e u r o p e a n j o u r n a l o f p h a r m a c e u t i c a l s c i e n c e s 3 1 ( 2 0 0 7 ) 62–67

avai lab le at www.sc iencedi rec t .com

journa l homepage: www.e lsev ier .com/ locate /e jps

Commentary

Effective integration of systems biology, biomarkers,biosimulation and modelling in streamlining drugdevelopment

Rajesh Krishnaa,∗, Hans Guenter Schaeferb, Ole J. Bjerrumc

a

Merck Research Laboratories, Merck & Co., Inc., 126 East Lincoln Avenue, Rahway, NJ, USAb Boehringer-Ingelheim Pharma GmbH & Co. KG, Biberach an der Riss, Germanyc Faculty of Pharmaceutical Sciences, University of Copenhagen, Denmark

a r t i c l e i n f o

Article history:

Received 12 February 2007

Accepted 14 February 2007

Published on line 20 February 2007

Keywords:

Systems biology

Conference report

Modeling

Simulations

Basel

a b s t r a c t

The European Federation of Pharmaceutical Sciences (EUFEPS) has long established itself

as leaders in the field of interdisciplinary meetings to discuss issues that face drug devel-

opment. It’s ever popular and well attended “Optimizing Drug Development” series has

tackled numerous issues, most recent of which have been drug interactions, getting the

dose right, candidate selection, and biomarkers (Lesko et al., 2000; Rolan et al., 2003; Stanski

et al., 2005; Tucker et al., 2001). Over a course of 3 productive days, the meeting on “Effective

Integration of Systems Biology, Biomarkers, Biosimulation and Modelling in Streamlining

Drug Development”, held in Basel, Switzerland was jointly sponsored by EUFEPS, European

Biosimulation Network of Excellence (BioSim), American College of Clinical Pharmacol-

ogy (ACCP), European Centre of Pharmaceutical Medicine (ECPM), and Swiss Society of

Pharmaceutical Sciences (SGRW). The meeting was focused on emerging aspects related

to the quantitative understanding of underlying pathways in drug discovery and clinical

development, i.e. moving from an empirical to a model-based, quantitative drug develop-

ment process. The objectives of the meeting were: (1) to highlight the current state of the

art on biomarkers (as they relate to quantitative fingerprinting of disease), systems biol-

ogy, modelling and simulation; (2) to illustrate the applications of these emerging tools in

increasing the efficiency and productivity of new drug development by case examples; (3)

to understand the gaps in the technology and organizational implementations in gover-

nance, and (4) allow an opportunity for cross-disciplinary interaction, i.e., scientists with

more theoretical and technical modelling and simulation expertise of the BioSim network

and researchers experienced in applying modelling and simulation techniques in day-to-

day drug development were drawn together. This report summarizes the outcome from this

meeting.

© 2007 Rajesh Krishna. Published by Elsevier B.V. All rights reserved.

∗ Corresponding author. Tel.: +1 732 594 1484; fax: +1 732 594 5405.E-mail address: rajesh krishna@merck.com (R. Krishna).

0928-0987/$ – see front matter © 2007 Rajesh Krishna. Published by Elsevier B.V. All rights reserved.doi:10.1016/j.ejps.2007.02.003

u t i c

1

Tcdi2MitmKmmbommilecommdnafeteiauvctifueamld

e u r o p e a n j o u r n a l o f p h a r m a c e

. Introduction

here are numerous reports outlining that current drug dis-overy and development are far from meeting the increasedemands of new safe and effective medicines for unmet med-

cal needs (Bjerrum, 2000; Kola and Landis, 2004; Krishna,006a,b; US Food and Drug Administration, 2004; New Safeedicines Faster, 2000). Recently several initiatives have been

ntroduced to improve drug development (see Table 1). Par-icularly for drug development candidates targeted to a new

echanism of action, the success rate is far from optimal.ola and Landis (2004) indicate that this probability, of newechanisms, reaches just 11%. Thus, the Scientific Program-ing Committee identified the mission of the conference to

e that, in order to increase the efficiency and productivityf new drug development, new tools for predictive humanodels of disease and of efficient drug discovery, develop-ent and evaluation are necessary. These new technologies

nclude, “mathematical modelling”, “biomarkers”, “biosimu-ation” and “systems biology”. A key aspect was while they allxisted as individual breakthrough technologies, it was notlear, prior to the conference, how much cross-fertilizationccurred and how much integration were there to collectivelyitigate the apparent gaps in drug discovery and develop-ent. For example, mathematical modelling is applied in drug

