this article reprinted from the official journal of ispe focuses on pharmaceutical … ·...

Commissioning and Qualification

JULY/AUGUST 2005 PHARMACEUTICAL ENGINEERING 1©Copyright ISPE 2005

A Practical Guide to Construction,Commissioning and QualificationDocumentation – and its Critical Rolein Achieving Complianceby Wael Allan and Andrew D. Skibo

This articlefocuses onconstructioncontractors whocan deliverqualifiedfacilities andpresents theadvantages ofusing aconstructioncontractor whocan performCommissioningandQualification.

Good documentation is essential forpharmaceutical and biotech manufac-turing facilities to achieve regulatorycompliance. As defined by the Food

and Drug Administration (FDA), “Validation isa documented program providing a high degreeof assurance that a process/system consistentlymeets pre-determined specifications.” The clearpresumption is that if the required activitieshave not been properly documented, then theyhave not been performed.

Naturally, the initial focus must be on theproper construction and start-up of the facility.Yet, a well-constructed facility that is on time,within budget, and whose every system is per-forming to specifications is of no value to theoperating company if the associated documen-tation does not effectively support the qualifi-cation process – it is a classic case of failing tosee the forest through trees. In fact, the con-struction contractor must appreciate the sig-nificance of documentation, and make it an

integral part of the con-struction planning, im-plementation, and com-missioning process fromday one.

Note: the term “Con-struction Contractor” isused to indicate a thirdparty company respon-sible for construction,commissioning and qual-ification. This article fo-cuses on constructioncontractors who can de-liver qualified facilities.It also presents the ad-vantages of using a con-struction contractor whocan perform C&Q, as wellas construction led ratherthan design led projects.A “full service provider”could provide all of theseactivities. These activi-ties could be provided in

Figure 1. Validation lifecycle.

Reprinted from The Official Journal of ISPE

PHARMACEUTICAL ENGINEERING® July/August 2005, Vol. 25 No. 4


2 PHARMACEUTICAL ENGINEERING JULY/AUGUST 2005 ©Copyright ISPE 2005

part by a Construction Management (CM) contractor, acommissioning contractor, a qualification contractor and theoperating company.

Turn Over Package (TOP): Laying the Foundation

Good documentation is an essential part of the QualityAssurance (QA) system. For new and renovated facilities,Commissioning and Qualification (C&Q) are key aspects ofcost and schedule. Therefore, documentation – in the form ofthe TOP, C&Q protocols, and other required documents –plays a pivotal role in ensuring a compliant facility. Everyoperating company has specific standards and methodologiesfor C&Q. The construction contractor’s job is to make surethat the TOP enables the operating company to effectivelycarry out its QA strategy. Therefore, the TOP must be well-organized, meet the operating company’s expectations, andprovide the proper level of documentation quality.

To be successful, an integrated documentation processmust start very early in the project in conjunction with theplanning of C&Q. It must involve engineering, construction,and qualification/validation. Moreover, project documenta-tion requirements, and the roles and responsibilities for theoperating company, construction contractors, and vendorsalike, must be clearly defined.

Too often, documentation work is delayed until the later

stages of the project, or the effort required to do it successfullyis underestimated. For example, on an 18-month project, theQA/documentation officer should be involved by the firstquarter of the project – during the planning phase — to beginto organize the TOP to meet the operating company’s needsand expectations. It is well worth the small investment tobring that individual on board early to avoid problems duringthe qualification process, and ultimately, reduce time tomarket. This approach eliminates duplication of effort byleveraging documentation from commissioning into Installa-tion Qualification (IQ). IQ is approximately 30% of the quali-fication effort; by leveraging commissioning documentation,a savings of approximately 50% of IQ effort could be made,resulting in an overall saving of 15% of qualification time.

The Validation Life CycleJust as good documentation is required to achieve regulatorycompliance, clear user specifications that are fully under-stood by the construction contractor are required to meet theoperating company’s unique needs and expectations. If theconstructor fails to understand these from the outset, theTOP is likely to be inadequate. Therefore, for every step takenby the operating company, engineers and construction con-tractor, a high level of integration is required to ensure thatthe resulting documentation is appropriate, compliant, andstructured in a meaningful manner - Figure 1. The diagram

Figure 2. Document hierarchy.



illustrates the importance of the “definition” stage for boththe operating company and the contractor. In all aspects ofthe life cycle, documentation/data is a very critical element.

The operating company’s definitions of specifications andprocess requirements drive the whole project. Specifically,the operating company defines the process, develops proce-dures and specifications, and verifies these. In turn, theconstruction contractor specifies and installs equipment, andtests and qualifies making systems ready for validation andoperation. This process requires that the construction con-tractor accurately “translate” the operating company’s defi-nitions and expectations into systems, bricks, and mortar.Otherwise, the succeeding steps – development of the valida-tion protocol through change control – will be fraught withproblems, delaying time to market.

A Practical Guide to Project PlanningEffective project planning is the key to a successful and cost-effective documentation effort. More importantly, is the inte-gration of construction with C&Q. Operating companies havedifferent preferences: some would award construction andcommissioning to the same construction contractor; otherswould go further and award construction and C&Q to thesame construction contractor, ensuring a successful, seam-less integration and quality documentation, resulting in thehandover of a qualified facility rather than a merely mechani-cally completed one. (There are additional variations andoperating company contracting preferences, which are out-side the scope of this article.) The construction contractor’srole in a C&Q project can be broken down into three steps:

• develop and agree upon the scope of work and responsibili-ties

• agree upon a methodology and develop the project execu-tion plan and quality manual

• develop a project management plan incorporating docu-ment management

Each of these steps has components with a significant impacton documentation; therefore, the construction contractorshould fully integrate these steps into project planning. Thefollowing is a practical guide to each step and its essentialcomponents.

Step 1: Develop and Agree Uponthe Scope of Work and ResponsibilitiesAt the start of a project, it is essential for the operatingcompany and construction contractor to agree on the scope ofwork, perform a risk analysis to identify critical and non-critical systems, and identify who is responsible for thevarious project deliverables, including documentation; theconstruction contractor also must identify systems and bound-aries.

Identify Systems and BoundariesWorking from the operating company’s definitions and engi-

neers’ drawings, the construction contractor must graphi-cally delineate each system and its boundaries. This hassignificance for the organization of the TOP, for example, if itis to be organized by system; it also identifies which subcon-tractor and/or vendor has responsibility for what element.

Perform Risk AnalysisThe risk analysis is an extremely important activity involv-ing the operating company and construction contractor. Theresult will capture operating company expectations by clas-sifying systems into critical and/or direct impact systems andnon-critical and/or indirect impact systems. This is signifi-cant because critical systems, such as fermentation, requirea higher level of documentation. Thus, the risk analysisculminates in establishing the project documentation re-quirements - Table A. Documents should be categorized intobaseline documents, controlled documents, and validationdocuments. For non-critical and/or indirect impact systems,Good Engineering Practice (GEP) is considered sufficientwhile critical/direct impact systems are earmarked for com-pilation of enhanced documentation packages.

PHASE DOCUMENT ALL CRITICALTYPE SYSTEMS SYSTEMS

Good EnhanceEngineering Documentation

Practice

Engineering Design Intent B VP&IDs C VSpecifications B BDrawings C* V

Construction Purchasing B BVendor B CInstallation C C

Commissioning Lists B BFactory Testing B BField Testing B VStart-up Forms B V

Operating Training B VMaintenance C VProcedures B V

Regulatory Applicable Codes B CInspection Certificates C* CPermit Documents B B

Validation Validation Master Plan --- VValidation Protocols --- VValidation Summary --- V Report

B: BASELINE DOCUMENT - These documents are created and retained inproject records with appropriate project approvals.

C: CONTROLLED DOCUMENT - These documents are created “as built”retained in project files with appropriate project approvals and maintainedafter commissioning. (Note C* means that only some of the documents inthis category require document maintenance)

V: VALIDATION DOCUMENT - These documents are treated as controlleddocuments, but with the added approvals required for validation programs.Additional approvals usually are limited to Quality Assurance and operations.Document maintenance is a formal change control process.

Table A. Documentation requirements.



Figure 3. Role of QA/documentation officer.

Name Responsible Parties/Individualsfor Specific Activities/DeliverablesWith respect to documentation, a hierarchy from “designintent” through to “validation summary reporting” should beestablished based on the delineation of systems and bound-aries - Figure 2. Responsibility for specific documents isassigned to various firms and individuals. Moreover, certaindeliverables must be completed before others can start,requiring that these individuals clearly understand theirresponsibilities, integrate their activities, and effectivelycommunicate with one another. A critical project also ben-efits from a graphic diagram of the project document hierar-chy, which is distributed to each team member. Otherwise,false assumptions might be made and critical documentscould “fall through the cracks.” This aspect of Step 1 culmi-nates in two sets of documents: non-FDA regulatory docu-ments (i.e., for reference, archiving, etc.) and regulatorydocuments.

Identify the Project QA/Documentation OfficerThe project QA/documentation officer is a key resource withrespect to documentation and must be carefully selected.That is, the project QA/documentation officer’s capabilitiesmust far exceed organizing and distributing documents – heor she must fully understand the regulatory requirements fordocumentation - Figure 3.

The QA/documentation officer should report to the con-struction or overall project manager and/or the C&Q man-ager (either as a member of construction staff or undersubcontract if C&Q is performed by a party other than theconstruction firm) and liaise with the operating company’svalidation, regulatory compliance, and QA groups. This indi-vidual is charged with ensuring that the operating company’srequirements, policies, and expectations are transferred toall parties involved; confirming that all parties know whichdocuments to produce, and when; and ensuring all the docu-ments are consistent with one another and fulfil the regula-tory requirements of compliance, validation, and QA. TheQA/documentation officer audits all documentation, includ-ing the TOP, ensuring that it can be leveraged to help theoperating company streamline qualification.

In essence, the QA/documentation officer must effectivelyalign the contractual expectations of the operating companyand construction contractor. While some operating compa-

nies prefer to hire a third party, the advantages of hiring aconstruction contractor with this in-house capability aregreater consistency, compliance, and speed.

Step 2: Agree Upon a Methodology, and Developthe Project Execution Plan and Quality ManualWell before design is completed, the construction contractorand operating company should agree upon a qualificationmethodology. Based on this, the construction contractor willidentify standard operating procedures; develop a schedule,commissioning plan and validation master plan; and alsoagree on the level of leverage from commissioning into quali-fication. These, among other strategies, will be used todetermine the organization of the TOP. Leveraging can bedefined as the utilization of properly documented activitiescarried out during construction and commissioning whichcan be used in support of qualification (IQ and OQ) resultingin the avoidance of unnecessary repetitions hence reducingqualification time.

Identify Standard Operating Procedures (SOPs)The construction contractor must identify and define allSOPs, including protocol formats, numbering systems, andthe review and approval process. This ensures that properprocedures are in place to document all systems and compo-nents for the TOP. The construction contractor must main-tain consistency among numbering systems used by thevarious suppliers and align these with the operating company’srequirements for the project. Ultimately this important com-ponent of the methodology and project execution plan en-sures a smooth compilation of the documentation and TOP.

Develop a Project Schedule CapturingCommissioning, Qualification, and ValidationDevelopment of a project timeline by the construction con-tractor identifies the required resources and critical timingsfor data or documentation that are required for subsequentactivities to proceed. This is critical to streamlining theprocess.

Develop Commissioning PlanAn important component of the project methodology is thecommissioning plan, which identifies the level of documenta-tion appropriate for commissioning, level of quality, andrequired signatories based on the operating company’s quali-fication strategy. For example, some operating companiesuse an Integrated Commissioning Qualification (ICQ) strat-egy, in which much of the commissioning documents will beleveraged into the qualification effort, thereby reducing timeto market - Figure 4.

Develop Validation Master PlanThe validation master plan, including computer systemsvalidation, must be aligned with the commissioning plan andoperating company’s qualification strategy.



Figure 4. Leveraging commissioning into qualification.

Determine Organization of TOPBased on the strategies of C&Q, the construction contractormust determine the proper organization of the TOP and whatlevel of documentation is required from the various parties.The organization of the TOP is critical to the success of thequalification process – that is, it must be organized in such away that the operating company is able to leverage thedocumentation and data into qualification. In effect, the TOPaligns the contractual responsibilities of the operating com-pany and construction contractor. The TOP signifies the endof an important phase and a handover to the operatingcompany. In some cases, the TOP could signify the end ofmechanical completion, where design and construction datais turned over to the operating company; in others, it signifiesthe completion of commissioning and qualification - Figure 5.

The TOP consists of specifications, manuals, drawings,and other documentation that fully characterizes each sys-tem or piece of equipment installed in the facility. Theconstruction contractor should prepare a TOP matrix for eachsystem, which defines the documentation required for all itscomponents. Compiled in a formal and organized package,the TOP serves as part of the basis for Installation Qualifica-tion (IQ), which verifies that the physical components of thesystem have been installed according to design specifica-tions. As the final major component and system quality auditprior to Operation Qualification (OQ), the IQ is a critical stepthat lays the foundation for compliance and testing of thefacility.

In fact, the TOP is critical throughout all phases of theproject – from design and procurement to handover at me-chanical completion, commissioning, or qualification. Thus,each of the following activities must be carried out to ensuredocumentation of compliance:

• Design and Procurement Phase: the constructioncontractor’s team must review the design of the project forsystem boundary demarcation, regulatory requirements,commissionability, Good Automated Manufacturing Prac-tice (GAMP), and Code of Federal Regulations (CFR) 21Part 11 compliance, and to develop the templates forcommissioning, Factory Acceptance Testing (FAT), SiteAcceptance Testing (SAT), and qualification documents.The commissioning plan and validation master plan aredeveloped and delivered in this phase. Members of theteam responsible for commissioning must review Pipingand Instrumentation Diagrams (P&IDs) and 3D designmodels for safety and the inclusion of commissioningrequirements. They also will attend Hazard and Operabil-ity (HAZOP) and constructability reviews, as well asdesign specifications and procurement documentation toassure that the requirements of the operating companyare included and delivered.

• FAT: during the procurement phase, the constructioncontractor should ensure that a determination is made asto which equipment will undergo formal FAT. Items which



Figure 5. TOP’s critical role in all project phases.

cannot undergo “FAT” should be subjected to a “documentand component verification.” For the selected items, theconstruction contractor must develop and issue FAT testprotocols. The construction contractor’s FAT team, led bythe individual who is responsible for commissioning, willexecute the FAT in the vendor’s facility. The commission-ing head must verify that the equipment complies with theUser Requirement Specifications (URS) and design speci-fication, and that all documentation for operation, main-tenance, and qualification are complete and available.Functional testing is generally undertaken for informa-tion and engineering verification purposes only. Program-mable Logic Controller (PLC) functionality must be veri-fied, as well as the data communications to the plantsupervisory systems to ensure their compatibility andtransmission. Finally, documentation, controls hardware,and the component schedule must be formally verifiedbecause these will be subsequently used to leverage theIQ.

• Construction: during construction, the project team willaudit construction of each system and witness primaryconstruction activities, such as loop testing, pressure test-ing, and flushing. The team also coordinates vendor instal-lation checks and verification of SAT readiness. In addi-

tion, the bulk of the commission and qualification docu-ments are generated and issued for approval during theconstruction phase. A number of construction activitiesmust be documented, as well as witnessed, to meet GoodManufacturing Practice (GMP) requirements (e.g., weld-ing of sterile piping and verification of slopes) and will beused to leverage the IQ. By working closely together, theQA/documentation officer, supporting staff, and construc-tion team can expedite the handover from construction tocommissioning – speeding time to market.

• SAT: the construction contractor typically splits the SATinto two phases – an installation and documentationverification phase and a functional testing phase. An auditof the FAT documentation and schedule checks is per-formed to verify that no changes have occurred since theFAT (i.e., the FAT data is still valid for the IQ). Duringfunctional testing, the system is tested, coupled with theactual site utilities, and linked to the site supervisorysystem for full data transmission functionality.

• Commissioning: in the pre-commissioning phase, theconstruction contractor performs final hot-loop checks andinstrument calibrations, as well as motor and device run-in tests. The construction contractor is responsible for the



Figure 6. Integrated documentation (project and qualification files).

execution of the commissioning procedure and IQ protocol.During commissioning, each system is tested in conjunc-tion with other interface systems for the functionality asrequired by the URS and specifications. Toward the end ofthe commissioning phase, the construction contractor per-forms a pre-run of the OQ testing to “de-bug” and provideassurance of a successful OQ execution. During commis-sioning, the construction contractor may perform an ex-panded level of testing of system functionality (i.e., inexcess of the OQ requirements) to test the limits of systemcapabilities. It is important to note that not all of thesetests will necessarily need to be documented – the con-struction contractor and operating company should dis-cuss and agree upon the extent of documentation requiredas part of the commissioning plan and its documentationrequirements.

• IQ: following the SAT and pre-commissioning, the con-struction contractor performs system IQ. Much of theverification data from FAT, construction, SAT, and pre-commissioning can be leveraged into the IQ if it has beeneffectively documented, which saves the operating com-pany valuable time - Figure 6. The QA/documentationofficer is responsible for the verification of the IQ supportdocumentation files from the various project phases.

• OQ: a separate standalone OQ should be performed toenable the commissioning procedure to be segregatedfrom the formal qualification files. A pre-run of OQ testingin the commissioning phase ensures a successful OQ,which will reduce deviations and streamline the OQ pro-cess and its documentation.

Plan for Integration of Construction,Commissioning, Qualification and ValidationThe construction contractor normally is focused on gettingthe facility built and started up. However, the constructioncontractor’s overriding goal must be the integration of allfield activities with respect to documentation – the success ofthe qualification and validation process ultimately dependson the quality of documentation at each step. Typically,documentation is divided into two categories — project filesand qualification files - Figure 6.

The project files are handed over to the operating companyfor reference and archiving; they will not compose part of thequalification package. The qualification files are controlleddocuments, which fall under the change control procedure;therefore, these must reflect greater attention to the URS,specs, vendor audit, validation master plan, and project plan.From a regulatory perspective, if the engineer, constructioncontractor, or operating company makes any changes to the



systems, the qualification files must be updated so that theyreflect the validated standards of the operating company’sfacility. Associated with the handover of documents to theoperating company, the construction contractor should set upoperations and maintenance training for the operatingcompany’s key personnel, as well as document training forthe entire project team. The construction contractor alsoshould be involved in change control and deviation manage-ment.

Step 3: Develop a Project ManagementPlan Incorporating Document ManagementConsidering the voluminous documentation gathered anddeveloped during the course of constructing, commissioning,qualifying and validating a modern pharmaceutical or biotechmanufacturing facility, the control, tracking, and data stor-age mechanisms employed to amass these materials havebecome indispensable tools for efficient and successful projectmanagement.

Document Control,Review Cycles, and ApprovalsIn a successful project, document control is not the soleresponsibility of a single individual. Instead, it is a teameffort that requires early involvement of all the participants– engineers, project managers, vendors and sub-constructioncontractors, commissioning, qualification and validationgroups, including their technology leads. Policies, practices,and procedures developed by this team must be rigorouslyadhered to by all parties throughout the project. The con-struction contractor’s role cannot be understated — evenprocurement documents gathered in the early stages of theeffort will become an important part of the TOP that willsupport the licensure and regulatory approval process.

If the operating company already has established docu-ment practices, procedures, and change control policies inplace, the construction contractor should establish documentmethods for distribution, review cycles, version control, num-bering, and approvals that mesh with these practices. Theconstruction contractor may use an electronic documentcontrol platform to create distribution groups for recipients ofvarious types of documents and control different versions ofdocuments as they are developed, recording all relevantcomments and the names of their originators.

Data SecurityThe operating company should ensure that the constructioncontractor is prepared to physically secure and store criticaloriginal hard-copy documents in fireproof cabinets, and en-sure that electronic versions of these documents are backedup regularly on two systems in two separate locations. Often,the construction contractor’s project managers will ensurethat copies of critical project documents are maintained atthree locations.

Reducing Time to MarketHigh-quality documentation is essential to achieve regula-

tory compliance. The construction contractor must appreci-ate the significance of documentation, and make it an inte-gral part of the construction planning, implementation, andcommissioning process from the inception of the project. Theconstruction contractor’s goal is not only to build a facility ontime and within budget with systems that perform to speci-fications – it is also to develop a TOP that is well organized,meets the operating company’s unique needs and expecta-tions, and provides the proper level of documentation quality.When the construction contractor effectively manages theconstruction and commissioning documentation process, theoperating company can leverage the resulting documenta-tion for the qualification process, reducing time to market.

References1. ISPE Baseline® Pharmaceutical Engineering Guide, Vol-

ume 5 - Commissioning and Qualification, InternationalSociety for Pharmaceutical Engineering (ISPE), First Edi-tion, March 2001, www.ispe.org.

AcknowledgementThe authors would like to thank Mr. Wim Vanhoutte forgiving permission to adapt his diagram (on leveraging com-missioning into qualification).

About the AuthorsWael Allan is President and Managing Di-rector of the Skanska Pharmaceutical Group,leading its commissioning, qualification/vali-dation, and regulatory compliance servicesglobally. He has substantial internationalexperience and strong bioprocess backgroundinvolving the design, commissioning, quali-fication, and validation of biopharmaceutical

facilities. He has obtained international status performingon projects for high-end pharmaceutical and biotech compa-nies worldwide with knowledge of biochemical and chemicalengineering. Allan holds a master’s degree in biochemicalengineering and a bachelor’s degree in chemical engineering,graduating with honors from the University of Swansea inthe United Kingdom. He can be contacted by telephone inUSA (973) 390-9219 or via email:[email protected].

Skanska Pharmaceutical Group, Skanska USA BuildingInc., 1633 Littleton Rd., Parsippany, New Jersey 07054.

Andrew Skibo is Vice President of Corpo-rate Engineering & Capital Projects forAmgen. He joined Amgen in August 2004with more than 25 years of experience inengineering and construction managementat Monsanto, Fluor-Daniel, Genentech, LifeSciences International, Foster Wheeler Cor-poration, and Skanska Pharmaceutical

Group. Skibo holds a master’s degree in chemical engineeringand a bachelor’s degree in Chemistry from MIT.

Amgen Inc., One Amgen Center Dr., 90-2-B, ThousandOaks, California 91320-1799.

Risk Assessment


Elicitation of Expert Knowledge aboutRisks Associated with PharmaceuticalManufacturing Processesby Dr. Nga L. Tran, Brian Hasselbalch,Dr. Kara Morgan, and Dr. Gregg Claycamp

This articledescribes thedevelopment,implementation,and results ofan expertelicitationsurvey aboutrisks associatedwithpharmaceuticalmanufacturingprocesses, anddiscussespotentialapplication ofthis datacollectionmethodology toa broader rangeof experts.

Introduction

Recently, the FDA-Center for DrugEvaluation and Research (CDER) con-ducted a survey to elicit expert knowl-edge about risks associated with the

manufacturing processes of a number of phar-maceutical product types. This survey was car-ried out as part of the Center’s ongoing effort todevelop and implement a systematic approachto prioritize sites for routine cGMP inspection.

The International Conference on Harmoni-zation (ICH) in the current draft of ICH docu-ment Q9, Quality of Risk Management, definesrisk as a combination of the probability of theoccurrence of harm and the severity of thatharm. As an ICH participant, CDER recog-nizes this definition of risk.

Prioritizing sites for inspection has been along-standing challenge for Agency managers.Historically, FDA district offices have identi-fied sites for annual inspection based on avariety of informally applied factors, includ-ing, for example, a district manager’s knowl-edge of the inspectional history and corporateculture of the district as well as the perceivedrisk to the public health of manufacturingerrors. More recently, under the cGMP Initia-tive, FDA-CDER has implemented a system-atic approach to prioritize sites for inspection

in order to ensure that FDA inspectional re-sources and oversight achieve the maximumpublic health impact. This effort thus far hasled to a risk ranking framework that is based onthree principal components: Product, Process,and Facility-Table A. A more detailed descrip-tion of the CDER-risk-ranking model has beendescribed in a white paper published by theAgency.1

To implement this risk-ranking framework,a risk estimate, rank, or weight must be as-signed to the factor associated with each top-level component (Product, Process, and Facil-ity). Such weight assignments ultimately de-termined the final site score, which would beused to rank and select site for inspection. Assuch, whenever possible, the weight assign-ment would be objectively based on empiricaldata. In order to estimate the relative contribu-tion to risk for the product and facility scores,we used the available information on productrecall, inspection, and compliance histories tooperationalize these aspects of the risk-rank-ing framework. However, such data do not existfor factors relating to the process component.

The key issues in the implementation of theprocess factor of the risk-ranking model involvequestions concerning the relevant inherent pro-cess risk factors, the relevant process control

Table A. Top-levelcomponents for the siteselection model.1

Factor Category Description Example(s)

Product Factors pertaining to the intrinsic properties of drug products such Dosage form; intrinsic chemicalthat quality deficiencies could potentially and adversely impact public propertieshealth.

Facility Factors relating to characteristics of a manufacturing site believed to Poor CGMP compliance historybe predictive of potential quality risks, such as the lack of effectivequality systems.

Process Factors pertaining to aspects of drug manufacturing operations that Measuring; mixing; compression;may predict potential difficulties with process control and/or fillingvulnerability to various forms of contamination.



Risk Assessment


and risk mitigation factors, and how to weight the importanceor rank them. Although the Agency does not have the infor-mation needed to answer these questions, the Agency doeshave a large number of staff with expertise in this area. Anexpert elicitation survey was developed by an Agency-wideworking group to systematically capture this body of knowl-edge, and formulate the key process-related factors andweights for inclusion in the current risk-ranking model.

Although it is preferred that data used in decision-makingare empirically derived, it is widely recognized that theneeded data are sometimes not available or, if available, areincomplete, unreliable, or only indirectly applicable. In suchcases, expert judgment is the only way to complete therequired knowledge. Expert data obtained under rigorousmethodological rules are increasingly being recognized as avaluable asset in numerous scientific fields, including chem-istry, nuclear sciences, seismic, and civil applications.

MethodsExpert Elicitation Survey DevelopmentAn FDA working group that included expertise in pharma-

Figure 1. Box plots of product rankings based on potential for a loss of state of control.

ceutical manufacturing sciences, chemistry, risk analyses,and expert elicitation was established to develop the expertelicitation survey. The working group was initially con-fronted with several broad questions including:

• What are the relevant process-related risk factors?• What are the sources of variability and poor quality?• What, if any, units of operation and/or products are more

liable to a loss of control or at risk to contamination?

Working group members agreed that answers to these chal-lenging questions would depend on the type of productsinvolved. However, it also was acknowledged that given thelarge number of potential products, it would not be feasible toconduct a survey that would elicit answers for every possiblecombination of product and manufacturing step (or unit ofoperation). To facilitate the survey, we recognized the need toidentify “mutually exclusive” categories of products and unitsof operation and were encouraged by ISPE’s approach pub-lished in its Baseline® Guide on Oral Solid Dosage Forms.4 Inthis Guide, ISPE characterizes levels of effort and difficulty

Risk Assessment


across a variety of areas of consideration in constructing anew production facility broadly by unit operation and equip-ment level.

In general, the survey we used was designed to elicit fromrespondent a relative ranking of the likelihood of a loss of astate of control and of the vulnerability of the process tocontamination for a product category and for each individualprocessing operation associated with that product category.Experts were asked to rate the manufacturing steps accord-ing to the commonly employed manufacturing operations(e.g., measuring, mixing, compression, and filling) and for avariety of product categories (e.g., immediate and modifiedrelease solid-oral drugs, sterile liquids, metered dose inhal-ers, and active ingredients by chemical and fermentationprocesses).

Subsequent to the initial discussion, the Working Groupmet on several occasions to discuss and identify variablesthat would be used to evaluate the risk to product failure andvariability, “mutually exclusive” categories of products andkey units of operation typically associated with these productgroups. The following sections describe these steps.

Step 1: Identifying Variablesof Interest and Developing Survey QuestionsA list of potential variables that could be used to evaluate riskof product failures and variability were first generated by theWorking Group members. Among the initial list were: con-tamination (product to product and environment to product),protecting operators (if operators could be harmed by expo-sure to material under process, it could result in less controlor attentiveness to quality), yield, changeover, cleanability,validation/qualification (validation to be defined as inclusiveof qualification), and maintenance. From this initial list ofvariables, the Working Group identified two broad types ofprocess-related factors:

• factors associated with maintaining process control, i.e.process control variables

• factors associated with potential vulnerability to productor environmental contamination, i.e., contamination vari-ables

Figure 2. Box plots of product rankings based on potential for contamination.

Risk Assessment


1. API Fermentation2. API Synthesized3. Biotech4. Liquids, Non-Sterile, Solution5. Liquids, Non-Sterile, Suspension/Emulsion6. Liquids, Sterile, Solution7. Liquids, Sterile, Suspension/Emulsion8. Metered Dose Inhaler (MDI), High Active9. Metered Dose Inhaler (MDI), Low Active10. Powders, High Active11. Powders, Low Active12. Semisolid (Ointment/Cream), High Active13. Semisolid (Ointment/Cream), Low Active14. Solid Oral Drugs, Immediate Release, High Active15. Solid Oral Drugs, Immediate Release, Low Active16. Solid Oral Drugs, Modified Release, High Active17. Solid Oral Drugs, Modified Release, Low Active18. Transdermal

Table B. Product categories in process elicitation survey.

As these two main factors were crystallized as the centralfocus of the expert elicitation survey, questions were devel-oped to capture the important concepts underlying each ofthese factors. The following three questions were constructedto capture the experts’ input on the three mutually exclusiveelements of risk to loss of control deemed to be critical by theWorking Group. Response options are shown after eachquestion.

1. To what degree does this unit of operation contribute tovariability in quality of the final product?1. minimal; 2. minimal to moderate; 3. moderate to high;4. high to very high; 5. very high

2. How difficult is it to maintain this unit of operation in astate of control?1. slightly; 2. slightly to moderately; 3. moderately; 4.moderately to very; 5. very

3. If a problem does occur, how reliable are the currentdetection methods?1. very; 2. very to moderately; 3. moderately; 4. moderatelyto slightly; 5. slightly

And the next two questions were developed to capture theexpert judgment on the two mutually exclusive elementsdeemed critical by the Working Group regarding contamina-tion:

4. Is this unit of operation more or less vulnerable to con-tamination from previous product?1. slightly; 2. slightly to moderately; 3. moderately; 4.moderately to very; 5. very

5. Is this unit of operation more or less vulnerable to con-tamination from the environment?1. slightly vulnerable; 2. slightly to moderately vulner-able; 3. moderately vulnerable; 4. moderately to veryvulnerable; 5. very vulnerable

Step 2: Identifying Product Categories and Unitsof OperationBecause the manufacturing of pharmaceutical products closelytrack product dosage form, products were categorized bydosage form, i.e., tablets, liquids, and metered dose inhalers.For each dosage form, additional distinction would be madeif it was determined by the Working Group that such distinc-tion would lead to a different answer to the questions listedin Step 1. For example, higher and lower active weightcontent was used to further categorize similar dosage formssince the Working Group members believed that the respond-ers would need to make these distinctions in order to be ableto accurately answer the posed questions. Using this ap-proach, the Working Group identified 18 mutually exclusiveproduct categories to be included in the expert elicitationsurvey. Table B lists these product categories.

To identify the manufacturing steps that are typicallyassociated with the majority of the above product categories,the Working Group relied on its own expertise as well as thefollowing references:

• Remington: Pharmaceutical Sciences, 18th edition5

• Modern Pharmaceutics, 3rd edition6

• Pharmaceutical Process Validation, 3rd edition7

• ISPE Baseline® Pharmaceutical Engineering Guide, Vol.2, Oral Solid Dosage Forms, 1st edition4

Expert Selection and Survey DeliveryPrior to the full implementation of the survey, a pilot surveywas conducted in December 2003 through in-person inter-views with five FDA experts. Feedback on the clarity of thesurvey instructions, questions, options for answering thequestions, product categories, and units of operation wereobtained from the pilot survey. In general, the pilot surveyshowed that the survey was clear and questions were answer-able. Based on comments received from the pilot survey,minor refinements were made and the survey was finalizedprior to final delivery to a full panel of experts.

The panel of FDA experts to whom the survey was deliv-ered was selected from the following groups: 1) reviewersfrom CDER, 2) senior CDER staff in the Office of Compliance,and 3) senior Office of Regulatory Affairs (ORA) field staff.Fifty experts were selected for the survey, based on theexpertise needed and the level of experience of the individu-als. The overall response rate was 100%. The survey wasconducted in May 2004. The survey was sent to the expertsvia email, and the experts were instructed to print andcomplete the survey by hand. Data from the completed andreturned surveys were entered by the Office of Compliancestaff. Data quality assurance was conducted by the staff of theOffice of Compliance.

Analyses and ResultsAverage Summary of ResponsesRanking responses on a 5-point scale (from 1 as the lowest to5 as the highest rank) as elicited from the survey for questions

Risk Assessment


Potential for a Loss of Control Potential for Contamination

Product Category Average Ranking Standard Deviation Average Ranking Standard Deviation Percent(Questions 1, 2, &3) (Questions 4 & 5) Response

Biotech 3.1 0.5 3.0 0.7 48%Liquids, Sterile, Solution 3.0 0.8 2.7 0.8 92%Liquids, Sterile, Suspension/Emulsion 3.1 0.7 2.7 0.8 100%Metered Dose Inhaler, High Active 3.0 0.6 2.6 0.7 62%Metered Dose Inhaler, Low Active 3.2 0.6 2.6 0.7 62%Liquids, Non-Sterile, solution 2.0 0.6 2.1 0.9 94%Powders, High Active 2.3 0.6 2.4 0.9 92%Solid Oral Drugs, Immediate Release, High Active 2.3 0.5 2.1 0.6 94%Liquids, Non-Sterile, Suspension/Emulsion 2.5 0.7 2.3 0.9 100%Semisolid (Ointment/Cream), High Active 2.5 0.5 2.2 0.7 84%API, Synthesized 2.6 0.6 2.2 0.6 100%Solid Oral Dose, Immediate Release, Low Active 2.6 0.5 2.2 0.6 94%Solid Oral Drugs, Modified Release, High Active 2.6 0.5 2.1 0.6 92%Powders, Low Active 2.7 0.6 2.5 0.9 92%Semisolid (Ointment/Cream), Low Active 2.7 0.6 2.2 0.7 82%API, Fermentation 2.8 0.6 2.3 0.7 100%Solid Oral Drugs, Modified Release, Low Active 2.8 0.5 2.2 0.6 92%Transdermal 2.8 0.6 2.2 0.7 66%

Table C. Average ranking and response rate for each product category.

1, 2, and 3 were averaged together to represent averagerating of risk for the potential loss of control. Then, they wereaveraged across units of operation and respondents to deter-mine the average risk ranking for potential loss of state ofcontrol for each product category. Similarly, responses toquestions 4 and 5 were averaged across units of operation andrespondents to determine the average rank of potential forcontamination for each product category. Table C summa-rizes the average ranks and standard deviations for potentialloss of state of control and potential for contamination foreach product category. Biotech, MDI (both high and lowactive), and sterile liquid (both solution and suspension/emulsions) product categories have the highest average rank-ing for both potential for loss of a state of control and potentialfor contamination.

Experiences with product categories were not equal amongthe surveyed experts. As such, not all respondents providedanswers to all product categories included in the survey.Biotech, MDI, and Transdermal product categories have thelowest response rates, 48%, 62%, and 66%, respectively.Response rate for each product category also is summarizedin Table C.

Box-plots of the ranking responses for questions 1, 2, and3, which were averaged together and across units of opera-tions to represent average rating of potential for a loss ofcontrol for each product category are shown in Figure 1.Similarly, box plots of responses for questions 4 and 5 torepresent potential risk of contamination for each productcategory are shown in Figure 2. Biotech, MDI (both high andlow active), and sterile liquid (both solution and suspension/emulsions) product categories remain the top ranked productcategories based on median scores.

Cluster Analyses of Responses on Combinationsof Product Categories and Units of OperationIn addition to averaging the responses, multivariate K-Meanclustering analyses of responses to the combinations of prod-uct category and unit of operation also were carried out usingS-Plus.8 Responses to questions 1, 2, and 3 for the productcategory and unit of operation combinations were clusteredinto five groups. Each cluster was assigned a ranking basedon the rank-order of the clusters’ centers, i.e., cluster with thehighest center was given the highest rank of five and clusterwith the lowest center is given the lowest rank of one. Aproduct category and unit of operation combination belong-ing to a cluster would assume its cluster rank. A similarclustering approach also was applied to questions 4 and 5. Asin previous averaging analysis, the cluster ranks based onquestions 1, 2, and 3 provide ranking of potential risk of lossof state of control, and the cluster ranks based on questions4 and 5 provide the ranking of potential risk of contamina-tion.

Ranking of Potentialfor Loss of a State of ControlThe cluster ranking of the combinations of product categoriesand units of operation resulted in the same top five rankedproduct categories (biotech, liquid sterile solution, liquidsterile suspension/emulsion, MDI low active, and MDI highactive) as those ranked based on averaging responses. Withineach product category, ranking varied between units of op-eration. While most of the processing steps associated withthe top five ranked product categories also are ranked high,the measuring step is typically ranked lower. For productcategories with overall low ranking, such as the solid oral

Risk Assessment


Table D. Cluster ranking of product categories and units of operation for potential loss of a state of control.

Product Product Units of Operation Unit ofCategories Category Operation

Ranking Rankings

Biotech 5 Bioreaction, Seed; Bioreaction, Production; Cell Bank Maintenance; Isolation Recovery; Pasteurization;Purification; Viral Clearance ........................................................................................................................................... 5Filling; Formulation ........................................................................................................................................................ 4Measuring ..................................................................................................................................................................... 3

Liquid, Sterile, 5 Aseptic Filling, Form-Fill-Seal; Aseptic Filling, Isolator; Aseptic Filling-Traditional Method; Mixing Blending;Suspension/Emulsion Terminal Sterilization ..................................................................................................................................................... 5

Measuring ..................................................................................................................................................................... 3

Liquid, Sterile, 5 Aseptic Filling, Form-Fill-Seal; Aseptic Filling, Isolator; Aseptic Filling-Traditional Method; Filtration;Solution Lyophilization; Terminal Sterilization .............................................................................................................................. 5

Measuring ..................................................................................................................................................................... 3Mixing Blending ............................................................................................................................................................. 2

Metered Dose Inhaler 5 Assembly; Filling; Micronization of components; Mixing Blending .................................................................................... 5(MDI), Low Active Measuring ..................................................................................................................................................................... 4

Metered Dose Inhaler 5 Assembly; Filling; Micronization of components; Mixing Blending .................................................................................... 5(MDI), High Active Measuring ..................................................................................................................................................................... 3

API Fermentation 4 Fermentation; Inactivation; Isolation; Purification ........................................................................................................... 5Processing ..................................................................................................................................................................... 4Primary Packaging; Weighing ......................................................................................................................................... 2

API Synthesized 4 Isolation; Purification; Reaction ..................................................................................................................................... 5Processing; Workup ....................................................................................................................................................... 4Weighing ....................................................................................................................................................................... 3Primary Packaging ......................................................................................................................................................... 2

Powders, Low Active 4 Mixing Blending ............................................................................................................................................................. 5Milling ........................................................................................................................................................................... 4Measuring ..................................................................................................................................................................... 3Primary Packaging ......................................................................................................................................................... 2

Semisolids (Ointment/ 4 Emulsification; Mixing Blending ...................................................................................................................................... 5Cream), Low Active Deaeration; Heating Cooling ........................................................................................................................................... 4

Measuring ..................................................................................................................................................................... 3Primary Packaging ......................................................................................................................................................... 2

Solid Oral, Modified 4 Coating; Pelleting .......................................................................................................................................................... 5Release, Low Active Compression (tablet); Drying; Encapsulation (hard gel); Granulation (dry and wet); Milling; Mixing Blending ..................... 4


Transdermal 4 Active Deposition, Coating; Extrusion ............................................................................................................................ 5Cutting; Drying; Mixing Blending; Primary Packaging ...................................................................................................... 4Measuring ..................................................................................................................................................................... 3

Liquid, Non-Sterile, 3 Emulsification; Mixing Blending ...................................................................................................................................... 5Suspension/ Emulsion Measuring ..................................................................................................................................................................... 3

Primary Packaging ......................................................................................................................................................... 1

Semisolids (Ointment/ 3 Emulsification ................................................................................................................................................................ 5Cream), High Active Deaeration; Heating Cooling; Mixing Blending ................................................................................................................. 4


Solid Oral, Immediate 3 Mixing Blending; Granulation (dry and wet) .................................................................................................................... 5Release, Low Active Coating; Compression (tablet); Drying; Encapsulation (hard gel); Milling; Pelleting ........................................................... 4


Solid Oral, Modified 3 Coating; Pelleting .......................................................................................................................................................... 5Release, High Active Compression (tablet); Drying; Encapsulation (hard gel); Granulation (dry and wet); Milling; Mixing Blending ..................... 4


Solid Oral, Immediate 2 Compression (tablet); Granulation (dry and wet); Mixing Blending; Pelleting .................................................................... 4Release, High Active Measuring ..................................................................................................................................................................... 3

Coating; Drying; Encapsulation (hard gel); Milling; Primary Packaging ............................................................................. 2

Powders, High Active 2 Milling; Mixing Blending ................................................................................................................................................. 4Measuring ..................................................................................................................................................................... 3Primary Packaging ......................................................................................................................................................... 2

Liquid, Non-Sterile, 1 Measuring ..................................................................................................................................................................... 3Solution Mixing Blending ............................................................................................................................................................. 2

Primary Packaging ......................................................................................................................................................... 2

Risk Assessment


immediate release high active category, most processingsteps are ranked low; however, there are several processingsteps that are ranked high, such as compression (tablet), wetand dry granulation, mixing-blending, and pelleting. Productand unit of operation rankings for potential loss of a state ofcontrol based on K-mean cluster analysis are summarized inTable D.

Potential for ContaminationCluster ranking of contamination risks for the combinationsof product categories and units of operation also resulted inbiotech, liquid sterile solution, liquid sterile suspension/emulsion, MDI low active, and MDI high active as the topranked product categories. Ranking also varied betweenunits of operation within each product category. Product andunit of operation rankings of contamination risks based on K-mean cluster analysis are summarized in Table E.

Discussion and RecommendationsSurvey ProtocolFormal methods for obtaining the judgments of experts havebeen evolving since their inception after World War II.Despite its long history of application, standardized protocolsfor the selection, preparation, and elicitation of experts do notand should not exist.9 Analysts in the field of expert elicita-tion have consistently argued that rather than standardizedprocedures, protocols should be crafted to suit the particularproblem under investigation.3,10,11 In accordance with conven-tional practice, an FDA team developed a protocol, in a formof a survey and detailed instructions, to elicit expert judg-ments about the potential for a pharmaceutical manufactur-ing process to be subject to loss of process controls or contami-nation. As such, it should be noted that the scope of theelicitation is limited to obtaining expert judgments about thelikelihood relating to the manufacturing processes such thatif a product category is judged to involve more risky manufac-turing steps it would then have a higher potential for poorquality. This protocol does not extend to judgments about riskto public health.

ExpertsExpert judgment studies make use of a panel of experts whobring in different information, arising from different inter-pretations, different analytical methods, and/or differentexperiences.2 Fifty-five experienced FDA officials were cho-sen to participate in this survey and 50 responses werereceived. Nearly half of the participants were senior drugprogram investigators from the Office of Regulatory Affairswith the remaining being senior review and drug cGMPcompliance officials in the Centers for Drug Evaluation andResearch and Veterinary Medicine. Review staff representeddisciplines such as chemistry, engineering, biochemistry,microbiology, pharmacology, and pharmacy. Nearly all re-sponders reported having 10 or more years combined experi-ence in FDA and the drug industry.

UtilityInformation obtained from the survey has been of greatutility in the implementation of the risk ranking model toprioritize pharmaceutical sites for cGMP inspection. To imple-ment this risk-ranking framework, a risk ranking (or weight)is first assigned to the factors associated with each top-levelcomponent (Product, Process, and Facility) and subsequently,the combination of these factor-ranks (weights) would deter-mine the site overall potential risk scores, which would beused to rank and target inspection. As previously indicated,the Agency has systematically compiled product and facilityrelated information such as product recall, inspection, andcompliance histories that could be used to operationalizethese aspects of the risk ranking framework. However, suchdata do not exist for factors relating to the process component.The expert elicitation survey provides a systematic means ofgathering knowledge and an objective approach to assignranks to the factors associated with the process component ofthe risk-ranking model. Ranking results also provide a basisfor investigators to better focus their product quality inspec-tion. For example, once a site has been chosen for inspectionbased on overall site risk score, variability in the ranking ofunits of operation within each product category (Tables D andE) could help the inspector to focus on units of operation thathave been ranked as more vulnerable to potential loss ofprocess controls or contamination.

The results from this formal and systematic approach ofcollating judgments from a broad range of experts also couldprovide the pharmaceutical industry with benchmark data,which can be used to examine a company’s risk assessmentpractices. If a company’s assessment leads to conclusionsthat are different from the experts’ norm then additionalevaluation can be carried out to determine reasons for differ-ences.

LimitationsThere are a number of limitations associated with this sur-vey. First and foremost, since the expert elicitation surveywas only delivered to FDA experts, the results reported inthis article do not capture the broad range of expertise thatexists outside of FDA. The response rate to the surveys for theBiotech Product Category was only 48% (only 24 respondedout of 50 surveyed experts). While the expert elicitation wasnot a random survey and statistical validity is not at issue,the low response rate presents some concern with regard tothe potential lack of expertise in the biotech area among thepool of experts included in this survey. An additional consid-eration is the fact that the survey was designed to elicitjudgments about the manufacturing risks associated withvery broad product categories (Table B) and not specificproduct. As such, experts were forced to average their an-swers across a broad range of products that fall into suchproduct category. While broad aggregation of products helpedto facilitate the delivering of the survey i.e., reducerespondent’s time spent on the survey and fatigue, the conse-quence could be a loss of a significant amount of information.

Risk Assessment


Product Product Units of Operation Unit ofCategories Category Operation

Ranking Rankings

Biotech 5 Bioreaction, Production; Bioreaction, Seed; Filling; Formulation; Isolation Recovery; Purification; Viral Clearance ........... 4Cell Bank Maintenance; Measuring; Pasteurization ......................................................................................................... 3

Liquid, Sterile, 4 Aseptic Filling-Traditional Method .................................................................................................................................. 5Solution Aseptic Filling, Form-Fill-Seal; Aseptic Filling, Isolator; Filtration; Lyophilization; Measuring; Mixing Blending .................. 3

Terminal Sterilization ..................................................................................................................................................... 1

Liquid, Sterile, 4 Aseptic Filling-Traditional Method .................................................................................................................................. 5Suspension/Emulsion Aseptic Filling, Form-Fill-Seal; Aseptic Filling, Isolator; Measuring; Mixing Blending ........................................................ 3

Terminal Sterilization ..................................................................................................................................................... 1

Metered Dose Inhaler 3 Micronization of components ......................................................................................................................................... 4(MDI), High and Filling; Measuring; Mixing Blending ................................................................................................................................ 3Low Active Assembly ...................................................................................................................................................................... 1

Powders, High and 2 Milling; Mixing Blending ................................................................................................................................................. 2Low Active Measuring; Primary Packaging ....................................................................................................................................... 1

API Fermentation 2 Fermentation ................................................................................................................................................................. 3Inactivation; Isolation; Processing; Purification .............................................................................................................. 2Primary Packaging; Weighing ......................................................................................................................................... 1

API Synthesized 1 Processing; Purification ................................................................................................................................................. 2Isolation; Primary Packaging; Reaction; Weighing; Workup ............................................................................................. 1

Liquid, Non-Sterile, 1 Mixing Blending ............................................................................................................................................................. 2Solution Measuring; Primary Packaging ....................................................................................................................................... 1

Liquid, Non-Sterile, 1 Mixing Blending Emulsification ....................................................................................................................................... 2Suspension/Emulsion Measuring; Primary Packaging ....................................................................................................................................... 1

Semisolids (Ointment/ 1 Emulsification; Mixing Blending ...................................................................................................................................... 2Cream), High and Deaeration; Heating Cooling; Measuring; Primary Packaging .......................................................................................... 1Low Active

Solid Oral, Immediate 1 Granulation (dry and wet) Milling; Mixing Blending ......................................................................................................... 2Release, High and Coating; Compression (tablet); Drying; Encapsulation (hard gel); Measuring; Pelleting; Primary Packaging ....................... 1Low Active

Solid Oral, Modified 1 Compression (tablet); Granulation (dry and wet); Milling; Measuring; Mixing Blending ..................................................... 2Release, Low Active Coating; Drying; Encapsulation (hard gel); Pelleting; Primary Packaging .......................................................................... 1

Solid Oral, Modified 1 Granulation (dry and wet); Milling; Mixing Blending ........................................................................................................ 2Release, High Active Compression (tablet); Coating; Drying; Encapsulation (hard gel); Measuring; Pelleting; Primary Packaging ....................... 1

Transdermal 1 Active deposition, coating .............................................................................................................................................. 3Extrusion; Mixing Blending ............................................................................................................................................. 2Cutting; Drying; Measuring; Primary Packaging .............................................................................................................. 1

Table E. Cluster ranking of product categories and units of operation for contamination risks.

Question 3 in the survey (“If a problem does occur, howreliable are the current detection methods?) would requireexperts to account for the average rate of firms implementingexpected process controls. As such, the results from thissurvey do not reflect risk associated with firms that areperforming below average expectation or standard industrypractices, i.e. not implementing minimal in-process controls;nor do results reflect firms exceeding expectations, e.g., firmswith Process Analytical Technologies (PATs). Nevertheless,for the purpose of selecting a site for cGMP inspection, suchdeviation from average/expected practices would be capturedduring the actual inspection. As such, using the results fromthis survey for the site-selection model does not preclude theinspector’s ability to differentiate between firms with en-hanced controls from those performing below averages.

RecommendationsIn light of the limitations described above, the following

recommendations are provided to improve the expert elicita-tion survey:

• Expand the expert panel to include expertise outside ofFDA such as ISPE working members.

ISPE has a broad range of members who would havecurrent working knowledge of the robustness and capa-bilities for a variety of products. They are likely to befamiliar with units of operation that require frequentattention and in-process monitoring and maintenance.Hence, inclusion of expert judgments from this groupwould greatly enhance knowledge about risk associatedwith various pharmaceutical manufacturing processes.

• Future revision of the survey protocol should considerfurther differentiation of the existing product categoriesand units of operation.

Risk Assessment


In the current survey, products are categorized based onbroad dosage forms. These broad dosage forms could befurther differentiated. For example, the oral solid dosageform could be differentiated into several product group-ings, including hard and soft capsules and tablets. Fur-ther, for products with additional processing steps that arenot captured in the current survey, these additional stepsshould be identified and included in future revision of thesurvey. Differentiations should be made where expertsbelieve there are true differences.

• Uncertainty

Expert knowledge is not a certainty, but it is entertainedwith an implicit level of confidence or degree of belief.2

Survey methods that allow the experts to express theirdegree of confidence in their responses will also permit adetermination of the level of confidence in models that usethese data. As such, future surveys should allow experts toexpress uncertainties in their responses.

References1. FDA-Center for Drug Evaluation and Research (FDA-

CDER), Risk-Based Method for Prioritizing CGMP In-spections of Pharmaceutical Manufacturing Sites – APilot Risk Ranking Model. Department of Health andHuman Services U.S. Food and Drug Administration.Rockville, M.D., September 2004, http://www.fda.gov.

2. Van Der Fels-Klerx, I. H. J., L.H.J. Goossens, H.W.Saatkampm and S.H.S. Horst, “Elicitation of Quantita-tive Data from a Heterogeneous Expert Panel: FormalProcess and Application in Animal Health,” Risk Analy-sis, Vol. 22, No. 1, 2002, pp. 67- 81.

3. Morgan, M.G. and M. Herion, Uncertainty: A Guide toDealing with Uncertainty in Quantitative Risk and PolicyAnalysis. N.Y., Cambridge University Press, 1990.

4. International Society for Pharmaceutical EngineeringInc. (ISPE), Baseline® Pharmaceutical Engineering Guide,Vol. 2, Oral Solid Dosage Forms, 1st edition, 1998.

5. Remington’s Pharmaceutical Sciences, 18th edition, P.A.,Mack Publishing Company, 1990, p. 2000.

6. Modern Pharmaceutics, 3rd edition, Revised and Ex-panded, N.Y., Marcel Dekker, Inc., 1996, p. 943.

7. Pharmaceutical Process Validation, 3rd edition, Revisedand Expanded. N.Y., Marcel and Dekker, Inc., 2003, p.860.

8. S-Plus for Windows Professional Version 4.5 Release 1.MathSoft, Inc. 1998.

9. Walker, K, J.S. Evans, and D. MacIntosh, “Use of ExpertJudgment in Exposure Assessment Part I. Characteriza-tion of Personal Exposure to Benzene,” Journal of Expo-sure Analysis and Environmental Epidemiology, Vol. 11,2001, p. 308-322.

10. Hora, S.C. 1992. Acquisition of Expert Judgment: Ex-amples from Risk Assessment. J. Energy Eng. 188(2):136-148.

11. Otway, H. and D. von Winterfeldt, “Expert Judgment inRisk Analysis and Management Process, Context, andPitfalls,” Risk Analysis, Vol. 12, No. 1, 1992, p. 83-93.

DisclaimerThe views expressed herein do not represent official FDA orUS government policy. No official support or endorsement bythe FDA is intended or should be inferred.

AcknowledgementWe would like to thank the many FDA staff who helpeddesign and develop this survey, especially Charles Gray ofthe Center for Veterinary Medicine, Diane Kelley and Tho-mas Arista from the Office of Regulatory Affairs, Elaine Colefrom the Center for Biologics Evaluation and Research, andNicholas Buhay, Jon Clark, Lindsay Cobbs, Stephen Mahoney,Vilayat Sayeed, Rajendra Uppoor, and Christopher Wattsfrom the Center for Drug Evaluation and Research. Wegreatly appreciate the participation of our 50 FDA expertrespondents who will remain anonymous. Last but not least,we would like to thank CDER-Director of the Office of Com-pliance, David Horowitz, without whose support and forwardthinking this project could not have happened.

About the AuthorsDr. Nga Tran is a Senior Managing Scien-tist at Exponent’s Food and Chemicals prac-tice in Washington, DC. She earned herDr.P.H. in environmental health sciences atJohns Hopkins University, Bloomberg Schoolof Public Health in 1994, and her Masters inpublic health at Yale University, Depart-ment of Epidemiology and Public Health in

1985. Dr. Tran has more than 15 years of experience inenvironmental and occupational health risk assessment fromthe private and public sectors. Prior to joining Exponent, Dr.Tran was a faculty member at the Johns Hopkins University,Bloomberg School of Public Health where she conductedresearch and taught exposure and risk assessment, riskprioritization, and risk harmonization. Dr. Tran remains anAdjunct Assistant Professor at the University.

Exponent, 1730 Rhode Island Ave. NW, Suite 1100, Wash-ington, DC 20036.

Brian Hasselbalch is a Senior Officer withthe U.S. Food and Drug Administration’sCenter for Drug Evaluation and Research,where he works on cGMP related guidanceand policy matters, and reviews recommen-dations for regulatory action. He is a formerFDA drug process investigator. Hasselbalchreceived his BS from University of California

at Riverside. He has been with FDA for 15 years.

Risk Assessment


Dr. Kara Morgan is the Senior Advisor forRisk Analysis in the Office of Policy andPlanning in the Office of the Commissionerat the U.S. Food and Drug Administration.She earned her PhD in engineering and pub-lic policy at Carnegie Mellon University in1999, and her Masters in environmental sci-ence from the School of Public and Environ-

mental Affairs at Indiana University in 1995. She has 13years of experience in risk and decision analysis. Dr. Morgan’sresearch interest focuses on developing tools to supporteffective risk management decisions in the face of uncer-tainty.

Dr. Gregg Claycamp is Director of Scien-tific Support Staff at the U.S. Food and DrugAdministration’s Center for Veterinary Medi-cine, Office of New Animal Drug Evaluation.He received his MS and PhD in radiologicalhealth engineering from Northwestern Uni-versity in 1977 and 1982, respectively. Dr.Claycamp has more than 25 years of experi-

ence in risk assessment and decision analysis. Prior to FDA,he was Professor and Associate Chairman at the Universityof Pittsburg, Department of Environmental and Occupa-tional Health.

Risk-Based Regulation


A Precedent for Risk-BasedRegulation?by Jonathan Coburn, Stanley H. Levinson, PhD, PE,and Greg Weddle

This articlecompares “risk-based”regulation as anew concept inthepharmaceuticalindustry against30 years ofexperience inthe nuclearpower industry.

Introduction

The life sciences industry has impres-sive capabilities for identifying and ana-lyzing risk. Yet, the industry’s applica-tion of this knowledge and expertise is

skewed. The industry’s risk assessment capa-

bility is unsurpassed when it comes to assess-ing the risks drug compounds pose to humans;however it has not applied this risk assessmentexpertise to the process of making those drugs.Instead, drug manufacturing is governed bythe “deterministic” cGMP regulations of 21

CFR 210/211. However, the FDA in-tends to change this. It recently pub-lished “Pharmaceutical cGMPS forthe 21st Century – A Risk-Based Ap-proach,”1 which stated “the FDA isimplementing a risk-based approachto regulating pharmaceutical manu-facturing.” As FDA and the industryembark upon this initiative, it mightbe interesting to consider, on a gen-eral basis, how regulation of produc-tion might transition from a deter-ministic to a risk-based approach.Here’s one possible scenario.

A Risk StoryConsider an industry whose activi-ties are deemed to be “risky,” and istherefore regulated by a governmentagency whose mission is to protectthe health and safety of the public.Historically, the regulator hasachieved its mission by using a de-terministic regulatory approachbased on good engineering practices,safety factors, and experience.

Industry events, economic forces,and changing technology drive theregulator to consider risk-assessmenttechniques as a way to improve theregulatory structure. While suitabletechniques exist, the industry hasonly slowly gained experience in us-ing them, and new technological ad-vancements now allow them to bebroadly applied. So the regulator

Figure 1. Nucleartimeline.





states its intention to move toward a risk-based regulatorystructure.

The industry perception that risk-based regulation mightsubstantially ease regulatory burden, coupled with the open-ness of the regulator to explicitly focus on risk, cause a flurryof risk-based experimentation. The industry begins to applyvarious risk assessment techniques in creative ways acrossthe entire regulatory spectrum. But soon the regulator facessome tough, fundamental issues.

Explicit consideration of risk shatters the black/white,safe/unsafe paradigm of deterministic regulation. The regu-lator knows there is no such thing as zero risk, even thoughits mission statement promises to “ensure” the health andsafety of the public. But can anything really be “ensured” ina risk-based world? The regulatory paradigm now changesfrom safe/unsafe to safe enough. The new paradigm haspotentially far reaching public relations, political, and legalramifications. The regulator and industry are faced withmany new questions, such as:

• How will risk-based regulation be viewed from outside theindustry?

• Is this a new way of regulating or a different slant on whatwe have always done?

• Will risk-based regulation replace existing regulation orcreate a second “layer” of regulation?

• How should risk-based regulation be presented to thepublic?

• How safe is “safe enough?”

The industry faces considerable economic pressures, includ-ing a high regulatory compliance cost burden. The industrysees risk-based regulation as a potential way to reducecompliance costs by eliminating existing deterministic regu-lations that impose costs without a commensurate contribu-tion to safety, or by identifying alternatives to existingregulatory requirements that are both less costly and lessrisky. The risk assessments show that it is possible to bothimprove safety and lower costs, a clear win-win-win proposi-tion for the industry, the regulator, and the public.

But all this activity gives the regulator concern. First,there is a credibility problem. For decades, the regulator hastold the public the deterministic regulatory approach was“safe,” but now this new risk-based approach seeminglychallenges the basis on which safe operation is based. Andthere are other issues. Regulators are conservative by nature,and the thought of a complete regulatory change is unset-tling. And the regulator knows that conservatism in regula-tion is not a bad thing. The existing deterministic regulatorystructure has generally served the public well, and simplyabandoning it for the promise of a better way imposes risk inand of itself. Now the regulator must ask:

• What will the new regulatory structure look like? Will it beentirely risk-based, entirely deterministic, or a mixture ofboth?

• How rapidly should the industry adopt risk-based tech-niques in making changes to a regulatory structure thathas, in the main, served the public well?

Figure 2. GAMP risk assessment approach.

A Broader Industry Risk ComparisonThe life sciences industry is unsurpassed in its ability todetermine design risk; that is, in assessing the risk to thepublic of “designing” and introducing a new drug. Thenuclear power industry, through sophisticated risk as-sessment techniques like PRA, is unsurpassed in its abilityto assess production risk. In contrast, PRA was not usedto a great extent in the design of the nuclear power plantsoperating today (though it is being used in new designs),and the application of risk assessment techniques topharmaceutical production is just beginning. Thus, eachindustry is strong where the other is weak.

This article focuses on production risk in light of theFDA’s intention to pursue risk-based cGMPs. However,a broader comparison of design vs. production and theassociated risks and regulatory approaches is instruc-tive. Table A offers a brief comparison of these issues.



After considering all the issues, the regulator issues a policystatement regarding the use of risk-assessments in regula-tory activities. Unwilling to completely abandon the existingbody of deterministic regulations, the regulator states thatrisk assessments will augment, rather than replace, thisexisting regulatory structure. To denote this shift in empha-sis, the regulator stops referring to “risk-based” regulation infavor of a new term: “risk-informed” regulation.

A policy is easily stated at a high level, but the policy mustbe implemented at the working level. Moreover, it must beimplemented in such a way that the regulator has sufficientconfidence in the results of the risk analyses to rely on them

when public safety is involved. So the following questionsarise regarding applicability, consistency, and technical ad-equacy/quality of risk assessments.

• What are the attributes of acceptable risk analysis tech-niques?

• How deeply will a regulator become involved in the detailsof risk assessment techniques when they are used forregulatory compliance?

• How important will it be to apply risk assessment tech-niques consistently across the industry? Will consistencybe audited? If so, how?

LIFE SCIENCES

Drug failures have higher probabilities but lower consequences than a nuclear accident. Someconsequences (e.g., adverse side effects) are expected for every drug, so the pharmaceutical industry hasa higher observable risk than the nuclear industry. Also, drug effects on the population develop moreslowly than could be the case for a nuclear accident, providing time for the consequences to be mitigated.

Pharmaceuticals face an inherent dilemma regarding risk. Drugs are approved considering theirrisk to the public in a statistical sense. But drug risk is considered in a personal sense by the millions ofindividuals who take prescription drugs every day. The public has a near zero risk threshold on anindividual basis, but the approval threshold is not so stringent. The public has been exposed to thewarnings and possibilities of “side effects” and relies, perhaps somewhat skeptically, on drugcompanies and physicians to help them weigh the personal risk of use. In this there is an inherentunderstanding that the individual has some control over drug risk.

Drugs are “designed” in an R&D process undertaken by the drug producers themselves, or by smallercompanies with intellectual property or who specialize in drug discovery. Innovation in the form of drugdiscovery is an important driver of competitive advantage. Designs (i.e., drug compounds) are patented,and patent protection plays a very prominent role in shaping the competitive dynamics of the industry.

The risk and impact of deploying a new medicine to the public must be evaluated at each newdrug application. Benefit versus adverse reaction, ability to diagnose, target, and exclude will beweighed against benefit. The impact to the GMP process, changes to standards, or in some cases, newproduction methods will be introduced. The impact of these new processes to the overall risk is a newconsideration.

Drugs pose inherent risk to the people who take them, and some level of adverse effects on the publicis expected for every drug. Drug approval depends on whether these risks are outweighed by thepotential benefit of the drug. Much of the legal exposure of the industry arises from individuals orgroups of individuals who have been harmed by a drug. Both anticipated and unanticipated adversehealth effects on the public reflect drug design risks. That is, these risks arise from the inherent natureof the drug, even if they are produced exactly as intended.

Drugs are produced in regulated production (GMP) facilities. GMP regulations are very generalcompared to the regulations that govern nuclear power plant operation. These high-level regulationsare used by the manufacturers to develop plant SOPs and quality programs in a manner similar to thenuclear industry, but drug producers have significantly more latitude in applying the regulations to theirown production processes. Today, they use historical best practices and subject matter experts in adeterministic approach to production. Risk-based regulatory approaches to drug production are in theirinfancy, and current risk assessment techniques are strictly qualitative.

The public safety risk of production basically comes down to not shipping defective product from thesite. This is managed in two ways. First, stringent measures are taken to ensure that drugs areproduced in accordance with the appropriate regulatory and process standards. Second, if somedefective product is produced, it must be detected before being shipped off the site for sale. Theindustry currently relies heavily on post-production testing to ensure safe products are distributed.Risk-based approaches to managing production risk are in their infancy.

Regulation aims to ensure that a drug is reasonably safe before it can be sold to the public and that itis produced consistently. Some harm to the public is expected from every drug; the FDA must weighthose costs against the benefits of the drug. The FDA does not have the resources to oversee allproduction and R&D facilities. The FDA must rely on data provided by the drug companies for drugapprovals. It performs spot inspections of production facilities. A company’s history will reflect onFDA’s inspection regime and frequency.

NUCLEAR

Risk ProfileA nuclear reactor accident is a low probability, high consequence event. Risk, defined as the product ofprobability and consequences, may be similar to that of pharmaceutical industry. The NRC, through thesafety goal policy, has established Quantitative Health Objectives (QHOs) to try to address thequestion: how safe is safe enough? The QHOs state, in part, that the risk of death from nuclear powerwill not exceed 0.1% of chance of dying from other risks to which the public is exposed. In practice,the NRC uses risk thresholds for risk-informed applications based on surrogate risk metrics (such ascore damage frequency) at which they are confident the QHOs are met.

The public views nuclear power risk in a general sense rather than an individual sense. Theobservable risk from nuclear power is low. Like commercial aviation, there is risk to individuals in astatistical sense, but everyone thinks that they will never be involved in a crash. The public has nopersonal experience base of nuclear power risk.

Design ProcessNuclear power is produced in the U.S. in only two ways; the processes and science behind them arewell understood. Nuclear power plants are not designed by the power producers, but rather by a smallgroup of large, global engineering companies. Patent protection does not play a prominent role; adesign type (e.g., pressurized water reactor) cannot be patented in its entirety like a new drugcompound, although the design companies possess numerous patents for various design aspects of theplants themselves.

Design RiskNuclear power plants are designed to withstand specific design basis accidents. Design risk exists tothe extent that the design basis accidents do not adequately envelope the entire range of risks actuallyexperienced during operations, and by uncertainties in the analysis of the effects of these events. Asthis paper discusses, events at nuclear plants have revealed gaps in the design bases, and plantmodifications and other corrective measures have been taken. By quantifying actual operational risk,PRAs have identified and mitigated much of the nuclear design risk.

Design risk has historically been addressed through deterministic General Design Criteria. Existingplants were designed using only deterministic risk considerations, though extensive modificationsdriven by risk insights have subsequently been made. New reactor designs incorporate risk-informedinsights.

Production ProcessSafe production of nuclear energy is the responsibility of an operating company (utility) that selects adesign, applies for and receives a license, builds a power plant, and then operates the plant. Nuclearpower production is heavily regulated through a number of mechanisms; day-to-day operation isgoverned by a set of Technical Specifications that define acceptable operating parameters andconditions for the plant. Detailed Operating Procedures are developed for all aspects of operation,driven by the Technical Specifications and plant design. Nuclear plants have pervasive qualityprograms with extremely high quality standards. A “risk-informed” regulatory approach is used fornuclear plant operation in which risk assessment techniques (PRA) can be used extensively to improvethe safety impact of operations, as well as to provide economic benefits (e.g., shorter outages, on-linemaintenance).

Production RiskThe public safety risk of nuclear power production consists of radiation exposure due to an accident.As this article discusses, the industry and the regulator use very sophisticated risk analysis techniquesto scrutinize how the plant is operated. The results of risk assessments find their way into plantTechnical Specifications, SOPs and maintenance programs such that production risk is mitigated asmuch as possible.

Regulatory ReviewRegulation is focused on mitigating and/or preventing nuclear accidents. Each nuclear power plant hasa license that can be revoked. Regulation pervades all aspects of operations, and inspections areconstant. The NRC maintains Resident Inspectors at all nuclear power plants. Additional inspectionintensity is focused on plants or their parent companies with a sub-par history of safe operation.

Table A. Broader industry risk comparison.



• What will the role of industry groups be?• Should risk-assessment standards be developed? If so,

what should their scope be and who should develop them?

To deal with issues of consistency and technical adequacy,the regulator suggests risk assessment requirements. Prac-titioners who perform risk analyses must be adequatelytrained. Risk assessment techniques must accurately con-sider all relevant risk variables for the regulatory applica-tions for which they are used. Risk models must be consis-tently applied. Professional and industry organizations canhelp by developing standards and conducting independentaudits and peer reviews.

After a long (25-year) evolutionary process, the regulatorand industry settle into a consistent and predictable risk-informed regulatory process. The regulator relies on both

deterministic and risk considerations to make changes to theexisting regulatory structure. The regulator approves the useof risk assessments for situations where it deems themsufficiently robust, and has confidence when industry usesthese techniques because the models adhere to certain stan-dards.

Not Just a StoryOf course, this is not just a story. It is a very broad descriptionof what transpired when the Nuclear Regulatory Commis-sion (NRC) and the commercial nuclear power industrytransitioned from a deterministic to a “risk-informed” regu-latory approach. Figure 1 presents a nuclear industry timelinethat includes major industry events (in blue) and majorevents related to the use of risk analysis in the industry (inred).

What is a Probabilistic Risk Assessment in the Nuclear Power Industry?A Probabilistic Risk Assessment (PRA) is a structuredanalysis using a combination of probabilistic and determin-istic techniques to estimate the risk associated with nuclearpower plant design, operation, and maintenance. The threerisk metrics generally used are core damage frequency(Level 1 PRA), radioactive material release (Level 2 PRA),and health effects on the public (Level 3 PRA).9

A PRA systematically analyzes potential sequences ofevents (called accident sequences) to determine theircontribution to one or more of the risk metrics. Eachsequence starts with an initiating event (e.g., a reactortrip). Specific systems are designed to actuate to mitigatethe consequences of an initiating event so that safety isensured; relays trip, pumps start, valves open, etc. Each ofthese events in a sequence has a probability of success orfailure that can be quantified from historical experienceand failure data sources.

In a Level 1 PRA, system models are used to depict thecombination of system successes and failures that consti-tute accident sequences. Although there are many differ-ent modeling techniques, event tree analysis and fault treeanalysis are used almost exclusively in PRAs.12

An event tree model traces accident sequences in termsof the functional or system successes or failures that makeup those sequences. An event tree model provides end-to-end traceability from the accident initiating event to theaccident sequence outcome.

Figure 3 is an example of an event tree with fourmitigative systems labeled A, B, C, and D. The event treestarts at the left with an initiating event and ends at theright with an outcome related to a risk metric. After theinitiating event, each mitigative system either succeeds(i.e., operates as designed) or fails. (throughout the eventtree, system successes take the upper branch and systemfailures take the lower branch). The logic of the event tree,the various permutations of system success and failure,depend on how the systems are designed and configured

to work together. One permutation shown in the event treereveals that if both systems A and B fail, systems C andD cannot prevent core damage. Other permutations showthat successful response of systems C and D will preventcore damage if either system A or system B fails, but theoutcome varies if one of systems A and B fails and one ofsystems C and D also fails.

Fault tree analysis is performed to identify the potentialevents or combinations of events that can make the plantsafety system unavailable to respond to initiating events.The system fault trees can be quantified to obtain esti-mates of the probability of system unavailability. Usingthese system unavailability probabilities as inputs to theevent tree models, the probabilities of the various accidentsequences can be estimated. These probabilities can becombined to estimate the core damage frequency of theplant, which will depend on the design of the plant, howit is operated, and how it is maintained.

A simple system of two pumps and two valves is shownin Figure 4. Assume that for a given accident sequence,success for this system is defined as providing flow fromboth of the pumps. Therefore, success requires that bothpumps start and their associated valves open. A corre-sponding fault tree is shown in Figure 5.

Developing and running PRA models can provide riskinsights that might otherwise be missed. For example,system redundancy can reduce risk, but also it addscomplexity so that more failures must be considered. Whatis the tradeoff? Using event tree and fault tree models, aPRA can assess how various systems and componentswork together to mitigate risk across system/humanperformance boundaries. PRA also facilitates risk rankingof systems, components, and human actions. The impor-tance indicates which items, if they fail, would have themost significant impact on plant risk. This allows plantoperators and regulators to focus on the equipment andsystems with the most risk impact.



This article contends that any industry that attempts toincorporate risk into an existing deterministic regulatorystructure will face fundamental questions similar to thosepresented above for the NRC and the nuclear power industry.All regulators are conservative and will find it hard to shifttheir regulatory paradigm. All regulators must use an ap-proach in which they have confidence and that is defendableto the public. All regulators want to apply regulations consis-tently. Because the regulatory dynamics are similar, itwould be wise to consider what lessons the FDA and the lifesciences industry can take from the nuclear industry experi-ence.

This article examines the nuclear experience and identi-fies seven themes that are likely to be of importance for a risk-based/risk-informed regulatory structure for production inthe life sciences industry. A full comparison of all the riskissues faced by these industries is beyond the scope of thisarticle, though some ideas are presented in the sidebar (“ABroader Industry Risk Comparison”) for interest. It is not thepurpose of this article to predict exactly how the transitionwill occur, nor is it to make a case that this transition willexactly mirror the nuclear industry experience. Rather, thisarticle attempts to help structure the risk-based cGMP de-bate in terms of fundamental regulatory issues. Hopefullythis article will help the life sciences industry more rapidlydevelop common expectations for the transition to risk-basedcGMPs.

A Brief Note aboutRisk Analysis Terminology

Before proceeding with the seven themes, a brief note isneeded regarding the terminology of risk assessment in thenuclear industry. The primary risk assessment techniqueused in the nuclear industry is a Probabilistic Risk Assess-ment (PRA). Please refer to the sidebar for a brief overview ofPRA. This article quotes from several NRC publications thatinclude the term “PRA,” both in generic and specific riskassessment contexts. Substituting “risk assessment” for “PRA”provides the proper context for this discussion. This articleargues that quantitative risk assessment techniques willbecome necessary within any risk-based/risk-informed regu-latory framework, but it does not argue for wholesale adop-tion of PRA techniques per se.

Seven Themes of Risk-Based RegulationMaking Risk ExplicitConsidering risk in a regulatory structure is not new. Thepurpose of regulatory agencies like the NRC and the FDA isto protect the health and safety of the public. In other words,they exist to manage risk. They protect the public from thepotentially severe adverse consequences from beneficial, butrisky, societal activities such as making drugs and operatingnuclear power plants.

If regulatory agencies exist to manage risk, why isn’t “risk-based” regulation already the preferred approach for regulat-ing? While both the FDA and NRC have traditionally reliedupon “deterministic” regulations to ensure safety, determin-

istic regulations have always been based on implicit consid-erations of risk. For example, cGMP regulations exist formanufacturing space because that is where the risk is; theFDA does not regulate how a pharmaceutical company oper-ates its corporate headquarters facility. Similarly, the activi-ties governed within the cGMP regulations are those thatcould affect product quality, i.e., that are risky. By extension,one can easily conclude that all public safety regulationimplicitly considers risk.

If it is accepted that there is an implicit risk-based founda-tion underlying the existing deterministic regulatory frame-work, then taking a “risk-based approach to regulating phar-maceutical manufacturing” does not appear to be a funda-mental philosophical change. Rather, it becomes a matter ofmaking the existing implicit risk assumptions explicit, exam-ining them in the light of new processes, analytical tools, andtechnology, and using the insights gained to improve safety.If there is a concurrent economic benefit, so much the better.

But making risk explicit can give the impression that newrisk elements are being injected into the regulatory frame-work that were previously overlooked. Many risk elementsare not new and have already been considered implicitly, butsome new risk insights are likely to be found by using riskassessment techniques (this is discussed below under “Un-foreseen Risk Drivers.”) This is a credibility trap for theregulator. How can these new risk insights be consideredwithout casting doubt on the adequacy of the entire existingregulatory structure? To minimize the effects of the credibil-ity trap, the FDA will need a well-considered, defendableapproach to risk-based regulation and a corresponding strat-egy to communicate it consistently and effectively.

Risk-Informed, Not Risk-BasedImagine a motorist driving down a street late at night.Authorities have placed a red light at an intersection in themotorist’s path, and deterministic traffic regulations say thedriver must stop regardless of the presence of other traffic orthe actual risk that running the light might pose to the driveror others. In a purely risk-based world, there would be notraffic light. Each driver would evaluate the probability andconsequences of operating a vehicle in various situations, andwould continue at speed through the light if the situation is“safe enough” or stop if the risk is too high. The driver’sactions are based on the perceived risk compared to anacceptable threshold, whether set by the authorities or deter-mined individually by each driver. In a “risk-informed” regu-latory approach, the authorities recognize the intersection asa risky situation that calls for additional caution so they haveinstalled a flashing yellow light. A driver approaching theintersection must slow down and check for traffic in alldirections. The driver stops if necessary, or continues throughthe light at a safe speed if the “coast is clear.” The driver’sactions are a consequence of considering actual real-time risk(is traffic coming?) and following deterministic regulations,i.e., the traffic laws governing driver behavior at a flashingyellow light.

In a risk-informed approach, risk considerations comple-



ment a deterministic regulatory framework. Deciding whethera risk-informed regulatory framework will replace or comple-ment an existing deterministic framework is a fundamentalquestion of policy. The NRC adopted the risk-informed ap-proach in the 1995 NRC Policy Statement6 regarding the useof PRA.

...the Commission believes that the use of PRA technol-ogy in NRC regulatory activities should be increased tothe extent supported by the state-of-the-art in PRAmethods and data and in a manner that complementsthe NRC’s deterministic approach.

What path will the FDA take? What the FDA means by theterm “risk-based” is open to some interpretation. What is thedifference between a new “risk-based” approach and anapproach that makes informed decisions about risk andimparts them in a deterministic regulatory structure? It canbe argued that the FDA’s approach1 is risk-informed, and itis hard to argue that any regulatory agency would opt foranything other than the risk-informed approach. To chooseotherwise would go against both precedent and the conserva-tive nature of regulation itself, and would put the agency’spolitical capital at risk.

Evolution, Not RevolutionFigure 1 shows that the first use of PRA in the nuclear

industry was WASH-14002 in 1975. The NRC presentedgeneral principles for using risk assessments in a regulatorycontext when it published Regulatory Guide 1.1747 in 2000.This is a span of 25 years. Truly, the nuclear industry saw anevolutionary, not a revolutionary, transition to a risk-in-formed regulatory structure. Evolutionary change is in keep-ing with the risk-informed approach, which takes as a premisethe retention of some form of deterministic structure and acontrolled transition.

The lag in computer technology was certainly a factorcontributing to the length of the evolutionary transition inthe nuclear industry. Even though it was not until the mid-1990s that computer capabilities advanced sufficiently toallow use of risk assessment techniques on a wide scale basis,much of the delay should be attributed to the nature ofregulation itself. The evolutionary pace of change allowed theindustry and the NRC to identify, investigate, and confi-dently address standards, validation, training, education,public perception, and the many other issues associated withthe change in regulatory approach. The need to treat theseissues right the first time in the life sciences industry arguesstrongly for the FDA to adopt an evolutionary approach.

There should be no such technological barrier to wide-spread application of quantitative risk assessment tech-niques in the life sciences industry. However, the FDA faceschallenges of both scale and scope as compared to their NRCcounterparts. There are 103 commercial nuclear power plants

Figure 3. Example of an event tree.



Figure 4. Example system.

operating in the United States; the number of cGMP facilitiesis many times greater. There are only two basic commercialnuclear power plant designs in the U.S. (pressurized waterreactors and boiling water reactors); the regulator under-stands them very well. In the life sciences industry, there aremany processes for which there are varying degrees of under-standing. How to frame a workable risk-informed structureto govern all this activity without a larger use of resourceswill be a substantial challenge for the FDA.

Unforeseen Risk DriversThe accident at Three Mile Island (TMI) in 1979 overturnedsome of the basic assumptions about risks associated withnuclear power plant operation. The most severe accidentpostulated for nuclear power plants prior to the accident, andthe one on which much of the safety design of the plants wasoriginally based, was the large break loss of coolant accident(LOCA). This was thought to be the limiting accident becauseit results in the most rapid conceivable loss of cooling waterfrom the reactor cooling system, and thus, the greatestpotential for melting the reactor core. But extensive riskanalyses after TMI showed that a small break LOCA waspotentially more severe than a large break LOCA. The smallbreak LOCA depressurizes the core less rapidly; the higherpressure limits the ability of the plant safety systems to pumpemergency cooling water into the reactor core. The lesson tobe learned is that formal risk analysis can providecounterintuitive insights that disprove conventional wisdomabout risk embedded in deterministic regulatory structures.

It is also interesting to note that risk analysis foundsignificant risk associated with seemingly benign engineer-ing support systems. For example, an instrument air systemmight actuate a valve that, because of unintended systemdependencies due to common cause failure, may have animportant effect on safety.

It is quite likely that taking a risk-informed approach inthe life sciences industry will reveal risk components that arecurrently underemphasized or simply not present in tradi-tional cGMP space. Bringing science into the cGMPs throughrobust risk analysis will require reconsideration of the con-ventional wisdom about where risk exists.

Quantification, Consistency, and AdequacyRisk assessment technologies used in the nuclear powerindustry use both quantitative and qualitative methods.Because qualitative methods tend to involve potentiallygreater uncertainties and inconsistencies, quantitative meth-ods are heavily relied on in risk-informed nuclear regulation.Most risk-informed applications in the nuclear power indus-try rely on the ability to estimate (at least) core damagefrequency quantitatively. In contrast, risk assessment tech-niques applied to GMP space, such as the GAMP approach(outlined below) have so far been entirely qualitative. Ifregulatory (i.e., safety) decisions are going to be made anddefended to the public based on risk assessments, there willbe irresistible pressure to develop and use consistent quanti-tative risk assessment methods.

Making risk explicit has broken the safe/unsafe paradigm.How does a regulatory agency provide “adequate” protectionof public health and safety within the new paradigm? Theanswer, of course, depends on what “adequate” means. Theuse of qualitative, risk-informed approaches helps to deter-mine adequacy. For example, consider the general GAMPrisk assessment approach in which potential hazards and/orrisk events are postulated and then assessed using a two-stepqualitative approach as shown in Figure 2.

1. Plot Severity (low-moderate-high) against Probability (low-moderate-high) to obtain a Risk Class (1-2-3) for eachevent

2. Plot Risk Class (1-2-3) vs. Detectability (low-moderate-high) to obtain a Risk Priority (high-medium-low) for eachevent.

The outcome of this process is a risk priority of low, medium,or high assigned to the hazard/event. Different technical oradministrative controls are put in place to mitigate the riskbased on the assigned risk priority.

There is considerable fuzziness associated with this ap-proach simply because it is qualitative and reasonably in-volved (it has 27 combinations and two sorting/binning pro-cesses to achieve three possible outcomes). Despite the fuzziness,qualitative risk assessments such as the GAMP approach canbe very useful in the context of guides or best practices. Theyare a simple and logical first step to making risk explicit.

But fuzziness can be harder to defend when it is used, notmerely as a guide or best practice, but as a basis for a newregulatory framework that will continue to “adequately en-sure” public health and safety. Then the FDA might beconcerned with questions such as the following:

• What exactly determines the boundaries between low,medium, and high business impact/risk likelihood/prob-ability of detection?



Figure 5. Example system fault tree.

• What if Pharmaceutical Company A deems the risk like-lihood of one event to be low and Pharmaceutical CompanyB deems the risk likelihood of the same event to be high?In such cases, does the FDA have a duty to investigate thediscrepancy? What if two different sites operated by thesame company show similar discrepancies?

• Do the outcomes of these assessments increase or decreaserisk compared to the existing deterministic regulatorystructure? Are these changes small enough that the regu-lator can continue to “adequately ensure” safety? Is therea risk threshold below which the risk level is alwaysacceptable? If so, how does one know when this thresholdis crossed?

• What is the cumulative effect of conducting multiple,independent, qualitative risk assessments across a broadspectrum of applications within a pharmaceutical manu-facturing facility? Would there be a danger of missingsystem and process interdependencies that would bringtotal risk to an unacceptably high level?

These questions can be answered in terms of quantifica-tion, consistency, and adequacy.

Quantification can be desirable in so far as it makes riskpriorities less fuzzy, and therefore more defensible, both intechnical and regulatory space. It may be in the best interest

of both the FDA and the industry to establish quantitativetechniques to avoid the resource drain and scrutiny thatcould be involved with answering questions like those pre-sented above for every qualitative analysis.

Although risk-informed regulation in the pharmaceuticalindustry is relatively new, the pressure for greater riskquantification is already evident. It is interesting that theentirely qualitative GAMP framework was one of the firstapproaches to explicitly consider risk in the industry. Thefirst stage of quantification might be a risk-weighting ap-proach in which numerical values are assigned to the differ-ent outcomes in a qualitative risk assessment process to yieldnumerical risk results. Risk thresholds can then be definedfor different risk rankings. The FDA has published a risk-weighting approach similar to this to prioritize inspections.10

The next stage of quantification might involve assessing therisk associated with production processes, such as risk-basedapplications of Process Analytical Technology (PAT). It willbe hard to keep these applications qualitative because theyare inherently quantitative; they are measured and con-trolled using numerical (quantitative) information that lendsitself well to statistical analysis of risk. So, as the FDA triesto bring the science into the regulatory structure, risk quan-tification is likely to expand.

There also are issues of consistency, scope, and technicaladequacy. The NRC looks at the issues this way.



In any regulatory decision, the goal is to make a soundsafety decision based on technically defensible informa-tion. Therefore, for a regulatory decision relying uponrisk insights as one source of information, there needs tobe confidence in the PRA results from which the insightsare derived. Consequently, the PRA needs to have theproper scope and technical attributes to give an appro-priate level of confidence in the results used in theregulatory decision-making.11

If risk assessments are to be included as part of a formal,stable regulatory structure, the regulator must have confi-dence in the results of those assessments. It would be be-trayal of its public trust for the FDA to simply rubberstampevery attempt at regulatory change that calls itself risk-based or risk-informed. So the FDA will eventually have toset some standards for risk assessments. The FDA is likely tobecome involved in certifying the methodologies, insisting onthe use of consistent data sets, setting training and profi-ciency requirements for those who conduct the analyses, etc.Following standards will go a long way in establishing consis-tency among the many risk assessment practitioners withinthe industry and the FDA. Professional groups could play amajor role by helping to establish consistency within theindustry.

Risk Thresholds vs. Incremental ChangesIf risk is quantified through consistent, robust risk-analysismethodologies, the next logical step might be to set a quanti-tative threshold that represents an acceptable level of risk.Conceptually, a quantitative risk model for a pharmaceuticalmanufacturing site could drive a “risk meter” that indicatesthe current risk state of the operation. If the overall risk is atan acceptable level, is it necessary to individually regulateeach activity that supports the process? Should cumulativerisk be tracked? If so, is there a limit to the maximumaccumulated risk? Or, if a process change can be shown toimpose no additional risk, should the change be subject toregulatory scrutiny?

The nuclear experience is described above. The NRC refusedto abandon its traditional approach for an exclusively quantita-tive, risk-based approach. Accordingly, the NRC has not set anabsolute risk threshold. However, relative risk thresholds andincremental risk thresholds are commonly used in the nuclearindustry’s risk-informed regulatory approach.

But implementation difficulties are lessened when quan-titative risk techniques are used to improve the regulatorystructure incrementally, on the margin. This minimizes theconcern over bias in the risk models by relating changes inrisk to an initial state that everyone agrees is “safe.” One neednot know an absolutely accurate absolute risk value to assesschanges in risk from a given initial state. Quantitative riskanalyses can be used to assess the incremental benefits ofnew and existing regulations (the amount of risk the regula-tions reduce) against the associated compliance costs. With-out robust, quantitative risk analysis techniques, it is diffi-cult to know how much the industry gains in terms of risk

reduction for each dollar spent on compliance. The FDA doesnot want to expend resources to enforce regulations that havelittle marginal risk impact, and the industry does not want tospend money on compliance that does not result in apprecia-bly lower risk. Quantitative analysis provides a frameworkfor discussion between the industry and the FDA to strike theright balance between cost and risk.

Establish General PrincipleTwenty-five years of experience with risk assessments in thenuclear industry culminated with the publication of Regula-tory Guide 1.1747 containing five general principles to governapplications of risk assessments within the nuclear powerindustry regulatory framework. These five principles areparaphrased below.

A risk-informed application:

1. Cannot be used as an avenue to violate existingregulations. That is not to say that regulations can not bechanged through a formal rulemaking process that takesrisk principles into account, but industry must complywith all regulations currently in force, seek a specificexemption, or change the regulation.

2. Must consider defense-in-depth. Defense-in-depth is aregulatory philosophy the NRC has used since the begin-ning of nuclear power regulation. It holds that multiplemeans to accomplish safety functions must be provided,i.e., no one measure should be relied upon completely toensure safety.

3. Must maintain sufficient safety margin.

4. Can be implemented only if the change in risk issmall. The regulatory guideline contains guidance onwhat constitutes a small change in risk.

5. Must be monitored as part of an overall program toassess the aggregate risk impact of (potentially)many minor risk changes. The concern is that manysmall changes in risk, each individually assessed, mightlead to an unacceptable total risk increase, especiallywhen common cause failure mechanisms are considered.

Of these specific principles, one could quite easily see some-thing similar to numbers 1, 4, and 5 adopted in the lifesciences industry. This points to common regulatory dynam-ics inherent to public safety regulation that this article hastaken pains to illustrate. But the larger point is to recognizethat general principles will evolve, which fit with both gen-eral regulatory dynamics and the specific challenges of lifesciences regulation. The ultimate intent of the discussion inthe life sciences industry should be to establish these basicprinciples. The sooner this is done, the sooner a stable andpredictable risk-informed regulatory framework can be es-tablished.



ConclusionsThis article presents a very high-level overview of the expe-rience of the nuclear power industry in moving away from astrictly deterministic regulatory approach to production toone that involves explicit considerations of risk. The regula-tory dynamics inherent to public safety regulation argue thatthe nuclear power industry experience should be consideredclosely when considering a similar transition for the lifesciences industry and the FDA.

The nuclear power industry experience suggests thatexplicit risk-informed regulation will complement, ratherthan replace, existing (risk-implicit) deterministic regula-tions. There will be pressure to use quantitative risk assess-ment techniques that are applied consistently throughoutthe industry and adhere to common standards. Regulatorychange will accelerate as general principles of risk-informedregulation are established, as risk analysis techniques usedin the industry become more robust, and as the FDA becomesmore confident in the accuracy of their results. These changeswill be evolutionary, not revolutionary, and will include bothchanges to existing regulations and implementation of newregulations.

Perhaps the evolution of risk-informed regulation of phar-maceutical production can be streamlined by consideringexperience from the nuclear power industry. This experienceshows that improved safety, lower costs, and better qualityare not mutually exclusive in a risk-informed world. Theseoutcomes can be achieved most rapidly by recognizing thepressures exerted by regulatory dynamics on both the regu-lator and industry, and how these are likely to play out.Working cooperatively within these constraints will reducethe uncertainty over what “risk-based regulation” will bring.If nothing else, answering the questions that confronted thenuclear power industry can provide a strategic framework forregulatory changes and help identify the logical next steps inthe regulatory evolution.

References1. “Pharmaceutical cGMPs for the 21st Century – A Risk-

Based Approach: Final Report,” U.S. Food and DrugAdministration, September 2004.

2. “Reactor Safety Study – An Assessment of Accident Risksin U.S. Commercial Nuclear Power Plants,” WASH-1400(NUREG-75/014), 1975.

3. “Clarification of TMI Action Plan Requirements,” NUREG-0737, U.S. Nuclear Regulatory Commission, 1980.

4. “Individual Plant Examination for Severe Accident Vul-nerabilities – 10 CFR 50.54(f),” Generic Letter 88-20, U.S.Nuclear Regulatory Commission, November 23, 1988.

5. “Severe Accident Risks: An Assessment for 5 U.S. NuclearPower Plants,” NUREG-1150, U.S. Nuclear RegulatoryCommission, 1990.

6. “Use of Probabilistic Risk Assessment Methods in NuclearRegulatory Activities; Final Policy Statement,” 60 FR42622, U.S. Nuclear Regulatory Commission, August 16,1995.

7. “An Approach for Using Probabilistic Risk Assessment in

Risk-Informed Decisions on Plant-Specific Changes tothe Licensing Basis,” Regulatory Guide 1.174, Revision 1,U.S. Nuclear Regulatory Commission, November 2002.

8. “An Approach for Determining the Technical Adequacy ofProbabilistic Risk Assessment Results for Risk-InformedActivities,” Regulatory Guide 1.200 for Trial Use, U.S.Nuclear Regulatory Commission, February 2004.

9. “Standard for Probabilistic Risk Assessment for NuclearPower Plant Applications,” American Society of Mechani-cal Engineers (ASME), Standard RA-S-2002, April 5,2002. Also Addenda to RA-S-2002, ASME RA-Sa-2003,December 5, 2003.

10. “Risk-Based Method for Prioritizing cGMP Inspections ofPharmaceutical Manufacturing Sites – A Pilot Risk Rank-ing Model,” U.S. Food and Drug Administration, Septem-ber 2004.

11. “Addressing PRA Quality in Risk-Informed Activities,”SECY-00-0162, U.S. Nuclear Regulatory Commission,July 28, 2000.

12. Breeding, Roger J., et. al., “Probabilistic Risk AssessmentCourse Documentation, Volume 1: PRA Fundamentals,”U.S. Nuclear Regulatory Commission, NUREG/CR-4350/1 of 7, August 1985.

13. Good Automated Manufacturing Practice Guide for Vali-dation of Automated Systems, GAMP4, Appendix M3,ISPE, 2001.

About the AuthorsJonathan Coburn is Business DevelopmentDirector - Life Sciences for Johnson Controls,and is responsible for managing businessrelationships with several large pharmaceu-tical clients in North America. Coburn servedas an officer in the US Navy submarine forcefrom 1983 to 1990 and in the US NavalReserve until his retirement in 2003. From

1990 to 1997, Coburn was employed by Framatome Technolo-gies, a supplier of field services, engineering, and technologyto the commercial nuclear power industry. While atFramatome, Coburn was involved in a number of companyand industry activities associated with risk-informed regula-tion. An ISPE member, Coburn holds a BS in mechanicalengineering from Rensselaer Polytechnic Institute, an MS inmechanical engineering from the University of Connecticut,and an MBA from Auburn University. He can be contacted bye-mail at [email protected].

Johnson Controls, 633-104 Hutton St., Raleigh, NorthCarolina 27606.

Stanley H. Levinson, PhD, PE is an Advi-sory Engineer with Framatome ANP inLynchburg, VA. Dr. Levinson received hisBS, ME, and PhD in nuclear engineeringfrom Rensselaer Polytechnic Institute in Troy,NY. He has more than 22 years of experiencein Probabilistic Risk Assessment (PRA), reli-ability/availability analysis, risk-informed



regulation, risk-informed applications, and Failure Modesand Effects Analysis (FMEA). Dr. Levinson’s primary areasof technical expertise are applying risk assessment tools andtechniques on a variety of mechanical, fluid, and electricalsystems with varying objectives related to safety, perfor-mance improvement, and licensing issues. He also has per-formed risk analyses in applications as diverse as shippingcasks for spent nuclear fuel, communication systems, super-conducting quadruple magnets, and the Yucca Mountainnuclear waste repository. He is a member of the NuclearEnergy Institute’s (NEI’s) Risk Applications Task Force(RATF), Option 2 Task Force, and Risk-Informed TechnicalSpecification Task Force (RITSTF). He is also a member ofthe American Nuclear Society’s (ANS’s) Risk-Informed Con-sensus Standards (RISC) Committee, the American Societyof Mechanical Engineer’s (ASME’s) Committee on NuclearRisk Management (CNRM), and was a member of the writingteam that developed ASME’s PRA Standard (Internal Events).Dr. Levinson is a member of the ANS, the Society for RiskAnalysis (SRA), and the National Society of ProfessionalEngineers (NSPE). He is a registered Professional Engineerin the Commonwealth of Virginia. He can be contacted by e-mail at [email protected].

Framatome ANP, 3315 Old Forest Rd., Lynchburg, Vir-ginia 24506-0935.

Greg Weddle is Global Manager of CriticalEnvironments for Johnson Controls. In thatrole he is responsible for direction and lead-ership for the Global Validation ServicesBusiness Unit, providing validation programdevelopment, solutions development, salessupport, and technical support world wide.In his 21 years with Johnson Controls, Weddle

has worked as a systems engineer, project manager, qualityassurance manager, bio-pharm team leader, and instructor.He is a trained Rummler Brache Facilitator in ProcessChange and Implementation; he also holds a HPAC license.Weddle has served as guest instructor for the FDA Office ofRegulatory Affairs (ORAU) and at Purdue University in thearea of Critical Facility Design and Validation. He is amember of ISPE and the American Society of Heating, Refrig-erating and Air-Conditioning Engineers (ASHRAE). Weddleholds a BS in mechanical engineering from Purdue Univer-sity. He can be contacted by e-mail at [email protected].

Johnson Controls, 1255 N. Senate Ave., Indianapolis,Indiana 46202.

Design Qualification


Design Qualification in Practice –from Strategy to Reportby Yossi Chvaicer

This articlepresents apractical modelfor designqualificationunder a pre-establishedstrategyframework.

Introduction

Much work has been written enhanc-ing the advantages of good inter-face between mechanical comple-tion, automation implementation,

and validation activities. Project managers areencouraged to follow ingenious flowcharts, V-models,1 W-models,2 and project life-cycle strat-egies aiming to accomplish overall quality,cost-benefit achievements, and time-to-mar-ket reductions. This truth reflects two excep-tional facts in the pharmaceutical industry.The first is the vast amount of thought put intothe Engineering Commissioning efforts. Thesecond is that Installation, Operation, andPerformance Qualification (IQ/OQ/PQ) are notonly common practice, but also spread knowl-edge in this industry today.

In spite of the benefits introduced by these

modern approaches, they lack a clear and use-ful methodology for the preceding Design Quali-fication (DQ) work. For instance, what exactlymust a validation engineer write down in his/her DQ reports to provide successful evidenceof achievement for these advanced models?What should be verified, and to what extent inthe design, when a plant capacity and layoutupdate occur six months after the HVAC sys-tem design has been completed? How will thatbe incorporated under design control? Simi-larly, what happens when the contact surfacefinishing of a new vessel is downgraded byprocurement personnel due to sudden projectbudget cut-backs? What traceability level isdesired? How can all these changes be reason-ably supported during an inspection focused ondesign?

Eventually, qualifying the design of a new

Figure 1. The threesteps of designqualification.





Table A. Numerical values for the justification factors.

Justification Factor Sub-group Description Numerical ValueAssociated

SystemTechnological Highly Complex Any piece of machinery or system that is operated by one or more 3Complexity of the following sub-systems: SCADA, HMI, Controllers

Semi Complex The system or equipment has a controller but does not have a 2SCADA or an independent HMI.

Low Complex The system is manually controlled by mechanical of simple 1electrical panel buttons

Business Value Highly expensive projects Project budget over $100K. 3

Medium Sized projects Project budget between $10K and $100K. 2

Low expense sized project Project budget below $10K. 1

Product Quality Impact Direct impact Equipment or system that has a direct impact on the safety, 3efficacy or integrity of products.

Indirect impact Any system or equipment involved or supporting the production 2process, but does not determine its character.

No impact No effect is found from the system to the product quality. 1

facility, its systems, and equipment, will demand not onlyunderstanding the concept and goals of these theoreticalmodels at an early stage, but also creating the practicaldocumented evidence of the facts, adaptations, and reasonsduring design changes that will help to achieve full compli-ance.

In this way, this article proposes a practical three-layerplan with the framework methodology at its upper leveldescribing how to perform Design Qualification, step by step.

Background Surveyand Motivation

Design qualification has been addressed in four approaches:

1. Enhanced Design Review (ISPE)The ISPE Commissioning and Qualification Baseline® Guide3

describes in detail a review activity for design aiming toaccomplish conformity with operational requirements andregulatory expectations. Although not quoted as an FDArequirement, it brings to light the benefits of an enhancedstructure review for any project. The results will mainlyassure that the design is carried out in a controlled manner,facilitating, as per the text, a good start to any inspection byauthorities. However, it must be noted, that the list ofqualifications proposed in this publication remains at abusiness-driven level rather than a regulatory-based deci-sion. How to do such work, as well as what exactly is abusiness-related issue in the face of inspections remainsundefined.

2. Adequate Design (CFR)This approach, stated by the 21 CFR Part 211,4 refers todesign within a statement of appropriateness. Its generalityis so wide one can easily be confused by the extent of regula-tion requirements, and as a result, find difficulties in deter-mining the scope of a design qualification. The specificationsstand for requirements regarding construction, cleaning,maintenance, calibration, identification, procedures, and other

issues rather than for the design itself. Consequently, thoseresponsible for dealing with the qualifications are driven togather all statements and translate them into “appropriate”design requirements under their best available judgment. Inyet another guideline also published by the FDA on processvalidation,5 further confusion by suggesting an examinationof design at the Installation Qualification stage is revealed.This means reviewing the design when the system hasalready been installed – a meaningless exercise unmatchedin the project life cycle. After all, what can be done to thedesign activity, other than post-mortem site corrections, if atthe installation qualification phase the design requirementsdo not comply?

In short, in this approach the design qualification conceptis somewhat hidden, whereas the major hint in the text is theterm “intended use.” Such a quote is the core motive forqualifying the design.

3. GMP Compliance Demonstration (EU)This approach, adopted by the European Commission onQualification and Validation,6 clearly suggests a qualifica-tion exercise for the design phase, as the first validationactivity. It also aims to demonstrate documented compliancewith GMP.

In retrospect, this emphasizes the need to assess engineer-ing, quality and process issues during the initial design.However, in practical terms, how far should one go for asuccessful GMP review during design? For instance, howmuch is plant space utilization– an issue extensively stressedin design for its budget related matters – a GMP criterion atthe planning phase? What about production plant capacities,environmental safety, or automation levels? How to demon-strate their GMP compliance at the design phase? This is notclear from the text, which suggests that if authorities applyadditional insight to this Guide it could help the EuropeanPharmaceutical Industry achieve better qualification suc-cess at the design phase. Quite the opposite, the glossary ofthis document surprisingly reveals a more detailed definition



for design qualification, very much like the approach adoptedby the International Committee on Harmonization (ICH)which is described below.

4. Verification Suitability (ICH)This approach, stated in ICH Q7A,7 introduces the concept of“intended purpose suitability” for a facility, system, or equip-ment. It actually defines a status for the design itself ratherthan a list of enhancing reviews, clued requirements, orunclear GMP inferences, as assumed by the other approaches.The ICH approach undoubtedly provides for the validationprofessional the base to develop the logical sequence forqualification activities at the design phase. Such an ap-proach, although much more understandable, still leaves the“how-to-do-it” issue largely open.

From the above, all the approaches share a common situa-tion. They leave ample space for interpretations, which in acomplex project could induce diversity, more than standard-ization. For this reason, when starting design qualification ina new plant, it is recommended at the outset to establish asupportive DQ structured plan for this activity and only thenconcentrate on the actual detailed activity.

Developing a DQ Structured PlanEffective design qualification for a new facility is achieved bydeveloping in the organization a structure to deal with thefollowing basic questions early in the project.

To begin with, what systems or equipment are designedfor the envisioned production processes? What reasons jus-tify qualifying their design? This means not only deciding thetechnology, the material flow concept, associated controls orsupporting services, but also establishing clear-cut criteriathat sustain decisions of what will or will not undergo designqualification. Risk assessment of critical organization issuesused as a strategic tool will help selecting the technologiesthat will undergo DQ.

Next, at what time during project evolution should DesignQualification be activated for the selected technologies? Whichevents or phases can trigger this activity? This matter re-quires establishing when it is most appropriate to aligndesign deliverables with their predetermined specifications.Choosing design related milestones in the natural course ofthe project will lead to the correct timing for DQ.

Finally, after the “What, Why, and When” are defined, theDQ structured plan will create the “How” is the qualificationgoing to be performed? This issue relates to the work itself. Itis setting up the hands-on methodology, responsibilities,reports, documentation requirements, formats, and stan-dard operating procedures for all the do’s and don’ts.

A DQ structure consisting of three layers was developedbased on the concepts above to provide the answers for theexposed questions.

The Basic Layer – DQ JustificationInitially, suppose you are starting the design of a new facility.Should the following projects undergo design qualifications?

• a jacketed, double-cone processor with size reduction andvacuum drying capabilities, external HMI pre-set recipes,historical trends, security management, alarms, and up-stream CIP interfaces

• an HVAC system which is planned to serve a 12-roomgranulation department with a common corridor and twoair-lock sections in between

• an off-the-shelf, direct contact peristaltic transfer pump

Next, admitting that market shifting is a desktop variable,what about the following?

• For a budget approved 18 months ago, is the facility beingbuilt today the correct one for the business?

• Within the atmosphere of constant SUPAC and R&Dinnovation demands on the one hand and daily technologi-cal advances on the other, are you acquiring the rightequipment for your processes?

It is clear from the above that any “Yes or No” answer is notas simple as it seems. Additionally, the entire project inven-tory will need a supporting rationale for justifying which ofthe equipment or systems will or will not undergo DQ.

Consequently, the justification process starts by using arisk assessment technique. First, identify the three majoraspects in the company that will establish the decisionmaking approach to perform Design Qualifications. They canbe:

a. technological complexity: indicating the level of automa-tion complexity of a project

b. business value: nominal budgets allocated or equipmentcost, when known

c. quality impact: The degree of impact in the product qual-ity, as described in the ISPE Guide3

Then, subdivide each factor into category levels with numeri-cal values each, e.g., 1, 2, or 3 as shown in Table A. Afterexamination of which levels best fit for each selected technol-ogy, their multiplication becomes your justified decisionindex. Finally, establish a criterion. For example, a systemwill undergo Design Qualification if the Complexity BusinessImpact (CBI) index equals 27, as the justification part inTable B illustrates for different types of equipment andutilities. In time, this justification tool will become suffi-ciently flexible in order to allow frequent updates during thedesign phase. They emerge from different sectors, such asprocess or technological development, finance and capacitytrends, or different quality approaches altogether in a busi-ness. The result of such an assessment becomes the company’smethodically justifiable DQ program, which also should be anintegral part of the Validation Master Plan.

The Middle Layer – DQ MilestonesProjects following budget approval start by the design phase.For qualifying purposes, this phase is broken down intological milestones with potential workable deliverables. The



first one, from the validation standpoint, is the User Require-ment Specifications (URS). This is a document considered tobe the foundation of the whole Design Qualification process.It contains for each equipment or system the requisites of theproduction process and it expresses the fundamental needs ofall involved users in one text. A responsible committeeformally approves the contents by involving representativesof different areas such as Engineering, QA, Process, Opera-tions, and Validation.

For utilities such as HVAC or water systems, such adocument is sometimes called Base of Design (BOD), whichserves the same purpose. The URS contents are examined ina subsequent section of this article, as well as three otherdeliverables described hereafter.

Next, as the project progresses, procurement activitiescombine URS controlled copies with commercial require-ments. A formal requisition for an offer is generally sent topotential vendors which return quotations for bidding pur-poses of the project purchasing phase. These quotations arepresumably based on the original URS, and also have techni-cal details attached. If a supplier is selected, a formal finan-cially approved Purchase Order (PO) will be issued, contain-ing the agreed commercial terms accompanied by the sametechnical specifications in the attachments received from thesupplier. Such documentation represents a clear and tan-gible milestone for the project, and it is then called theRequest for Purchase (RFP). This is when the next designqualification step finds its biggest potential to take place,becoming as a result, the second natural milestone.

The third target for the qualification work is the detaileddesign. For the project – here is when the manufacturer orsupplier delivers detailed documentation that waits for theuser’s official approval so that the manufacturing of theequipment at the workshops can start running. These include

engineering blueprints, mechanical drawings, detailed lay-outs, instrumentation diagrams, and software developmentdocumentation having the Functional Requirement Specifi-cations (FRS) as the leading records. In general, these docu-ments consist of a design package called Detailed DesignSpecifications (DDS).

The last milestone on the project design phase is theFactory Acceptance Test (FAT). In FAT the user expects tosee (and test) the major items of the URS fulfilled before theactual system or equipment is shipped for installation. Inother words, at this point, nothing remains to be designed sothat the design phase of the project is considered to becomplete.

All these four consecutive events, defined as project mile-stones, are workable turning points that trigger the DesignQualification exercise.

The Upper Layer – DQ WorkThe difference between DQ and the classical IQ/OQ/PQ orProcess Validation is the fact that when qualifying a designthere are no clearly defined measurable acceptance criteria.There is a lack of pre-determined parameters to be comparedto simply because the facility, equipment, and its processesdo not yet exist. Therefore, qualification of the design phaseis performed in three critical steps by crosschecking thedelivered documentation along the project design advance-ment, as depicted in Figure 1, using the URS as the leadingreference. The qualification work actually links every twopre-established milestones.

URS x RFP – Crosscheck 1The first crosscheck qualifies the RFP by comparing it withthe URS. Chances are that quotations will arrive in theformat developed by the potential vendors. This means, the

Table B. Justification index and qualification results.

DESIGN QUALIFICATION

JUSTIFICATION Step 1 Step 2 Step 3URS x RFP RFP x DDS DDS x FAT EstimatedDeviations Deviations Deviations Added

No. Equipment, System Technological Business Quality CBI Technical Process Technical Process Technical Process Valueor Utility Complexity Value Impact Rate

1 Fluid bed Drier 1200L 3 3 3 27 10 6 6 3 12 5 8%2 Fluid bed Drier 800L (1) 3 3 3 27 16 12 4 5 10 4 9%3 Fluid bed Drier 800L (2) 3 3 3 27 16 12 4 5 10 4 9%4 High Shear Mixer 900L 3 3 3 27 10 6 1 2 5 4 12%5 Coater Machine 65" 3 3 3 27 14 8 1 4 8 2 5%6 Diffusion granulator 50 c.f. 3 3 3 27 13 1 4 1 3 2 5%7 Diffusion processor 20 c.f. 3 3 3 27 15 5 8 2 4 1 2%8 Diffusion granulator 75 c.f. 3 3 3 27 10 1 9 6 5 0 1%9 CIP Skids 500L 3 3 3 27 16 5 2 1 7 1 10%10 R&D Fluid bed granulator 120L 3 3 3 27 4 7 8 0 6 1 4%11 Compressing Machine 36 st 3 3 3 27 43 23 Not avail. Not avail. 7 5 Undetermined12 Purified Water System 3 3 3 27 20 10 7 2 3 1 5%13 Capsule polisher 3 2 3 1814 Conical Mill 2 2 3 12 Design Qualifications submitted to policy constraints15 Metal Detector 1 1 3 3



quotations or their technical annexes do not follow the URSsequence arrangement at the paragraph level, as approvedby the responsible committee. So the first item to qualify isthe traceability of the URS itself. Placing the RFP side by sidewith the URS, qualification starts by carefully checkingwhether all URS paragraphs appear in the vendor specifica-tions. At the same time, each item also is scrutinized for itscontents to match the URS characterizations. If the items arefound to be identical in both documents, the compliance isrecorded. Here, paragraph citations, as well as the contentsof the quote, are inspected with the same level of importancefor the qualification work. This procedure for the URS x RFPcrosscheck is illustrated in the flowchart of Figure 2.

Every difference revealed between the URS and the RFPgenerates a Design Qualification Deviation Report - Figure 3.This report indicates to which of the three categories thediscrepancy belongs, namely an omitted-by-the-supplier item,an existent-but-modified-by-the-supplier item, or a new addeditem. In the example shown, an RFP modification was iden-tified on the contact surface finish grade for the tank of thepurified water project, generating as a result the DQDR #4.

Traceability is then assured by a record in a VerificationTable, which follows all URS sections, as depicted in Figure4. This table also will track the Design Qualification processfrom the start to its end, throughout the remaining two crosschecking steps explained below. The table example is a real-life Design Qualification project for a 65" [165 cm] tablet filmcoater machine, showing 15 out of 96 traceable URS para-graphs developed for this system (characterizing the spraysystem, the inlet AHU, the CIP and control features, thesafety, the direct and indirect contact parts).

Each individual deviation report, issued by the validationengineer in charge of the project, includes a description of thefinding, which is presented in a DQ meeting of the respon-sible committee. This forum will decide whether or not theinconsistency revealed is acceptable for the design carry-over. Furthermore, the committee shall agree on two addi-tional issues: The rationale justifying the deviation cause,and the corrective action to be taken. Such a procedure isperformed for all deviation reports issued during the crosscheck qualifications. All deviations and associated requiredactions are recorded in a summary report for that stepincluding the findings, paragraph numbers, deviation cat-egory and action to be taken by both user and vendor.Subsequently, this summary is addressed with the vendor forconfirmation, necessary changes, and corrections at the de-sign and project follow-up.

The vendor responses are expected to include correctivedetails and a timeframe to be finally post approved by thecommittee. The most appropriate time to send the DQDRsummary report to the supplier is approximately a weekbefore the RFP is legally issued by the PO placement. Thereare two reasons for this: first, any corrective action made tothe design after PO placement is strongly unadvisable for allparties concerned. Second, the supplier is usually prone toagree to corrective changes without extra charge in the fewdays preceding receipt of the order from the customer. In

time, when a budget sum is associated to each deviationfound, it is optionally included in the DQDR becoming thedesign qualification added value of this item for the project.

The above methodology leads to a worth mentioning mid-way conclusion: there is no need for retroactive changes in theoriginal URS. The URS is a document that reflects the bestknowledge of all the issues agreed among the committeemembers involved by the time it is approved. As such, itshould not be changed back when a new issue is raised duringthe actual design process. Redefining the URS on a docu-mented controlled environment would lead to cumbersomechange control procedures, unnecessary at the early phase ofthe project. Nevertheless, all inevitable changes during de-sign, their justifications, and corrective actions required areregistered in the DQ deviation reports. The full traceabilityfor each case is ensured by recording it in the DQ VerificationTable (Figure 4), as part of the DQ protocol described below.

RFP x DDS – Crosscheck 2In a similar way to the first check, this step verifies if all theDDS contents can be traceable to the RFP paragraphs, whichin turn were tracked down to the prior URS. In this way, thisnew phase of design qualification focuses simultaneously ondrawings and on the FRS. Table C illustrates an omissiondetected on the FRS for the atomizing pressure control of the

Figure 2. Flow chart for the design qualification URS x RFPcrosscheck step.



spray system (See “complies” column for the FRS documenta-tion, at the URS paragraph line 3.3.4 - spray system - wherean “N” is circled for “not complies”). This finding has gener-ated a DQDR#23, where all three main issues are recorded,namely the description of the finding (missing atomizing airpressure attribute as an acquired feature in the RFP approveddocumentation), the justification for no design compliance (itis a process related variable, part of the production steps andrecords), and the corrective action necessary to be imple-mented at the design (to include the feature in the SCADA –Supervisory Control and Data Acquisition - system). Thisqualification procedure is similar to the DQDR #4 example atthe URS x RFP prior qualification stage for the purified watertank project as described in the previous section.

At the end of this second crosschecking step, a new set ofDQDRs are filled with rationales and action items. Thesummary of the new DQDRs is submitted to the vendor afterapproval of the reactivated committee so that the detaileddesign can be adjusted accordingly without necessarily chang-ing the URS retroactively. However, it is recommended toidentify the new set of deviation reports by a sequentialnumber which follows the first crosschecking step (URS xRFP). Where needed, notes are introduced to the verificationtable referencing further annexes of extra detailed workperformed.

DDS x FAT – Crosscheck 3The last stage of qualifications for the design deals with theFAT results. The same tracing principle is used. The FAT DQdeviation reports are summarized as well, but the correctiveactions agreed at this time will be decided to be implementedeither at the commissioning phase during the Site Accep-tance Test (SAT) or directly into the IQ/OQ/PQs.

DQ Protocol and ReportThe DQ report is a document which will provide evidence thatthe system or equipment is designed for its intended use. Itmust be accurate and have supportive documentation for allthe verification steps conducted during the qualificationeffort. It also should reflect the natural progress of design forthe project in question. Reports with the following sevensections summarize this concept and respond in a high degreeof compliance to any regulatory demand:

1. The approval section. It contains the names, functions andsignatures of the representatives for the project team,usually Engineering, Operations (the User), Process, QA,and Validation.

2. The description section. A brief explanation of the equip-ment or system, its intended purpose, planned site loca-tion, phase in the production chain, process capacities,basic functional and automatic capabilities, cleaning fea-tures, and peripheral items associated, where applicable.

3. The URS section. This section contains the approved UserRequirement Specifications, preferably the original setsubmitted to vendors for the initialization of the procure-ment phase. If electronic documents are the only meansused, a read-only copy should be kept in the report file withpertinent details such as dates, revisions, approvals, andreply confirmations.

4. The attachments section. This section includes the list ofall attached supporting documents, for example, engi-neering drawings, layouts, production process flowcharts,a summary of the bid analysis, design review meetingminutes, and FAT reports. It is a known fact, not allvendors do have the required records or certificates avail-able on time. In such cases, a notification for projectmanagement must be issued detailing the missing data.This note is part of the DQ protocol, and when received,qualification work will continue for that part.

5. The project follow-up section. This is a general synoptictable with major completion dates for the URS approval,PO placement, DDS approvals, FAT, and shipment.

6. The qualification documentation section. This includesthe Design Verification Table, all the associated DQDRs,and summary reports. This makes the design phase fol-low-up completely traceable for the project - Figure 4. It is

Figure 3. Design qualification deviation report.



actually the documented evidence that the design for theequipment or system was carried out to suit its intendeduse as defined at the URS.

7. The GMP assessment section. This is a review section thatincludes a questionnaire considering items such as iden-tifications of drawings and instrumentation, contact ma-terials suitability, calibrations, safety, and an adequacyconcluding statement.

Qualification ExpertiseURS ContentsThe approved URS is the basis for the entire design qualifi-cation process. When distributed among potential vendors, itbecomes the acceptance criteria used by the future owner notonly for qualifying the quotations received for that specificequipment, but also along the qualification steps in thedesign phase that will be continued for the entire validationlife-cycle ahead. Yet, at early stages before approving it, thevalidation representative role is to support the URS consoli-dation phase, keeping in mind the subsequent qualificationactivities, as depicted in Figure 1. These activities will beessentially performed on the process functionality, engineer-ing technical issues, and regulatory aspects characterized bythe URS approving committee.

Afterwards an effective design qualification is betterattained across the organization when the URS is formattedin a standard way for all different technologies or systems

comprising the project. An example of the URS format and itscontents is as follows:

• Cover Page: Company name and logo, project title, sys-tem identification by name and number, URS revisionnumber, issuing date, and signatures dated from all partsof the approving team.

• Revision History: A page with a place to insert revisionnumbers, details, revision responsible, and recording ofcontrolled distributed copies.

• Table of Contents: A page reference with all numeratedsections and appendices.

• General: Aspects of the project such as the use, locations,access, safety statements, and an introductory sectionbriefly describing the production process.

• Scope of Supply: This defines to the vendor demarcationplans, delivery, manufacturing and testing responsibili-ties, expected participations in training, qualifications,and start-up assistance.

• User Specifications: The user requirements encom-passing all possible technical and process related issuessuch as batch sizes, manufacturing rates, auxiliary equip-ment involved, material loading and unloading features,

Figure 4. The verification table.



process and final product quality parameters character-ization, a list of monitoring parameters required withtheir specific measuring units, as well as start and endpoint restrictions for the process. In addition, all sorts ofprocess related connections like vacuum, atomization, gasblanketing, treated air and steam, all with their associ-ated cleanliness classifications.

• Cleanability: Definitions of required cleaning methodsand options such as Cleaning in Place (CIP) or Washing inPlace (WIP). Separated or integrated skids, as well asdrainable features and additional side storage tanks arepart of the requirements.

• Automation: Internal company standards, instrumenta-tion required for the control parameters. Special featuresdescription such as alarm management, historical andbatch data recording with HMI languages, and accesslevel definition for automatic processes.

• Materials of Construction: Specifications for surfacefinishing grades of direct contact parts, gaskets, andlubricants where necessary.

• List of Drawings: This section contains a list of allnecessary drawings for the project such as a floor plan(initial layout), process flow diagrams, and data sheetsspecifying services and surrounding conditions for sup-port systems at the room level, where appropriate.

• Glossary: A straightforward list of all used acronyms inthe entire text.

• Appendices: Technical and commercial appendices aspart of the URS specify site conditions of work, plantutilities, maintenance and piping requirements, electricaland safety standards, packaging, marking, shipping, in-voicing instructions, and all non-disclosure agreementconditions.

Moreover, it is highly recommended to insert a dedicatedappendix for Validation Requirements. Vendors should beencouraged to fully address the Validation Requirements, asto commercial terms alike. This special appendix shouldcontain:

• a Project Quality plan from the vendor• a compliance declaration to 21 CFR Part 210, 211, and 11

for software controlled systems• traceability for material certificates for all direct contact

parts and surface finish tests, where applicable• welder credentials, welding certificates, and procedures

for high purity tubing systems under boroscopy inspectionregime

• layouts, elevation, and instrumentation drawings• demarcation diagrams

• IQ/OQ/PQ and FAT protocols• calibration certificates for critical instrumentation• operating and maintenance manuals with spare parts list

Additional guidelines for a successful URS:1. Process Related Issues

The effective text will clearly distinguish between theactual required specifications from the product or processlimitations, which are most likely known by the user.When delineation is not comprehensible, vendors mayoverlook the specifications due to their own responsibilityrestrictions upon the product, as further illustrated inreal-life cases.

2. FormattingIn order to make it qualifiedly workable, the URS contentsmust be systematically marked making them traceable bythe paragraph level. This means not bulleting, but num-bering all sections, items, and sub-items in a simple, butlogical way. It is worth mentioning that well establishedvendors in the market have their documentation practicesnot often changed. Consequently, differences between anowner’s well organized URS and the vendor quote formatwill be found. As was pointed out in section URS x RFP, thestructured URS approach will definitely help the qualifi-cation exercise making it an efficient traceability work, asrequired.

What to look for in the RFP?There are additional important issues, neither technical norprocess related, to be addressed before placing the PO. Theyare a mixture of purchasing and logistics activities, i.e., aftersales service, payment conditions, transportation and crat-ing, non-disclosure agreements, and so forth. As mentionedbefore, all these items are out of the design qualification scopeof the present model. In practice, after bid analysis, thetechnical proposal of the selected vendor is usually attachedto the commercial PO, thereby transforming it into an officialRFP. With that in mind, attention should be placed before,but close to the PO placement on the following:

• working volumes, capacities, minimum, and maximumbatch sizes

• safety issues, dust explosion classes, equipment dimen-sions, and system interfaces

• RPMs, speeds, loading, unloading, flow, and nominalproduction rates

• filtering efficiency, heating, cooling, and pressure specifi-cations

• materials of construction for direct contact parts• alarms, interlocks, manual, and automatic control fea-

tures• data recording, HMI languages, and parameter exporta-

tion• all peripheral equipment specifications



What to look for in the DDS?By following the verification table, but before approving theproposed detailed design, the qualification exercise shouldverify the following:

• The detailed manufacturing drawings in all mechanicalfeatures and legends. Where possible, refer them to thespecifications of previous documentation.

• The Process and Instrumentation Diagrams (P&ID) intheir details namely, numbers, codes, symbols, names,units, and process.

• An instrument list by tracking it on working ranges, units,calibration traceability, accuracy, and physical accessibil-ity (production or technical areas) of each item, as it wouldseem defined in the original URS and confirmed at theRFP documentation.

• A detailed description of all the Sequence Of Operations(SOO), sometimes called Functional Requirement Specifi-cations (FRS) that reflects the automation and controlconditions of the URS in all logical aspects. This documentshould be provided in a traceable matrix or available indexas the base for software development and qualification.

• All possible out-of-specification situations with associatedalarm messages quoted at the written word level.

• HMI screen definitions according to the company stan-dards where available for words, colors, fonts, and lan-guage.

• Hardware, software, and firmware detailed specifications(each of these should be a separate document).

• Software development procedures and a CFR Part 11compliance gap analysis with software deliverables as perthe validation appendix of the URS.

What to look for in the FAT?This is the last design qualification step. It takes place afterthe actual FAT is performed so that it has the advantage tobe based on the actual testing results. A detailed FAT checksdocumentation, equipment operating ranges, surface finish-ing grades, software controlled outputs, and process out-comes in cases where a placebo/product test is performed.Previous work published8 is recommended for those readerswho are interested in getting prepared for a detailed FAT.Since an acceptance criterion is tangible and measurable atthis point of the project for its features, it becomes straight-forward to make the GO/NO-GO decision during every test.However, as a preparation sub step, the design qualificationshould focus on the extent of the tests by previously checkingthe FAT vendor protocol versus the initial URS, as well as allthe deviations that occurred during the earlier RFP and DDSdesign phases. If the vendor has no FAT protocol available, itshould be developed by the user. At last, the FAT report isqualified by the crosschecking technique. Those tests consid-ered to have failed or not performed due to tight budget/scheduling constraints will be recorded in the DQDR andtransferred to the IQ/OQ/PQ protocols. Such procedure con-cludes the design qualification lifecycle.

A New Pharmaceutical Plant Case Study:DQ Real-Life Examples and Lessons

A DQ project using the present model was implemented overa span of one and half years on 13 major pieces of processingequipment, three critical systems, and two building facilitiesfor a new 250,000 sq. ft. [23,000 m2] OSD, liquids, andointments plant in the Israeli pharmaceutical industry. Thequalification staff comprised of two validation engineers (onefor equipment and one for critical systems and facilities), anda team leader. The responsible committee had representa-tives from QA, Maintenance, Engineering, Production, andTechnology. Table B shows in the Design Qualification partthe number of deviations for 11 pieces of processing equip-ment and one critical system. Some cases had minor devia-tions gathered into only one DQ report. A total of 478 reportswere recorded on the finalized projects listed in the table bythe qualification start date.

After a case-by-case examination, all the reports wereintentionally dichotomized into two categories: technicaldeviations and process related deviations. Records for thetechnical deviations category related to mechanical and manu-facturing discrepancies such as incorrect drawings, outstand-ing dimensions or fittings at detailed design, unexpecteddifferent construction materials, change in surface finishgrades, omitted safety standards, modified filtration levels,missing parts or tooling, and incomplete engineering docu-mentation, and even substitution of pre-approved models forparts used during assembly at pre-delivery inspections. Theprocess related category, in turn, recorded deviation reportsfor items such as omission on control features, modifiedautomation aspects, unfinished software, misplaced instru-mentation, incorrect performance capabilities, inaccurateflow specifications, partial sequence of operations, lack ofinterlocks, missing alarms, HMI inconsistent screens, al-tered failure safe definitions, inflexible feeding rates, non-adjustable filling volumes or rotating speeds, pressures,loading/unloading rates, inappropriate working ranges andscales.

An extensive evaluation of these real-life examples re-sulted in the following remarks which are well worth men-tioning:

Inappropriateness and Similar Deviations amongDifferent SuppliersVendors alike did not confirm specific system output perfor-mances such as spray rates, air velocities, differential pres-sures, speeds, yields, and have overlooked explicit workingranges for product temperature. Moreover, cycle times, end-point of process control, or interlock fail safe protections havebeen omitted from the technical part of most proposals, evenwith full product characteristics information available. Thisrange and number of incompatibilities do not necessarilyimply low quality level on the work performed by the vendorside. It is the combination of two facts: the first is thatscreening out the project design under magnifying lenses atits earliest stage may generate premature DQ reports. Thesecond is the exaggeration on requirements set by users that



naively expect to find solutions for their current productiontroubles on the new technology. This situation calls forsharpening the URS contents, as previously mentioned. Theparticipation of technological personnel in the DQ respon-sible committee ensures process expertise inputs and avoidssetting requirements out of the vendor’s capabilities andscope.

Converging TrendThe number of deviations found decreases as the projectadvances for each qualification project. This expected, butquantitatively concluded, converging trend occurs due to thefact that fewer design modifications are raised as the projectnears the installation phase. It is also likely that internal oroutsourced design engineers will keep to the user definitionsas they participate in the screening qualification committee.The result is fewer deviations along the project pipeline.

Steady Ratio between the Number of Technicaland Process Related Design DeviationsWith two exceptions only, all three qualification steps had thetotal number of deviation reports of the technical categorysuperseded by a twofold trend in the number of deviationreports for the process related category. This ratio suggestsemphasizing the attention on the technical manufacturingissues more then to its production and process features atleast during the early stages of the project.

Model Consistency and AdaptationsTable B contains conclusive facts for projects with finalizeddesign qualification. Other projects like Compressed ProcessAir, Material Handling, Process and Stability Rooms under-went the same roadmap model of Design Qualification. Theirsimilar, midway results obtained by the time this article waswritten confirmed the usefulness of the model.

However, the model presented, due to its flexible natureand simplicity, can be used for complex supporting utilitiesand API production systems as well as with very smalladaptations. For example, mechanical issues of HVAC sys-tems are tested separately from their off-line software devel-opment, while the duct subsystems are not tested until airbalancing starts close to OQ. Since all these issues are beinghandled by different suppliers at different dates, it ends upwith a multidimensional DQ inside one single project. Thisillustrates the need for further break down of the three majorcrosschecks into sub steps with different milestones for each.Added ValueDQ is the first step in the qualification process. When prop-erly performed, it has the biggest cost saving potential for theproject. Table B also shows the percentage estimated ondirect and indirect costs associated with every DQ deviationfor each project budget, reflecting the cost benefit achieved bythis activity. A 12% added value shows the impact for savings,giving a clue to management decisions when to start involv-ing validation personnel in the project. The figures haveproven – the sooner the better.

OrganizationTo help accomplish the DQ program at the completion of anew facility design, it is highly recommended to employ askilled validation engineer to lead the DQ responsible com-mittee. However, he must work independently of projectmanagers, who may jeopardize the design qualification workin favor of tight schedule interests.

Future Qualification PhasesIn the major sense, the technical and processing issues raisedduring design qualification correspond to installation andoperational issues, respectively. The deviations found forboth aspects of the system provide at an early stage potentialcontribution for developing protocols aiming the ‘en-route’validation phases: Installation and Operation Qualifications.In a quality oriented environment the information transferfrom DQ to IQ/OQ/PQs leaves motivation and space forextension of the presented methodology.

Concluding RemarksThe objective of this article was to provide a model with toolscapable of supporting the design qualification process fromstrategy to report. Although changes are likely to be inevi-table during design, the “right first time” aspiration conceptin the whole pharmaceutical industry will benefit from theuse of this model in the following:

1. the basics for establishing a flexible and justified policy fordesign qualifications

2. the practical way of dealing with design deviations by asimple and methodical technique

3. the approach that keeps the essence of the facts and avoidsunwanted retroactive URS corrections or cumbersomechange control procedures early in the project, but ensuresfull traceability at the design phase

4. the controlled documentation of rationales for requiredregulatory justifications of design changes

5. the documented evidence that the defined “intended use”of a system or equipment is maintained from the start andis followed all the way through the natural project designmilestones

In the long run, the methodical approach for DQ as presentedwill provide evidence of early overall compliance and maxi-mize the cost benefits of the efforts put into validation fromthe beginning of the project.

References1. Wood, C., “Commissioning and Qualification,” Pharma-

ceutical Engineering, Volume 21, Number 2, March/April2001, pp. 50-54.

2. Blackburn, T.D., “A Practical Approach to Commissioningand Qualification – A Symbiotic Relationship,” Pharma-ceutical Engineering, Volume 24, Number 4, July/August2004, pp. 8-18.

3. ISPE Baseline® Pharmaceutical Engineering Guide, Vol-ume 5 - Commissioning and Qualification, International



Society for Pharmaceutical Engineering (ISPE), First Edi-tion, March 2001, www.ispe.org.

4. Code of Federal Regulations, Part 210 and 211 cGMP inManufacturing, Processing and Packaging of Holding ofDrugs.

5. Guideline on General Principles of Process Validation, USFood and Drug Administration, May 1987.

6. European Commission, Final version of Annex 15 to theEU Guide to Good Manufacturing Practices, July 2001.

7. Guidance for Industry: ICH Q7A Good ManufacturingPractice Guidance for Active Pharmaceuticals, August2001.

8. Roberge, M.G., “Factory Acceptance Test (FAT) of Phar-maceutical Equipment,” Pharmaceutical Engineering, Vol-ume 20, Number 6, November/December 2000, pp. 8-16.

About the AuthorYossi Chvaicer is Assistant Director Man-ager of the Validation Department for Sys-tems and Equipment in Taro Pharmaceuti-cals Ltd., where he has worked since 1996. In1981, he obtained his BSME from the Uni-versity of the State of Rio de Janeiro, Brazil,and since 1989 holds an M.Sc. degree inindustrial engineering from the Technion,

Haifa Institute of Technology, Israel. Chvaicer has headedvalidation projects worldwide in the pharmaceutical indus-try, covering qualification, quality assessments, and interna-tional compliance audits in Canada, USA, Ireland and Israel.During his professional career, he has worked in Brazil andin Israel in the major chemical process and pharmaceuticalindustries. His validation and engineering experience hasranged from design, construction, commissioning to valida-tion of existing and new cGMP pharmaceutical plants, in-cluding HVAC, PW, and WFI systems, warehouses, produc-tion, packaging, and process equipment. Chvaicer can becontacted by tel: +972-48475737 or by e-mail: [email protected].

Taro Pharmaceutical Industries Ltd., 14 Hakitor St., POB10347, Haifa Bay 26110, Israel.

Validating Automation Systems


Project Management of AutomatedSystems Validationby Jim Verhulst

This articlefocuses onvalidatingprocessautomationsystems for atypical batchbiopharmaceuticalproject andpresents newways to thinkabout organizingthe effort. Inaddition, itdefines a typicalframework ofrelateddocumentsneeded forsuccess, anddescribes newpractices andtools needed tomanage theinherentcomplexity ofthese projects.

Introduction

Several years ago, the project directorfor a large construction firm told methat automation work really scaredhim. He knew a great deal about build-

ing the other aspects of the project from previ-ous experience, but how to control the automa-tion team was a mystery. Worse yet, he contin-ued, the automation system would have to bevalidated1 in some way. This unfortunately, isthe case for many project managers I have met.Because of all the technical jargon, they feelvery uncomfortable with an important part ofthe construction process, not realizing that it isin many ways similar to putting together theconcrete, steel, and process piping with whichthey are familiar. Automation systems also arespecified, designed, and tested prior to theiruse in the facility. Data flows and storagefacilities are similar to fluid flows and holdingtanks in that there is a need to provide capacityand to prevent contamination in both worlds.The big difference to project managers is vis-

ibility. They can readily see progress and testquality when dealing with tangible components.Software is too often invisible and its statusand quality assessment requires the opinionsof yet more “software types.” The ready avail-ability of straightforward tools and techniquesthat make software visible and testable wouldgo a long way toward bringing software into theproject mainstream.

I also have met Quality Assurance and Vali-dation Managers who, because they did notreally understand how automation systems getdesigned and built, felt uncomfortable withdeveloping an automated system validationstrategy.

Because of the specialized and technicalnature of the work, automation projects, whichare usually sub-components of much largerconstruction projects, are often assigned to com-puter programmer types. These people may benew to project management and have onlylimited experience with processing equipment.The QA and validation personnel assigned to

Figure 1. Feedback andthe GAMP waterfall.





the project also may lack experience, especially with large,integrated automation installations.

Automation ProjectsLike other kinds of construction projects, an automationproject has to meet defined schedule, cost, and quality stan-dards. Because automation equipment and systems becomeintegral parts of the physical manufacturing equipment theycontrol, they can significantly impact the profitability,throughput, operability, maintainability, and safe operationof the finished production facility.

Projects we have studied vary greatly in achieving thoseobjectives. The usual complaint is that the system is late,closely followed by validation of the system is both late andover budget. Both of these problems can be traced to incom-plete and hurried project planning. It is fairly common to hearthings like, “Let’s get started, we’ll figure out the validationpart later.” These are crucial mistakes and will become clearbelow.

Generally accepted project management practice orga-nizes project activities into phases such as design/review,procurement/delivery, and commissioning. Each project phasehas its tasks and resource requirements. Project manage-ment for the automation area should be consistent with therest of the project. Each project phase has its deliverables.With software, these deliverables are usually documents.The preparation of a document for every specification and

every test of that specification is actually a project taskrequiring appropriate resources.

The project also must interface with the existing organiza-tion. In general, the corporation for which the project is beingbuilt (the Owner) has in addition to the Engineering, Main-tenance, and Production functions, Quality Assurance, Regu-latory Affairs and Validation Departments. These three havesomewhat different responsibilities. Although this may bedone differently from one Owner to another, in general:

Quality Assurance (QA) is responsible for:• ISO 9000 style auditing and review that asks the ques-

tions: “Do the entities being audited have appropriate andcomprehensive procedures and are they following them?”

• batch records review prior to product release• Quality Control (QC) - primarily laboratory analytical bio/

chemistry tests and controls• document management and control

Regulatory affairs is the primary interface to the FDA,especially with regard to notification of process or equipmentchanges and their impact on sterility, purity, potency, stabil-ity, and efficacy.

Validation is often a part of Production or Engineeringalthough it may report to QA. Cases could be made to haveValidation report to Regulatory Affairs (more familiar withFDA) or even the Owner’s Legal Department because the

Figure 2. Additional necessary project documents.



Validation Group is, in fact, developing the basis of a legaldefense, should it ever be required. Validation responsibili-ties include:

• developing documented evidence that will stand up in acourt of law by writing and executing IQs, OQs, and PQs

• enforcing Change Controls to maintain the validated state

The capital project has its own Quality Management functionthat is responsible for:

• design and documentation reviews• monitoring construction activities• project change control/developmental change control (usu-

ally considered pre-validation change control and goodengineering practice)

• Factory Acceptance Test (FAT) and Site Acceptance Test(SAT) commissioning tests

Obviously, a capital project will proceed more smoothly ifeach group understands the charter and scope of the others.Commissioning (done by the project team) should lead logi-cally to Validation (the responsibility of the Owner). The QAbatch records review group should work closely with theproject programmers who are developing the batch reports.QA should be allowed to critique the software developmentactivities to assure that proper procedures are being followed.

Writing automation software for a biopharmaceutical fa-cility is, in the simplest case, the task of configuring aConfigurable, Off-The-Shelf (COTS) software application toproduce a system with the desired characteristics. Simpleconfiguration is rarely the case. Most systems are combina-tions of COTS configurations and custom software scriptsand code. But for the most part, the project’s engineers andprogrammers get to stand on the shoulders of giants, so tospeak, which are the vendor companies who have investedconsiderable time and effort to produce the COTS software.

This article does not address the development or testing ofthe COTS software itself. However, the tools used by firmsbuilding COTS software are worth considering. Most large-scale software development efforts use complex and verycompetent tools that have been around for decades. Theproblem for a capital project manager is that these tools areoften too expensive and take too long to learn to use effec-tively. There are very few mid-sized tools available. Cost andschedule constraints often leave the project manager withhomegrown applications that may be incomplete and lackvendor support.

It’s time to re-think current practices and to find betterways of doing things. Better tools are clearly needed, espe-cially for larger capital projects. Project managers must beable to quickly measure construction progress against sched-ule. They also want to be able to determine the suitability ofthe software designs and code modules that have been pro-duced to date so that if rework is required, it can be put intothe project schedule and budgeted.

Several firms are starting to supply new and right-sized

software tools that help make automation software develop-ment more accessible. The tools do this by providing thecomplete list of tasks, specifications, schedule milestones,and the progress of software testing. This visibility is the firststep in controlling the project. Project managers and automa-tion engineers both know that you can’t control what you can’tmeasure. As you also will discover, managing the complexrelationships between documents and sections of documentsis the key especially when the inevitable changes occur.Careful selection of tools is required. The real gains are notto be found in reducing the clerical effort with new tools, butrather in having a real-time overview of project progress.

One of the advantages to the use of automated tools is thatthey can become a “leave-behind” that enable operations andmaintenance personnel to continue to track and control anautomation project throughout its entire lifecycle.

Having said all that, better tools are only a part of thesolution. The path forward also requires a clear understand-ing of the tasks at hand and the practices needed to organizeand execute them. Many project managers only learn the fullextent of the problem sometime during the project. By then itis too late.

GAMP® 4 to the RescueISPE’s GAMP 4, a Guide to Good Automated ManufacturingPractice,2 is a big step forward in formalizing a framework forthe design and qualification processes. It promotes a consis-tent, practical approach to developing, installing, and operat-ing compliant systems, which meet regulatory requirements.Here is a little background on the approach.

The Design-Test Result Feedback ConceptFeedback is a core concept in both GAMP 4 and in automationdesign although it may be known by different names. Feed-back is well known to all automation engineers because theterm is the original basis of automated control systems. Forcontrol systems, the idea is to measure a process variable,compare it to a desired value, and calculate the differencebetween the two. This difference is known as the error. Thecontrol system then changes the value of a separate systemvariable, (one that affects the process variable), to reduce theerror to zero. The concept in lay terms goes like this:

• If the shower feels too hot, reduce the flow of hot wateruntil the desired temperature is reached.

The use of feedback in GAMP 4 is indicated in Figure 1. Here,the term is related to the pass/fail status of the validation testexecution and the related deviation and remediation reports.It is important to note that validation and automation engi-neers are both employing the same feedback concept, eventhough it is being expressed using different words.

Extending GAMP 4 ThinkingAs good as GAMP 4 is, there is still more work to be done. Tobegin with, GAMP 4 has a narrow perspective, that is, it tendsto define all process automation systems as isolated services,



much the way a database server operates.Next, and this is hardly a criticism, GAMP 4 is limited in

scope. The GAMP Forum wisely chose to address only regu-latory issues (a good decision or they’d probably still bewriting GAMP 4). Most projects involve many more documenttypes beyond those pertaining solely to regulatory concerns.Figure 2 illustrates this point. (Consider this diagram, whichexplains a medium sized batch control project to be a supersetof the things a smaller project should at least consider to seeif any of the documents apply. However, Figure 2 also couldbe a sub-set of a large project with many integrated systems.)

Note the limited coverage (indicated with the light graybackground in Figure 2) that GAMP 4 provides for a typicalbatch automation project. Although additional specificationsand recommendations have been published since, the origi-nal GAMP 4 V-Model had significant limitations for a realworld batch control project. Among these:

• The methodology works well for a single system, but itdoes not address system-system integration. There is nomention of specifications that are meant to cover a numberof related systems.- All systems should be built from the same parts and

pieces if possible. For example, using the same versionof only one database manager, even though the data-bases and database applications may be different.

- where systems are meant to work together as compli-mentary parts of the same puzzle

- where systems must co-exist on the same network orcomputer hardware

• No mention of rapid prototyping practices,3 iterative de-sign and development, or other life cycle models.

• Little mention of safety, maintainability, operability, ordesign reviews.

• Little mention of commissioning tests.4

• No interface with S88 batch systems components. (ANSI/ ISA S88.01 is an international standard describing batchprocess control terminology).

• No mention of corporate and project policies that impactwriting the URS and the test plans.

• Limited advice on change control procedures. All projectsmust accommodate several levels of revision and changesas the project is defined.

The good news is that the recently published GAMP GoodPractice Guide: Validation of Process Control Systems, ad-dresses some of these issues:

• Factory and Site Acceptance Tests (FAT and SAT)

• safety reviews

• design reviews

• system interface design issues

The ISPE Baseline® Guide on Commissioning and Qualifica-tion5 also helps to fill some of these gaps. It “focuses on theengineering approaches and practices involved in providingfacilities in a timely manner that meet their intended pur-poses.” The Guide describes the organization and content ofthe Commissioning Plan document and provides guidance onthe management and execution of commissioning activitiessuch as inspection, start-up, adjustments, performance test-ing, training, turnover, and close-out.

It should be mentioned here that software testing is acommissioning activity that should occur even in a non-GMPimplementation. The GAMP V-Waterfall works well here toobecause it develops the relationships between specificationsand test content and provides the basic acceptance criteria forthe commissioning tests. Recognizing this, many firms nowemploy rigorous commissioning prior to validation. Problemsfound during commissioning are handled under developmen-tal change control and do not generate the complex paper-work of a qualification incident or validation deviation.

Project Management ChallengesAutomation project managers are often confronted with chal-lenges that are the result of fast-track execution. Recognizingthese issues early on and developing strategies to deal with

Figure 3. Project bookshelf.



them are key to success. As explained in the paragraphs thatfollow, these include:

Incomplete Design SpecificationsAutomation project managers often need to work with incom-plete and preliminary information. Some of this is caused bythe process design itself. Process development specialistsworking toward yield and purity goals in the lab often havethe latitude to change conditions, even late in the game tomeet these targets.

Design Policies Not Clearly EstablishedManufacturing facilities are complex, integrated entities inwhich the interests of all the various stakeholders must bebalanced for efficient operation. The role and importance ofhigh level policies in achieving this balance are explained indetail later in this article.

Change Control StrategySpecification changes ripple down through the constructionproject as either specifications that are late or specificationsthat change once they have been defined. These changes oftenaffect other facets of the project design and impact cost andschedule. Obviously, a well established and efficient changecontrol process is mandatory for a well run project.

Testable Requirements and SpecificationsWriting test plans should be, in large measure, the task ofefficiently organizing the test execution so that the testsdemonstrate the desired functionality of the equipment. Infact, most of the time is spent figuring out what the specifica-tions actually mean so that comprehensive and quantitativetests can be written. We sometimes find tests that have nobasis in the specifications, and often that is the result ofincomplete or vague specifications. The GAMP 4 Require-ments Traceability Matrix (RTM) addresses this issue. Theproblem is keeping the RTM up to date in the face of designchanges. Knowing this up-front, the project manager canassign sufficient staffing and buy the right software tools tokeep the RTM current and useable.

Automation and the Project Critical PathAnother reason project managers worry about automation isthat experience has taught them that automation commis-sioning is almost always on the project critical path. And thismakes sense. Simulation can help, but automation and con-trol strategies cannot really be finished until all the wiring isdone, all the transmitters calibrated, and all the controlparameters are adjusted for the irregularities of the realworld. The goal of the automation project manager is to makethat part of the critical path as short and predictable aspossible. No offense meant, but most computer programmertypes are not very good at commissioning since they usuallylack the process and equipment experience. Again no offenseintended, but QA and validation types often lack both thecomputer science and process backgrounds. Commissioningand qualification of automation systems will fall to a rather

specialized, multi-discipline group if the project schedule is tobe maintained. These resources may be difficult to find.

Project Bookends ConceptThis is a very simple, yet often overlooked principal. Everywell run project has clearly defined objectives. Satisfyingthese objectives is the finish line for the project. Knowingwhen the objectives have been accomplished is a key mea-surement for any project. Obviously, the specifications andthe acceptance tests should come in pairs, which can bethought of as bookends. The project’s defining documents arethe left bookends on the project timeline. The commissioningtests, validation tests, and project evaluation documents arethe right bookends.

Automation tasks rather naturally form the specification- test pairs described. For example, if a design specificationsays that the display screen background color shall be navyblue, the test asks whether the screen background color isnavy blue. If a control sequence dictates that valve XV-1003should open at a particular point, the test verifies that thevalve actually opens. (This is not to imply that writing thespecifics of just how to perform the test is simple; it oftentakes skilled and knowledgeable people to design these tests.But this process starts with knowing which tests must beperformed).

Bookends also bring accuracy and closure to project esti-mates. Project managers can now see all the tasks betweenthe start and finish lines and can deal with them appropri-ately. Automation tasks and tests are now de-mystified andbrought into clear view where they can be staffed, scheduled,cost estimated, and managed.

The Project BookshelfCompiling all the specifications and tests into usable docu-ments results in a project bookshelf that looks something likeFigure 3. Getting consensus on not only the specifications,but also for the tests and the acceptance criteria by which thetests pass or fail, is of critical importance.

• Completing the IQ is important to sub-contractors whowant to be paid at mechanical completion.

• Completing the OQ is normally the most financially sig-nificant milestone for the construction project managers.It is at this point that ownership of the facility usuallypasses to the operations and maintenance groups.

• The PQ and the corresponding process validation are themost significant for the Owner team that will obtain alicense to operate the facility to be able to sell product.

These tests and their acceptance criteria have to be decidedup-front. Needless to say, a project manager will be in a poornegotiating position with the sub-contractors if the accep-tance tests are added or changed later on in the project.

The URS is not the Top DesignLevel Document

Because the User Requirements Specification for an automa-



tion system is such an important document, design engineersand project managers often hurry to start writing it. How-ever, all User Requirements Specifications need to have atraceable basis and solid justification or they will be subjectto change once the project is underway.

A URS is actually driven by the documents in three levelsabove it. I will call them the Project Definition Documents,the Technical Strategy Documents, and the Master Specifica-tions, System Architecture Design, and the CoordinationDocuments. However, there is no consensus on this nomen-clature. The new levels are shown in Figure 4.

The feedback concept is used in these upper levels as well.In the biopharmaceutical world, feedback at the upper levelsmay be called review, critique, lessons learned, project evalu-ation, or benchmarking.

These three new upper levels need to be defined andcodified before the writing of the URS can begin. A draft URSmay identify where additional upper level decisions, docu-ments, and policies need to be defined.

For the project manager, these layers can be thought of asa project workflow that progresses from top to bottom asshown in Figure 5. Please note that the FS and DDS levelsshown in Figure 4 have not been shown in Figure 5, only for

clarity.As can be seen in Figure 5, there are many more docu-

ments to be considered beyond those described by GAMP 4.(Note how the URS for the Process Control System, located inthe lower center of the diagram, is overshadowed by all theother required tasks, policies, and documents). Here is a listof typical documents:

Project Definition Documents• business operational requirements• enterprise-wide design policies• project authorization, organization, and schedule

Technical Strategy Documents• Business operational strategies, e.g., the project will in-

clude a production scheduling functionality.

• Project specific design and operating policies. These poli-cies are derived from the appropriate enterprise-widepolicies and adapted to this specific project. There could beas many as 50 different policy documents that may beneeded for a large project. Some of these may already existand may be re-used from other projects. Others may have

Figure 4. Six layer waterfall diagram.



to be created from scratch. There are at least 12 differenttypes, or classes, of policies. For example, control systemdesign, record-keeping, and simulation policy classes mayall be needed. Policies are working documents for all thestakeholders where agreed-upon decisions that need to beapplied uniformly across the project can be stipulated.These policies form the basis and justification for deci-sions made in the layers below.

• Sized process and utilities definitions (sometimes calledthe basis of design). These usually start as sized ProcessFlow Diagrams (PFDs) and Utilities Flow Diagrams(UFDs). The Process and Utilities Piping and InstrumentDiagrams (P&IDs and Utilities P&IDs) are derived fromthe PFDs and UFDs. The design engineers recognize thatprocess and utilities designs have much in common, havenumerous mechanical interfaces, and should employ thesame construction techniques if possible.

• The plant master electrical design describes, among otherthings, the distribution and location of motor controlcenters and whether bus communication technologies willbe used. It also describes the standby power system (oftencalled emergency generators), and the UninterruptiblePower System (UPS), and the plant electrical and instru-ment grounding systems.

• Automation and control strategies are normally brokeninto three parts according to ISA/ANSI S88.01 conven-tions: S88 recipes, the S88 physical model design conceptsfor both process and utilities (closely related to the P&IDs),and the S88 procedural model. This last model describesthe control algorithms and processing sequences to beused by the automation system.

Master Specifications and System Coordination• The Master Specifications and System Coordination (MS/

SC) documents level is a meta-layer that defines therelationships between multiple automation systems andspecifies practices and components that should be used byall systems if possible. Documents in this layer:- Establish common specifications for equipment and

systems to simplify training and reduce the variety andnumber of spare parts.

- Specify compatible components of a larger design, e.g.,a Supervisory Data Acquisition and Control (SCADA)system and a data historian.

- List components from different systems that must beable to coexist, e.g., the same version of a databasemanager installed on shared hardware.

• Master Specifications exist in two forms, each with theirown revision histories:- Master Specification templates that are used to derive

Master Specification instances. Templates are revisedbased on project experience so that the templates areready to use for the next project. This is analogous to the

control of most of the other specifications - one wants tolearn from experience.

- Master Specification instances initially contain short-lists of products and services from approved vendors.These will become part of the Request for Quotation(RFQ) to these vendors. These general specificationswill be refined to detailed specifications, suitable forprocurement from the selected vendor, later in theproject once the technical and commercial reviews ofthe bids have been completed.

• Examples of Documents in the MS/SC meta-layer:- system architecture diagrams (often modeled as UML

diagrams to show system-system interactions)- list of preferred or allowed database management soft-

ware- preferred or allowed document management systems- network architecture diagram- preferred or allowed computer hardware suppliers- preferred motor control center equipment

New Tools, New ThinkingSeveral companies are building and selling new documentmanagement tools to help manage automation project com-plexity. However, some re-thinking of current practices andproject execution strategies also may be needed.

Clarify the Meaning of “Quality”An old joke among project managers, when referring to thecost/schedule/quality triad is “Pick two of three.” You canhave price and quality at the expense of schedule, and so on.New thinking on project quality leads one to a simple rule forquality - that all quality measurements must be quantifiable.This in turn leads to the conclusion that quality becomes “Itmeets the spec, end of discussion.” There are many advan-tages to this way of thinking:

• tight specifications can be bid accurately by knowledge-able sub-contractors

• accurate bids yield realistic schedules• commissioning and validation personnel cannot introduce

new or vague interpretations

The price, of course, is that this puts a significant burden onspecification and test writers early in the project. This defini-tion is also at odds with most rapid prototyping concepts. Butknowing this in advance, resources can be found and allow-ances can be made. More discussion of this point is includedbelow.

A New Paradigm for Project DesignThe planning for any new automation project should considermulti-system design to be the norm. This implies a cleardefinition of how the systems are related and how data passesfrom system to system. Multi-system design should definetopics like system interfaces, data custody and quality, reli-able operation and communications (especially when one or



Figure 5. Examples of upper layer documents.



more components may be temporarily unavailable), and USFDA 21/CFR Part 11 compliance.

New project planning also should extend the GAMP 4Requirements Traceability Matrix (RTM) to the upper leveldocuments. This will assure that each specification hasproper definition and justification.

Emphasize writing testable specifications. Avoid usingvague terms that are difficult or impossible to test. Forexample, “Report generation will not affect the operation ofthe rest of the system.” Substitute the following, for example,“The server shall be sized so that the Central Processing Unit(CPU) Usage, as shown on the system performance screen,does not exceed 70% load while generating the XYZ report”.

Staff up early to be able to do all the top down planningrequired during the design phase. It often takes time toassemble a project team. Many project managers, aware oflong delivery times for some equipment items, favor puttingresources on design activities and wind up in the position ofsaying “Let’s get the design and procurement phases started,we’ll figure out the commissioning and validation later.” Theleft bookend is in place, but right bookend is undefined. Thisunfortunately can lead to open-ended costs, uncertain sched-ules, and vague staffing requirements for commissioning andvalidation; not what anybody wants.

Get consensus from stake-holders early on in the designphase. Focus on the primary documents first. For example,the use of policy templates to formulate project policies raisesquestions early in the design stage that may otherwise ariselate in a project. While the project is running, these policiesserve as impartial referees when disputes occur betweenstake-holders. Master Specifications and Coordination docu-ments make clear how the various parts of the project puzzleall fit together and avoid incompatibilities once the multiplesystems are integrated together.

Include all commissioning and validation activities in thelist of project tasks to produce a complete, reliable, andmanageable project schedule. If every specification has acorresponding test, project commissioning and validationcosts and schedules can be accurately estimated. Having acomplete project schedule is very important to upper manage-ment and to project managers who have to staff the project.

Recognize that in an automation software project, thepreparation of a document for every specification and everytest of that specification is actually a project task. Eachdocument is a quantifiable entity.

Most document preparation tasks can be estimated accu-rately from past experience, but in some cases, a rapidprototyping approach is justified. Special handling is re-quired for these.

• Significant allowances must be used for poorly defined, orfirst-of-a-kind efforts, and it is important to know whichdocuments these are

• Identifying difficult tasks and their place in the criticalpath, often leads project designers to substitute standard-ized, proven technologies for critical tasks

• If there are instances where rapid prototyping is appropri-

ate, these need to be identified early, and placed near thebeginning of the schedule so that there is time to write thedetailed specifications and tests.

Knowing which documents are what degree of difficulty helpsproject managers to prioritize tasks and focus on justifyingand obtaining experienced resources to reduce risks of thesedocuments being late or inaccurate.

A New Validation ParadigmA new paradigm for validation may be required to reduce thetime and cost of validation. This will change the roles andtiming of the validation effort somewhat.

• Quality/validation personnel must be involved from thebeginning of the project.

• Project management must find ways to break down the“engineering discipline” walls between QA and engineer-ing if these exist.

• The design engineers should write most of the tests be-cause they understand the underlying principals andassumptions.

• Quality personnel will move to support and audit rolewhere their input is primarily on the “testability” of thespecifications and auditing the procedures used by thedesign engineers.

• Validation personnel should assist with the commission-ing effort, as this is in many ways a “pre-function” for thevalidation activities.

• Validation personnel should be in charge of the executionof the qualification activities. Obviously, this has to bedone in coordination with Engineering and Productionpersonnel to ensure safe operation during the tests.

New Group Collaboration ToolsMost biopharmaceutical concerns have been using the indus-try-standard document preparation and management tools6

for some time. Many firms also use newer group-orientedsoftware tools and databases. These are useful and severalhave gained broad acceptance. However, the next generationof tools needs to go further to:

• fully Web-enable collaboration• support the full system lifecycle7

• allow the requirements traceability matrices to be gener-ated automatically as contributors from multiple loca-tions work on the web of interrelated specifications andacceptance tests

• manage the six (or possibly more) layers of documentsdescribed in Figure 4

• be extensible and adaptable to be able to handle futurerevisions of GAMP (but that’s a topic for another article)

ConclusionsAt first glance, project management of automation systemsdevelopment and validation, (and by association, GAMP 4),appears to be complex and arcane. The purpose of this article



has been to help explain the concepts involved. Automationvalidation projects can be considered variations of morecommonly understood projects, such as piping and equip-ment installation, because the same basic principals apply.Some new thinking (or re-thinking) is required, but experi-enced design engineers and commissioning and validationpersonnel are already familiar with the processes involved.

New tools will be helpful in promoting needed visibilityand these will become available as the industry moves tomore standard practices. One of the advantages to the use ofstandardized procedures and automated tools is that theycan become a “leave-behind” that enable operations andmaintenance personnel to continue to track and control anautomation project throughout its entire lifecycle.

The author does not claim to have the final word on thiscomplex topic. On the contrary, my hope is to inspire furtherdialog and discussion that will move the field forward.

Many thanks to the reviewers who helped to direct andrefine this article.

Glossary of TermsCBER - US FDA Center for Biologics Evaluation and Research

CDRH - US FDA Center for Devices and Radiological Health

GAMP 4 - A Guide to Good Automated ManufacturingPractice produced by the GAMP Forum, ISPE Community ofPractice

ISPE - International Society for Pharmaceutical Engineering

S88 Batch - ANSI/ISA S88.01 standard describing batchprocess control terminology

URS - User Requirements Specification

FS - Functional Specification

DDS - Detailed Design Specification

IQ - Installation Qualification

OQ - Operational Qualification

PQ - Performance Qualification

UML - Unified Modeling Language

Software Validation - In the FDA document “GeneralPrinciples of Software Validation; Final Guidance for Indus-try and FDA Staff,” dated January 11, 2002 and issued byCDRH and CBER, the FDA defines software validation as:‘confirmation by examination and provision of objective evi-dence that software specifications conform to user needs andintended uses, and that the particular requirements imple-mented through software can be consistently fulfilled.’

References1. The term Validation is used here to mean Installation,

Operational, and Performance Qualification.

2. GAMP 4, Good Automated Manufacturing Practice (GAMP)Guide for Validation of Automated Systems, InternationalSociety for Pharmaceutical Engineering (ISPE), FourthEdition, December 2001, www.ispe.org.

3. Rapid prototyping can be used to develop a concept, testmachine performance, data throughput, and to refine thespecifications. However, to be successful, rapid prototypingshould only be used to develop and refine specifications.Otherwise, rapid prototyping can be at odds with top-downdesign.

4. The term Commissioning is meant to be a broad parallelof the mechanical commissioning activities. These nor-mally include the pre-operation activities (e.g. inspection,mechanical adjustment, loading of lubricants, motor rota-tion checks, piping leak tests, etc.), and the start upactivities that involve the gradual application of load andthe measurements made while running under load for aperiod of time.

5. ISPE Baseline® Pharmaceutical Engineering Guide, Vol-ume 5 - Commissioning and Qualification, InternationalSociety for Pharmaceutical Engineering (ISPE), First Edi-tion, March 2001, www.ispe.org.

6. Microsoft® WordTM and ExcelTM are commonly used.DocumentumTM from EMC Corp. is perhaps the best-known document management system, but there are manymore, some of which are more appropriate for smallerprojects. Adobe® AcrobatTM with its Portable DocumentFormat (PDF) also is widely used for document review anddistribution.

7. Software and system lifecycle normally includes the fol-lowing phases: Planning, Specification, Design, Imple-mentation, Testing, Deployment and Acceptance, Docu-mentation Turnover, Ongoing Operation, and Archivingwhen the system is retired.

About the AuthorJim Verhulst is the Vice President of Engi-neering at Innovative Process Solutions, Inc.in Acton, MA. Innovative provides automa-tion and validation solutions to thebiopharmaceutical manufacturing commu-nity. Verhulst holds a BSChE from the Uni-versity of Wisconsin. He has been doing com-puter automation projects for 25 years. He

learned biotech processes and equipment on the job duringthe 10 years he worked at Biogen as a technical manager.Previously, he had worked at Shell Chemicals, Syntex, andthe Foxboro Company. Verhulst is a member of ISA, AIChE,and has been an active member of ISPE since 1994. He can bereached at 1-978/266-0149, ext. 12.

Innovative Process Solutions, Inc., 130 Main St., Acton,Massachusetts 01720.

Microwave Extraction


Microwave Extraction of AntioxidantComponents from Rice Branby W.H. Duvernay, J.M. Assad, C.M. Sabliov,M. Lima, and Z. Xu

Background

The effects of extraction parameters, ex-traction temperature and extractiontime were assessed for microwave ex-traction of rice bran oil. The objectives

of the research were: to effectively extract ricebran oil from rice bran using microwave as-sisted extraction and to analyze the influenceof temperature and extraction time on the ricebran oil and vitamin E yield. Results showedthat the extraction time had a minimal effecton the vitamin E and rice bran oil yield at alltemperatures. Temperature, on the other hand,had a significant effect on the oil and vitamin Eyield. More vitamin E was extracted at 140°C(P less than 0.05.

IntroductionRice is a primary source of food and nutritionfor billions of people around the world and is animportant global economic factor. Louisiana isone of the largest rice producing states in thenation. The total value of rice produced onLouisiana rice farms in 2002 was 159.6 milliondollars. Of this amount, 122.8 million was thegross farm income, the money made directlyfrom selling the processed rice. The remaining

36.8 million was gained from using the by-products of the rice milling process.1 The husksand the bran account for the largest amount ofthese by-products. Today, most rice farmerssell their bran as animal feed. Farmers couldincrease their rate of return if they could sellthe bran to the food industry and/or pharma-ceutical companies which would be able toutilize the antioxidants for health purposes. Ifthe vitamin E and other antioxidants can beextracted from rice bran, this would maximizeprofits and allow for a more complete utiliza-tion of the rice by-products.

Rice bran is of interest to the pharmaceuti-cal industry because it contains 15-20% oil byweight with high concentrations of vitamin Ecomponents (tocopherols and tocotrienols).Crude rice bran oil is composed of 88 to 89%neutral lipids, 3 to 4% waxes, 2 to 4% free fattyacids, and approximately 4% unsaponifiables.2

Vitamin E components exhibit antioxidant ac-tivity and are proven to have various medicalbenefits which include reducing cholesterol lev-els, decreasing early arteriosclerosis, and pre-venting heart disease.3 The components of vita-min E include alpha-tocopherol (aT), alpha-tocotrienol (aT3), gamma-tocopherol (gT),

gamma-tocotrienol (gT3),beta-tocopherol (bT), delta-tocopherol (dT), and delta-tocotrienol (dT3). The vita-mer of greatest interest innutrition is the mostbiopotent one, alpha-toco-pherol.4

Two major extractionmethods conventionallyused to extract vitamin Ecomponents from rice branare Soxhlet extraction andsolvent extraction. Solventextraction is the most com-

Figure 1. Oil and vitaminE components extractedper gram of rice bran atdifferent temperatures(80, 110, 140°C) for 15minutes at a 3:1solvent:rice bran ratio.

This articlediscusses theassessment ofextractiontemperature andtime for themicrowaveextraction ofrice bran oil andvitamin E yield.

This articlerepresented theSouth CentralChapter as afinalist in theInternationalStudent PosterCompetition/UndergraduateLevel Divisionheld at the2004 ISPEAnnual Meetingin San Antonio.





Figure 3. Vitamin E components extracted per gram of rice branfor different times (5, 10, 15 minutes) at 110°C with a 3:1solvent:bran ratio.

monly used batch method that is applied to extraction oflipids from foods. It involves mixing of the substrate matrixwith the solvent at the extraction temperature for a predeter-mined extraction time. Soxhlet extraction is performed byrepeatedly washing the solid with an organic solvent, usuallyhexane or petroleum ether, in a glass apparatus duringreflux. Disadvantages of the solvent and Soxhlet proceduresinclude long extraction times, and large solvent volumesneeded.

Microwave-assisted extraction is a novel process that usesmicrowave energy to heat the solvents and the sample toincrease the mass transfer rate of the solutes from the samplematrix into the solvent. The combination of solvents and heatis expected to increase the extraction yield as compared toother methods.5 Because of the heat produced, microwave-assisted solvent extraction takes 10-30 minutes, whereas theother methods can take hours or days to complete. Theamount of solvent used in microwave extraction is alsoconsiderably less than the amount used in the other extrac-tion processes.6

Microwave-assisted extraction is a relatively new extrac-tion technique and has been successfully employed to extracttea polyphenols and tea caffeine from green tea leaves,7

piperine from black pepper,8 capsaivinoids from capsicum,9

phenolic compounds from grape seeds,10 puerarin from Radixpuerariae,11 and color pigments from paprika.12

The goal of this project is to use microwave extraction asmeans to effectively extract vitamin E components from ricebran. The objectives of this research were 1. to extract ricebran oil and constitutive vitamin E components from ricebran using microwave extraction and 2. to study the influenceof processing parameters, temperature and time, on thevitamin E yield.

Materials and MethodsRice bran was obtained by milling 11.4 kg samples of Cypressrice with a pilot scale rice mill located within the Biologicaland Agricultural Engineering Department at Louisiana StateUniversity. The pilot scale system of unit operations in order

consists of a paddy husker with separator, a rice whiteningmachine, a wet polishing machine, and a color sorter. Thefirst two unit operations of the mill were used to produce ricebran samples with a roll gap distance of 1.5 mm (husker) anda flow rate of 123 g/s (whitening machine). Samples were usedimmediately after milling.

Samples of 20 g freshly ground rice bran were prepared in3:1 isopropanol to bran ratios. Isopropanol (a polar com-pound) was used as the solvent because polar solvents wereproven to absorb microwave energy better than non-polarsolvents. The samples were placed in one of three pressurecontrolled Teflon vessels included in the Microwave Extrac-tion System along with a magnetic stirring rod. Once theextraction was complete, the samples were allowed to cool for10 minutes. A vacuum pump was used to filter the solvent andoil mixture from the rice bran and the volume of each samplewas recorded.

A 5 ml portion of each sample was stored in a sealed,opaque container and placed in a refrigerator to minimizeexposure to oxygen, light, and heat. Finally, the isopropanolwas evaporated from the sample and the remaining oil wasanalyzed for Vitamin E components by normal phase HighPressure Liquid Chromatography.13

Microwave SystemThe Microwave Extraction System chosen in the presentresearch allows for control of pressure, temperature, andenergy input. It contains up to six pressure controlled Teflonvessels which are placed on a rotating platform. A motordriven magnet is used to spin stirring rods inside each of thevessels and consequently mix the sample. Multimode irra-diation evenly distributes heat throughout the samples asthey rotate inside the cavity. The temperature inside themicrowave system is measured using a fiber optic sensor witha gallium arsenide crystal tip and a protective tube inserteddirectly into one of the three vessels.

Statistical AnalysisThe experiment was designed as a two-factor (time and




temperature) factorial treatment structure with three levelsfor each factor and two replications for each treatment com-bination. The experimental data was analyzed by a two-wayprocedure. Multiple comparison tests were performed todetermine the significant difference between treatments at P< 0.05.

Results and DiscussionThe influence of processing parameters, extraction tempera-ture and extraction time on the amount of oil and vitamin Ecomponents, alpha-tocopherol, alpha-tocotrienol, gamma-to-copherol, and gamma-tocotrienol extracted was assessed -Figures 1, 2, 3, and 4.

Influence of Temperatureon Oil and Vitamin E YieldOil and total vitamin E components extracted per gram of ricebran at different temperatures (80, 110, 140°C) for 15 min-utes at a 3:1 solvent: rice bran ratio were measured - Figure1. As temperature increased, oil yield and vitamin E yieldincreased. As temperature increased from 80°C to 110°C, oilyield increased by 0.02g (from 0.11g to 0.13g), an 18% in-crease. A 31% improvement (from 0.13g to 0.17g) occurredwhen the temperature was increased to 140ºC. Vitamin Eyield increased across this temperature range as well. From80ºC to 110ºC, 0.2mg (33%) more vitamin E was extracted. Amore significant change of 0.5 mg (63%) was observed whentemperature was increased from 110°C to 140°C.

Influence of Extraction Time on Vitamin E YieldThe influence of extraction time on the amount of vitamin Ecomponents, extracted per gram of rice bran at 80°C wasanalyzed. The mass of all four vitamin E components, includ-ing the target component alpha-tocopherol, showed minimalchanges as the extraction time increased from 5 to 15 minutesfor all temperatures. This lack of increase shows that maxi-mum yield can be achieved in only 5 minutes at this tempera-ture.

Vitamin E components extracted per gram of rice bran fordifferent times at 110°C is illustrated in Figure 3. Over the 10minute time range the vitamin E yield improved from 0.70

mg to 0.82 mg at 110ºC, an increase of 17%. This differs fromthe samples extracted at 80°C because of the noticeable risein vitamin E extracted at each time interval. However, therewas no significant difference in the amount of vitamin Eextracted between 10 and 15 minutes, indicating that a 10minute exposure time is ideal at this temperature.

Figure 4 shows the amounts of vitamin E componentsextracted per gram of rice bran for different times at 140°C.The vitamin E yield increased from 1.03 mg to 1.29 mg byincreasing extraction time from 5 to 15 minutes. This is a 25%increase compared to that observed at 110°C. The maximumyield of 1.29 mg was found to occur at 15 minutes. The meanyield at 140°C was determined to be 1.15 mg, the largest of thetemperatures tested. At lower temperatures, extraction timehad minimal affects on the vitamin E yield. However, asextraction times were increased at higher temperatures,greater vitamin E yields were produced. The higher amountof vitamin E extracted at higher temperatures may be ex-plained by a disruption of the physical structure of the sourcematerial which would increase diffusion rates and result in ayield increase.

Statistical AnalysisStatistical analysis of the results (Table A) showed a signifi-cant increase (P<0.05) in the amount of oil extracted at 140°Cas compared to the 110 and 80°C treatments (for 5 and 15minutes extraction time). Vitamin E yield increases signifi-cantly by increasing the temperature from 110 to 140°C forall extraction times. The 80 and 110°C treatments were notsignificantly different from each other.

Solvent Extraction versus Microwave ExtractionThe oil extraction yields obtained in this study were compa-rable to those found in the literature. Isopropanol extractedoil from rice bran was reported to be in the range of 16.3% and17.2% (% of rice bran).14 By microwave extraction, 10.40% oilwas extracted at 80°C for 5 min and 17.12% was extracted at140°C for 15 min. In terms of vitamin E yield, the minimumamount of tocopherols extracted by microwave extraction

Table A. Crude Oil and Vitamin E Extracted with Different Timesand Temperatures using a 3:1 solvent:bran ratio.

Time Temp. Crude Oil Vitamin E(min) (ºC) (g/g rice bran) (mg/g rice bran)

5 80 0.104±0.004a 0.624±0.002a

110 0.105±0.024a 0.697±0.191a

140 0.150±0.027b 1.029±0.108b

10 80 0.108±0.004a 0.686±0.012a

110 0.140±0.020a 0.762±0.006a

140 0.149±0.015a 1.135±0.100b

15 80 0.106±0.008a 0.636±0.021a

110 0.128±0.004a 0.820±0.027a

140 0.171±0.010b 1.294±0.159b

The results are expressed as g crude oil/g rice bran and mg Vitamin E/g ricebran. Significantly different values (P < 0.05) of oil and vitamin E in thesame column of each time are indicated by different letters a,b.




was 261.8 ppm at 80°C for 5 minutes and the maximum was453 ppm at 140°C and 15 minutes, as compared to the solventextracted 343 ppm tocopherols reported by Hu, et al.2 Theamount of tocotrienol extracted by microwave extractionvaried between 361 ppm (80°C and 5 minutes) and 841 ppm(140° for 15 min) by microwave extraction as compared toonly 265 ppm extracted by solvent extraction.2 These resultsshow that microwave extraction results in comparable oilyields and higher tocopherol and tocotrienol yields withrespect to conventional extraction methods.

ConclusionMicrowave extraction was used to successfully extract vita-min E from rice bran. Over a temperature range of 80ºC to140ºC, an increase of 104% vitamin E yield occurred. At lowertemperature, time had little to no effect on yield, but as thetemperature increased, so did the effect of extraction time onthe vitamin E yield. Optimum parameters for extraction ofvitamin E from rice bran were 140ºC for 15 minutes at a 3:1isopropanol to rice bran ratio. Microwave extraction wasfound superior to other methods for obtaining high yields oftocopherols and tocotrienols from rice bran.

References1. LSU Ag Center Research and Extension. 2002. Progress

Report.2. Hu, W., J. Wells, T. Shin, and J. Godber. 1996. Compari-

son of Isopropanol and Hexane for Extraction of VitaminE and Oryzanols from Stabilized Rice Bran.

3. Dunford, N. T. 2001. Health Benefits and Processing ofLipid-Based Nutritionals. Food Technology. 55(11): 38-43.

4. Combs, G. 1998. The Vitamins: Fundamental Aspects inNutrition and Health, 2nd Ed., Academic Press, London.

5. Nuchter, M., B. Ondruschka, W. Bonrath, and A. Gum.2004. Microwave Assisted Synthesis- A Critical Technol-ogy Overview.

6. Shah, S., R. C. Richter, and H. M. S. Kingston. 2002.Microwave-Assisted Organic Extraction and Evapora-tion: An Integrated Approach. LCGC North America.20(3): 280-286.

7. Pan X. J., G. G. Niu, and H. Z. Liu. 2003. Microwave-Assisted Extraction of Tea Polyphenols and Tea CaffeineFrom Green Tea Leaves. Chemical Engineering and Pro-cessing. 42(2): 129-133.

8. Raman, G., and V. G. Gaikar. 2002. Microwave-AssistedExtraction of Piperine From Piperi nigrum. Industrialand Engineering Chemistry Research. 41(10):2521-2528.

9. Williams, O. J., V. G. S. Raghavan, V. Orsat, and J. Dai.2002. Microwave-Assisted Extraction of Capsaicinoidsfrom Capsicum Species. Paper no. 026003 an ASAEMeeting Presentation. ASAE Annual International Meet-ing/ CIGR XVth World Congress. Chicago, IL, USA.

10. Hong, N., V. A. Yaylayan, G. S. V. Raghavan, J. R. J. Pare,and J. M .R. Belanger. 2001. Microwave-Assisted Extrac-tion of Phenolic Compounds From Grape Seed. NaturalProduct Letters. 15(3):197-204.

11. Guo, Z. K., Q. H. Jin, G. Q. Fan, Y. P. Duan, C. Qin, andM. J. Wen. 2001. Microwave-Assisted Extraction of Effec-tive Constituents From a Chinese Herbal Medicine RadixPuerariae. Analytica Chimica Acta. 436(1) 41-47.

12. Kiss, G. A. C., E. Forgacs, T. Cserhati, T. Mota, H. Morais,and A. Ramos. 2000. Optimisation of the Microwave-Assisted Extraction of Pigments From Paprika (Capsi-cum Annuum L.) Powders. Journal of Chromatography A889(1-2):41-49.

13. Zhimin Xu. 2002. Analysis of Tocopherols and Tocotrienols.In Current Protocols in Food Analytical Chemistry. JohnWiley & Sons, Inc., New York.

14. Proctor, A. and D. J. Bowen. 1996. Ambient-TemperatureExtraction of Rice Bran Oil with Hexane and Isopropanol.JAOCS. 73(6) 811-813.

About the AuthorsWilliam Hauser Duvernay is currentlyenrolled as a junior at Louisiana State Uni-versity in the Department of Biological andAgricultural Engineering. He is employed asa Chancellor’ Aid student worker in thebioprocess engineering lab. His interests in-clude pharmaceutical and environmentalengineering. Duvernay belongs to both ISPE

and AICHE and is currently starting a Student Chapter ofISPE at LSU. He is also the Ag Council Representative for theBiological Engineering Student Organization on campus.

John Michael Assad is a junior in theDepartment of Biological and AgriculturalEngineering at Louisiana State University.He works under a Chancellor’s Aid scholar-ship in the bioprocess engineering lab. Assadis a member of AICHE and ISPE. He success-fully obtained a grant from the LouisanaState University College of Agriculture to

work on microwave extraction of vitamin E. In general, he isinterested the biological applications of engineering andintends to pursue a masters degree in this field.

Dr. C. Sabliov is an Assistant Professor inthe Biological and Agricultural EngineeringDepartment at Louisiana State University.She came to LSU from North Carolina StateUniversity, where she received a double-major PhD in food science and biological andagricultural engineering and acquired oneyear of post-doctoral experience. Dr. Sabliov’s

areas of interest include equipment design, process develop-ment, image processing, and mathematical modeling appliedto the field of bioprocessing. She is actively involved inprofessional organizations such as ISPE, American Instituteof Chemical Engineers, and Institute of Biological Engineers.She is presently the Faculty Advisor for two student groups,the Society of Women Engineers and the Biological Engineer-



ing Student Organization and she is in the process of startingan ISPE Student Chapter at Louisiana State University. Asa new faculty member, Dr. Sabliov plans to build a bioengi-neering program internationally well known for its innova-tion, fundamental approach, and industrial applications.

Marybeth Lima is an Associate Professorin the Biological and Agricultural Engineer-ing Department at Louisiana State Univer-sity; she began her faculty career at LSU in1996. Her current responsibilities includeteaching undergraduate courses in biologi-cal engineering and engineering design.Lima’s research contributions have included

the broad areas of food processing and engineering education.She has authored or co-authored 22 refereed publicationsconcerning food and bioprocess engineering and engineeringeducation. Lima has performed extensive work in engineer-ing education and has pioneered the use of educationaltechniques such as service-learning and the integration ofcommunication and teaming skills across the biological engi-neering curriculum. Lima is passionately committed to im-proving the community through engineering-community part-nerships that bring the principles of engineering alive forchildren and tangibly use engineering to enhance democraticsociety.

Dr. Z. Xu is an Assistant Professor in theFood Science Department at Louisiana StateUniversity. His research interests are in theareas of food lipids and chemistry analysis.Dr. Xu received a PhD in 1998 from Louisi-ana State University, and has worked as aResearch Associate Post-doc at the Univer-sity of Massachusetts. He is a professional

member of Phi Tau Sigma Honorary Society, and Institute ofFood Technologists.

Please address all correspondence to:Dr. C. Sabliov

Assistant ProfessorDepartment of Biological and Agricultural Engineering

141 E.B. Doran Building Louisiana State University

Baton Rouge, Louisiana 70803-4505Tel: 1-225/578-1055Fax: 1-225/578-3492

e-mail: [email protected]

ISPE Professional Certification


ISPE Professional CertificationProgram: Raising the Image of thePharmaceutical Professionby Jerry Roth, P.E., ISPE Director of ProfessionalCertification

This articledescribes theChPP credentialand the nextmilestone in itsdevelopment.

Introduction

The pharmaceutical industry is facingsignificant issues in its quest to en-hance product safety and quality andimprove the health of the world popula-

tion. Some of these issues include regulatoryharmonization and compliance, business glo-balization, and facilitating innovation and effi-ciency in the research, product development,and manufacturing areas. Consequently, tech-nical pharmaceutical professionals are beingchallenged to seek new technologies and applythem in a risk-based approach to achieve qual-ity by design. Academia is tuning in to thepharmaceutical industry’s specific needs andadapting curricula to better prepare studentsfor making significant contributions to theiremployers’ mission. Government regulators areencouraging a risk-based approach in applyingnew innovative technologies.

Pharmaceutical professionals involved inthe design and operation of product develop-ment and manufacturing processes and facili-ties aspire to attaining greater strides in theindustry value chain. For industry to success-fully achieve real innovation, these profession-als need to merge science with engineering,and apply risk-based approaches to raise thebar on product quality and safety. The profes-sionals that possess a good science-based aca-demic background in addition to diverse indus-try experience and knowledge can make a sig-nificant contribution, if not lead the way, toindustry innovation.

ISPE supports the industry innovation trendand recognizes that the pharmaceutical profes-sional can make a difference in the speed andquality of innovation. ISPE is acting as a cata-lyst for innovation in a number of ways. First,

the Society is collaborating with academia todevelop curricula focused on industry needs. Byproviding a platform for teamwork betweenindustry and government, ISPE is promotingbetter understanding of regulatory require-ments and helping companies achieve a higherdegree of compliance. In addition, ISPE is imple-menting a professional certification program toraise the image of pharmaceutical profession-als, increase their competency, and providegreater value for their employers.

ObjectiveProfessional certification is the process of con-ferring a time-limited credential to a profes-sional once it is verified that she or he meets theestablished criteria. One type of credentialinginvolves certifying individuals who have beenfound to meet specified competence and quali-fication criteria. With its new certification pro-gram, ISPE’s fundamental objective is to im-prove competence, quality, and effectivenessfor its Members and other pharmaceutical in-dustry stakeholders.

ISPE’s Professional Certification Commis-sion (PCC) was established to create a certifica-tion program and develop credentials that ben-efit credential holders and the companies theirwork for. The first professional certificationcredential to be offered is the Chartered Phar-maceutical Professional (ChPP).

A Chartered Pharmaceutical Professional:

• Has command of the objectives and methodsof the following areas of knowledge: drugdevelopment process, quality systems, pro-duction systems, facilities and equipment,regulatory issues, materials management



ISPE Professional Certification


For more information, contact Jerry Roth,ISPE Director of Professional Certification at e-mail: [email protected].

and economics, and information data management andcontrol

• Delivers innovation by synthesizing and integrating di-verse input into productive and cost-effective, cross-func-tional approaches and risk-based solutions.

The next step in developing the ChPP credential is referredto as a job analysis: the process of identifying the knowledge,skills, and abilities of professionals in the pharmaceuticalindustry in order to determine standards of performance,experience, and knowledge required for the ChPP credential.The PCC will solicit proposals from qualified firms and awarda professional services contract for the job analysis by mid-August 2005. Then, the PCC will work closely with theseconsultants to select the most appropriate methodology forconducting the job analysis. Generally, the job analysis isconducted in two phases: Phase 1, Role Delineation andPhase 2, Assessment Specification.

Phase 1 - