Drug Discovery Today Volume 19, Number 8 August 2014
PERSPECTIVE
feature An overview of FDA-approved new molecular entities: 1827–2013
The pharmaceutical industry is undergoing fundamental change and its future is unclear. We performed a meta-analysis by cataloging FDA-approved legacy drugs and new molecular entities (NMEs). Objective information regarding scientific, medical and commercial activities was captured and provides insight into processes governing drug development. In this report, we review the rates of NME introduction through to the end of 2013. Recent trends show the emergence of a handful of companies that controls two-thirds of NMEs. We also report growth in the number of NMEs controlled by marketing organizations that have little or no internal drug discovery or development activities. This trend has increased dramatically since 2000 and could raise important questions about the future landscape and viability of drug discovery and development.
Introduction The enterprise of drug discovery and development is fundamentally shifting. Dramatic and irreversible changes are reshaping the roles of the pharmaceutical, biotechnology and academic realms. The general view is that the process whereby future drugs are discovered and developed will be fundamentally different to how these activities were performed in the past. With an eye to the future, we began a meta-analysis of the drug development enterprise from its inception through to the present day. As a first step, we cataloged and detailed all legacy drugs and recent new molecular entities (NMEs) approved for use in the USA. In doing so, we sought to determine how the enterprise has changed over time, with emphasis on trends and evaluation of the key players responsible for the discovery of new medicines.
As a brief background, US Congress passed the Drug Price Competition and Patent Term Restoration Act (better known as the Hatch– Waxman Act) in 1982 to bolster an emerging generic drug industry [1]. As part of this process, the FDA was required to identify safe and efficacious drug products through the publication of Approved Drug Products with Therapeutic Equivalence Evaluations or, as it is more commonly known, the ‘Orange Book’ (http://www.accessdata.fda.gov/scripts/Cder/ob/default.cfm). The Orange Book conveys a running list of all drugs, NMEs as well as their generic equivalents, approved by the FDA for use in the USA. Although this serves as an important resource for medical practice, the Orange Book excludes drugs that are no longer marketed or withdrawn owing to concerns about safety or lack of efficacy. Thus, the first step was to supplement information from the Orange Book with compounds
1359-6446/06/$ - see front matter ß 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.drudis.2014.03.018
that had been FDA-approved but are no longer marketed in the USA. These compounds include drugs that were withdrawn owing to lack of efficacy, obsolescence, manufacturing or marketing issues. Representative examples range from the myriad first-generation antibiotics that have been rendered obsolescent to Vioxx1, which was withdrawn owing to apparent safety concerns [2], and through to Bexxar1, which was withdrawn in 2013 based on production or demand limitations at GlaxoSmithKline [3]. All NMEs were included in our analysis, excluding imaging and diagnostic agents. It was important to identify objective information about each NME. This was accomplished by referencing information available from the FDA website and cross-referencing this information with additional sources. When available (generally for NMEs approved after 1990), the drug approval package (including medical, www.drugdiscoverytoday.com
1033
Features PERSPECTIVE
Michael S. Kinch, [email protected], [email protected], Austin Haynesworth, Sarah L. Kinch and Denton Hoyer
PERSPECTIVE
Drug Discovery Today Volume 19, Number 8 August 2014
chemistry and pharmacology reviews as well as published correspondence pre- and postapproval) was captured and analyzed in detail. Additional information was obtained from searches of the American Chemical Society (SciFinder1) based on the unique Chemical Abstracts Service (CAS) number for each NME. Likewise, the CAS number, generic name and structures of the compounds were used to search the US Patent and Trade Office (PatFT for patents; TESS for trademarks), the World Intellectual Property Organization (WIPO), The National Library of Science (MedLine1) as well as publically available information (websites and press releases) from pharmaceutical and biotechnology sources. In addition to analyzing the scientific and medical information about each NME, considerable efforts were invested to identify the organizations involved in the drug discovery process, ranging from the initial discovery (e.g. patent or publication) through to FDA approval and post-marketing obligations. In doing so, the fate of each organization was determined to (a)
evaluate the impact of mergers, acquisitions, bankruptcies or other outcomes. This information was compiled by assessing FDA communications about changes in sponsorship as well as corporate press releases and reviews of the US Securities and Exchange Commission (Edgar) databases and by employing conventional internet search engines. The information captured included the organization that gained the initial approval for each NME as well as the date of approval. For all compounds approved before 1982, this information generally required more-extensive analyses using SciFinder1 and MedLine1. Information was also captured as to the clinical indication, mechanism of action, patent priority dates (US and international), first peer-reviewed publication, investigational new drug (IND) submission as well as the dates of clinical trials (Phase I to Phase III).
