oncology indications, which have dominated orphan approvals since pas- sage of this critical legislation during the 1980s. The ODA was passed to create incentives for unmet diseases, which afflict fewer than 200 000 Americans.17 In practice, many organizations have taken advantage of these requirements to identify patient subsets (of a larger disease) that meet this requirement. This practice has been decried as disingenu- ous because the intended population for many of these oncology drugs is for a population that would not qualify for orphan drug status.14
An increase in drugs meant to intervene against truly rare and under-served indications would be welcome by many. Thus, we will be interested to see whether indications other than oncology will continue to increase in future years. A second and still lingering question surrounds the impact of industry consolidation. The continued erosion in the num- ber of companies participating in research and development is a topic of possible concern.
Although some acquisitions rep- resent a combination of large organiza- tions, we have previously reported that most of the consolidation activity over the past 20 years has involved the acquisi- tion of a small-to-medium-sized biotech- nology company by a more established and well-capitalized company.2,3 Such trends have encouraged academic organi- zations to favor innovation. Although such outcomes are promising, it appears unlikely that the rate of new company foundation will be sufficient to meet the need for start-ups in the future.3
A decline in the number can be viewed in a negative sense as limiting the breadth of participation in development activities and, therefore, might portend a slowdown in an aging enterprise, with massive eco- nomic and public health implications.18 Indeed, our research over the past decade, including observations herein, suggest that a combination of consolidation and business decision-making has reduced the number of organizations actively participating in drug development.2,6 Recent trends are also consistent with a Wall Street Journal article published a decade ago with the headline, ‘In Drug Mergers, There’s One Sure Bet: The Layoffs’.19
Recent evidence confirms that layoffs in the biotechnology sector have been grow- ing, impacting both established and startup companies.20 Biotechnology com- panies faced a particularly challenging year in 2022 as declining initial public offerings decreased interest by investors in drug development. 21 Although the comparatively high number of product approvals since 2013 has been promising (e.g., Figure 1), the sustainability of this momentum is uncertain.
Data will be made available on request. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
This work was generously supported by Arnold Ventures.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Research reported in this publication was solely supported by Washington University in St Louis.