public and private sector contributions to the discovery ...ipasa.co.za/downloads/value of medicines...

Public and Private Sector Contributionsto the Discovery and Development

of ‘Impact’ Drugs

Janice M. Reichert, PhD and Christopher-Paul Milne, DVM, MPH, JD

A Tufts Center for the Study of Drug Development White Paper

M A Y 2 0 0 2

Executive SummaryRecently, well-publicized reports by PublicCitizen and the Joint Economic Committee(JEC) of the U.S. Congress questioned the roleof the drug industry in the discovery anddevelopment of therapeutically importantdrugs. To gain a better understanding of therelative roles of the public and private sectorsin pharmaceutical innovation, the Tufts Centerfor the Study of Drug Development (TuftsCSDD) evaluated the underlying NationalInstitutes of Health (NIH) and academicresearch cited in the Public Citizen and JECreports, and performed its own assessment of the relationship of the private and publicsectors in drug discovery and development of21 ‘impact’ drugs. The following are the majorfindings of this analysis:

� In August 2001, NIH released A Plan toEnsure Taxpayer’s Interests Are Protected inresponse to Congress’s concern that NIH issecuring appropriate return on its invest-ment in basic research for therapeuticdrugs with more than $500 million peryear in U.S. sales. NIH identified 47 drugsthat fit the sales threshold criteria, but only4 drugs for which the government couldclaim “use or ownership rights.” This resultconflicts with the findings of the PublicCitizen and JEC reports because differentmethods were used to assess NIH “owner-ship” of the drugs.

� In February 2000, the NIH completed theadministrative document NIH Contributionsto Pharmaceutical Development — Case StudyAnalysis of the Top-Selling Drugs to deter-mine the extent to which publicly fundedresearch contributed to the development ofcertain medically or commercially success-ful products. Their methodology consistedof reviewing the scientific literature for

the names and affiliations of scientists responsible for basic research that led todrug discovery. This methodology overesti-mated the public sector contributionbecause of its focus on basic research, andunderestimated the private sector contribu-tion because industry scientists have lessincentive to publish. Also, information onfunding and co-authorship by academiaand industry were incomplete.

� Economists Iain Cockburn and RebeccaHenderson have written several papers eval-uating the relative contributions of the private and public sectors to the discoveryand development of 21 drugs “having hadthe most impact” on therapeutics from1965 to 1992. In their analysis, the authorsdetermined which sector was responsiblefor the key enabling discovery and firstsynthesis of the drug. While the PublicCitizen and JEC reports both note thatapproximately 75% of the key enabling dis-coveries were made in the public sector,they do not acknowledge that 78% of thesedrugs were first synthesized by pharmaceu-tical industry scientists.

� Tufts CSDD’s investigation of the 21 ‘impact’drugs found that using simple publicationcounts is not a reliable method to quantifythe relative contributions of the public andprivate sector for the following reasons:industry contributions tend to be underes-timated; the relevance of NIH-funded clini-cal studies cannot be determined; and, thesubstantial amount of non-U.S. researchand development undertaken on thesedrugs complicates the valuation of theroles of the U.S. public and private sectors.

Ultimately, any attempt to measure the relativecontribution of the public and private sectorsto the R&D of therapeutically important drugsby output alone, such as counting publicationsor even product approvals, is flawed. NIH is

1

not in the business of marketing drugs, anymore than pharmaceutical industry scientists are employed solely to author jour-nal publications. By the same token, it isequally flawed to measure relative contribu-tions to the R&D of innovative drugssolely by the source and size of themonetary investment. Several keyfactors — e.g., the degree ofuncertainty, the expected marketvalue, and the potential socialbenefit — affect investment

decisions, and determine whether public orprivate sector funds, or both, are most appro-priate. Because of the competitiveness andcomplexity of today’s R&D environment, bothsectors are increasingly challenged to show

returns on their investment, and thetraditional boundaries separatingthe roles of the private and publicresearch spheres have becomeincreasingly blurred. What

remains clear, however, is thatthe process still starts withgood science and ends withgood medicine.

2

Ultimately, any attempt to measure the relative

contribution of the public and private sectors

to the R&D of therapeutically important drugs

by output alone, such as counting publications or

even product approvals, is flawed. NIH is not in

the business of marketing drugs, any more than

pharmaceutical industry scientists are employed

solely to author journal publications.

IntroductionA recent, widely publicized report has dispar-aged the contribution of the drug industry tothe development of important new drugs. InJuly 2001, Public Citizen, a consumer watch-dog group, released a report called Rx R&DMyths: The Case Against the Drug Industry’sR&D “Scare Card,” which stated that, “IndustryR&D risks and costs are significantly reducedby taxpayer-funded research, which hashelped launch the most medically importantdrugs in recent years….”1 Industry critics havealso used a report entitled The Benefits of Medical Research and theRole of the NIH, which wasissued in May 2000 by theJoint Economic Committee(JEC) of the U.S. Congress, tominimize the role of industryin the discovery and develop-ment of innovative drugs. TheJEC report focused on a groupof 21 drugs introduced from1965 to 1992 “…that were con-sidered by experts to have hadthe highest therapeutic impacton society.…”2 Public Citizen and the JECcited research done by the National Institutesof Health (NIH), as well as a study done byeconomists Iain Cockburn and RebeccaHenderson, in support of their claims.

The findings of the Public Citizen reportregarding the relative contribution of the public and private sector to drug discoveryand development were based on one docu-ment in particular — an internal NIH studyentitled NIH Contributions to PharmaceuticalDevelopment — Case Study Analysis of the Top-Selling Drugs. This document, which wasmade widely available by Public Citizen inJuly 2001, purported to demonstrate therespective roles of the public and private sectors in the development of five top-sellingdrugs through analyses of published literature.However, a fundamental defect underlyingstudies of this kind is the subjective nature

of assigning relevance. A count of publishedstudies of specific drugs cannot be used toassign values to relative public and privatecontributions to their development because therelevance of the studies to the ultimate approvalof the drugs often cannot be determined.