USA as of 31 December 2013. Chronologically, the earliest NMEs pre-date the creation of the FDA. Specifically, the earliest compound identified in our assessment was morphine, which was first introduced into the USA after Merck initiated commercial sales in Germany in 1827 [4]. Aspirin, first introduced in 1899 [5], also predates the signing of the Food and Drug Act by President Theodore Roosevelt in 1906. All other NMEs were subject to review by the Food, Drug and Insecticide Organization and its successor, the FDA. The modern pharmaceutical industry began in earnest during the 1930s. The rate of NME approvals of this nascent industry remained at a relatively low level (averaging fewer than four NMEs per year) until 1950 (Fig. 1a). Since that time, the rate of new approvals increased from an average of 15 NMEs per year in the 1950s and remained at this rate through the 1970s. In the 1980s, the average rate increased to 25–30 NMEs per year, where it remains today. In general, approval rates have been rather stable with the exception of considerable volatility in the
NMEs: 1827–2013 A total of 1453 NMEs was identified as having been approved for use as therapeutics in the
1600
Features PERSPECTIVE
Cumulative NMEs
1400 1200 1000 800 600 400 200 0
30
(b)
40
35
19
45
19
19
19
50
19
55
19
65
60
19
19
75
80
19
19
Average Annual NME Rate
50 40 30 20 10
85
90
19
19
00
95
05
20
19
20
10
20
40
(c)
60 Annual NME Approval Rate
70
19
35 30 25 20 15 10 5 0
0
1
0 93
1
0 94
0 95
1
1
0 96
1
0 97
80 19
1
0 99
2
0 00
2
0 01
0s
3 19
0s
4 19
0s
5 19
0s
6 19
0s
7 19
0s
8 19
0s
9 19
0s
0 20
0s
1 20
Drug Discovery Today
FIGURE 1
FDA-approved new molecular entities. (a) The accumulation of FDA-approved new molecular entities (NMEs) over time since 1930 is indicated. Please note that the approval of two molecules, morphine and aspirin, pre-dated the creation of the FDA and its precursors. (b) The number of annual approvals since 1930 is shown, as are (c) the average annual rates of approval by decade. 1034
www.drugdiscoverytoday.com
Drug Discovery Today Volume 19, Number 8 August 2014
(b)
nizations developed at least one NME. As a consequence of mergers, acquisitions, bankruptcies or strategic repositioning, 93 (33.8%) remain active and independent as of the end of 2013 (Fig. 2a). This reduction in number led us to track the number of organizations over time and to assess the relative rate of entries and exits of organizations that gained an approval from the FDA. The number of different organizations with a successful NME grew steadily from the 1930s through the 1950s (Fig. 2a). By the end of 1960, 50 different companies had developed at least one NME. This equates to an average entry of 18.7 new organizations per decade (Fig. 2b). There were relatively few mergers and acquisitions and thus the exit rate was one company per decade. The period 1960–1980 saw a decrease in new entries (averaging eight per decade) coupled with rising exits (4.3 per decade) owing to mergers, acquisitions and strategic repositioning. This volatility expanded dramatically in the 1980s (38 entries and 21 exits) and 1990s (85
entries and 67 exits). The first decade of the new millennium witnessed the highest levels of volatility thus far (118 entries offset by 124 exits) and these trends are likely to continue based on information from the early part of the current decade (2011–2013).
Organizations controlling NMEs An increasing number of NMEs coupled with stagnant growth in the number of organizations with NME experience revealed that the average number of NMEs per organization increased. Whereas the average number of NMEs held per organization was 1.3 NMEs in 1940, the average today is greater than 14.3 (data not shown). This led us to ask which organizations developed or controlled the most NMEs. Specifically, we asked which organizations have received NMEs and classified these based on the number received (e.g. 1, 2–5, 6–10, etc.). Most organizations (152 of 275) developed only one NME approval from the FDA (Fig. 3a; red bars). Fewer organizations earned between two
120 100 80
Features PERSPECTIVE
(a)
Companies granted at least one NME
mid-to-late 1990s (Fig. 1b). For example, a peak of 55 NMEs was approved in 1997 whereas three years before there were only 19 approvals. For ease of simplicity (and to eliminate such year-to-year volatility), we adopt hereafter an approach of comparing the data by averaging the information for each decade (e.g. 1931–1940; Fig. 1c). Once the list of approved drugs was compiled, we asked which organizations gained approval for each NME and determined the fate of these organizations. For our purposes, an organization is defined as a business or, rarely, a governmental entity that retains the same general corporate structure. Based on these criteria, a merger of equals creates one new organization and eliminates two existing organizations. For example, the merger in 1989 created one new organization (Bristol-Myers Squibb) and triggered the exit of two prior organizations (Bristol Myers and Squibb). By contrast, the acquisition of Onyx by Amgen in 2013 simply led to the exit of one (Onyx). Using these criteria, 275 different orga-
PERSPECTIVE
60 40 20 0
30
19
35
19
40
19
45
19
50
19
55
19
60
19
65
19
70
75 19
19
70
75 19
19
19
80
85
19
90 19
95
19
00
20
05
20
10
20
Entry
20 15 10 5 0 Exit
-5 -10 -15 -20 30
19
35
19
40
19
45
19
50
19
55
19
60
19
65
19
19
80
85
19
90
19
95
19
00
20
05
20
10
20
Drug Discovery Today
FIGURE 2
Changes in the number of companies with at least one FDA-approved new molecular entity. (a) The number of different companies that have gained approval for at least one new molecular entity (NME) is indicated – as organizations merge or are acquired the numbers can decrease, as indicated in the text. (b) The annual numbers of entries (green) or exits (red) of companies with NME experience is indicated. Note the increasing volatility (entries and exits) since 1980. www.drugdiscoverytoday.com
1035
PERSPECTIVE
(a)
Drug Discovery Today Volume 19, Number 8 August 2014
(b)
160
Sandoz Upjohn
Organizations with indicated number of NMEs
Schering-Plough Abbott
140
At time of approval
American home products Pfizer
Current
Eli Lilly
120
J&J Roche
100
Merck
0
50
80
100
150
200
250
NMEs Granted (c)
60
Valeant Eli Lilly Johnson & Johnson
40
Bristol-Myers Squibb Roche GlaxoSmithKline
20
Sanofi Aventis Novartis Merck
0+ 10
0 to
10
50
Pfizer
51
to 21
to
20
10 11
to 6
2
to
5
1
0
0
50
100 150 NME Controlled
200
Drug Discovery Today
FIGURE 3
Features PERSPECTIVE
Distribution of new molecular entities and growth of ‘Big Pharma’. (a) The number of new molecular entities (NMEs) was tabulated per organization and the distribution is indicated. The results are shown as of the initial time of approval for each NME (in effect, the creator of the NME; blue bars) as well as the current owner of the NME (in effect, the controllers; red bars). The top ten organizations that have been granted FDA approval (b) or currently control NMEs (c) are indicated.
