SPECIAL ARTICLES
`
`Translation of Highly Promising Basic Science
`Research into Clinical Applications
`
`Despina G. Contopoulos-Ioannidis, MD, Evangelia E. Ntzani, MD, John P. A. Ioannidis, MD
`
`PURPOSE: To evaluate the predictors of and time taken for the
`translation of highly promising basic research into clinical ex-
`perimentation and use.
`METHODS: We identified 101 articles, published between
`1979 and 1983 in six major basic science journals, which clearly
`stated that the technology studied had novel therapeutic or pre-
`ventive promises. Each case was evaluated for whether the
`promising finding resulted in relevant randomized controlled
`trials and clinical use. Main outcomes included the time to pub-
`lished trials, time to published trials with favorable results
`("positive" trials), and licensed clinical use.
`RESULTS: By October 2002, 27 of the promising technologies
`had resulted in at least one published randomized trial, 19 of
`which had led to the publication of at least one positive random-
`
`ized trial. Five basic science findings are currently licensed for
`clinical use, but only one has been used extensively for the li-
`censed indications. Promising technologies that did not lead to
`a published human study within 10 to 12 years were unlikely to
`be tested in humans subsequently. Some form of industry in-
`volvement in the basic science publication was the strongest
`predictor of clinical experimentation, accelerating the process
`by about eightfold (95% confidence interval: 3 to 19) when an
`author had industry affiliations.
`CONCLUSION: Even the most promising findings of basic re-
`search take a long time to translate into clinical experimenta-
`tion, and adoption in clinical practice is rare. Am J Med. 2003;
`114:477-484. ©2003 by Excerpta Medica Inc.
`
`Medical progress is highly dependent on the
`
`products of basic research (1), which occasion-
`ally lead to discoveries that have clinical prom-
`ise. However, it is not known how often and how fast
`original basic research findings translate into clinical de-
`velopment and use, as well as what are the predictors of
`and obstacles to realization of these findings. To address
`these questions, we evaluated a sample of basic research
`publications in highly cited journals that had presented
`findings showing a clear clinical promise, and studied
`whether the original expectations materialized over a pe-
`riod of 20 years.
`
`METHODS
`Inclusion Criteria
`We searched PubMed for articles published from 1979 to
`1983 in six highly cited basic science journals: Science,
`
`From the Clinical Trials and Evidence-Based Medicine Unit, Depart-
`ment of Hygiene and Epidemiology (DCI, EN, II), Department of
`Pediatrics (DCI), University of Ioannina School of Medicine, Ioannina,
`Greece; Biomedical Research Institute (JI), Foundation for Research
`and Technology-Hellas, Greece; Department of Pediatrics (DCI),
`George Washington University School of Medicine and Health Sci-
`ences, Washington, D.C.; and Division of Clinical Care Research (JI),
`Tufts—New England Medical Center, Boston, Massachusetts.
`Requests for reprints should be addressed to John P. A. Ioannidis,
`MD, Department of Hygiene and Epidemiology, University of Ioannina
`School of Medicine, Ioannina 45110, Greece, or jioannid@cc.uoi.gr.
`Manuscript submitted August 31, 2002, and accepted in revised form
`November 12, 2002.
`
`Nature, Cell, the Journal of Experimental Medicine, and
`the Journal of Clinical Investigation, which had the highest
`impact factors in 2000, and the Journal of Biological
`Chemistry, which receives the most citations. We identi-
`fied all articles that contained the word therapy, therapies,
`therapeutic, therapeutical, prevention, preventive, vaccine,
`vaccines, or clinical. From these articles, we retained all
`original publications that clearly stated that the studied
`technology might have future clinical therapeutic or pre-
`ventive application. The 5-year period (1979 to 1983) al-
`lowed a meaningful time of approximately 20 years to
`elapse for examining the translation of basic science re-
`search into clinical research and practice. Eligible tech-
`nologies included substances, antibodies, vaccines, gene
`therapies, technical devices and other nonpharmacologic
`interventions, combination therapies, or novel tech-
`niques for production of the above technologies. We only
`considered technologies that were still at an experimental
`stage (molecular, cellular, animal, and early nonrandom-
`ized human studies) that did not have prior application in
`humans for the specific promise. We also included arti-
`cles that focused on a novel application (different disease
`or indication) of a technology already in use in humans or
`on a novel strategy combining technologies already in
`use. We excluded articles that did not describe a clear
`clinical promise in the abstract; editorials; commentaries;
`reviews; news articles; articles that focused on mecha-
`nisms of action, pathophysiology, or diagnosis; and arti-
`cles on agricultural or veterinary applications. Initial
`
`02003 by Excerpta Medica Inc.