The impact of publicly funded biomedicalresearch on private sector drug discovery anddevelopment efforts was also investigated byCockburn and Henderson. The focus ofCockburn and Henderson’s work was the relationship between private sector firms’ participation in “open science,” as measured

by counting coauthorship ofscientific papers with publicsector scientists, and the pro-ductivity of their in-houseresearch.3 Cockburn andHenderson used case historiesof 21 drugs to illustrate theirpoint, but they did not attemptquantitative analyses of therelative contributions of thepublic and private sectors tothe discovery and developmentof these drugs. In fact,

Cockburn and Henderson acknowledged intheir paper that it was difficult to quantify thepublic sector’s specific contribution to theindustry’s pool of knowledge capital. Theauthors surmised that this difficulty was dueto the long lags between fundamental discov-eries and consequent marketed products, andthe “…complex and often bidirectional rela-tionship between the public and private sec-tors….”4 The authors emphasized that theiranalysis was an attempt to measure the extentand nature of the “connectedness” betweenthe public and private sectors. Specifically,they examined the ability of industry to accessthe common pool of useful knowledge gener-ated by public sector research, therebyenhancing industry’s productivity.5 Despitemethodological limitations noted by theauthors, Cockburn and Henderson’s work has

3

A count of published

studies of specific

drugs cannot be used

to assign values to

relative public and

private contributions

to their development.

been used to paint an unflattering picture ofthe contribution of industry to the researchand development of many breakthrough drugs.

A clear counterpoint to the JEC and PublicCitizen reports emerged from the most recentreport by NIH on this contentious and com-plex subject. In August 2001, the NIH releasedA Plan to Ensure Taxpayers’ Interests AreProtected.6 The plan was devised in responseto the Committee Report for FY 2001Department of Health and Human Services(DHHS) Appropriation, in which the NIH wasinstructed to pre-pare a plan toensure taxpayers’interests are protected whenNIH invests inbasic research.Specifically, NIHwas directed toreview a list ofFDA-approvedtherapeutic medicines that had $500million/year sales in the United States, andhad received NIH funding, when preparingthe plan.

According to NIH, only four of 47 drugs meet-ing the criteria were developed with patentedtechnologies for which the government has useor ownership rights. This finding was derivedfrom a review methodology that focused onpractical considerations of the NIH’s mission,and on legal aspects of the Stevenson-WydlerTechnology Innovation Act and the Patent andTrademark Amendments of 1980 (known asthe Bayh-Dole Act), as implemented by NIH.The picture of the relative contribution of thepublic and private sector to drug developmentwhen this methodology is employed bearssharp contrast to the image portrayed in theJEC and Public Citizen reports.

In this white paper, we first review the perti-nent points of the recent analysis conducted

to formulate the NIH plan to protect taxpayers’interests regarding NIH-funded research. Wedemonstrate that assessment of governmentpatent rights as a measure of NIH “owner-ship” of a drug brings a new perspective to thequestion of whether the public is getting anappropriate return on the NIH investment inbasic research. We next examine the studiesby NIH (as cited in the Public Citizen report)and Cockburn and Henderson, which usedpublications as a measure of public and privatecontributions to the discovery and developmentof specific best-selling, ‘impact’ drugs. These

studies in partic-ular have beenused to implythat NIH has“ownership,” atleast in part, ofthe drugs. Weexplore themethodologyused in thesestudies, and dis-

cuss their limitations. Finally, we provide anindependent assessment of the very complexand interdependent relationship of the publicand private sectors in drug discovery anddevelopment.

NIH response to the Committee Reportfor the FY 2001 DHHS Appropriationinstruction:A Plan to Ensure Taxpayers’ Interests AreProtected (released August 2001)

The following instructions were given to NIHin the Committee Report for the FY 2001DHHS Appropriation:

“The conferees have been made aware of thepublic interest in securing an appropriatereturn on the NIH investment in basicresearch. The conferees are also aware of themounting concern over the cost to patients oftherapeutic drugs. By July 2001, based on alist of such therapeutic drugs which are FDA

4

. . . assessment of government patent rights

as a measure of NIH “ownership” of a drug

brings a new perspective to the question

of whether the public is getting an

appropriate return on the NIH investment

in basic research.

approved, have reached $500 million per yearin sales in the United States, and have receivedNIH funding, NIH will prepare a plan toensure that taxpayers’ interests are protected.”7

In responding to the instructions, NIH notedthe legal framework into which it is bound,described the process NIH uses to fundresearch, discussed the methodology and find-ings, then proposed a plan. The overall con-clusion was that, “NIH and its recipient insti-tutions apply the provisions of Bayh-Dole tobest advantage in seeking the optimal returnon investment in terms of public health bene-fit.”8 After noting the difficulty in associatingparticular NIH grants and contracts that gaverise to inventions with patents or licenses forthe final product, NIH proposed a plan to initi-ate better information collection, and develop-ment of a web-based database that wouldallow NIH to make the required associations.

The key phrase of the conclusion was “applythe provisions of Bayh-Dole.” The NIH hasstrict limitations under Bayh-Dole — NIH doesnot have title to grants-supported or contracts-supported (extramural) research discoveries,and cannot dictate terms for licensing or com-mercialization of the work. Support for thisextramural research, which is done by non-Federal researchers at academic, medical, andresearch institutions in the U.S. and abroad,accounts for nearly 84% of the NIH budget.9

As per the instructions, NIH identified 47drugs that fit the designated criteria, and thendetermined whether the government, eitherdirectly or through a grantee or contractor,held patent rights, or was designated as hav-ing an interest on the patents of the 47 drugs.NIH found that, “NIH has Government use orownership rights to patented technologies usedin the development of four of those drugs.”10

NIH case study report: NIH Contributions to PharmaceuticalDevelopment — Case Study Analysis of the Top-Selling Drugs (Administrative documentprepared by NIH staff, February 2000)i

The NIH case study report was undertaken todetermine whether and to what extent publicfunding of research enabled the developmentof certain medically or commercially success-ful products. Additionally, this study began tolay a basis for discussing the specific ways bywhich those who expand fundamental under-standing of the workings of the natural worldare as important to technological advance asthose who implement that knowledge.11

The NIH case study report was prepared froman analysis of review articles identified byMedline searches of the chemical name of thedrugs and original research articles cited by thereviews. The literature selected for inclusionin the report was focused on basic research,as stated in the methodology section: “The scientific discoveries that led to the necessaryconcepts and techniques were identified,along with the names and affiliations of thescientists performing the work. Rather thanattempt to identify a small number of “keypapers,” which does not accurately representthe way scientific ideas develop in the researchcommunity, the approach taken was to identifymajor areas of research which led to drug dis-covery and the individuals or laboratories whowere significantly involved.” [emphasis added]Despite the focus on basic research, the reportincluded sections listing publications in theareas of “drug development and testing” and“clinical trials.”