and five (67 organizations) or six and 20 NMEs (45 organizations). Only 14 organizations have received more than 20 NMEs and fewer still (six) received more than 50. The most successful organization thus far is Merck, which has received 63 approvals for NMEs (Fig. 3b). Merck is followed by Roche, Johnson & Johnson, Eli Lilly, Pfizer and American Home Products. Of the top ten organizations granted NMEs, one-half (American Home Products, Abbott, Schering-Plough, Upjohn and Sandoz) does not currently exist as active and independent drug development organizations. This led us to ask who currently controls these NMEs. To do so, we assessed the chain of custody of each NME after changes due to mergers, acquisitions and licenses. For the purpose of this study, the owner is the organization to which the NME was granted as identified in FDA communications. This approach avoided confusion that would occur with collaborative projects involving multiple partners. In the event that such information was unclear or unavailable, a search of US 1036
www.drugdiscoverytoday.com
Government patent and copyright databases was conducted to identify the NME owner. Notably, holding an NME does not necessarily indicate that the organization is actively marketing the drug because a subset of NMEs is no longer marketed as a result of toxicity, lack of efficacy, obsolescence, production issues or lack of demand. In total, 115 organizations currently control at least one NME. These companies can be divided into three groups based on the number of NMEs held. Almost one-half of the organizations (54 of 115) currently control only one NME (Fig. 3a; blue bars). An additional 30 organizations control two to five NMEs. A smaller number (19) of mid-tier holders control between six and 20 NMEs. Twelve companies control more than 20 NMEs each. The five largest holders of NMEs are Pfizer (198) followed by Merck (106), Novartis (98), SanofiAventis (84) and GlaxoSmithKline (79) (Fig. 3c). Altogether, the top five companies control more than 40% of all NMEs and this fraction expands to more
than two-thirds of all NMEs if one includes the top ten organizations. To determine how these top companies came to dominate the drug development enterprise, we evaluated their expansion over time. We postulated that leading drug companies grew to prominence more as a result of mergers and acquisitions than internal R&D. To determine this, we assessed how the top five companies have changed over time (Fig. 4). In 1950 the company with the largest number of NMEs was Eli Lilly, followed by Merck, Abbott and Ciba. American Home Products increased its presence in the 1950s and, along with the aforementioned companies, remained as one of the leading holders of NMEs throughout the 1960s and 1970s. The 1980s and 1990s witnessed multiple pharmaceutical industry mergers and only American Home Products remained as a member of the top five NMEholding companies in the year 2000. At present, all of the top five companies gained their positions as a result of larger mergers or
Drug Discovery Today Volume 19, Number 8 August 2014
PERSPECTIVE
1950
1960
1970
1980
1990
2000
2010
2013
1
Eli Lilly
American Home Products
Eli Lilly
Eli Lilly
Eli Lilly
Glaxo SmithKline
Pfizer
Pfizer
2
Merck
Eli Lilly
American Home Products
American Home Products
Bristol-Myers Squibb
Novartis
Merck
Merck
3
Abbott
Merck
Merck
Abbott
American Home Products
American Home products
Novartis
Novartis
4
Ciba
Abbott
Abbott
Merck
Merck
Pfizer
5
Burroughs wellcome
Ciba
Ciba
Ciba-Geigy
Roche
Bristol-Myers Squibb
Sanofi-Aventis Sanofi-Aventis Glaxo SmithKline
Glaxo SmithKline
Drug Discovery Today
FIGURE 4
New molecular entity leaders over time. The identity of the top ten new molecular entity (NME)-controlling organizations is shown at the beginning of each decade. Bold font indicates companies that entered or remained in the top ten based on a merger or major acquisition.
200
(a)
150
Pfizer
Features PERSPECTIVE
NMEs Controlled
Eli Lilly
100
50
0 1953
1963
1973
(b) 50
(c)
30 20 10
1993
2003
2013
250
NMEs Controlled
NMEs Controlled
40
0 1963
1983
200 150 100 50
1973
1983
1993
2003
2013
0 1963
1973
1983
1993
2003
2013
Drug Discovery Today
FIGURE 5
NMEs controlled by innovating and marketing-based pharmaceutical companies. (a) The number of NMEs controlled by Eli Lilly, which largely has avoided large mergers, is compared with Pfizer, which actively pursued a strategy of mergers and acquisitions. The number of NMEs controlled by Valeant (b) and companies lacking direct NME approval experience (c) reveals a growing trend in NMEs controlled by organizations that do not participate in new molecular entity research and development. www.drugdiscoverytoday.com
1037
PERSPECTIVE
acquisitions. Merck rejoined the top companies as a result of its takeover of ScheringPlough in 2009.
Spotlight: the evolution of key organizations
Features PERSPECTIVE
To assess these changes in greater detail, we compared the evolution of two different organizations: Eli Lilly and Merck. Eli Lilly relied almost entirely on internal R&D and held the most NMEs in four out of five decades from 1950 until 2000 (Fig. 4). The number of Eli Lilly NMEs peaked at 46 in 1990 and declined somewhat owing to corporate decisions to out-license or sell NMEs to competitors (Fig 5a; red line). This analysis demonstrated a growth rate at Pfizer that largely paralleled that of Eli Lilly throughout the 1950s through to 1990, albeit starting with a lower initial base (Fig. 5a; blue line). From the 1990s onwards, Pfizer acquired multiple NMEs, mostly as a result of mergers or acquisitions of large pharmaceutical companies (Warner-Lambert, Pharmacia-Upjohn, American Home Products). Today, Pfizer’s portfolio of NMEs is more than fourfold larger than that of Eli Lilly. We conjectured that the growth achieved throughout the 1990s and 2000s is not sustainable because this would require the development or acquisition of more than 100 NMEs per decade. Indeed, the growth in NMEs at Pfizer has been more modest since its acquisition of King Pharmaceuticals in 2010. Pfizer remains a potent R&D leader in drug discovery and marketing, being directly awarded five NMEs during the past two years. A very different situation is represented by the growth of Valeant Pharmaceuticals (Fig. 5b). In 2010, Valeant entered the top ten NME holders (with a total of 44 NMEs). The company utilized an approach of acquiring other companies that had directly obtained an NME from the FDA. Although Valeant has not directly been awarded an NME since its inception in the early 1990s, the collection of NMEs under its control has impressively grown over the past ten years. Valeant is but one example of a prominent change in the drug discovery and development enterprise. The growth of companies with a similar strategy is recent and robust. As of 1990, only two NMEs (0.26% of all NMEs awarded up to that date) were controlled by organizations that had not been directly awarded an NME (Fig. 5c). By the end of 2013, this numbers had swelled to 215 NMEs (or 14.8% of all NMEs granted to date). Indeed, of the 118 companies that hold at least one NME today, 25 (21.1%) did not receive an NME approval from the FDA. 1038
