Bioreactor-based roadmap for thetranslation of tissue engineeringstrategies into clinical productsIvan Martin1, Timothy Smith2 and David Wendt1

1 Departments of Surgery and Biomedicine, University Hospital Basel, Hebelstrasse 20, 4031 Basel, Switzerland2 Octane Medical Group, 640 Cataraqui Woods Drive, Kingston, Ontario K7P 2Y5, Canada

Opinion

Despite the compelling clinical need to regeneratedamaged tissues/organs, impressive advances in thefield of tissue engineering have yet to result in viableengineered tissue products with widespread therapeuticadoption. Although bioreactor systems have been pro-posed as a key factor in the manufacture of standardizedand cost-effective engineered products, this conceptappears slow to be embraced and implemented. Herewe address scientific, regulatory and commercial chal-lenges intrinsic to the bioreactor-based translation oftissue engineering models into clinical products, propos-ing a roadmap for the implementation of a new para-digm. The roadmap highlights that bioreactors must beimplemented throughout product development, allow-ing scientific, medical, industrial and regulatory partiesto address basic research questions, conduct sound pre-clinical studies and ultimately facilitating effective com-mercialization of engineered clinical products.

IntroductionIn the last couple of decades, the appealing possibility ofcombining living cells with suitable carriers for the regen-eration of damaged or lost tissues and organs has promotedinteraction among scientists, engineers, clinicians andbusiness people, leading to the establishment and pro-gressive consolidation of the field of tissue engineering.Examples of successful clinical implementation of the con-cepts developed include restoration of corneal surfaces[1,2], replacement of a bronchus segment [3] and recon-struction of bone [4] and cartilage defects [5] and diseasedbladder [6]. Despite significant achievements and enor-mous clinical demand, the clear need for viable engineeredtissue products that are broadly available to patients aspart of the routine toolkit of medical treatments stillremains unsatisfied. On the one hand, this might bebecause of relatively limited establishment of prospective,randomized clinical studies demonstrating reproduciblysuperior effectiveness of engineered grafts compared toconventional treatments. On the other hand, it has beenproposed that addressing manufacturing-related issues iskey for the success of cell-based engineered products [7].Indeed, as in other biotechnology applications (e.g. theproduction of antibodies or molecular vaccines), successfulclinical use of engineered tissue products, as well as their

Corresponding author: Martin, I. (imartin@uhbs.ch).

0167-7799/$ – see front matter � 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.tibtech.20

commercial exploitation, might be critically dependent onthe introduction of bioreactor-based manufacturing sys-tems [8]. Bioreactors, intended as a means to generate andmaintain a controlled physicochemical culture environ-ment, indeed represent a key element in the automated,standardized, traceable, cost-effective, safe and regulat-ory-compliantmanufacture of cell-based products or engin-eered grafts for clinical applications [9,10] (Figure 1).However, notwithstanding the promise of a few pioneeringsystems currently under development or clinical testing(e.g. Octane Biotech Inc., Canada, http://www.octaneco.com/biotech and Aastrom Biosciences Inc., USA, http://www.aastrom.com), this concept has not yet broadly facili-tated the transfer of cell-based processes into clinically andcommercially viable therapeutic solutions. In this paper weaddress the scientific, regulatory and commercial chal-lenges that are hampering the bioreactor-based trans-lation of tissue engineering strategies into clinicalproducts and, based on these issues, propose a roadmapfor the implementation of a new paradigm. The new para-digm is fundamentally rooted in the perspective that sen-sor-based bioreactors must be deployed and validatedthroughout the product development pipeline, from theinitial concept until manufacture of the implant.