5

i Public Citizen acquired the NIH case study report through a Freedom of Information Act request andposted the report on their website on July 23, 2001. The NIH case study report is cited in the JEC reportreleased in May 2000.

6

We examined the NIH case study report andfound the study to be limited in the followingways:

� The methodology is inherently biasedtoward public sector input. Specifically,public sector scientists have a muchgreater incentive topublish than doindustry scientists.

� Industry contribu-tion is underesti-mated because anumber of publica-tions co-authored by academic andindustry scientistsare assigned only topublic sector affilia-tion. A category for the co-authored publi-cations, the main focus of Cockburn andHenderson’s work, does not exist in theNIH case study report.

� The list of published studies included inthe sections on “drug development andtesting” and “clinical trials” is not com-plete. Therefore these sections cannot beused for quantitative analyses.

� Industry contribution could be underesti-mated, since complete information on thefunding of the published studies is not pro-vided. Thus, though the scientist’s affiliationwas academic, funding for the work mighthave been wholly, or in part, from industry.

� No information is given as to whether theNIH grants were for extramural or intra-mural research. The distinction has impli-cations due to the legal limitations on NIH“ownership” of technology resulting fromextramural work.

Cockburn and Henderson papers:Public-Private Interaction and the Productivityof Pharmaceutical Research (4/97 working paper)

Absorptive Capacity, Coauthoring Behavior, andthe Organization of Research in Drug Discovery(6/98)

Iain Cockburn andRebecca Henderson havewritten a number of paperson the topic of public-private interaction inpharmaceutical research.Their work has been citedby NIH, the JEC, and byPublic Citizen. As part oftheir work, Cockburn andHenderson constructed

case histories for a group of 21 drugs that were considered to have hadthe most impact upon therapeutic practicebetween 1965 and 1992. They then attemptedto answer two questions concerning the drugs.These two questions were:

1. Was the key enabling discovery made by ascientist working in the public sector? (Yes/No)

2. Was the drug first synthesized or isolated bya scientist working in the public sector? (Yes/No)

Answers to the first question led Cockburnand Henderson to state that “only 5 of thesedrugs, or 24%, were developed with essentiallyno input from the public sector.”12,13

This particular statement has been cited invarious ways. The JEC cites the Cockburn andHenderson working paper and states, “71 per-cent (15 drugs) were developed with inputfrom the public sector.”14 Public Citizen citesthe same paper and states, “A study by aMassachusetts Institute of Technology (MIT)scholar of the 21 most important drugs intro-duced between 1965 and 1992 found that

publicly funded research played a part in dis-covering and developing 14 of the 21 drugs(67 percent).”15 The NIH case study reportcites an unpublished version of the work pre-pared for a conference in June 1996 anddeclares, “among these 21 drugs, publicly-funded research was instrumental to thedevelopment of 16, or 76%.”16

These statements by JEC, Public Citizen, and NIH are alike in one important way —none of them mention the fact that the “publicly-funded research” was for keyenabling discoveries only. Cockburn andHenderson provide this information for 19 of the 21 drugs in the working paper (key enabling discoveries for 14 are assignedto public sector scientists). Interestingly, JEC,Public Citizen, andNIH make no refer-ence to the answersto their second ques-tion. Cockburn andHenderson’s second result was that 14 (78%)of the drugs were first synthesized by industrialscientists (information provided for 18 of the21 drugs).

Use and misuse of the NIH case study report, and Cockburn andHenderson’s workThe NIH case study report and Cockburn andHenderson’s work are in agreement in theirfinding that publicly funded research effortsin medicine and the pharmaceutical industry’sresearch and development efforts are, on thewhole, complementary. Publicly fundedresearch tends to focus on basic science, whilethe pharmaceutical industry tends to focus onapplied research and clinical studies of prom-ising drugs, though there is some overlap ofthese areas.

The NIH should be recognized for its long history of achievements and contributions to

medicine, and for its five Nobel laureate scien-tists. In addition, NIH has funded excellentscience over the years, as evidenced by the 93Nobel Prizes won by NIH-funded scientists.17

The pharmaceutical industry should also berecognized for its excellent research anddevelopment efforts, which have resulted inthe approval for marketing of 438 new medi-cines in the U.S. during the last 20 years,18

and of the three scientists who won NobelPrizes while working in the private sector.

Both NIH and the research-based pharmaceu-tical industry have sponsored studies designedto illustrate the importance of their work.Issues arise when “facts,” some of which areincorrect or used out of context, from thesestudies are used in an attempt to demonstrate

the greater value ofone group’s effortscompared to theother. Both the NIHand the pharmaceuti-

cal industry are vulnerable to criticism. NIH issensitive to criticism that a proportion of itsfunding goes to projects that do not producetangible results. The pharmaceutical industryis equally sensitive to criticism that all thehard work of drug discovery is done by pub-licly funded researchers, and that industryunfairly profits by simply marketing theresulting drugs.

Counting publications that involve use of thedrug is a biased way of assigning “ownership”of the drug. NIH’s “ownership” of medicalresearch cannot be determined by countingthe number of new medicines it marketsbecause NIH is not in the business of market-ing drugs. By the same token, the pharmaceu-tical industry’s “ownership” of marketed drugscannot be determined by counting publica-tions because industrial scientists typically areemployed to produce commercial products,not publications.

7

Both the NIH and the pharmaceutical

industry are vulnerable to criticism.

Tufts CSDD investigation: Public and private sector contributionsto the discovery and development of‘impact’ drugsWe independently studied the case histories ofthe 21 ‘impact’ drugs included in the Cockburnand Henderson work. Using an approach sim-ilar to that described in the NIH case studyreport, we searched the published literaturefor information on the discovery and develop-ment of the drugs. We noted the affiliationand location of the authors, the source offunding for the work (if provided by authors),and the date of the study (if provided) or pub-lication. We restricted our search to papers thatpertained to a given drug and were publishedbetween the year of the synthesis of the drugand two years after U.S. approval of the drug.ii

We did not duplicate the study of the enablingdiscoveries done both by NIH and by Cockburnand Henderson, but focusedonly on published studiesthat reported properties(e.g., chemical, physical,toxicologic, pharmacoki-netic/dynamic, therapeutic)of the actual drugs. The 21drugs we studied are listedin Table I.