www.drugdiscoverytoday.com
Drug Discovery Today Volume 19, Number 8 August 2014
Concluding remarks: findings and implications The major finding of this report is the creation of a comprehensive database of all NMEs approved for use in the USA. The data include historical information about the drug itself as well as the organizations involved in the development and marketing of each novel molecule and changes that have occurred over time. This database provides an objective and updatable means to identify and track trends governing drug R&D. We also seek to identify opportunities to continue to improve and expand the information and make it available for research purposes. This is particularly important for older NMEs, where required information about the development process could be provided by individuals or organizations with knowledge of past development activities. The pharmaceutical enterprise has been subject to profound changes. The early years of the industry (1930–1960) were characterized by a steady growth in the number of approved NMEs and a proportional expansion in the breadth of organizations that had achieved an NME. From 1960 to 1980, the number of NMEs continued to increase but the number of organizations that had obtained an NME reached a plateau, with few new entries or exits. Starting in the 1980s, a rather dramatic volatility began in which new companies entered and exited the industry. The entries largely represented start-up companies, largely biotechnology companies, whereas the exits were largely caused by consolidations from mergers and acquisitions (rather than bankruptcies). Such changes have given rise to an industry characterized by four types of organizations. Most organizations were awarded or control only one NME. These small ‘singlet’ companies are often the target of acquisitions, usually by mediumsized peers. These medium companies, which control between two and 20 NMEs, are occasionally acquired whereas a few merged to form larger corporations. The third tier consists of large, established companies, which hold a disproportionately large number of NMEs. These companies invariably rose from the middle tier as a result of mergers and none grew without at least one major merger. Notably, the top ten companies control two-thirds of all NMEs (as of the end of 2013). This represents a 65% rise from 1950, when the top ten companies controlled roughly 40% of all NMEs. The fourth type of company (which can include examples from the others) consists of companies that primarily or only function as sales, marketing and manufacturing engines. These companies have
minimized internal discovery and development activities and, instead, focus on obtaining NMEs from licensing, mergers and acquisitions. Although volatility continues, two potentially important changes were noted. First, the number of exits from the drug discovery industry is beginning to exceed the number of entries. It is generally recognized that the larger pharmaceutical organizations have been scaling back their internal R&D activities. Our data suggest that this rebalancing began more than two decades ago and, in part, gave rise to the aforementioned rise of biotechnology and increasing turnover of companies awarded NMEs. As will be detailed in a future report, some of the early scaling-back of the pharmaceutical industry arose as a result of increased risk awareness as pharmaceutical companies licensed or acquired projects from biotechnology companies. The present findings suggest that organizations that obtain NMEs are efficiently eliminated by acquisitions. This has generally been a hallmark of the biotechnology industry. However, the second notable change is the rapid rise of acquiring organizations with limited internal R&D capabilities. In light of tightening venture funding for early-stage biotechnology companies, this raises concerns that the R&D infrastructure for drug development could be progressively and perhaps irreversibly shrinking and that our ability to discover and/or develop new medicines is being progressively dismantled. Clearly, the need for new medicines persists. It will be important to help identify and utilize such trends to develop new mechanisms to address unmet medical and commercial opportunities. Much impressive prior work has shed considerable insight into changing trends in regulatory control and processes [6,7]. We seek to utilize this new database of FDAapproved NMEs as a ‘big data’ means to analyze long-term trends based on medical or scientific criteria. Specifically, the information gathered will provide future opportunities to compare and contrast the changes in different therapeutic areas (e.g. changes in oncology or infectious diseases), drug types (small molecule and biologics), contributors (academic, government, biotechnology, pharmaceutical) and mechanisms of action.
Acknowledgements The authors thank Drs Jon Soderstrom and Richard Foster for helpful discussions and inspiration; and Drs Jane Merkel and Mark Plummer for critical reading of the manuscript. The work herein was performed as part of an initiative by the Yale Center for Molecular
Drug Discovery Today Volume 19, Number 8 August 2014
References 1 Mossinghoff, G.J. (1999) Overview of the Hatch–Waxman Act and its impact on the drug development process. Food Drug Law J. 54, 187–194 2 Katz, J.A. (2013) COX-2 inhibition: what we learned – a controversial update on safety data. Pain Med. 14 (Suppl. 1), 29–34
3 GlaxoSmithKline Press release (2014) GSK to discontinue manufacture and sale of the BEXXARAˆ1 therapeutic regimen (tositumomab and iodine I 131 tositumomab). Available at: http://us.gsk.com/html/media-news/ pressreleases/2013/GSK-to-discontinue-manufactureand-sale-of-the-BEXXAR-therapeutic-regimen.html 4 Hayes, A.N. and Gilbert, S.G. (2009) Historical milestones and discoveries that shaped the toxicology sciences. EXS 99, 1–35 5 Jeffreys, D., ed. (2008) Aspirin: The Remarkable Story of a Wonder Drug, Bloomsbury Publishing 6 Kaitin, K.I. and DiMasi, J.A. (2011) Pharmaceutical innovation in the 21st century: new drug approvals in the first decade, 2000–2009. Clin. Pharmacol. Ther. 89, 183–188
7 Kaitin, K.I. (2010) Deconstructing the drug development process: the new face of innovation. Clin. Pharmacol. Ther. 87, 356–361
Michael S. Kinch, Austin Haynesworth, Sarah L. Kinch, Denton Hoyer Yale Center for Molecular Discovery, Yale University, 600 West Campus Drive, West Haven, CT 06516, USA
Features PERSPECTIVE
Discovery (YCMD) to develop a comprehensive screening strategy for drug re-purposing. Please contact Dr. Michael Kinch if you or your institution would like to investigate potential interactions with YCMD.