Scientific aspectsThe introduction of bioreactors in the field of tissue engin-eering was initially advocated as a means to apply definedregimes of physical forces, with the ultimate goal of reg-ulating and possibly improving the mechanical function-ality of the resulting engineered grafts [11]. This approachhas advanced scientific understanding of mechanoregula-tion of developing tissues and emphasized the importanceof controlled physical conditioning in tissue regeneration.However, the effective need to apply mechanical forces invitro to generate more functional grafts is still controver-sial. Indeed, beyond a few cases for which graft function-ality might need to be fully developed prior toimplantation, such as for engineered blood vessels [12]and heart valves [13], it is becoming increasingly clearthat tissue maturation could be more efficiently induced bythe physiological, biochemical and mechanical cues of thebody as an ‘in vivo bioreactor’ [14], provided that startingmaterial of appropriate quality is implanted. Moreover,use of a mechanical loading regime that correctly emulates

09.06.002 Available online 3 August 2009 495

Figure 1. Potential model of a commercial bioreactor-based system. Translation of

bioreactor-based tissue engineering into the clinic would benefit from integration

of bioreactor technology into state-of-the-art systems that meet the safety,

traceability and efficiency expected in healthcare environments. To eliminate the

prospect of cross-contamination, the bioreactor and associated media

management would be ideally isolated on a patient-specific level through the

use of disposable bioreactor cartridges or cassettes. Automated operational

control over the cassette would occur via a space-efficient, multi-channel (i.e.

multi-patient) instrument that operates under password-protected security access.

Initiation and monitoring of the process for each patient would occur via an

intuitive graphical interface that provides an accurate and timely indication of

status operation, along with secure data archiving. As the process advances within

the cassette, biosensors associated with the bioreactors and media management

system provide feedback on tissue culture conditions and enable the control

system to maintain critical process parameters. On completion of the process, the

cassette would enable removal of an output cartridge that inherently protects the

internal vial during the final stages of transport to the surgical center for

implantation. Automated cell-processing systems have been developed by

several groups, such as Aastrom Biosciences Inc. (http://www.aastrom.com) and

Tissue Genesis Inc. (http://www.tissuegenesis.com). The example illustrated is

under development by Octane Biotech Inc. (http://www.octaneco.com/biotech)

with the essential goal of enabling full GMP processing on a miniature and cost-

effective scale.

Opinion Trends in Biotechnology Vol.27 No.9

the dynamic physiology of the body to generate a fullyfunctional tissue is likely to make the manufacturingprocess too complicated and lengthy and, as a result, tooexpensive and impractical.

In principle, compared to smart functionalizedmaterials and drug delivery devices, cell-based grafts havethe potential to provide superior clinical outcomes becauseof the multivalent biological activity of cells (multiplegrowth factor release, self-contribution to tissue regener-

496

ation) and/or of the extracellular matrix deposited (effi-cient storage and release of morphogens, physiologicallyfunctional mechanical properties). However, engineeredproducts will only be a viable and competitive alternativeto upcoming off-the-shelf innovations in regenerativemedicine [7] if they are manufactured with reproducibleproperties, a prerequisite for consistent clinical outcomes.This important target is mainly challenged by the intrinsicvariability, often not sufficiently highlighted, in the beha-vior of human cells from different batches or donors [15], aswell as by the sensitivity of cells to perturbations in theculture environment. Whereas the focus in bioreactor-based tissue engineering has typically been on enhancingthe functionality or maturation stage of the resultingengineered tissues, the potential for controlled sensor-based bioreactor systems to minimize process and productvariability should receive equal attention. If the physico-chemical culture parameters aremonitored and controlled,bioreactors can help to standardize both the bioprocessesrequired and the resulting engineered graft and ensurethat automated protocols are compatible with regulatoryand commercialization requirements (Box 1). Moreover,because of streamlining of culture processes and bypassingof operator/handling-dependent procedures, bioreactor-based systems have the potential to increase the robust-ness and stability of the graft manufacturing process[16,17].

It is clear that to deliver the opportunities described,continued progress in overcoming the predominant chal-lenges is essential. These challenges are related not only tothe limited availability of on-line, non-invasive sensingtechniques for important culture parameters (e.g. cellnumber, differentiation stage, metabolic activity), but alsoto the limited fundamental understanding of the cellularand molecular processes underlying the regeneration andfunction of specific tissues and organs. Indeed, to bettercontrol and standardize key output product characteristicssuch as cell identity, quality, purity and potency, the modeof action associated with tissue regeneration should beidentified and validated. For example, in most cases it isnot yet established whether the performance of a specificengineered product depends on the number of cellsimplanted, the cell phenotype at the time of implantation,the amount of specific extracellular matrix proteins depos-ited, or the profile of cytokines released. Understandingthese scientific uncertainties will help to identify relevantbiomarkers, including extracellular matrix components,proteins indicative of proliferative/differentiation stageand metabolites, that can serve as critical process qual-ity-control points and predict the potency of an engineeredtissue product [18]. With this fundamental knowledge thenext challenge will be to define the specific biochemical,metabolic and/or environmental parameter ranges (e.g.,growth factors, glucose, pH, etc.) required to guaranteereproducible potency. In this context it is clear that themonitoring and control capabilities provided by bioreactorsystems, in conjunction with scientific knowledge obtainedthrough application of defined and controlled sequences,provides a valuable feedback loop to refine the cultureprocesses and optimize the properties of the engineeredproduct.