evelopment to predict and/or simulate the outcome of theext steps to be taken in various processes, thus allowingmore informed decision, when progressing a compound

rom one development phase to the next. Biomarkers willnable translational data integration from preclinical specieso humans, hence helping to understand drug effects, sideffects and toxicity in a more predictive manner. Data qual-ty, rigor in analyses, and documentation of data analysesre critical components for successful submission to reg-latory authorities, arrived at by careful study design andalidation. Biosimulation is an old dream of being able toonduct in silico pharmaceutical sciences to assist with predic-ion. Core activities would include translation of experimentalnsights into consistent mathematical frameworks, includingormal models to predict the course of physiological processesnder varying conditions. Systems biology approaches arexpected to link genomic, proteomic and metabolomic data

nd bioinformatics, to provide better understanding throughodelling and simulation of the function and dysfunction of

iving systems. New insights will change paradigms of drugiscovery, development and evaluation. The Conference tar-

Table 1 – International initiatives on improving drug developm

Innovative Medicines Initiative (EU) http:Critical Path Initiative (US FDA) http:Medicamentos Innovadores (Spain) http:Top Institute Pharma (Netherlands) http:European Clinical Research Infrastructure Network http:Safety Biomarkers (UK) http:Critical Path Institute (University of Arizona) http:Center for Biomedical Innovation (MIT)) http:Toxicogenomics Project (JPMA) http:Proteome Factory Consortium (JPMA) http:Large Scale Clinical Trial Network (Japan) http:

a l s c i e n c e s 3 1 ( 2 0 0 7 ) 62–67 63

geted drug discoverers, developers and evaluators, includingin silico, in vitro and in vivo scientists, clinical pharmacolo-gists and other scientists exploiting modelling, biomarkers,biosimulation and systems biology, as well as for academic sci-entists wanting better insight into how industry apply theseapproaches, and finally for regulatory scientists evaluatingnew medicines for therapy.

In total, there were one keynote session and eight scientificsessions, which included six sessions dedicated to model-based drug development, and two to lessons learned. Thesescientific sessions included:

• Model-Based Drug Development (I): Role of Systems Biologyin Drug Discovery

• Model-Based Drug Development (II): Pharmacokinetic/Pharmacodynamic Principles in Preclinical Development

• Model-Based Drug Development (III): Case studies in DrugDiscovery and Preclinical Development

• Model-Based Drug Development (IV): Modelling and Simu-lation in Clinical Development—The Concept

• Model-Based Drug Development and Biomarkers (I): CaseStudies in Clinical Development

• Model-Based Drug Development and Biomarkers (II): Clini-cal Trial Designs and Statistical Aspects

and sessions devoted to lessons learned were:

• Learning from Failures: Could Biomarkers and Model-BasedDrug Development have Prevented Phase III Failures?

• Implementation of Systems Biology, Biomarkers, Biosimu-lation and Modelling in Pharmaceutical Industry: CurrentStatus and Lessons Learned

Importantly, these sessions addressed the gaps and oppor-tunities of each of the tools in various segments of the drugdiscovery and development pipeline.

2. The foundation for more effectiveintegration

The keynote session addressed the foundation for the meet-ing. The landmark regulatory initiatives across two continents

were reviewed, as they relate to the US FDA critical pathinitiative (US Food and Drug Administration, 2004) and theEU innovative medicines initiative (IMI) (New Safe MedicinesFaster, 2000; EFPIA, 2004). Whereas the former is a govern-

ent

//www.cordis.europa.eu/FP7/home.html//www.fda.gov/oc/initiatives/criticalpath//www.medicamentosinnovadores.org///www.tipharma.org.nl//www.ecrin.org///www.nc3rs.org.uk/downloaddoc.asp?id=367&page=241&skin=0//www.c-path.org///web.mit.edu/cbi///wwwtgp.nibio.go.jp/index-e.html//www.jhsf.or.jp/English/topic 2002.html//www.pharm.kitasato-u.ac.jp/biostatis/presentation2003/Hirota.ppt

e u t

64 e u r o p e a n j o u r n a l o f p h a r m a c

ment funded initiative by a regulatory agency, the latter is anEFPIA/EU Commission joint initiative. Both initiatives focuson biomarkers and disease pathways, modelling and simu-lation or systems biology, and clinical trial designs. The IMIis being proposed as a joint technology initiative under theEU Framework Programme 7. The goals of both initiativeswere to enhance the predictive ability in discerning effecton diseases while providing a framework for drug/diseasemodels and reducing risk in late stage clinical development.Both initiatives have provided effective foundations in defin-ing the problem and both organizations are now in theprocess of defining the path forward in the respective initia-tives including funding of necessary research. However, bothteams contended that it was the aspect of implementationthat represented the major challenges due to the associatedchanges in pharmaceutical companies, regulatory agencies,and federal health institutions. Improvements through effec-tive funding of precompetitive research in public-privatepartnerships will be one of the routes forward (EFPIA, 2004).