PERSPECTIVE
www.drugdiscoverytoday.com
1039
PERSPECTIVE
feature An overview of FDA-approved new molecular entities: 1827–2013
The pharmaceutical industry is undergoing fundamental change and its future is unclear. We performed a meta-analysis by cataloging FDA-approved legacy drugs and new molecular entities (NMEs). Objective information regarding scientific, medical and commercial activities was captured and provides insight into processes governing drug development. In this report, we review the rates of NME introduction through to the end of 2013. Recent trends show the emergence of a handful of companies that controls two-thirds of NMEs. We also report growth in the number of NMEs controlled by marketing organizations that have little or no internal drug discovery or development activities. This trend has increased dramatically since 2000 and could raise important questions about the future landscape and viability of drug discovery and development.
Introduction The enterprise of drug discovery and development is fundamentally shifting. Dramatic and irreversible changes are reshaping the roles of the pharmaceutical, biotechnology and academic realms. The general view is that the process whereby future drugs are discovered and developed will be fundamentally different to how these activities were performed in the past. With an eye to the future, we began a meta-analysis of the drug development enterprise from its inception through to the present day. As a first step, we cataloged and detailed all legacy drugs and recent new molecular entities (NMEs) approved for use in the USA. In doing so, we sought to determine how the enterprise has changed over time, with emphasis on trends and evaluation of the key players responsible for the discovery of new medicines.
As a brief background, US Congress passed the Drug Price Competition and Patent Term Restoration Act (better known as the Hatch– Waxman Act) in 1982 to bolster an emerging generic drug industry [1]. As part of this process, the FDA was required to identify safe and efficacious drug products through the publication of Approved Drug Products with Therapeutic Equivalence Evaluations or, as it is more commonly known, the ‘Orange Book’ (http://www.accessdata.fda.gov/scripts/Cder/ob/default.cfm). The Orange Book conveys a running list of all drugs, NMEs as well as their generic equivalents, approved by the FDA for use in the USA. Although this serves as an important resource for medical practice, the Orange Book excludes drugs that are no longer marketed or withdrawn owing to concerns about safety or lack of efficacy. Thus, the first step was to supplement information from the Orange Book with compounds
1359-6446/06/$ - see front matter ß 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.drudis.2014.03.018
that had been FDA-approved but are no longer marketed in the USA. These compounds include drugs that were withdrawn owing to lack of efficacy, obsolescence, manufacturing or marketing issues. Representative examples range from the myriad first-generation antibiotics that have been rendered obsolescent to Vioxx1, which was withdrawn owing to apparent safety concerns [2], and through to Bexxar1, which was withdrawn in 2013 based on production or demand limitations at GlaxoSmithKline [3]. All NMEs were included in our analysis, excluding imaging and diagnostic agents. It was important to identify objective information about each NME. This was accomplished by referencing information available from the FDA website and cross-referencing this information with additional sources. When available (generally for NMEs approved after 1990), the drug approval package (including medical, www.drugdiscoverytoday.com
1033
Features PERSPECTIVE
Michael S. Kinch, [email protected], [email protected], Austin Haynesworth, Sarah L. Kinch and Denton Hoyer
PERSPECTIVE
Drug Discovery Today Volume 19, Number 8 August 2014
chemistry and pharmacology reviews as well as published correspondence pre- and postapproval) was captured and analyzed in detail. Additional information was obtained from searches of the American Chemical Society (SciFinder1) based on the unique Chemical Abstracts Service (CAS) number for each NME. Likewise, the CAS number, generic name and structures of the compounds were used to search the US Patent and Trade Office (PatFT for patents; TESS for trademarks), the World Intellectual Property Organization (WIPO), The National Library of Science (MedLine1) as well as publically available information (websites and press releases) from pharmaceutical and biotechnology sources. In addition to analyzing the scientific and medical information about each NME, considerable efforts were invested to identify the organizations involved in the drug discovery process, ranging from the initial discovery (e.g. patent or publication) through to FDA approval and post-marketing obligations. In doing so, the fate of each organization was determined to (a)
evaluate the impact of mergers, acquisitions, bankruptcies or other outcomes. This information was compiled by assessing FDA communications about changes in sponsorship as well as corporate press releases and reviews of the US Securities and Exchange Commission (Edgar) databases and by employing conventional internet search engines. The information captured included the organization that gained the initial approval for each NME as well as the date of approval. For all compounds approved before 1982, this information generally required more-extensive analyses using SciFinder1 and MedLine1. Information was also captured as to the clinical indication, mechanism of action, patent priority dates (US and international), first peer-reviewed publication, investigational new drug (IND) submission as well as the dates of clinical trials (Phase I to Phase III).