Box 1. Key opportunities and challenges for implementing

bioreactors

Scientific aspects

Key opportunity: Sensor-based bioreactors provide unmatched

ability to regulate bioprocesses to minimize process and product

variability.

Key challenge: Additional scientific insight is required to max-

imize the utility of bioreactor technology in providing engineered

tissue grafts with consistent properties.

Regulatory aspects

Key opportunity: Bioreactor-based processes, which include mon-

itoring and data management systems, can offer a high level of

traceability and compliance to safety guidelines.

Key challenge: Ambiguous regulatory guidelines currently hinder

the deliberate design of bioreactors that comply with specific and

clear specifications.

Commercial aspects

Key opportunity: Automated bioreactor systems can deliver safe

and standardized production of engineered tissue grafts, maximiz-

ing prospective scale-up and cost-effectiveness in the long term.

Key challenge: Models for commercialization of engineered tissue

products are not well established, resulting in uncertainties related

to markets, regulatory approval, reimbursement and overall clinical

adoption.

Opinion Trends in Biotechnology Vol.27 No.9

Regulatory aspectsThe clinical introduction of engineered tissue products islikely to involve significant regulatory oversight with anadditional degree of complexity compared to policies andguidance recently implemented at different internationallevels for cell therapy [19–21]. Whereas manufacturingstrategies based on conventional manual cell-culture tech-niques might face difficulties in complying with the newregulatory framework, bioreactor-based manufacturingprocesses inherently involve automation and reproducibil-ity that facilitate compliance with regulatory objectives. Inparticular, monitoring of process parameters and specificproperties of the developing/final graft provide a higherlevel of traceability of key manufacturing data related toidentity, purity and potency of the implant. Moreover, byminimizing the manual procedures required for operatorhandling of cells and tissue constructs, automated bio-reactor systems with control platforms and software-baseddata management will facilitate compliance to safety regu-lations. Ultimately, a stand-alone, fully automated andclosed system would provide a good manufacturing prac-tice (GMP)-compatible manufacturing unit complete withenvironment control, full operational traceability and fail-safe protection measures that go well beyond simplisticautomated cell culture. In essence, this is a type of GMP-in-a-box concept with the potential to reduce dependence onlarge, costly and sometimes not easily available regulat-ory-compatible facilities (Box 1).

To translate these opportunities into effective ways tofacilitate more widespread use of engineered tissue pro-ducts in the clinic, we foresee commonly underestimateddifficulties and challenges related to the practical details ofimplementation. First, the often rather undefined mode ofaction of engineered tissues and the limited availability ofpotency markers at a tissue level mean that pragmatictesting protocols that can be realistically deployed in theproduction process are difficult (see Scientific aspects).

Issues related to the lack of certainty that the productoutput provides the expected patient benefit could hampercredible plans for validation, which guarantees perform-ance with a high degree of confidence in the absence ofdirect data. Validation of production processes is pivotal toregulatory compliance and needs considerable attention inpreparation for regulatory review.

Clinical deployment of bioreactors for the preparation ofengineered tissues inherently depends on safe, reliable anduser-friendly operation of the system in productionenvironments. Fundamental to any regulatory assessmentis a comprehensive analysis of the product, process, andenvironment risk, along with specific implementation ofrisk mitigation practices to address critical risks. Riskassessment is based on occurrence, severity and detection,with all factors playing a crucial role in identifying keynodes for active risk management. As for other fields ofbiotechnology, the use of standardized and automatedsystems (in this case bioreactor-based manufacturing sys-tems) means that critical risk factors associated withoperator handling and the production process can be suc-cessfully mitigated. However, an overview of risks and riskcontrol for production of engineered tissues in bioreactorsystems (Table 1) indicates the complexity and wide rangeof processes involved. Indeed, regulatory authority skepti-cism on the implementation of bioreactor-based graft man-ufacturing could be tempered by introducing a sound riskassessment early enough in product development.

Beyond the characterization and mitigation of risksassociated with automated tissue engineering, regulatoryapproval of bioreactor systems for the production of cell-based implants for clinical use will be highly dependent onthe compliance of the surrounding production environ-ment. For facilities with pre-approval for cell manipula-tions, integration of additional equipment to furtherculture the implant would be significantly more straight-forward (although appropriately challenging on the detailsof the process) than implementation in the more openenvironment of a specialized clinic. Uncertainties in theinterpretation of a closed system, especially consideringthe sampling necessary for off-line monitoring (e.g. forsterility tests) and the need for direct intervention basedon out-of-specification signals, make it difficult to identifyprecise details of the practicality of the GMP-in-a-boxconcept from a regulatory or safety standpoint.

Commercial aspectsWhereas the promise of tissue engineering has captivatedmany enthusiasts and has generated significant inter-national investment in research, commercially engineeredproducts are confronted with tough economic realitiesrelated to uncertain cost–benefit performance and ultimateeligibility for reimbursement. The limited and perhapsdiscouraging commercial progress to date could be relatedto the fact that the basic procedures for generating engin-eered tissues have generally been based on conventionalmanual bench-top cell and tissue culture techniques. Thesemanual techniques seem particularly appealing during theinitial stages of product development (particularly forstart-up initiatives) because the simple and widespreadmanual approach is generally viewed as a route to mini-

497

Table 1. Risks and mitigation for bioreactor-based manufacturing systems

Risk Risk control/mitigation

Biology

Contamination and infection � Closed sterilized system

� Inputs sterile/clean when introduced

� Contamination testing on every lot (individual patient level)

Toxicity/bioburden/pyrogens � Bioreactor and related fluid contact surfaces selected from USP grade

� Pre-test bioreactor system with target cell type

� Validate production bioburden and pyrogen levels

Shear stress-induced cell damage � Bioreactor system designed with benefit of fluid modeling

� Validate operational sequence and maximum flow rates with target cell type

Compromised cell vitality and

performance

� Biosensors for strict maintenance of culture conditions

� Media and reagent sources refrigerated while on-line

� Adaptive software to minimize effects of donor variability

Insufficient final cell population for

clinical objective

� Establish gateway specifications in multi-step bioprocesses

� Automate monitoring of cell behavior to trigger cell collection

Sterilization failure � Validate according to established international standards

Inadequate packaging shelf life � Validate according to established international standards

Biorector/instrumentation

Failure of disposable cassette in

ensuring consistent and error-free

operation

� Direct attachment of cell/tissue input container to minimize manual handling

� Multi-layers to input container to reduce contamination risks

� High level of process bioreactors visibility for operator verification of steps

� Biological pathway entirely closed within cassette, zero transfer to instrument

� Sampling ports for sterile withdrawal of samples (e.g. microbial testing)

� Output container designed for direct transfer to clinical setting

Bioreactor system integrity failure � Validate connections and methods

� 100% leak test

� Sampling program

Sensor reliability issues � Sensor selection enables accurate monitoring for ex-vivo period without fouling

� Operational back-up (dual and alternate monitoring)

� Validate sensor tolerance to sterilization protocol

� Validate sensor shelf life when incorporated into cassette

Electromagnetic interference (EMI) and

static discharge sensitivity

� Test EMI emissions and sensitivity

Electrical safety issues � Meet legislative standards for safety

Software

Code error � Design code according to standards

� Rigorous code validation procedures

Software corruption � Standard software error checking

� Reload from non-volatile memory

� System watchdog

Power failure recovery � Establish redundant memory

� Automated recovery routines

Compromised operating system � Code redundancy

User

Inadvertent misuse � Design of bioreactors and instrumentation anticipate and preclude misuse

� Failsafe operation

� Clear instructions for use

� Field maintenance program

� Remote service link-up

Inadequate operational records � Process tracking and storage in non-volatile memory

� Data output via computer link

� Compliant with regulations on electronic records

Unauthorized use � User security codes/redundancy

Opinion Trends in Biotechnology Vol.27 No.9

mize initial development time and investment costs forrapid entry into clinical trials and the marketplace. How-ever, as scientific, technical and commercial momentumbuilds, production costs associated with manual pro-duction quickly become a barrier. In addition, traditionaltechniques have inherent contamination risks and intra-/

498

inter-operator variability. Furthermore, pragmatic limitsto scale-up of unit production volumes and related ineffi-ciencies in traceability pose significant challenges tobusiness models based onmanual protocols. As an alterna-tive, a closed, standardized and operator-independent bio-reactor-based production system has great benefits in

Box 2. SWOT analysis for bioreactor-based manufacture of tissue-engineered (TE) products

Strengths Weaknesses

� Bioreactors enable consistent implementation of bioprocesses

under regulated culture conditions to maximize TE productivity

� No pre-existing market model to use as a commercialization

benchmark

� Smart bioreactor systems using integrated sensors provide a

foundation for validation of TE processes for implant

production

� Operational constraints associated with bioreactor design

potentially limit visualization of an ongoing process

� Intelligent system operation enables autonomous sequences

to be delivered without the need for continual staff intervention

� Automation requires initial investment in capital equipment;

however, follow-on costs are economical

� Bioreactor control systems use protective measures to ensure

the process output is safe and effective

� The automated system and related user interface controls are

easy to use within implant production facilities

� Pre-configured bioreactors and related system components

provide low-cost routine operation following initial set-up

� Automatic monitoring, collection and archiving of process

data enhance institutional compliance with regulatory

requirements

� TE production systems based on automated bioreactor

technology are ideally suited to progressive scale-up of

process throughput

Opportunities Threats

� Large regenerative medicine market [7] provides opportunity for

TE-based product derivatives once effective implant production

and delivery methods are available

� Cost–benefit of cell-based products might be insufficient to

compete with off-the-shelf alternatives

� Success of a few representative models of bioreactor-based

delivery of TE into the clinic will accelerate commercial and

healthcare provider interest

� Use of preconfigured bioreactor assemblies and automated

techniques might require a change in mindset for certain users

� Technical and administrative consistency of bioreactor-based

TE supports transition and expansion from research through

to the clinic

� Operation of a system to produce cell-based implants via

automated techniques potentially displaces specialized workforces

� Adoption of bioreactor-based standards will enhance clinical

data analysis and reinforce regulatory documentation

� Underlying science of bioreactor-based TE might be limited by

insufficient supporting data (e.g. mechanism of action)

� Inherent consistency and quality of bioreactor output maximize

the opportunity to pursue effective reimbursement for

TE procedures

� Innovative bioreactor technology might be constrained or delayed

by mismatch with established regulatory standards

� Continued refinement of scientific insights into tissue

engineering processes can be translated to the clinic via

bioreactor upgrades

� Implementation of bioreactor-based TE strategies could be

constrained or delayed by validation challenges inherent in the

adoption of automated processes (e.g. lack of appropriate quality

control markers)� Early focus for TE in clinical therapeutics is likely to be on

critical conditions that ensure recovery of bioreactor R&D costs

Opinion Trends in Biotechnology Vol.27 No.9

terms of safety and regulatory compliance. Despiteinitially incurring high product development costs, thisapproach has great potential to improve the cost-effective-ness of a manufacturing process in the long run andmaximize process scale-up (Box 1).

The advantages of a bioreactor-based approach appearconvincing and yet bioreactor technology is not widelyadopted. This cannot be because of a lack of competentbioreactor designs; a literature or patent search reveals aplethora of various bioreactors developed to date. So whyhas the drive to implement bioreactors for themanufactureof engineered products for clinical applications been lim-ited?

Perhaps the clinical vision in different areas of tissueengineering is still relatively immature and hence quality,validation and productivity demands are considered sec-ondary challenges. One factor contributing to the delayedintroduction of novel bioreactor technology could be thechange in mindset required to break from well-establishedcell culture methods. To make this leap and for bioreactorsto gain greater general acceptance, simple and more user-friendly bioreactor systems need to be introduced during

the research phase, given that the routine handling ofcomplex and technically challenging systems may discou-rage their introduction into laboratories that lack a special-ized and trained workforce. Moreover, even if it isacknowledged that bioreactors can generate engineeredtissues of higher quality, there are currently few datahighlighting the process reproducibility and – to the bestof our knowledge – no sound analysis that criticallyassesses the potential cost-effectiveness of a bioreactor-based production system. Indeed, economic tools to guideinvestment decisions and to evaluate the potential cost-effectiveness of an engineered product are complex, rely onstrong assumptions and are not yet well established. Asa first step, typical methods for the analysis of factorsinfluencing strategic planning need to be applied tothe commercialization of bioreactor-based tissue engineer-ing for the clinic. Without reference to a specific clinicalindication, the results of a SWOT (strengths, weaknesses,opportunities and threats) analysis (Box 2) indicate thatbioreactors for the clinical delivery of tissue engineeringhave attractive strengths and opportunities. Not surpris-ingly, the new bioreactor-based approach also has certain

499

Figure 2. Proposed roadmap for translating bioreactor-based engineered products into the clinic. Highlighted are the most critical scientific, regulatory, and commercial

challenges that will need to be addressed, irrespective of the particular tissue engineering application. On starting to develop a bioreactor-based tissue engineering system

and targeting a specific clinical application with clear goals in terms of projected clinical outcome, core biological criteria associated with the engineered product, the

relevant bioprocesses and basic regulatory guidelines must be established and verified. Planning and conducting sound pre-clinical studies will allow assessment and

validation of the criteria previously established for the core biology platform, most likely leading to refinement and optimization of the bioprocesses and to a better

understanding of the underlying biology. Interactions with industrial partners will facilitate the transition between platforms through the evolution and industrialization of

research-based technologies and by defining the commercial potential of the product. In the final stage of the roadmap, which clearly is the most challenging and critical,

translation to the clinic platform will require clinical, regulatory and reimbursement strategies [26].

Opinion Trends in Biotechnology Vol.27 No.9

weaknesses and poses threats that need to be addressed aspart of a well-developed clinical, regulatory and reimbur-sement plan. To provide more quantitative indications, theheadroom method has been proposed as a simple butrigorous approach to obtain preliminary conclusions onthe cost-effectiveness of a new tissue engineering treat-ment [22]. Difficulties in implementing such analyticaltools are related to the fact that some critical parametersrequired as input, for example the expected improvementin clinical performance, can only be arbitrarily assumedbecause of the uncertainties associated with an undeve-loped technology.

The lack of cost–benefit analyses in the context ofclinical efficacy precludes a plausible assessment of thelikelihood of reimbursement coverage for the productionand surgical implantation of engineered products, which inturn severely impacts the credibility of commercialbusiness plans founded on the widespread use of tissueengineering. Given the problems with the tissue engineer-ing business model and the variable demand for bioreactorsystems, it is not surprising that industry has been hesi-tant to embrace commercialization of engineered products

500

and invest significant resources to develop and/or usesophisticated manufacturing technology.

Closer interactions among scientists and engineers andthe other stakeholders involved (regulatory bodies, clin-icians, industrial partners, healthcare providers) arerequired for integrated bioreactor designs that can be usedin systems that are practical and economical for the clinic.Data from nearly two decades of bioreactor technology forengineered tissues are available. Although these research-oriented systems are generally too complex, user-unfriendly, unsafe, and expensive for direct use in clinicalapplications, their underlying principles could neverthe-less lay a solid foundation for more clinically compliantmanufacturing systems. This will require the identifi-cation of only the most essential processes, cultureparameters, and construct parameters that must be mon-itored and controlled to standardize production and pro-vide meaningful quality and traceability data, at the sametime as minimizing risks, costs, and user complexity.

In this context it is questionable whether complex andrather costly automated/robotic systems, which essentiallymimic established manual procedures, can actually

Opinion Trends in Biotechnology Vol.27 No.9

demonstrate real cost–benefit by replacing manual cellculture techniques in a manufacturing process. Moreover,current automation techniques might fail to capture theexpert nature and role of the cell culture technician.Instead, bioreactor designs could be dramatically simpli-fied and related development costs significantly reduced ifwe re-evaluate the conventional tissue engineering para-digms and implement novel concepts and techniques thatcould streamline the individual bioprocessing steps. Sim-plified tissue engineering processes could be key to futuremanufacturing strategies by requiring only a minimalnumber of bioprocesses and unit operations, thereby facil-itating simplified and compact bioreactor designs withlimited automation requirements, with the likely resultof lower product development and operating costs. Here weunderline the importance of simplicity and minimal costswhile maintaining the core attributes of implant consist-ency and overall manufacturing productivity and scalabil-ity. Although we have attempted to draw several parallelsto bioreactors in other fields of biotechnology, the analogybegins to break down during up-scaling, especially in thecontext of autologous implants. Given that cells from eachpatient are highly variable and must be processed ascompletely independent batches, we cannot simply scaleup the total volume as for a fermentation process or evenan allogeneic product. Instead the production of autogolousproducts must be ‘scaled out’ (i.e., a large number ofparallel bioreactor systems independently performingthe same bioprocesses). While scaling-out will require ahigh level of process reproducibility among each of thebioreactors, production could be readily up-scaled bysimply adding additional (low-cost) units as productdemand increases [23] (Figure 1).

Finally, we must ask where the potential exploitation ofbioreactors lies within the broad range of tissue engineeringapplications. Could there be amarket for a simple and user-friendly bioreactor for general use in basic research appli-cations? Are hospitals a viable market? A closed GMP-in-a-box systemwouldallow competenthospitals tomanufactureengineered grafts on site, without reliance on a centralizedindustrial manufacturing facility. Another consideration iswhether the predominant market would comprise indus-trial firmsproducing engineeredproducts in-house,with theperspective of introducing automated bioreactors withintheir manufacturing process. At present, none of thesemarkets canbeexcludedand it is likely thatbusinessmodelswill have to be adapted to some specific initial decisions. Forexample, it is possible that a centralized manufacturingfacility is appropriate for the engineering of allogeneicgrafts, whereas hospital-based production could become areality for autologous cell-based tissues.

ConclusionsThe use of bioreactor-based platforms for the translation oftissue engineering strategies into clinical products offersattractive opportunities for exploitation and potential forbroad implementation of cell-based grafts as therapeuticsolutions. However, for the success of this paradigm, sev-eral diverse challenges need to be realistically consideredand constructively addressed, as outlined in the roadmapproposed in Figure 2. Bioreactor-based manufacturing

concepts need to be introduced as early as possible inthe development of an engineered tissue product. This willenable academics, developers and industry participants to(i) properly address underlying research questions (e.g.testing the effect of controlled changes in discreteparameters), (ii) carry out sound pre-clinical studies (e.g.testing the in vivo performance of grafts with reproduciblefeatures) and (iii) maintain strong ties with regulatory andcommercial dimensions (e.g. implementation of scalingconcepts and compliance with safety requirements). Inthe proposed roadmap, scientific, regulatory and commer-cial aspects should all be considered in each of the transla-tional stages from the initial concept to finalimplementation of a clinical product, leading to continuousrefinements and beneficial corrections.

Finally, the success of bioreactor-based tissue engineer-ing products will also depend on acceptance of the proposedparadigms by the stakeholders involved (e.g. surgeons,industry alliance partners [24], investors, health insurancecompanies) and by the general public (i.e. the potentialpatient population) [25]. In fact, while the availability oftherapeutic products has the potential to develop themarket, it should not be underestimated that key for theirultimate success will closely depend on the generation of areceptive society, which needs to be educated by appro-priate dissemination activities.

AcknowledgmentsThe authors would like to thank Mrs. Anke Wixmerten for help withediting of the text and the figures.

References1 Pellegrini, G. et al. (1997) Long-term restoration of damaged corneal

surfaces with autologous cultivated corneal epithelium. Lancet 349,990–993

2 Tsai, R.J. et al. (2000) Reconstruction of damaged corneas bytransplantation of autologous limbal epithelial cells. N. Engl. J.Med. 343, 86–93

3 Macchiarini, P. et al. (2008) Clinical transplantation of a tissue-engineered airway. Lancet 372, 2023–2030

4 Quarto, R. et al. (2001) Repair of large bone defects with the use ofautologous bone marrow stromal cells. N. Engl. J. Med. 344, 385–386

5 Brun, P. et al. (2008) Characteristics of repair tissue in second-look andthird-look biopsies from patients treated with engineered cartilage:relationship to symptomatology and time after implantation. ArthritisRes. Ther. 10, R132

6 Atala, A. et al. (2006) Tissue-engineered autologous bladders forpatients needing cystoplasty. Lancet 367, 1241–1246

7 Lysaght, M.J. et al. (2008) Great expectations: private sector activity intissue engineering, regenerative medicine, and stem cell therapeutics.Tissue Eng. Part A 14, 305–315

8 Martin, I. et al. (2004) The role of bioreactors in tissue engineering.Trends Biotechnol. 22, 80–86

9 Kirouac, D.C. and Zandstra, P.W. (2008) The systematic production ofcells for cell therapies. Cell Stem Cell 3, 369–381

10 Ratcliffe, A. and Niklason, L.E. (2002) Bioreactors and bioprocessingfor tissue engineering. Ann. N. Y. Acad. Sci. 961, 210–215

11 Butler, D.L. et al. (2000) Functional tissue engineering: the role ofbiomechanics. J. Biomech. Eng. 122, 570–575

12 Niklason, L.E. et al. (1999) Functional arteries grown in vitro. Science284, 489–493

13 Hoerstrup, S.P. et al. (2001) Tissue engineering of small calibervascular grafts. Eur. J. Cardiothorac. Surg. 20, 164–169

14 Stevens, M.M. et al. (2005) In vivo engineering of organs: the bonebioreactor. Proc. Natl. Acad. Sci. U. S. A. 102, 11450–11455

15 Barbero, A. et al. (2004) Age related changes in human articularchondrocyte yield, proliferation and post-expansion chondrogeniccapacity. Osteoarthritis Cartilage 12, 476–484

501

Opinion Trends in Biotechnology Vol.27 No.9

16 Braccini, A. et al. (2005) Three-dimensional perfusion culture of humanbone marrow cells and generation of osteoinductive grafts. Stem Cells23, 1066–1072

17 Scherberich, A. et al. (2007) Three-dimensional perfusion culture ofhuman adipose tissue-derived endothelial and osteoblastic progenitorsgenerates osteogenic constructs with intrinsic vascularizationcapacity. Stem Cells 25, 1823–1829

18 De Bari, C. et al. (2008) A biomarker-based mathematical model topredict bone-forming potency of human synovial and periostealmesenchymal stem cells. Arthritis Rheum. 58, 240–250

19 European Medicines Agency (2008) Guideline on human cell-basedmedicinal products (http://www.emea.europa.eu/pdfs/human/cpwp/41086906enfin.pdf)

20 FDA (2008) Guidance for industry on potency tests for cellular and genetherapy products (http://www.fda.gov/CBER/gdlns/testcellgene.htm)

21 FDA (2008) Guidance for FDA reviewers and sponsors on content andreview of chemistry, manufacturing, and control (CMC) information for

502

human gene therapy investigational new drug applications (INDs)(http://www.fda.gov/cber/gdlns/gtindcmc.htm)

22 McAteer, H. et al. (2007) Cost-effectiveness analysis at thedevelopment phase of a potential health technology: examples basedon tissue engineering of bladder and urethra. J. Tissue Eng. Regen.Med. 1, 343–349

23 Mason, C. and Hoare, M. (2006) Regenerative medicine bioprocessing:the need to learn from the experience of other fields. Regen. Med. 1,615–623

24 Pangarkar, C. and Hutmacher, D.W. (2003) Invention and businessperformance in the tissue-engineering industry. Tissue Eng. 9, 1313–

132225 Mason, C. and Dunnill, P. (2008) The need for a regen industry voice.

Regen. Med. 3, 621–63126 Mason, C. and Dunnill, P. (2008) The crucial linkage required between

regenerative medicine bioprocessors and clinicians. Regen. Med. 3,435–442