`All rights reserved.
`
`0002-9343/03/$—see front matter 477
`doi:10.1016/S0002-9343 (03)00013-5
`
`Dr. Falk Ex. 2040
`GeneriCo v. Dr. Falk IPR2016-00297
`Page 1
`
`

`
`Translating Basic Science Research into Clinical Applications/Contopoulos-loannidis et al
`
`screening was based on the title and abstract. Two review-
`ers screened the full texts of potentially eligible articles,
`discussed the rules for selection, and independently
`screened articles for eligibility. Discrepancies were re-
`solved by consensus.
`Data Extraction
`The following information was collected from each eligi-
`ble publication: author name, publication year, journal
`name, study design, promising technology, whether a
`specific technology or a category of technologies was in-
`volved, anticipated application (therapeutic, preventive,
`or both), and disease target (single disease/condition vs.
`broader disease category). We also noted if there was in-
`volvement by the biotechnology or pharmaceutical in-
`dustry, defined as reported industry affiliation by an au-
`thor, financial support, or provision of the technology
`studied.
`
`Outcomes
`For each promising technology, we noted if any ran-
`domized controlled trials had been published, and
`whether the technology had shown favorable results
`("positive" trial), including statistically significant supe-
`riority (P <0.05) to placebo, no treatment, or established
`interventions; or stated equivalence compared with es-
`tablished interventions. We also determined whether any
`published research was performed in humans for any ap-
`plication (general human study) or for the specific appli-
`cation described in the basic science publication (specific
`human study).
`Identification of Human Studies and Trials
`We searched PubMed (to October 2002) using strategies
`that considered all alternative names of the experimental
`technology, including drug class and chemical substance,
`if applicable. Alternative names were identified by re-
`viewing the full text and references of the basic science
`publication, the medical subject heading in PubMed, ab-
`stracts of related articles (including subheadings), rele-
`vant publications by the same authors, relevant basic and
`clinical science textbooks, electronic books available
`from the National Center for Biotechnology Information
`(National Library of Medicine, Bethesda, Maryland), and
`drug reference guides. For antibodies and vaccines, we
`ascertained that the same technology discussed in the ba-
`sic science publication was used for production of these
`substances. For identification of human studies, we re-
`stricted the search to the "human study" group. For iden-
`tification of randomized controlled trials, we restricted
`the search by using the terms randomized controlled trial,
`randomized clinical trial, controlled clinical trial, random
`allocation, double blind, and single blind. Additional terms
`were used to focus the search on articles on prevention or
`therapy, if applicable. Disease-related terms were used
`when the promising technology pertained to a specific
`
`disease or disease category. Abstracts and full articles
`were examined to ascertain the study design and perti-
`nence of the retrieved studies. When the targeted search
`did not retrieve relevant articles, we used unrestricted
`search strategies.
`Determination of Current Clinical Use and
`Development Status
`For technologies that reached at least the stage of specific
`human study, we searched the most recent editions of the
`British National Formulary (2) and the Physician's Desk
`Reference (3). We performed an extensive PubMed
`search for trials, meta-analyses, guidelines, reviews, and
`other articles about the current development status of the
`technology. For technical devices, we searched the list of
`approved devices in the database of the Center for De-
`vices and Radiological Health, U.S. Food and Drug Ad-
`ministration (4).
`Statistical Analysis
`Kaplan-Meier curves were constructed for the time to
`publication of the first randomized and positive trials.
`We also examined which characteristics of the basic sci-
`ence article affected the publication rate. Comparisons
`were made using Cox proportional hazards models. Mul-
`tivariate models considered all variables with P <0.1 on
`univariate models and used a backward elimination ap-
`proach for final selection. There was no overt violation of
`the proportionality assumption. All analyses were per-
`formed with SPSS 10.0 (Chicago, Illinois). P values are
`two-tailed.