USA as of 31 December 2013. Chronologically, the earliest NMEs pre-date the creation of the FDA. Specifically, the earliest compound identified in our assessment was morphine, which was first introduced into the USA after Merck initiated commercial sales in Germany in 1827 [4]. Aspirin, first introduced in 1899 [5], also predates the signing of the Food and Drug Act by President Theodore Roosevelt in 1906. All other NMEs were subject to review by the Food, Drug and Insecticide Organization and its successor, the FDA. The modern pharmaceutical industry began in earnest during the 1930s. The rate of NME approvals of this nascent industry remained at a relatively low level (averaging fewer than four NMEs per year) until 1950 (Fig. 1a). Since that time, the rate of new approvals increased from an average of 15 NMEs per year in the 1950s and remained at this rate through the 1970s. In the 1980s, the average rate increased to 25–30 NMEs per year, where it remains today. In general, approval rates have been rather stable with the exception of considerable volatility in the
NMEs: 1827–2013 A total of 1453 NMEs was identified as having been approved for use as therapeutics in the
1600
Features PERSPECTIVE
Cumulative NMEs
1400 1200 1000 800 600 400 200 0
30
(b)
40
35
19
45
19
19
19
50
19
55
19
65
60
19
19
75
80
19
19
Average Annual NME Rate
50 40 30 20 10
85
90
19
19
00
95
05
20
19
20
10
20
40
(c)
60 Annual NME Approval Rate
70
19
35 30 25 20 15 10 5 0
0
1
0 93
1
0 94
0 95
1
1
0 96
1
0 97
80 19
1
0 99
2
0 00
2
0 01
0s
3 19
0s
4 19
0s
5 19
0s
6 19
0s
7 19
0s
8 19
0s
9 19
0s
0 20
0s
1 20
Drug Discovery Today
FIGURE 1
FDA-approved new molecular entities. (a) The accumulation of FDA-approved new molecular entities (NMEs) over time since 1930 is indicated. Please note that the approval of two molecules, morphine and aspirin, pre-dated the creation of the FDA and its precursors. (b) The number of annual approvals since 1930 is shown, as are (c) the average annual rates of approval by decade. 1034
www.drugdiscoverytoday.com
Drug Discovery Today Volume 19, Number 8 August 2014
(b)
nizations developed at least one NME. As a consequence of mergers, acquisitions, bankruptcies or strategic repositioning, 93 (33.8%) remain active and independent as of the end of 2013 (Fig. 2a). This reduction in number led us to track the number of organizations over time and to assess the relative rate of entries and exits of organizations that gained an approval from the FDA. The number of different organizations with a successful NME grew steadily from the 1930s through the 1950s (Fig. 2a). By the end of 1960, 50 different companies had developed at least one NME. This equates to an average entry of 18.7 new organizations per decade (Fig. 2b). There were relatively few mergers and acquisitions and thus the exit rate was one company per decade. The period 1960–1980 saw a decrease in new entries (averaging eight per decade) coupled with rising exits (4.3 per decade) owing to mergers, acquisitions and strategic repositioning. This volatility expanded dramatically in the 1980s (38 entries and 21 exits) and 1990s (85
entries and 67 exits). The first decade of the new millennium witnessed the highest levels of volatility thus far (118 entries offset by 124 exits) and these trends are likely to continue based on information from the early part of the current decade (2011–2013).
Organizations controlling NMEs An increasing number of NMEs coupled with stagnant growth in the number of organizations with NME experience revealed that the average number of NMEs per organization increased. Whereas the average number of NMEs held per organization was 1.3 NMEs in 1940, the average today is greater than 14.3 (data not shown). This led us to ask which organizations developed or controlled the most NMEs. Specifically, we asked which organizations have received NMEs and classified these based on the number received (e.g. 1, 2–5, 6–10, etc.). Most organizations (152 of 275) developed only one NME approval from the FDA (Fig. 3a; red bars). Fewer organizations earned between two
120 100 80
Features PERSPECTIVE
(a)
Companies granted at least one NME
mid-to-late 1990s (Fig. 1b). For example, a peak of 55 NMEs was approved in 1997 whereas three years before there were only 19 approvals. For ease of simplicity (and to eliminate such year-to-year volatility), we adopt hereafter an approach of comparing the data by averaging the information for each decade (e.g. 1931–1940; Fig. 1c). Once the list of approved drugs was compiled, we asked which organizations gained approval for each NME and determined the fate of these organizations. For our purposes, an organization is defined as a business or, rarely, a governmental entity that retains the same general corporate structure. Based on these criteria, a merger of equals creates one new organization and eliminates two existing organizations. For example, the merger in 1989 created one new organization (Bristol-Myers Squibb) and triggered the exit of two prior organizations (Bristol Myers and Squibb). By contrast, the acquisition of Onyx by Amgen in 2013 simply led to the exit of one (Onyx). Using these criteria, 275 different orga-
PERSPECTIVE
60 40 20 0
30
19
35
19
40
19
45
19
50
19
55
19
60
19
65
19
70
75 19
19
70
75 19
19
19
80
85
19
90 19
95
19
00
20
05
20
10
20
Entry
20 15 10 5 0 Exit
-5 -10 -15 -20 30
19
35
19
40
19
45
19
50
19
55
19
60
19
65
19
19
80
85
19
90
19
95
19
00
20
05
20
10
20
Drug Discovery Today
FIGURE 2
Changes in the number of companies with at least one FDA-approved new molecular entity. (a) The number of different companies that have gained approval for at least one new molecular entity (NME) is indicated – as organizations merge or are acquired the numbers can decrease, as indicated in the text. (b) The annual numbers of entries (green) or exits (red) of companies with NME experience is indicated. Note the increasing volatility (entries and exits) since 1980. www.drugdiscoverytoday.com
1035
PERSPECTIVE
(a)
Drug Discovery Today Volume 19, Number 8 August 2014
(b)
160
Sandoz Upjohn
Organizations with indicated number of NMEs
Schering-Plough Abbott
140
At time of approval
American home products Pfizer
Current
Eli Lilly
120
J&J Roche
100
Merck
0
50
80
100
150
200
250
NMEs Granted (c)
60
Valeant Eli Lilly Johnson & Johnson
40
Bristol-Myers Squibb Roche GlaxoSmithKline
20
Sanofi Aventis Novartis Merck
0+ 10
0 to
10
50
Pfizer
51
to 21
to
20
10 11
to 6
2
to
5
1
0
0
50
100 150 NME Controlled
200
Drug Discovery Today
FIGURE 3
Features PERSPECTIVE
Distribution of new molecular entities and growth of ‘Big Pharma’. (a) The number of new molecular entities (NMEs) was tabulated per organization and the distribution is indicated. The results are shown as of the initial time of approval for each NME (in effect, the creator of the NME; blue bars) as well as the current owner of the NME (in effect, the controllers; red bars). The top ten organizations that have been granted FDA approval (b) or currently control NMEs (c) are indicated.