The following pointsshould be noted:

� The limitations inherentin our methodologywere the same as for theNIH case study report — industry contri-butions tended to be underestimatedbecause contributions were assigned topublished work only.

� Our results, like those in the NIH case studyreport, cannot be used for quantitativeanalyses — we identified hundreds of pub-lications that met our criteria; some of thepublications were hundreds of pages inlength (these were proceedings of symposia).It was not possible to ascertain whether wefound all pertinent publications (e.g., onlypublications in English were included).

� There was substantial non-U.S. involvementin the discovery and development workdone for the 21 drugs — the majority of thedrugs were synthesized, patented, or firstlaunched outside of the U.S., which com-plicated any analyses of the relative contri-butions of the public and private sectors.

� The relevance of any clinical studies fundedwholly or in part by the NIH to the approvalof the drug could not be determined.

Our study illustrates thecomplex interactions of theU.S. public sector, especiallythe NIH, and the private sector in the discovery anddevelopment of drugs. Forthis set of 21 drugs, we notedthat the involvement of NIH,usually in the form of extra-mural research funding, wasgreatest in the preclinicaland clinical development ofdrugs that were treatmentsfor serious or life-threatening

diseases, such as AIDS or cancer. There wasclearly a public health benefit derived fromfacilitating the development of these drugs.NIH was also involved in the discovery and/ordevelopment of compounds that were in the“public domain,” i.e., knowledge of the exis-tence and method of preparation of the

8

ii Additional clinical testing funded by the U.S. public sector may have occurred after the drugs wereapproved for marketing. It is important to note that after approval, drugs may be tested in clinical studiesby any qualified investigator. Since the drug is commercially available, the FDA does not necessarily haveto be informed of the clinical studies (although IRB and informed consent regulations still apply), andinvolvement or consent of the manufacturer is not required.

9

Table IYear

of syn- Location First Year Year Name of thesis of company Non-US Non-US marketed of of sponsoring Generic Trade (bio. head- syn- patent outside first UScompany name name activity) quarters thesis priority US launch launch

Burroughs UKWellcome acyclovir Zovirax 1974 (US facility) N Y Y 1981 1982

Burroughs 1963 UK Wellcome AZT Retrovir (1980) (US facility) N Y N 1987 1987

Squibb captopril Capoten 1974 US N N N 1981 1981

Smith US Kline French cimetidine Tagamet 1972 (UK facility) Y Y Y 1976 1977

1845Bristol-Myers cisplatin Platinol (1965) US Y Y N 1979 1979

1970 Sandoz cyclosporin Sandimmune (1972) Switzerland Y Y N 1983 1983

1985 Amgen erythropoietin Epogen (1970) US N N Y 1988 1989

Merck finasteride Proscar 1983 US N N N 1992 1992

US Pfizer fluconazole Diflucan 1981 (UK facility) Y Y Y 1988 1990

Eli Lilly fluoxetine Prozac 1970 US N N Y 1986 1988

1924 Astra foscarnet Foscavir (1978) Sweden Y Y Y 1989 1991

USPfizer gemfibrozil Lopid 1968 (UK facility) N N N 1982 1982

Berlex/Chiron interferon ß-1b Betaseron 1980 Germany/US N N N 1993 1993

Janssen ketoconazole Nizoral 1976 Belgium Y N N 1981 1981

Merck lovastatin Mevacor 1979 US N N N 1987 1987

Bayer nifedipine Procardia 1967 Germany Y Y Y 1975 1982

Astra omeprazole Prilosec 1979 Sweden Y Y Y 1988 1989

Glaxo ondansetron Zofran 1983 UK Y Y Y 1990 1991

ICI (Zeneca) propranolol Inderol 1964 UK Y Y N 1967 1967

Glaxo sumatriptan Imitrex 1984 UK Y Y Y 1991 1993

ICI (Zeneca) tamoxifen Nolvadex 1962 UK Y Y Y 1973 1978

Note: Date of discovery of biological activity is listed separately if different from date of synthesis.

compounds was publicly available before therapeutic potential was identified. Three of the drugs (AZT, cisplatin, foscarnet) were synthesized well before biological activity wasobserved. These types of compounds initiallymight not have been of interest to the phar-maceutical industry because possible patentclaims were limited.

Having acknowledged the con-tribution of the public sector,in particular the NIH, in thedevelopment of the impactdrugs just discussed, it isimportant to note that NIHcannot patent or in any otherway claim “ownership” ofdrugs simply by funding studies. It is also necessary toplace the part played by NIHfunded trials in the overall picture of drugdevelopment. No one would argue that duringthe period in which many of the 21 drugsfrom our review were in active development,public funding in real terms in the U.S. forhealth-related research increased by 200%.19

However, the majority of that health-relatedresearch was not focused on pharmaceuticals.NIH’s definition of “clinical research” hasbeen criticized as being too inclusive, since itencompasses not only clinical trials but alsomechanisms of human disease, therapeuticinterventions, development of new technolo-gies, epidemiologic and behavioral studies, aswell as outcomes and health services research.20

According to a recent study by the GeneralAccounting Office (GAO), even NIH’s use of

the term “clinical trials” includes a range ofresearch activities encompassing testing of newapproaches to disease prevention, diagnosis, ortreatment.21 The GAO report further acknowl-edges that while both NIH and pharmaceuticalcompanies are the major sponsors of clinicaltrials that focus on drugs, devices and vaccines,the NIH-supported trials also address preven-

tion strategies and surgicalprocedures, and may targetspecial populations, such aspatients with rare diseases.22

The pharmaceutical industry,in contrast, typically supports the large clinical trials thatdetermine therapeutic efficacyof new drug products for conditions that affect largenumbers of people.23

“Nobel Prize winning” drug research:the product of time, capital, and manpower The importance of the contribution of the pri-vate sector to drug discovery and developmentwas acknowledged when the 1988 Nobel Prizein Physiology or Medicine was awarded to Dr.George Hitchings, Ms. Gertrude Elion, and SirJames Black for their discoveries of importantprinciples for drug treatment. Hitchings andElion, working at Burroughs Wellcome, con-tributed to the discovery and development ofacyclovir.iii Black contributed to the discoveryand development of propranolol and cimeti-dine, while working at Imperial ChemicalIndustries and later SmithKline & French.iv

10

iii Hitchings joined Burroughs Wellcome in 1942, and became Vice President in charge of research in1967. Elion joined Burroughs Wellcome in 1944 and was Head of the Department of ExperimentalTherapy from 1967 to 1983.iv Black was a research scientist at Imperial Chemical Industries (ICI), where he worked from 1958 to1964 on the ß-receptor antagonist program that resulted in the discovery and development of propranolol.He then moved to SmithKline & French where he worked from 1964 to 1972 on the H2-receptor antago-nist program that resulted in the discovery and development of cimetidine.