Despite less than desirable progress globally, key break-throughs have included the critical path biomarker consor-tiums, exemplified by the oncology biomarker qualificationinitiative, cardiac safety task force, nephrotoxicity biomarkers,drug-induced liver injury models, and serious adverse events.The FDA, in collaboration with Drug Information Association(DIA), Safe and Innovative Medicines (PhRMA) and other par-ticipating scientific associations will follow up to discuss theclinical, regulatory, legal and marketing issues surroundingthe area of pre-competitive space when sharing drug/diseasemodels. These approaches are aimed at further streamliningthe engagement of effective transparent workflows in the pre-competitive space.

A third aspect of the keynote session included an introduc-tion to systems biology, chartered by the EU supported BioSimNetwork of Excellence, and its integration in mainstreamdrug development. The network of 26 academic, 10 industrial,and 4 regulatory partners hopes to bridge biochemistry andgenomics from the drug discovery phase with trial planningand PK/PD in enabling a more detailed computational insightsinto normal and pathological disease processes. Presently,four therapeutic areas are targeted including cancer, hyper-tension, diabetes, and mental disorders.

Collectively, these three cross-continental initiativesaddress issues and challenges inherent in our understandingof disease pathways and point to routes on how to addressthem (see also Table 1).

3. Model-based drug discovery anddevelopment research

3.1. Drug discovery and preclinical development

The applications of systems biology in disease pathway fin-gerprinting were highlighted in this session. Systems biologyis an emerging discipline that evaluates how components of

a complex biological system dynamically interact within thesystem allowing the assessment of the influence of a givenstimuli whether that be a physiological or a pharmacologi-cal insult. Importantly, advances in genetics and molecular

i c a l s c i e n c e s 3 1 ( 2 0 0 7 ) 62–67

biology have afforded an extensive understanding of sin-gle components associated with complex pathophysiologicaldeterminants. However, only systems biology can allow thesynergy between such reductionist approaches aimed at partsor single components and provide a collective understand-ing of the whole system. One challenge was the integrationof systems biology in the early discovery phase which wasidentified as a key gap in drug development. For systems biol-ogy to succeed, it was argued that organizations be adaptiveto the cohesion between informatics, pathway analysis, andquantitative biology. When that is done, it was truly pos-sible to predict on and off-target side effects and enhanceour predictive understanding of human disease and effectof various insults in that system. Systems biology, if effec-tively integrated and hypothesis driven, could offer a goodunderstanding of the mechanism of action of new molecu-lar entities. One such example was the use of systems biologyin the development of pharmaceutical targets leveraging com-putational science in the understanding of complex signallingpathways. From a drug development perspective, this wouldhelp identify responder subset of patients and aid in the iden-tification of patients that might respond to the benefits of atherapeutic intervention favourably. Another aspect would bethe use of systems biology to select which targets to pursue.On the other hand, systems biology can also predict the patho-logical insult by pharmacological interventions and aid in theunderstanding of toxicity. Here, the systems biology approachallows the modelling of different effects individually on func-tional basis and thus estimate their overall contribution to thenet effect, allowing the multiple pathways to be delineated.From a drug discovery standpoint, this would help understandoff target toxicities of new molecular entities and allow thedelineation of risk and benefit early in the process.

Some of the gaps limiting systems biology in pathway mod-elling include:

• Adequacy of knowledge and impact of incomplete knowl-edge when building the model.

• Automating the model development.• Informed guesses on parameterization as compared with

experimentally generated information.• Computational time and speed in model execution and

visualization.• Educational aspects of this emerging science, requiring a

new skilled force, and enhanced organizational awareness,communication, and interactions.

The benefits of mechanism based models were empha-sized in the classical approaches to drug development suchas ADMET, physiologically based PK and PK/PD modelling.Given the conceptual similarities between systems biologyand physiologically based PK modelling, it was envisioned thatmore integrated interactions between system biologists andPK/PD modelling specialists would provide a more systematicevolution of model based drug development.