`
`RESULTS
`
`Of the 25,190 articles published from 1979 to 1983 in the
`six basic science journals, 562 contained the selected key
`words (Figure 1). Of those, 101 were original articles that
`clearly stated future clinical therapeutic or preventive ap-
`plications in humans for the studied technologies (Table
`1). None of the 101 articles were published in Cell, and
`only four were published in the Journal of Biological
`Chemistry, which likely reflects the focus of those journals
`on basic molecular rather than preclinical research, as
`compared with the other journals.
`Translation into Clinical Research
`Except for one outlier (n = 26,840 articles), two to 7715
`articles (median, 183 articles) were retrieved for screen-
`ing each of the 101 promising technologies. By October
`2002, 27 promising technologies had resulted in at least
`one published trial, 19 of which had at least one published
`positive trial (Table 1). Ten years after the initial basic
`science report, the probability of having a relevant publi-
`cation was 48% for a general human study, 37% for a
`specific human study, 18% for a randomized trial, and
`12% for a positive trial; at 20 years, the rates were 54%,
`
`478 (cid:9) April 15, 2003 THE AMERICAN JOURNAL OF MEDICINE® Volume 114
`
`Dr. Falk Ex. 2040
`GeneriCo v. Dr. Falk IPR2016-00297
`Page 2
`
`

`
`Translating Basic Science Research into Clinical Applications/Contopoulos-loann idis et al
`
`25,190 articles published in the six journals from 1979-1983
`
`562 articles retrieved with key word search
`
`•
`153 potential y eligible articles
`evaluated in full text
`
`101 articles included in the analysis
`
`409 articles rejected at abstract stage:
`Already known human application (n = 17)
`Reviews/commentaries (n = 134)
`Focus on pathophysiology (n = 190)
`Focus on diagnostic application (n = 38)
`No clinical promise in humans (n = 24)
`Randomized controlled trials (n = 6)
`
`52 articles rejected at full-text stage:
`Already known human application (n = 20)
`Reviews/commentaries (n = 5)
`Focus on pathophysiology (n = 10)
`Focus on diagnostic application (n = 5)
`No clinical promise in humans (n = 5)
`Randomized controlled trials (n = 7)
`
`Figure 1. Selection of basic science publications with promising therapeutic or preventive applications.
`
`45%, 27%, and 19% (Figure 2). When no human study
`was published 10 to 12 years from the index basic science
`publication, it was unlikely that one would be published
`subsequently. The likelihood of having a published ran-
`domized trial or positive trial also decreased after 12 to 15
`years (Figure 2).
`
`Factors Associated with Publication
`The rate at which the first randomized (or first positive)
`trial was published was considerably faster when there
`was industry involvement (Table 2). Having an author
`affiliated with the pharmaceutical or biotechnology in-
`dustry was associated with an eight- to 10-fold acceler-
`ated process, whereas commercial financial support or
`provision of the tested technology had a lesser effect.
`Promises of a vaccine were somewhat more likely to have
`a published positive trial, and more recent basic science
`findings were somewhat less likely to have a published
`positive trial. Multivariate models yielded similar results
`for author affiliations with the industry for published
`randomized trials (rate ratio [RR] = 5.7; 95% confidence
`interval [CI]: 2.6 to 13) and for published positive trials
`(RR = 8.3; 95% CI: 3.1 to 22). Similarly, after multivari-
`ate adjustment, more recent basic science articles were
`less likely to have published randomized trials (RR = 0.8
`per year of publication; 95% CI: 0.6 to 1.0) and published
`positive trials (RR = 0.5; 95% CI: 0.3 to 0.8), whereas
`vaccine studies were more likely to have published posi-
`tive trials (RR = 3.7; 95% CI: 1.3 to 11).