Drug development is

a resource-intensive

undertaking that

requires commitment

of time, capital, and

manpower.

The history of the discovery and developmentof the drugs that garnered their inventors thecoveted Nobel Prize illustrates another point— drug development is a resource-intensiveundertaking that requires commitment of time,capital, and manpower. For example, BurroughsWellcome’s antiviral discovery program wasactive for six years before acyclovir was syn-thesized in 1974 at the U.S. facility. In thedecade that followed, further R&D involvingbiological activity screens (plaque reductionand inhibition) and preclinical work (mouse,guinea pig, rabbit models) was done at thefacility in England, ultimately requiring 330‘scientist years’ and millions of pounds sterling.24

Similarly, Imperial Chemical Industries’ ß-receptor antagonist program was active forsix years before propranolol was synthesizedin 1964. Sir James Blacknoted that his co-workers’use of deductive organicchemistry, understandingof the link between drugdelivery and effect, anddevelopment of analyticalmethods for estimating thelevels and tissue distribu-tion of a drug and itsmetabolites were crucialto the discovery and devel-opment of propranolol.25

Finally, SmithKline & French’s histamine H2-receptor antagonist program was initiatedin 1964. By mid-1970, over 700 compoundshad been synthesized and tested for bioactivity;several (burimamide and metiamide) were tested in humans. Approximately 150 scientistswere involved in the program. Cimetidine wassynthesized in 1972, tested in normal volun-teers in March 1975, first given to patients inNovember 1975, and marketed in England inNovember of 1976. The rapid pace of thedevelopment program was a result of the

knowledge gained from the work that hadbeen done on the precursor compounds.

These examples illustrate the point that drugdiscovery is a time-consuming process requir-ing the synthesis and in-vitro screening of manycompounds before one with the best proper-ties is selected for clinical testing. A priori,there is no guarantee that a drug discoveryprogram will be successful. Moreover, evenafter a promising candidate is discovered,there are many more steps to the process (e.g., formulation, stability testing, preclinicaltesting) before the drug can be studied in theclinic (see Table II).

Success Has Many Fathers The composite history of the 21 impact drugsreveals that the public and private research

spheres are symbiotic, sus-taining each other in bothexpected and unexpectedways. The public sector isoften responsible for basic science research that laysthe groundwork for newdrugs (e.g., such as moderngenetics did for interferonbeta-1b), or even researchinto new uses for olddrugs. Still other publicsector research contributes

to the clinical knowledge used in the design ofefficacy tests for discovered drugs (e.g., finas-teride), or epidemiological and long-termhealth outcome studies (e.g., the interventionstudies of gemfibrozil by the U.S. Dept. ofVeterans Affairs) needed to establish newdirections for drug related studies, so-calledagenda-setting research.26

But even those functions are interchangeable,as private sector development of new thera-peutic agents often benefits basic research inthe public sector by enabling scientists toexplore the way the body works at the cellularlevel, such as occurred with cyclosporin and

11

12

Time

Thre

e to

Fiv

e Ye

ars

Production Processes, Equipment,Materials, and Personnel

NIH extramural grants support research in over2,000 universities, medical schools, hospitals,small businesses, and research institutions inthe U.S. and abroad.

Combinatorial chemistry using robots to synthesize thousands of molecules.

High through-put screening for toxicity and biological activity using robotics andminiaturized formats.

In vitro testing with human liver microsomes

Tests involve the use of animal tissues, isolatedcell cultures, isolated enzymes, cloned receptorsites and computer models.

Assay lead compound for strength, potency or impurities.

Separate API from impurities, degradationproducts and inert excipients.

From starting material, API needed only ingram quantities.

Involves use of animals, tissue culture, andother test systems to examine relationship of dose, frequency of administration, and duration of exposure to short and long-termsurvival; requires veterinarians, toxicologists,as well as animal and lab technicians.

Ramp-up production of API to kilogram levels;each of multiple steps in synthesis needs to besystematically investigated and scaled-up.

Early batches must be characterized for purity,impurity, physical and chemical attributes to ensure material to be tested in animals and humans is identical to previous lots withconsistent stability and bioavailability.

Costs, Risks, Delays andFailures

Can take decades to produce a drug concept.

Average NIH research grant for new clinical trials was$600,000 in 1996.

At one large pharma companyin 1997, even with the capacity to construct 2 millioncompounds and screen 35 million compounds per year,only 18 entered exploratorydevelopment.

There are multiple sequentialchemical or biological reactionsyielding desired intermediates;entire synthesis pathway mustbe charted and validated.

Time frame for completion ofmultiple sequential stepsincreases from 1 week to 1month for each reaction step.

Costs of animal studies rose 3-6 fold from 1980-1990,testing in 2000 would rangefrom $20 thousand for acuterat toxicity study to $2 millionfor 2-yr. rat bioassay.

Drug has to show an adequatesafety profile in animal toxicol-ogy testing.

Risk of new impurities due to larger scale operations,the addition of formulation excipients, and processingsteps can alter bioavailability in human.

Risk that short-term stabilitytests may not predict long-termresults.

Table IISteps in Discovery,Development and Regulatory Approval

Basic Research: Researchon fundamental mechanismof disease, biological process,and action of known thera-peutic agents � Drug concept.

Discovery: Random selection, broad biologicalscreening, structure/activityrelationship (SAR) analysis,bioinformatics, serendipity� Lead compound (investi-gational drug/biologic).

Synthesis & Early Testing:Chemical synthesis of smallmolecules.

Extraction from naturalsources by fermentation andbiotechnology.

Early pharmacology studies toexplore the pharmacologicalactivity and therapeuticpotential of compoundsbegins� Active PharmaceuticalIngredient (API).