3.2. Clinical development

Within the conceptual framework on the integration ofbiomarkers in drug development three specific aspects were

u t i c

daatmfnrwcAaafetsamb

Ptfahtisaiirdmmoegatadcafbeee

4is

4

Tc

e u r o p e a n j o u r n a l o f p h a r m a c e

iscussed, first, the role of biomarkers in the pathogenesisnd clinical progress of target disease in the presence andbsence of drug treatment, second, the specific considerationshat support linkage between biomarkers and clinical effect

easurements, and third, approaches to qualify biomarkersor surrogacy or support regulatory evidence of effective-ess. The goals of the innovative medicine initiative wereeviewed within the context of the biomarker consortium,herein EU Innomed integrated project was a demonstration

ase. The project included two specific initiatives PredTox andddNeuromed. The PredTox consortium was a conglomer-te of pharmaceutical companies and academic universitiesnd aimed at validating predictive models of toxicity by per-orming concerted analysis of 16 model compounds to enablearlier terminations of unsafe developmental compounds andhus reduce attrition due to toxicity. The AddNeuromed con-ortium, also an industry/academia network, focussing onlzheimer’s disease to aid in the development of diagnosticarkers, biomarkers or markers of disease, and responder

iomarkers.Case studies exemplifying the integration of biomarkers in

K/PD and drug/disease models showed how the results ofhe modelling provided clear value in experimental designsor innovative clinical trials as well as to validate the mech-nism of action of new molecular entities. Case studies alsoighlighted the value modelling and simulation would bringo clinical development, including selection of doses, predict-ng the degree of uncertainty in drug response, probability ofuccess in meeting a desired endpoint for an efficacy trial,ssessment of various clinical trial designs for maximizingnformation, and use of adaptive designs for maximizingnformation on dose/response. Collectively, these approacheseduce attrition related to drug efficacy and help achieve evi-ence of target engagement earlier than latter, allowing toake quick termination of compounds, and conserve invest-ent. An area of further opportunity was the application

f modelling and simulation to better understand diseaseffects. For example, for neurological disorders, it is almostiven that single point measures to differentiate protectivend symptomatic effects may not be successful. It is hopedhat quantitative clinical pharmacology or strategic modellingnd simulation can significantly modulate clinical risk/benefitecisions using quantitative data. Such drug/disease modelsould leverage the use of fast and slow biomarkers for ther-peutic areas as osteoporosis enabling efficacy (probability ofracture) prediction readout earlier than a traditional study onone mineral density could provide. Such drug/disease mod-ls may help delineate symptomatic effects from protectiveffects and thus aid in the understanding whether a drugxhibits a certain disease modifying behaviour.

. Challenges in implementing anntegrated biomarker and modelling/simulationtrategy

.1. Prospective clinical modelling

he use of modelling and simulation prospectively to informlinical decisions was discussed. There are three potential

a l s c i e n c e s 3 1 ( 2 0 0 7 ) 62–67 65

considerations. (1) While it was recognized that there wereseveral retrospective examples on the utility of modelling andsimulation in drug development, there were relatively fewerexamples where such strategies were applied prospectively. (2)A secondary aspect is in leveraging an approach across a dis-ease platform. For example, leveraging bisphosphonate datato inform a new mechanism of action for osteoporosis (suchas cathepsin K inhibitors). (3) Incorporating side effects andefficacy in the overall clinical utility assessment of a clinicaldrug candidate.

4.2. Knowledge management

Data mining and knowledge management is a critical aspect.Currently, several individual computational tools are used ina manual environment and data, be they clinical, pharma-cokinetic, safety, or efficacy often exist in separate databases.Given the density of data and advanced computational infras-tructure, it is clear that high throughput in computationalinfrastructure is desired. Current programmatic software andhardware requirements may need to be carefully assessed asthe broader implications of storage, archival, and processingof data become paramount. This may include a technol-ogy platform that incorporates multiple programs such asgrid computing. A few companies are addressing this gap byindustrializing computer processing and data storage infras-tructure. These types of multi-nodal computing environmentcan be customized to perform automated statistical cal-culations, automated generation of tables and graphs forreport writing, automated model evaluation including boot-strapping, and automated covariate searches. By automatingmundane portions of the work related to modelling and sim-ulation, scientists focus the extra time identifying and solvingproblems and less on operational aspects of data manage-ment. These approaches may also alleviate documentationand audit trail issues within a compliance perspective.