and five (67 organizations) or six and 20 NMEs (45 organizations). Only 14 organizations have received more than 20 NMEs and fewer still (six) received more than 50. The most successful organization thus far is Merck, which has received 63 approvals for NMEs (Fig. 3b). Merck is followed by Roche, Johnson & Johnson, Eli Lilly, Pfizer and American Home Products. Of the top ten organizations granted NMEs, one-half (American Home Products, Abbott, Schering-Plough, Upjohn and Sandoz) does not currently exist as active and independent drug development organizations. This led us to ask who currently controls these NMEs. To do so, we assessed the chain of custody of each NME after changes due to mergers, acquisitions and licenses. For the purpose of this study, the owner is the organization to which the NME was granted as identified in FDA communications. This approach avoided confusion that would occur with collaborative projects involving multiple partners. In the event that such information was unclear or unavailable, a search of US 1036
www.drugdiscoverytoday.com
Government patent and copyright databases was conducted to identify the NME owner. Notably, holding an NME does not necessarily indicate that the organization is actively marketing the drug because a subset of NMEs is no longer marketed as a result of toxicity, lack of efficacy, obsolescence, production issues or lack of demand. In total, 115 organizations currently control at least one NME. These companies can be divided into three groups based on the number of NMEs held. Almost one-half of the organizations (54 of 115) currently control only one NME (Fig. 3a; blue bars). An additional 30 organizations control two to five NMEs. A smaller number (19) of mid-tier holders control between six and 20 NMEs. Twelve companies control more than 20 NMEs each. The five largest holders of NMEs are Pfizer (198) followed by Merck (106), Novartis (98), SanofiAventis (84) and GlaxoSmithKline (79) (Fig. 3c). Altogether, the top five companies control more than 40% of all NMEs and this fraction expands to more
than two-thirds of all NMEs if one includes the top ten organizations. To determine how these top companies came to dominate the drug development enterprise, we evaluated their expansion over time. We postulated that leading drug companies grew to prominence more as a result of mergers and acquisitions than internal R&D. To determine this, we assessed how the top five companies have changed over time (Fig. 4). In 1950 the company with the largest number of NMEs was Eli Lilly, followed by Merck, Abbott and Ciba. American Home Products increased its presence in the 1950s and, along with the aforementioned companies, remained as one of the leading holders of NMEs throughout the 1960s and 1970s. The 1980s and 1990s witnessed multiple pharmaceutical industry mergers and only American Home Products remained as a member of the top five NMEholding companies in the year 2000. At present, all of the top five companies gained their positions as a result of larger mergers or
Drug Discovery Today Volume 19, Number 8 August 2014
PERSPECTIVE
1950
1960
1970
1980
1990
2000
2010
2013
1
Eli Lilly
American Home Products
Eli Lilly
Eli Lilly
Eli Lilly
Glaxo SmithKline
Pfizer
Pfizer
2
Merck
Eli Lilly
American Home Products
American Home Products
Bristol-Myers Squibb
Novartis
Merck
Merck
3
Abbott
Merck
Merck
Abbott
American Home Products
American Home products
Novartis
Novartis
4
Ciba
Abbott
Abbott
Merck
Merck
Pfizer
5
Burroughs wellcome
Ciba
Ciba
Ciba-Geigy
Roche
Bristol-Myers Squibb
Sanofi-Aventis Sanofi-Aventis Glaxo SmithKline
Glaxo SmithKline
Drug Discovery Today
FIGURE 4
New molecular entity leaders over time. The identity of the top ten new molecular entity (NME)-controlling organizations is shown at the beginning of each decade. Bold font indicates companies that entered or remained in the top ten based on a merger or major acquisition.
200
(a)
150
Pfizer
Features PERSPECTIVE
NMEs Controlled
Eli Lilly
100
50
0 1953
1963
1973
(b) 50
(c)
30 20 10
1993
2003
2013
250
NMEs Controlled
NMEs Controlled
40
0 1963
1983
200 150 100 50
1973
1983
1993
2003
2013
0 1963
1973
1983
1993
2003
2013
Drug Discovery Today
FIGURE 5
NMEs controlled by innovating and marketing-based pharmaceutical companies. (a) The number of NMEs controlled by Eli Lilly, which largely has avoided large mergers, is compared with Pfizer, which actively pursued a strategy of mergers and acquisitions. The number of NMEs controlled by Valeant (b) and companies lacking direct NME approval experience (c) reveals a growing trend in NMEs controlled by organizations that do not participate in new molecular entity research and development. www.drugdiscoverytoday.com
1037
PERSPECTIVE
acquisitions. Merck rejoined the top companies as a result of its takeover of ScheringPlough in 2009.
Spotlight: the evolution of key organizations
Features PERSPECTIVE
To assess these changes in greater detail, we compared the evolution of two different organizations: Eli Lilly and Merck. Eli Lilly relied almost entirely on internal R&D and held the most NMEs in four out of five decades from 1950 until 2000 (Fig. 4). The number of Eli Lilly NMEs peaked at 46 in 1990 and declined somewhat owing to corporate decisions to out-license or sell NMEs to competitors (Fig 5a; red line). This analysis demonstrated a growth rate at Pfizer that largely paralleled that of Eli Lilly throughout the 1950s through to 1990, albeit starting with a lower initial base (Fig. 5a; blue line). From the 1990s onwards, Pfizer acquired multiple NMEs, mostly as a result of mergers or acquisitions of large pharmaceutical companies (Warner-Lambert, Pharmacia-Upjohn, American Home Products). Today, Pfizer’s portfolio of NMEs is more than fourfold larger than that of Eli Lilly. We conjectured that the growth achieved throughout the 1990s and 2000s is not sustainable because this would require the development or acquisition of more than 100 NMEs per decade. Indeed, the growth in NMEs at Pfizer has been more modest since its acquisition of King Pharmaceuticals in 2010. Pfizer remains a potent R&D leader in drug discovery and marketing, being directly awarded five NMEs during the past two years. A very different situation is represented by the growth of Valeant Pharmaceuticals (Fig. 5b). In 2010, Valeant entered the top ten NME holders (with a total of 44 NMEs). The company utilized an approach of acquiring other companies that had directly obtained an NME from the FDA. Although Valeant has not directly been awarded an NME since its inception in the early 1990s, the collection of NMEs under its control has impressively grown over the past ten years. Valeant is but one example of a prominent change in the drug discovery and development enterprise. The growth of companies with a similar strategy is recent and robust. As of 1990, only two NMEs (0.26% of all NMEs awarded up to that date) were controlled by organizations that had not been directly awarded an NME (Fig. 5c). By the end of 2013, this numbers had swelled to 215 NMEs (or 14.8% of all NMEs granted to date). Indeed, of the 118 companies that hold at least one NME today, 25 (21.1%) did not receive an NME approval from the FDA. 1038
www.drugdiscoverytoday.com
Drug Discovery Today Volume 19, Number 8 August 2014
Concluding remarks: findings and implications The major finding of this report is the creation of a comprehensive database of all NMEs approved for use in the USA. The data include historical information about the drug itself as well as the organizations involved in the development and marketing of each novel molecule and changes that have occurred over time. This database provides an objective and updatable means to identify and track trends governing drug R&D. We also seek to identify opportunities to continue to improve and expand the information and make it available for research purposes. This is particularly important for older NMEs, where required information about the development process could be provided by individuals or organizations with knowledge of past development activities. The pharmaceutical enterprise has been subject to profound changes. The early years of the industry (1930–1960) were characterized by a steady growth in the number of approved NMEs and a proportional expansion in the breadth of organizations that had achieved an NME. From 1960 to 1980, the number of NMEs continued to increase but the number of organizations that had obtained an NME reached a plateau, with few new entries or exits. Starting in the 1980s, a rather dramatic volatility began in which new companies entered and exited the industry. The entries largely represented start-up companies, largely biotechnology companies, whereas the exits were largely caused by consolidations from mergers and acquisitions (rather than bankruptcies). Such changes have given rise to an industry characterized by four types of organizations. Most organizations were awarded or control only one NME. These small ‘singlet’ companies are often the target of acquisitions, usually by mediumsized peers. These medium companies, which control between two and 20 NMEs, are occasionally acquired whereas a few merged to form larger corporations. The third tier consists of large, established companies, which hold a disproportionately large number of NMEs. These companies invariably rose from the middle tier as a result of mergers and none grew without at least one major merger. Notably, the top ten companies control two-thirds of all NMEs (as of the end of 2013). This represents a 65% rise from 1950, when the top ten companies controlled roughly 40% of all NMEs. The fourth type of company (which can include examples from the others) consists of companies that primarily or only function as sales, marketing and manufacturing engines. These companies have
minimized internal discovery and development activities and, instead, focus on obtaining NMEs from licensing, mergers and acquisitions. Although volatility continues, two potentially important changes were noted. First, the number of exits from the drug discovery industry is beginning to exceed the number of entries. It is generally recognized that the larger pharmaceutical organizations have been scaling back their internal R&D activities. Our data suggest that this rebalancing began more than two decades ago and, in part, gave rise to the aforementioned rise of biotechnology and increasing turnover of companies awarded NMEs. As will be detailed in a future report, some of the early scaling-back of the pharmaceutical industry arose as a result of increased risk awareness as pharmaceutical companies licensed or acquired projects from biotechnology companies. The present findings suggest that organizations that obtain NMEs are efficiently eliminated by acquisitions. This has generally been a hallmark of the biotechnology industry. However, the second notable change is the rapid rise of acquiring organizations with limited internal R&D capabilities. In light of tightening venture funding for early-stage biotechnology companies, this raises concerns that the R&D infrastructure for drug development could be progressively and perhaps irreversibly shrinking and that our ability to discover and/or develop new medicines is being progressively dismantled. Clearly, the need for new medicines persists. It will be important to help identify and utilize such trends to develop new mechanisms to address unmet medical and commercial opportunities. Much impressive prior work has shed considerable insight into changing trends in regulatory control and processes [6,7]. We seek to utilize this new database of FDAapproved NMEs as a ‘big data’ means to analyze long-term trends based on medical or scientific criteria. Specifically, the information gathered will provide future opportunities to compare and contrast the changes in different therapeutic areas (e.g. changes in oncology or infectious diseases), drug types (small molecule and biologics), contributors (academic, government, biotechnology, pharmaceutical) and mechanisms of action.
Acknowledgements The authors thank Drs Jon Soderstrom and Richard Foster for helpful discussions and inspiration; and Drs Jane Merkel and Mark Plummer for critical reading of the manuscript. The work herein was performed as part of an initiative by the Yale Center for Molecular
Drug Discovery Today Volume 19, Number 8 August 2014
References 1 Mossinghoff, G.J. (1999) Overview of the Hatch–Waxman Act and its impact on the drug development process. Food Drug Law J. 54, 187–194 2 Katz, J.A. (2013) COX-2 inhibition: what we learned – a controversial update on safety data. Pain Med. 14 (Suppl. 1), 29–34
3 GlaxoSmithKline Press release (2014) GSK to discontinue manufacture and sale of the BEXXARAˆ1 therapeutic regimen (tositumomab and iodine I 131 tositumomab). Available at: http://us.gsk.com/html/media-news/ pressreleases/2013/GSK-to-discontinue-manufactureand-sale-of-the-BEXXAR-therapeutic-regimen.html 4 Hayes, A.N. and Gilbert, S.G. (2009) Historical milestones and discoveries that shaped the toxicology sciences. EXS 99, 1–35 5 Jeffreys, D., ed. (2008) Aspirin: The Remarkable Story of a Wonder Drug, Bloomsbury Publishing 6 Kaitin, K.I. and DiMasi, J.A. (2011) Pharmaceutical innovation in the 21st century: new drug approvals in the first decade, 2000–2009. Clin. Pharmacol. Ther. 89, 183–188
7 Kaitin, K.I. (2010) Deconstructing the drug development process: the new face of innovation. Clin. Pharmacol. Ther. 87, 356–361
Michael S. Kinch, Austin Haynesworth, Sarah L. Kinch, Denton Hoyer Yale Center for Molecular Discovery, Yale University, 600 West Campus Drive, West Haven, CT 06516, USA
Features PERSPECTIVE
Discovery (YCMD) to develop a comprehensive screening strategy for drug re-purposing. Please contact Dr. Michael Kinch if you or your institution would like to investigate potential interactions with YCMD.
PERSPECTIVE
www.drugdiscoverytoday.com
1039