Preclinical, Formulationand Stability Tests: preclin-ical animal pharmacology andtoxicology to determinepotential risks of API to manand the environment.

Manufacturer must demon-strate company’s capacity toproduce a product in largevolume and ensure chemicalstability, batch-to-batch uniformity, and overall product quality.

Dec

ades

13

Time

Thre

e to

Fiv

e Ye

ars

(con

t.)


Develop comprehensive database of physicaland chemical properties of API and biopharma-cological profile of molecule to evaluate impactof lot-to-lot variation on drug performance dueto evolutionary changes in synthesis scale (i.e.,the process of turning active compound into aform and strength suitable for human use).

Must be able to demonstrate sufficient stabilitythat 90% of the API at a minimum, and only1% of degradation products at a maximum, arepresent from time of manufacture until lastsubject, last dose.

Requires preformulation, analytical and formu-lation scientists.

Develop early formulation.

Updating rudimentary API to commercial product may require 6-9 months.

Equipment for typical clinical trial includesinstruments for physical exams & vital signs(e.g., sphygmomanometer), electrocardiogram,clinical lab tests, and miscellanea such as biohazard containers totaling as many as 30separate items.

Disposable materials required for typical clinicaltrial could include as many as 30 items rangingfrom syringes to patient slippers.

Phase I: 20-100 patients and 3 Fiscal TimeEquivalents (FTEs).

Phase II: 100-500 patients and 6-12 FTEs.

Phase III: 1,000-3,000 patients and 12-20FTEs.

Clinical trials involve data collection and management personnel, research physician and nurse time, tests performed for researchpurposes, doctor visits, hospital stays,laboratory tests, and X-rays.


Chemical and biological studies must also be conductedwhenever the dosage form,formulation, or manufacturingprocess is changed.

Formulation development alone can cost $1,000,000.

Because scope of clinical usage is unknown, have tooverproduce investigational product by 25-300%.

Equipment costs: $291-318/participant.

Disposables costs: $123-129/participant.

Phase I failure rate: 25%.

Phase II failure rate: 38.8%.

Phase III failure rate: 13.2%.

Overall, 77% of projects abandoned.

Only 11% of patients whorespond to clinical trial open-ings actually enroll; 80% ofcompanies miss enrollmentdeadlines.

Table II (cont.)Steps in Discovery,Development and Regulatory Approval

Preclinical, Formulationand Stability Tests (cont.):Develop clinical trial designand protocol; select sites andassess qualifications, willing-ness, availability and perform-ance of investigators; prepareinvestigational new drug (IND)application; package, label,and deliver product to sites.

Clinical Studies: First set of bioavailability studies areconducted on healthy humanvolunteers to document rateof absorption and excretionof active ingredients.

Phase I studies establish tolerance of human subjectsto different doses, define itspharmacologic effects atanticipated therapeutic levelsand study its absorption,distribution, metabolism andexcretion patterns in humans.

Phase II controlled clinicaltrials show a compound’spotential usefulness andshort-term risks.

Phase III controlled anduncontrolled clinical trials ofdrug safety and effectivenessfor specific indications,identifying range of adverseeffects, dose level, andmethod of use for labeling.

Four

to

Six

Year

s

Table continued

14

Time

Four

to

Six

Year

s (c

ont.

)


Need bioequivalence studies to prove thatinvestigational product will be equivalent tocommercial product, as well as appropriatelong-term stability testing (requiring 3-6 months).

Develop successor product to early formulationand prototype of commercial product onceproof-of-principle is completed, dose must beknown from clinical studies and dosing frommarket studies.

NDA phase: can involve 20-30 FTEs.

Documentation for drug application can requireas much as 23 linear feet of storage space andattendant costs for materials, labor, shipping,and storage can be $500,000 to $1 million.

Bioavailability studies must be repeated justprior to marketing to ensure that the formula-tion used to demonstrate safety and efficacy inclinical trials is equivalent to the product thatwill be distributed for sale.

Must design, produce, package and label theproduct before marketing.

Phase IV studies can require thousands ofpatients and 10-15 FTEs.


68 trials, 4,235 patients and50-80 FTEs required per newdrug application.

Fully loaded FTE costs$125,000/yr.

Each day of delay can cost $1.3 million in direct and lostopportunity costs.

Costs of pharmacoeconomicstudies can be so substantial as to render the overall cost of drug trials infeasible.

During NDA period, FDA mayask for additional information,clarification, or refuse to file or approve application.

Other countries have differentregulatory processes and requirements, drug may berejected for reimbursement or other reasons, or requireadditional studies.

1-in-5 drugs first tested inhumans receives approval.

Each company pays applica-tion, establishment and product fees to FDA every yeartotaling thousands to millionsof dollars.

Phase IV study can cost $20-30 million.

Steps in Discovery,Development and Regulatory Approval

Clinical Studies (cont.):Long-term animal testingmust now be done.

Conduct pharmacoeconomicstudies comparing cost-effectiveness of investiga-tional drug to competingdrugs or other health careinterventions.

Approval and PostApproval: Company mustsubmit new drug application(NDA) to FDA for marketingapproval.

All information about drugfrom discovery through devel-opment is assembled for NDAand subsequent labeling.

After approval, experimentalstudies and surveillance activ-ities including clinical trials(Phase IV studies) conductedto determine undetectableadverse outcomes (especiallyin sub-populations), and long-term morbidity and mortality profiles continue.

One

to

Two

Year

s

Sources: Bernstein H., Drug Inf J 2000;34(3). Mathieu M, Pharmaceutical R&D Statistical Sourcebook1999:55 & 2000:104. Day et al., Appl Clin Trials 1998 June:71. Pink Sheet, 2000 Apr 17:25. McSweegan,Appl Clin Trials 2000 June; Suppl:12. Shaw I, Scrip Magazine July/Aug 1999:6. Davies L, Appl Clin Trials1998 June:62. Aoki N, Boston Globe 10/11/2000. US GAO. May 2000: Doc. No. GAO/HEAS/AIMD-00-139:3.Hill T. Scrip Magazine 1994; No. 22:28-30. Other sources include various FDA documents and Tufts CSDD publications.