It is clear that for effective integration and knowledge basedmanagement of data, that there is a clear cultural changewithin an organization. This is particularly true of pharmaceu-tical companies which have dedicated departments and thusdeliver individual contributions to drug development. Manycompanies are in the midst of an organizational change toadapt to this emerging new science. The process will be sup-ported by the Innovative Medicines Initiative (EFPIA, 2004).

4.3. Regulatory decision making

Expectations of the integration of modelling and simulation inregulatory decision-making are high. However, there are twocentral issues related to this. The first is the expectation of dis-ease model libraries with the FDA’s vision of the critical pathinitiative. It is clear that cross functional/industry collabora-tion may be necessary to leverage the development of diseasemodel libraries. The clinical, legal, marketing, and regulatoryaspects of sharing data and disease progression informationacross industry, academia, and regulatory authorities in prec-

ompetitive space are unclear, but not insurmountable. Giventhe resource issues at the FDA, it is unclear whether the mora-torium on End-of-Phase 2A meetings, which offer better dialogbetween the sponsor and the regulator on matters central to

e u t

r

66 e u r o p e a n j o u r n a l o f p h a r m a c

exposure/response relationships, will be lifted in the future.The second aspect is how the European regulators amenableto using modelling and simulation for approvals, labels, andevidence of effectiveness. The presence at the meeting fromseveral European national agencies (Norway, Denmark andThe Netherlands) shows that the interest is rising.

4.4. Training and education

As with systems biology, there is an enormous need forskilled scientists in this emerging new area of science. Manyorganizations are not optimally staffed, with recruitmentand retention of well-trained and experienced modelling andsimulation scientists being an ongoing challenge. Presently,there is no university degree on pharmacometrics, whereasa few schools are gearing towards this area with increasedpharmaceutical company funding. Regardless, the concertedinvestigation of this education issue is warranted to sus-tain this field and interests in improving drug development.Scientific societies like EUFEPS are aware of the need andwork necessary for coordination in Europe. The EU Innova-tive Medicines Initiative has a budget for establishment of thenecessary courses.

4.5. Integration of theoretical academic scientists indrug development

The necessity of enhanced interactions between modellingand simulation scientists, system biologists, pharmacologists,clinicians, and regulators is evident. The strong technical andtheoretical focus in this emerging area (which are part engi-neering, mathematics, advanced computer science, appliedscience, pharmacokinetics, etc.) and the strong medical focusof mainstream clinical research may result in unproductivegaps which may hinder the effective integration of theseapproaches in drug development. Institutionalizing inter-face scientists, such as those who can speak the languagesof both the modelers and the clinicians will help bridgethis gap. This new multidisciplinary interface communityintegrates relationships between physiology, pharmacology,disease pathophysiology, clinical outcomes, pharmacokinet-ics, pharmacodynamics, and interindividual variability into anew quantitative clinical pharmacology mindset. The BioSimnetwork represents such a group of basic scientists whothrough the Conference were brought into contact to theapplied world of drug development.

5. Concluding remarks and futuredirections

It is clear that the quality, efficiency, and productivity of newdrug development need to improve. The fact that just about11% of all new molecular entities targeting new mechanismsreach clinical development is an appalling statistic (Kola andLandis, 2004). Furthermore, the bottleneck in drug develop-

ment which is considered to be Phase II stage as being thedefault decision making stage for continuance or terminationis prohibitive from a cost perspective, given late stage. It iswidely recognized that the industry needs to make decisions

i c a l s c i e n c e s 3 1 ( 2 0 0 7 ) 62–67

on terminations early enough once evidence of a lack of phar-macological benefit is better understood during translationaland early clinical development. Using biomarkers appropri-ately and using predictive human models of disease early inthe translational stage would help with the “quick-win/quick-kill” paradigm allowing resources to be directed to compoundswith high probability of success. Development of new toolsto link prediction to clinical outcomes would allow the accu-rate prediction of hazards and reduce uncertainty in risk/benefit.

The overall consensus at the Conference was that prospec-tive implementation of systems biology, modelling andsimulation can significantly improve our ability to quantita-tively understand pathways of disease such that mechanismsof action of new molecular entities are better understood,off target toxicities effectively delineated, and the risk/benefitprofile of clinical candidates profiled. Important critical areasas points for further engagements include infrastructuretechnology management and education and training oppor-tunities to sustain this emerging science. This is especiallyimportant given the diversity in educational background,organizational complexities, and cultural subtleties inherentin the effective implementation of these strategies. Theseissues highlight the need to effectively and openly collabo-rate, network, collectively learn, and implement best practiceswherever possible. The establishment of sufficient fundingfor these activities is equally important. Ultimately, it woulddepend on our ability to convert modelling information intouseful knowledge management that would play a central rolein reducing late stage attrition and fundamentally modulatedrug development.