`
`Current Status
`Of the 27 technologies with at least one published ran-
`domized trial (Table 3), only five are licensed for clinical
`use (2,3). Of them, four had at least one positive trial
`
`during their development. Only one—angiotensin-con-
`verting enzyme inhibitors (N-carboxymethyl dipep-
`tides)— has shown extensive clinical advantages with
`expanding indications (5). Pergolide mesylate is used as
`adjunctive treatment to reduce the dose of levodopa in
`patients with Parkinson's disease (3,6). Recombinant in-
`terleukin 2 is licensed in the United Kingdom for the
`treatment of metastatic renal cell carcinoma, with ex-
`panding indications in the United States (2). However, it
`is toxic; tumor shrinkage occurs only in few patients; and
`survival benefits are limited (2). Alpha-1 antitrypsin
`treatment and naloxone (for shock) are licensed in the
`United States (3), but there are questions about their ef-
`ficacy (7,8).
`Four other technologies had limited clinical use (Table
`3). The rotavirus vaccine was licensed in the United States
`in 1998 (9), but it was subsequently withdrawn because of
`adverse effects (intussusception). The acyl-enzyme anis-
`treplase, which showed no clear advantages over other
`thrombolytic agents, is no longer available (2). Eflorni-
`thine (difluoronethylornithine) may be used to treat
`trypanosomiasis on special request, but the drug has only
`been tested in nonrandomized studies for this indication
`(10). Finally, there is no licensed subunit vaccine with the
`F glycoprotein of respiratory syncytial virus (paramyxo-
`virus F glycoprotein). Monoclonal antibodies against this
`glycoprotein were licensed in the United States (3,11),
`but cost-effectiveness of the therapy is questioned in the
`United Kingdom (12).
`Ten technologies that had at least one published ran-
`domized trial are still in development, with some con-
`tinuing to show promise, although use is still limited to
`research purposes (Table 3). Eight have had at least one
`
`April 15, 2003 THE AMERICAN JOURNAL OF MEDICINE® Volume 114 479
`
`Dr. Falk Ex. 2040
`GeneriCo v. Dr. Falk IPR2016-00297
`Page 3
`
`

`
`Translating Basic Science Research into Clinical Applications/Contopoulos-Ioannidis et al
`
`Table 1. Characteristics of the Eligible Basic Science Publications*
`
`Promising Technology Leading to
`Publication of:
`
`Total
`(n = 101)
`
`Randomized Trial
`(n = 27)
`
`Positive Trial
`(n = 19)
`
`Number
`
`Journal
`Science
`Nature
`Journal of Clinical
`Investigation
`Journal of Biological Chemistry
`Journal of Experimental
`Medicine
`Cell
`Year
`1979
`1980
`1981
`1982
`1983
`Industry involvement
`Author affiliation
`Financial support
`Provision of technology*
`None reported
`Type of study
`Molecular
`Cellular
`Animal
`Human
`Promising technology
`Substance
`Antibody
`Vaccine
`Other*
`Type of promising technology
`Specific
`General
`Implication
`Therapy
`Prevention
`Vaccine
`Both therapy and prevention
`Target of potential application
`Single disease/condition
`Broader disease category
`
`47
`21
`18
`
`4
`11
`
`0
`
`11
`28
`21
`22
`19
`
`16
`5
`11
`69
`
`6
`22
`64
`9
`
`69
`9
`12
`11
`
`70
`31
`
`73
`10
`13
`5
`
`62
`39
`
`11
`8
`4
`
`1
`3
`
`5
`7
`7
`4
`4
`
`11
`
`5
`10
`
`7
`16
`3
`
`20
`1
`6
`0
`
`21
`6
`
`15
`4
`7
`1
`
`19
`8
`
`9
`5
`2
`
`0
`3
`
`0
`
`5
`6
`3
`3
`2
`
`9
`1
`3
`6
`
`1
`4
`12
`2
`
`12
`1
`6
`0
`
`16
`3
`
`9
`2
`7
`1
`
`13
`6
`
`Supplementary information on the 101 publications can be obtained from the authors.
`Without reported author affiliation with the industry.
`Combination of therapeutic interventions, including nonpharmacologic interventions (n = 7), technical
`devices (n = 1), and gene therapy (n = 3).
`
`published positive trial, three of which had the first pos-
`itive trial published after at least 12 years.
`Eight additional technologies that had at least one pub-
`lished randomized trial have had mostly discouraging
`overall results (Table 3). N-acetylcysteine pretreatment
`
`for doxorubicin cardiotoxicity (13), naloxone for spinal
`cord injury (14), relaxin (recombinant human relaxin)
`for dystocia (15), and vitamin D metabolites for leukemia
`(16) never had published positive trials, and their ran-
`domized trials suggest they are unlikely to be effective.
`
`480 (cid:9) April 15, 2003 THE AMERICAN JOURNAL OF MEDICINE° Volume 114
`
`Dr. Falk Ex. 2040
`GeneriCo v. Dr. Falk IPR2016-00297
`Page 4
`
`

`
`Translating Basic Science Research into Clinical Applications/Contopoulos-Joannidis et al
`
`100
`
`90
`
`80 -
`
`70 —
`
`60 —
`
`50 -
`
`40 -
`
`30 —
`
`20 -
`
`
`
`10 —
`
`O)
`co
`4E'
`a)
`'a;
`13-
`
`- - "Positive" trial
`
`— Any trial
`
`- (cid:9)
`
`_r
`
`- J-
`
`0,
`0
`
`2 (cid:9)
`
`-r
`4 (cid:9)
`
`6
`
`8 (cid:9)
`
`10 (cid:9)
`
`12 (cid:9)
`
`14 (cid:9)
`
`16 (cid:9)
`
`18 (cid:9)
`
`20 (cid:9)
`
`22 (cid:9)
`
`24
`
`Years after Basic Science Publication
`
`Figure 2. Proportion of promising technologies that were evaluated in at least one published randomized controlled trial and at least
`one published positive trial, by time since the index basic science publication.
`
`Even in the case of the other four technologies with at
`least one positive trial, most evidence has been unfavor-
`able. Preventive treatment with vitamin A analogs
`showed promise in decreasing mortality from mesotheli-
`oma (17), but larger trials reported poor rates for lung
`
`cancer (18). Similarly, thiorphan, which decreased post-
`myelography headache in a small trial (19), has not been
`used since evidence suggested that it does not affect spinal
`pain control (20). Aspirin, which was similar to ursode-
`oxycholic acid in prevention of gallstone and crystal for-
`
`Table 2. Factors Associated with Publication of Randomized Controlled Trials and Positive Trials
`Rate Ratio* (95% Confidence Interval)
`Randomized Trial
`Positive Trial
`(n = 27)
`(n = 19)
`
`Variable
`
`Year of basic science article (per year)
`Industry involvement in basic science article
`Author affiliation
`Financial support/provision of technology
`None reported
`Basic science article on humans or animals
`Promising technology
`Substance
`Antibody
`Vaccine
`Other
`Specific promising technology
`Preventive application proposed
`Specific disease target proposed
`
`0.8 (0.6-1.1)
`
`0.7 (0.5-1.0)
`
`8.0 (3.4-19)
`2.9 (1.0-7.9)
`1.00
`0.9 (0.4-2.1)
`
`1.00
`0.3 (0.04-2.5)
`1.9 (0.7-4.6)
`UndefinedF
`1.6 (0.7-4.0)
`2.1 (0.96-4.6)
`1.6 (0.7-3.6)
`
`9.9 (3.5-28)
`2.9 (0.81-10)
`1.00
`1.1 (0.4-3.0)
`
`1.00
`0.6 (0.08-4.6)
`3.3 (1.2-8.9)
`Undefined'.
`2.5 (0.7-8.7)
`2.5 (0.99-6.2)
`1.4 (0.5-3.6)
`
`* A rate ratio >1 indicates a faster publication.
`t None of the 11 other technologies had a published randomized trial.