Table II (cont.)

the immune system, and lovastatin and cho-lesterol biosynthesis.27 Even drugs that arecommercial failures for the drug companiescan benefit the public sector because the workdone during discovery and development canyield important insight into human physiologyand biochemistry.28

NIH has long recognized the integral role ofthe pharmaceutical industry. In its 1997Director’s Panel Report, the NIH stated as one of its 10 recommen-dations for the effectivecontinuance of clinicalresearch that NIH should“. . . sustain a productivedialogue on enhancingclinical research with itspartners: the academichealth centers, privatefoundations, and the phar-maceutical and managedhealth care industries.”29

The NIH report notes thatthe $15.1 billion spent byresearch-based pharmaceutical companies in1996 “makes the pharmaceutical industry thelargest funder in the aggregate of clinicalresearch in the United States.”30 It also pointsout “. . . that the percentage of NIH’s contribu-tion to clinical research as a whole, althoughconsiderable. . . may be smaller relative to theother large contributors than was originallythought.”31

Both the public and private sectors bear theburdens and share the benefits of pharmaceu-tical R&D. While the burdens and benefitsmay be different in nature, the risks areincurred in similar terms — time and money.Who should bear these risks? How should limited resources be apportioned amongseemingly limitless needs?

Funding of research, whether by the public or private sector, is an investment decision.The degree of uncertainty, the expected mar-ket value, and the potential social benefit arethe key factors in determining the appropriateroles for the public and private sectors inhealth capital investment decisions.32 Whethersuch funding is best supplied by the public orprivate sector involves the interplay of thesethree key factors. When there is a high level of uncertainty combined with potentially large

social benefits in theabsence of sufficientpatent protection, there is a strong imperativefor public investment.Conversely, when theexpected market value ishigh and future use ispredictable, private sectorfunding should suffice. Agray zone exists, however,whenever there is consid-erable uncertainty. Here a collaboration of public

and private funding is appropriate, with shar-ing of the risks and benefits.33

Pharmaceutical companies are hard-pressedto justify research when there is difficulty inobtaining exclusive economic benefits. Patentprotection is essential for companies investingin pharmaceutical R&D. Unlike many othertechnological advances, a drug product, oncediscovered, is relatively easy to reproduce.Without the period of market exclusivity thatpatents provide, companies would not havethe opportunity to recoup their R&D invest-ments.34 Yet, patents do not grant completemonopoly power in the pharmaceutical indus-try because competitors can discover andpatent similar drugs that use the same basicmechanism to treat an illness. The first drugusing the new mechanism to treat that illness— the pioneer drug — usually has between

15

Both the public and private

sectors bear the burdens

and share the benefits of

pharmaceutical R&D. While

the burdens and benefits may

be different in nature, the

risks are incurred in similar

terms — time and money.

one and six years on the market before a ther-apeutically similar patented drug (“me-too”drug) is introduced.35 In fact, seven of the 21impact drugs examined here had a period ofpioneer exclusivity of six years or less.36 Bothin the private sector and the public sector,“medical research, like medical practice, isincreasingly, and reasonably, challenged toshow value for money.”37 Pharmaceuticalcompanies are no exception, and preservingthe fruits of decades of labor, through patentsand regulatory grants of market exclusivity, isone way firms can show value for money torisk-averse investors.

The public sector alsohas begun to appreciatethe need to protect andprofit from its intellectu-al capital. Over the lasttwo decades, universitieshave made increasing useof the technology transfer

laws passed in the early 1980s, especially theBayh-Dole Act. Since its enactment, the num-ber of patents issued to universities as well asthe number of licenses granted by universitieshave increased ten-fold, and royalties paid touniversities have nearly quadrupled.38 For itspart, NIH recognizes the need to enhance itsstewardship of the public’s investment in drugdiscovery and development by enhancing datacollection and public access to information onNIH funding of inventive research. However,NIH emphasizes that “requiring direct finan-cial recoupment of the federal investment in

biomedical research canpotentially impede thedevelopment of promisingtechnologies.…”39

Moreover, NIH believesthat recoupment strategies“…would destabilize a successful balancebetween public and pri-vate needs for innovationand development.”40

16

SummaryOur analysis indicates that the researchreported in NIH’s case study report and inCockburn and Henderson’s papers has beenboth misinterpreted and inappropriately usedin quantitative analyses of the public and private contributions to drug discovery anddevelopment. Both the NIH’s and Cockburnand Henderson’s methodologies underestimatethe contributions ofthe private sector. Theoutcome of NIH’sanalysis of 47 top-selling drugs under-scores this fact.

By the same token,our review of the history of 21 impactdrugs further illus-trates that it is illusoryto assign “ownership”of drugs categorically. The biological bases ofthe diseases alleviated by the 21 impact drugs,as well as the chemical origins of the drugsthemselves, were the focus of decades of priorresearch efforts. The pieces of these researchpuzzles were pulled together over manydecades, by many researchers from manycountries, working in both the public and private sectors.

The “reality” of drug discovery is that it relieson a complex chain of interrelated events,41

and it involves an incremental learning processthat takes place over time.42 The basic researchthat underlies new therapeutic compounds is

a combination of publicly available biomed-ical knowledge and basic research conductedby firms.43 There is a high degree of complexityand creativity in the process of drug discovery.Nevertheless, there is a progression in researchand learning. To the extent that firms monitorand use publicly available medical knowledgein their research, they can begin the processof drug innovation with something other thana “blank chalkboard.”44

Enormous changeshave occurredwithin the drugR&D environmentsince the time period duringwhich the drugsdiscussed in thispaper first beganthe long road fromtest tube to phar-macy shelves. The

biotechnology industry has flourished as aconsequence of the passage of the Bayh-DoleAct and the availability of collaborations andalliances with major pharmaceutical firms.Today, the boundaries between publicly fundedand privately sponsored medical research,which were never sharply defined, are evenmore unclear.45 Now, as in the past, it is evi-dent that the private sector needs the publicsector “to do good science,”46 while the publicsector needs the private sector to transformthat scientific capital into products that bene-fit society, and thus to do good medicine.

17

It is evident that the private sector

needs the public sector “to do good

science,” while the public sector

needs the private sector to transform

that scientific capital into products

that benefit society, and thus to do

good medicine.