Acknowledgements

The individuals who served with us on the scientific program-ming committee and those to contributed to the presentationsare gratefully acknowledged. These individuals are: FritzR. Buhler, Anne Marie Clemensen, Meindert Danhof, Hart-mut Derendorf, Igor Goryanin, Lawrence J. Lesko, Hans H.Linden, Erik Mosekilde, Christian R. Noe, Carl Peck, Ian C.Ragan, Shashank Rohatagi, Malcolm Rowland, Donald R. Stan-ski, Steffen Thirstrup, Hans Westerhoff, Michael Zuhlsdorf,Didier Scherrer, Oleg Demin, Denis Noble, David Clark, MortenColding-Jorgensen, Paolo Vicini, Bernd Muller-Beckmann,Inaki Troconiz, Marc Pfister, Joga Gobburu, Robert Powell,Dinesh DeAlwis, Albert Goldbeter, Lars Sundstrom, PeterLockwood, Ariel Alonso, Richard Peck, Colin Pillai, Hans West-erhoff, Klaus Prank, Klaus Lindpaintner, Alexander Staab,Colin Spraggs, John Bloom.

The Conference was financially supported by The EUBioSim Network of Excellence.

e f e r e n c e s

Bjerrum, O.J., 2000. New safe medicines faster: a proposal for akey action within the European union’s 6th frameworkprogramme. Pharmacol. Toxicol. 86 (Suppl. 1), 23–26.

u t i c

E

K

K

K

L

N

N

e u r o p e a n j o u r n a l o f p h a r m a c e

FPIA, 2004. Innovative Medicines for Europe. Available at:http://ec.europa.eu/research/fp6/index en.cfm?p=1 innomed.Accessed December 29, 2006.

ola, I., Landis, J., 2004. Can the pharmaceutical industry reduceattrition rates? Nat. Rev. Drug Discov. 3 (8), 711–715.

rishna, R., 2006a. Introduction to dose optimization. In: Krishna,R. (Ed.), Dose Optimization in Drug Development, vol. 161.Taylor and Francis, pp. 1–13.

rishna, R., 2006b. Quantitative clinical pharmacology: makingparadigm shifts a reality. J. Clin. Pharmacol. 46 (9),966–967.

esko, L., Rowland, M., Peck, C.C., Blashcke, T., 2000. Optimizingthe science of drug development: opportunities for bettercandidate selection and accelerated evaluation in humans.Eur. J. Pharm. Sci. 10 (4), 4–14.

ew Safe Medicines Faster, 2000. Proposals for research topics,methodologies, techniques and other means of promoting for

the drug development process to the benefit of Europeancitizens. Available at http://eufeps.org. Accessed December 29,2006.

ew Safe Medicines Faster. How to rethink and accelerate thedevelopment and approval of innovative new medicines for

a l s c i e n c e s 3 1 ( 2 0 0 7 ) 62–67 67

faster patient relief, 2003. Available at http://eufeps.org.Accessed December 29. 2006.

Rolan, P., Atkinson, A., Lesko, L., 2003. Use of biomarkers fromdrug discovery through clinical practice: report of the NinthEuropean Federation of Pharmaceutical Sciences Conferenceon Optimizing Drug Development. Clin. Pharmacol. Ther. 73(4), 284–291.

Stanski, D.R., Rowland, M., Sheiner, L.B., 2005. Getting the doseright: report from the Tenth European Federation ofPharmaceutical Sciences (EUFEPS) conference on optimizingdrug development. J. Pharmacokinet. Pharmacodyn. 32 (2),199–211.

Tucker, G.T., Houston, J.B., Huang, S.M., 2001. EUFEPS conferencereport. Optimising drug development: strategies to assessdrug metabolism/transporter interaction potential—towards aconsensus. European Federation of Pharmaceutical Sciences.Eur. J. Pharm. Sci. 13 (4), 417–428.

US Food and Drug Administration, 2004. Challenge andOpportunity on the Critical Path to New Medical Products.Available at:http://www.fda.gov/oc/initiatives/criticalpath/whitepaper.html.Accessed December 29, 2006.