`
`April 15, 2003 THE AMERICAN JOURNAL OF MEDICINES Volume 114 481
`
`Dr. Falk Ex. 2040
`GeneriCo v. Dr. Falk IPR2016-00297
`Page 5
`
`(cid:9)
`

`
`Translating Basic Science Research into Clinical Applications/Contopoulos-Ioannidis et al
`
`Table 3. Characteristics of Promising Technologies in Basic Science Publications That Were Eventually Tested in At Least One
`Randomized Trial
`
`Technology/Implication for Use (Year)
`
`Naloxone (opiate antagonist)/shock (1979)
`Thymopoietin pentapeptide 32-36
`analog/immunodeficiency (1979)
`Native type III group B Streptococcus
`polysaccharide/vaccine (1979)
`Pergolide mesylate/dopamine deficiency disorders
`(1979)
`Bovine rotavirus/rotavirus vaccine (1979)
`Difluoromethylornithine/parasitic diseases (1980)
`Paramyxovirus F glycoprotein/paramyxovirus
`vaccine (1980)
`Synthetic vitamin A analog/occupational lung
`cancer (1980)
`N-carboxymethyl dipeptide (ACE
`inhibitor)/hypertension (1980)
`Gonadotropin-releasing hormone antagonists/male
`contraception (1980)
`Thiorphan/pain control (1980)
`Interleukin 2/diseases involving T lymphocytes
`(1980)
`Classic antidepressants plus a-adrenergic
`antagonists/depression (1981)
`N-acetylcysteine pretreatment/doxorubicin toxicity
`prevention (1981)
`Naloxone (opiate antagonist)/spinal cord injury
`(1981)
`Alpha-1 antitrypsin/substitution in emphysema
`(1981)
`Aspirin, NSAIDs/gallstone formation prevention
`(1981)
`Cyclosporin A/autoimmune uveitis (1981)
`Acyl-enzymes/fibrinolysis (1981)
`Ibuprofen/septic shock (1982)
`Avian-human influenza A reassortant
`virus/influenza A vaccine (1982)
`Free radical scavengers*/prevention of insulin-
`dependent diabetes (1982)
`Herpesvirus glycoprotein D gene
`sequenced/herpesvirus vaccine (1982)
`Escherichia coli—recombinant sporozoite surface
`antigen gene/malaria vaccine (1983)
`Recombinant human relaxin/dystocia (1983)
`Vitamin D3 dihydroxymetabolite/leukemia (1983)
`Recombinant cholera with mutant ctxA gene/live
`cholera vaccine (1983)
`
`Year of First Trial
`Any (Positive)
`
`Endpoint of First
`Positive Trial
`
`Development Stage
`(In October 2002)
`
`1984 (1984)
`1982 (1983)
`
`Hemodynamics, amines
`Atopic dermatitis
`
`1996 (1996)
`
`Immunogenicity
`
`In licensed use
`In development
`
`In development
`
`1981 (1983*)
`
`Prolactin levels
`
`In licensed use
`
`1984 (1984)
`1992 (—)
`1993 (1993)
`
`Rotavirus diarrhea
`
`Respiratory syncytial virus
`
`Withdrawn
`Use on request
`Antibody in use
`
`1995 (1998)
`
`Mesothelioma mortality
`
`Discouraging
`
`1981 (1981)
`
`Blood pressure
`
`In licensed use
`
`1992 (1992)
`
`Hormone levels
`
`In development
`
`1983 (1983)
`1992 (1992)
`
`Postmyelography headache
`Natural killer cells
`
`Discouraging
`In licensed use
`
`1986 (—)
`
`1983 (—)
`
`1990 (—)
`
`1999 (—)
`
`In development
`
`Discouraging
`
`Discouraging
`
`In licensed use
`
`1983 (1988*)
`
`Gallstone/crystal formation
`
`Discouraging
`
`1986 (1988)
`1986 (1986)
`1991 (1999)
`1988 (1988*)
`
`1993 (—)
`
`Visual acuity
`Indices of infarct size
`Mortality rate
`Influenza
`
`In development
`Not available
`In development
`Discouraging
`
`In development
`
`1994 (1994)
`
`Genital herpes recurrence
`
`In development
`
`1990 (1992)
`
`Malaria
`
`1993 (—)
`1990 (—)
`1988 (1988)
`
`Cholera
`
`In development
`
`Discouraging
`Discouraging
`In development
`
`* Showed no difference from other effective interventions (equivalence or no superiority).
`ACE = angiotensin-converting enzyme; NSAID = nonsteroidal anti-inflammatory drug.
`
`mation in one trial (21), had no association with the de-
`velopment of gallstones in larger epidemiologic studies
`(22). Avian-human influenza A reassortant virus initially
`tested favorably as a candidate influenza vaccine (23),
`but subsequent research suggested that the avian virus
`
`was not a suitable donor for attenuation of wild-type in-
`fluenza virus (24).
`Twenty-four technologies have been tested in specific
`human studies without having had published random-
`ized trials. Only sodium benzoate is licensed for use in
`
`482 (cid:9) April 15, 2003 THE AMERICAN JOURNAL OF MEDICINE® Volume 114
`
`Dr. Falk Ex. 2040
`GeneriCo v. Dr. Falk IPR2016-00297
`Page 6
`
`

`
`Translating Basic Science Research into Clinical Applications/Contopoulos-loannidis e al
`
`rare metabolic conditions (2). No randomized trials were
`conducted to support its use because of the uncommon-
`ness of these diseases; the indication is based on case se-
`ries. None of the 50 technologies without a published
`specific human study have been licensed.
`
`DISCUSSION
`
`In our study, only one in four promising technologies
`resulted in a published randomized trial and fewer than
`one in 10 entered routine clinical use within 20 years of
`the index basic science publication, supporting the no-
`tion that basic science research rarely translates into clin-
`ical research and clinical practice, even when they seem
`highly promising. Furthermore, only one technology has
`had a major clinical impact to date. Indeed, several factors
`may hinder the clinical development process. Findings
`may have been refuted in early phases of development by
`other biological or clinical evidence, and randomization
`may have been considered unethical. Even though non-
`randomized studies may have merits (25,26), random-
`ized trials are difficult to replace with other designs to
`achieve licensure for technologies that reach the stage of
`human experimentation. Moreover, an approved tech-
`nology may be shown to be harmful or less effective than
`its competitors and thus be discontinued. In fact, the
`strongest predictor of having a published randomized
`trial was industry involvement in the original basic sci-
`ence publication. Scientists without industry support
`rarely saw their discoveries materialize.
`We also found that there were considerable delays in
`the transition from basic research to clinical research and
`practice, regardless of the type of original study, promis-
`ing technology, and therapeutic and preventive implica-
`tion. Basic science promises are lost, refuted, or neglected
`at all stages of the clinical development process. Although
`it is difficult to state what constitutes a "good" rate of
`translation of basic research, current rates appear to be
`somewhat slow, and the process of rejecting or adopting
`new research findings should be accelerated. At the same
`time, caution should be exercised to avoid the implemen-
`tation of faulty ideas that may stem from hastiness.
`Only one in six of the promising technologies were
`"validated" in at least one published positive trial. Still,
`having a positive trial does not necessarily warrant adop-
`tion in clinical practice. "Negative" trials that report ad-
`verse effects or demonstrate irreproducibility of experi-
`ments should also be considered. We identified technol-
`ogies that were not approved for licensed clinical use
`despite continued promise and published positive trials.
`In three cases, the first published positive trial was at least
`12 years old, suggesting either a very slow development or
`a lack of reliability in results (27).
`
`L
`
`Our study has several limitations. Our estimates of the
`rate of translation of basic research into clinical applica-
`tions are probably optimistic. We selected top journals
`that were most likely to attract submissions on major ba-
`sic science breakthroughs. We used strict eligibility crite-
`ria to ensure that only promising technologies with a clear
`therapeutic or preventive implication were included. Our
`search algorithm limited subjective selection, yielding a
`reproducible sample of basic science promises. More-
`over, basic research often leads to subsequent clinical
`breakthroughs simply by answering fundamental ques-
`tions instead of targeting specific clinical problems
`(1,28). Hence, our eligibility criteria selected cases where
`translation into clinical application would have been
`most imminent. The study's retrospective design and its
`dependence on computer searches were other limita-
`tions. Since some human studies and randomized trials,
`especially those with negative results, may have remained
`unpublished, our data pertain to the time of published
`evidence rather than to the time of the conduct of the
`study. Nevertheless, unpublished studies or studies in less
`prominent journals are unlikely to make major contribu-
`tions towards the clinical adoption of a basic science
`promise. Finally, although it is unknown whether very
`recent basic research would translate faster into clinical
`experimentation and use in the near future, our data sug-
`gest that more recent promises in the period 1979-1983
`were actually less likely to result in positive trials.
`The gap between basic and clinical research needs to be
`narrowed. Promoting more interdisciplinary training
`will be challenging, given the demands of clinical prac-
`tice, the changing health environment in research-ori-
`ented countries (29-32), the rapid pace of basic and tech-
`nological research, and the competing resources for
`funding (33-37). Despite major improvements (38) since
`clinical investigators were called an "endangered species"
`(39), there is still a dearth of well-trained clinical re-
`searchers. Furthermore, outside of governmental fund-
`ing, clinical research is often conducted by or for the
`pharmaceutical and biotechnology industry. Our data
`suggest that investigators without links to the industry
`may have difficulties realizing their discoveries. The pri-
`vate sector is the major producer of therapies and preven-
`tive measures, and scientific merit and human needs may
`not always coincide with corporate profit (40). There is
`room for improvement in the translation of important
`basic research into clinical applications, and leaders in
`academic medicine and industry must develop strategies
`to enhance interdisciplinary work (41-43).
`
`REFERENCES
`
`1. Gannon F. Reaping the benefits of basic research. Nature. 1998;392:
`752.
`
`April 15, 2003 THE AMERICAN JOURNAL OF MEDICINE® Volume 114 483
`
`Dr. Falk Ex. 2040
`GeneriCo v. Dr. Falk IPR2016-00297
`Page 7
`
`

`
`Translating Basic Science Research into Clinical Applications/Contopoulos-Ioannidis et al
`
`2. British National Formulary. BNF 44. September 2002. British Med-
`ical Association and the Royal Pharmaceutical Society of Great Brit-
`ain. Available at: http://www.bnf.org. Accessed October 2002.
`3. Physician's Desk Reference. 56th ed. Montvale, New Jersey: Medical
`Economics Company Inc.; 2002. Available at: http://www.
`physician.pdr.net. Accessed October 2002.
`4. Center for Devices and Radiological Health, U.S. Food and Drug
`Administration. Available at: http://www.fda.gov/cdrh/index.html.
`Accessed October 2002.
`5. Neal B, MacMahon S, Chapman N. Effects of ACE inhibitors, cal-
`cium antagonists, and other blood-pressure-lowering drugs: results
`of prospectively designed overviews of randomized trials. Blood
`Pressure Lowering Treatment Trialists' Collaboration. Lancet.
`2000;356:1955-1964.
`6. Inzelberg R, Carasso RL, Schechtman E, Nisipeanu P. A compari-
`son of dopamine agonists and catechol-O-methyltransferase inhib-
`itors in Parkinson's disease. Clin Neuropharmacol. 2000;23:262-
`266.
`7. Abboud RT, Ford GT, Chapman KR. Alpha 1-antitrypsin
`deficiency: a position statement of the Canadian Thoracic Society.
`Can Respir J. 2001;8:81-88.
`8. Boeuf B, Gauvin F, Guerguerian AM, et al. Therapy of shock with
`naloxone: a meta-analysis. Crit Care Med. 1998;26:1910-1916.
`9. Zanardi LR, Haber P, Mootrey GT, et al. Intussusception among
`recipients of rotavirus vaccine: reports to the vaccine adverse event
`reporting system. Pediatrics. 2001;107:e97.
`10. Milord F, Pepin J, Loko L, et al. Efficacy and toxicity of eflornithine
`for treatment of Trypanosoma brucei gambiense sleeping sickness.
`Lancet. 1992;340:652-655.
`11. The IMpact-RSV Study Group. Palivizumab, a humanized respira-
`tory syncytial virus monoclonal antibody, reduces hospitalization
`from respiratory syncytial virus infection in high-risk infants. Pedi-
`atrics. 1998;102:531-537.
`12. Clark Si, Beresford MW, Subhedar NV, Shaw NJ. Respiratory syn-
`cytial virus infection in high risk infants and the potential impact of
`prophylaxis in a United Kingdom cohort. Arch Dis Child. 2000;83:
`313-316.
`13. Unverferth DV, Jagadeesh JM, Unverferth BJ, et al. Attempt to pre-
`vent doxorubicin-induced acute human myocardial morphologic
`damage with acetylcysteine. J Nail Cancer Inst. 1983;71:917-920.
`14. Bracken MB, Shepard MJ, Collins WF, et al. A randomized, con-
`trolled trial of methylprednisolone or naloxone in the treatment of
`acute spinal-cord injury. Results of the Second National Acute Spi-
`nal Cord Injury Study. N Engl J Med. 1990;322:1405-1411.
`15. Brennand JE, Ca