References

1 Public Citizen. Rx R&D myths: the case against the drug industry’s R&D “scare card.” New York:Public Citizen Congress Watch, July 2001:2 http://www.citizen.org/congress/drugs/R&Dscarecard.html

2 Joint Economic Committee, United States Senate. The benefits of medical research and the role ofthe NIH. 106th Congress 2d Session:27 http://jec.senate.gov

3 Cockburn I Henderson R. Private-public interaction and the productivity of pharmaceutical research.NBER working paper 6018. Cambridge MA: National Bureau of Economic Research, April 1997http://www.nber.org

4 Cockburn I Henderson R. Public-private interaction and the productivity of pharmaceuticalresearch. NBER Working Paper 6018. Cambridge MA: National Bureau of Economic Research, April 1997:abstract http://www.nber.org

5 Cockburn I Henderson R. Public-private interaction and the productivity of pharmaceuticalresearch. NBER Working Paper 6018. Cambridge MA: National Bureau of Economic Research, April 1997:2 http://www.nber.org

6 National Institutes of Health. NIH response to the conference report request for a plan to ensure taxpayers’ interests are protected: a plan to ensure taxpayers’ interests are protected. Rockville MD:National Institutes of Health, July 2001 http://www.nih.gov/news/070101wyden.htm

7 National Institutes of Health. NIH response to the conference report request for a plan to ensure taxpayers’ interests are protected: a plan to ensure taxpayers’ interests are protected. Rockville MD:National Institutes of Health, July 2001:2 http://www.nih.gov/news/070101wyden.htm



10 National Institutes of Health. NIH response to the conference report request for a plan to ensuretaxpayers’ interests are protected: a plan to ensure taxpayers’ interests are protected. Rockville MD:National Institutes of Health, July 2001:12 http://www.nih.gov/news/070101wyden.htm

11 National Institutes of Health. NIH response to the conference report request for a plan to ensuretaxpayers’ interests are protected: a plan to ensure taxpayers’ interests are protected. Rockville MD:National Institutes of Health, July 2001: Background section http://www.nih.gov/news/070101wyden.htm

12 Cockburn I Henderson R. Public-private interaction and the productivity of pharmaceutical research.NBER Working Paper 6018. Cambridge MA: National Bureau of Economic Research, April 1997:6http://www.nber.org

13 Cockburn I Henderson R. Absorptive capacity, coauthoring behavior, and the organization ofresearch in drug discovery. Journal of Industrial Economics 1998 Jun;XLVI(2):160


18

15 Public Citizen. Rx R&D myths: the case against the drug industry’s R&D “scare card.” New York:Public Citizen Congress Watch, July 2001:10 http://www.citizen.org/congress/drugs/R&Dscarecard.html

16 National Institutes of Health. NIH response to the conference report request for a plan to ensuretaxpayers’ interests are protected: a plan to ensure taxpayers’ interests are protected. Rockville MD:National Institutes of Health, July 2001 http://www.nih.gov/news/070101wyden.htm


18 Reichert J Chee J Kotzampaltiris C. The effects of the Prescription Drug User Fee Act and theFood and Drug Modernization Act on the development and approval of therapeutic medicines. Drug Information Journal 2001;35(1):91-2


20 National Institutes of Health. The NIH Director’s Panel on Clinical Research Report to the Advisory Committee to the NIH Director, December, 1997: executive summary. Rockville MD: NationalInstitutes of Health, July 2001:1 http://www.nih.gov/news/crp/97report/exexsum.htm

21 United States General Accounting Office. NIH clinical trials: various factors affect patient participation. GAO/HEHS-99-182. September 1999:3 http://www.gao.gov

22 United States General Accounting Office. NIH clinical trials: various factors affect patient participation. GAO/HEHS-99-182. September 1999:3 http://www.gao.gov

23 United States General Accounting Office. NIH clinical trials: various factors affect patient participation. GAO/HEHS-99-182. September 1999:3-4 http://www.gao.gov

24 Vane J. Pathways to New Drugs. Scandinavian Journal of Infectious Diseases Supplementum1985; 47: 5-8.

25 Black J. Autobiography. Nobel e-Museum. http://www.nobel.se/medicine/laureates/1988/

26 Cockburn I Henderson R. Public-private interaction and the productivity of pharmaceuticalresearch. NBER Working Paper 6018. Cambridge MA: National Bureau of Economic Research, April 1997:8-9 http://www.nber.org


28 Cockburn I Henderson R. Public-private interaction and the productivity of pharmaceuticalresearch. NBER Working Paper 6018. Cambridge MA: National Bureau of Economic Research,April 1997:12 http://www.nber.org



19


32 Gelijns AC Rosenberg N Moskowitz AJ. Capturing the unexpected benefits of medical research.New England Journal of Medicine 1998 Sep 3;339(10):695

33 Gelijns AC Rosenberg N Moskowitz AJ. Capturing the unexpected benefits of medical research.New England Journal of Medicine 1998 Sep 3;339(10):695-697

34 Strongin RJ. Rx marketplace dynamics. National Health Policy Forum Issue Brief No. 755,George Washington University 2000 May 31:5

35 Congressional Budget Office, United States Congress. How increased competition from genericdrugs has affected prices and returns in the pharmaceutical industry. Washington DC: GPO, July1998:xi http://www.cbo.gov

36 Boston Consulting Group. Sustaining innovation in U.S. pharmaceuticals. [New York]: BostonConsulting Group, January 1996:42


38 Valoir T. Government funded inventions: the Bayh-Dole Act and the Hopkins v. CellPro march-inrights controversy. Texas Intellectual Property Law Journal 2000 Winter;8:211+





43 Toole AA. The impact of federally funded basic research on industrial innovation: evidence from thepharmaceutical industry. Stanford CA: Stanford University Center for Economic Policy Research,June 1998:8

44 Toole AA. The impact of federally funded basic research on industrial innovation: evidence from thepharmaceutical industry. Stanford CA: Stanford University Center for Economic Policy Research,June 1998

45 Rettig RA. The industrialization of clinical research. Health Affairs 2000 Mar-Apr;19(2):131


20

21

Funding for this study was provided in part by a pharmaceutical industry grant. Theauthors thank Elaine Bergman, Dorothy McMindes, and Cherie Paquette for researchassistance and technical assistance in manuscript preparation, and Kenneth Kaitin forhelpful discussions and editorial suggestions.

Tufts Center for the Study of Drug Development

Tufts University192 South Street, Suite 550

Boston, MA 02111 USA

Tel 617-636-2170Fax 617-636-2425

Email [email protected] www.tufts.edu/med/csdd

public and private sector contributions to the discovery ...ipasa.co.za/downloads/value of medicines...

Documents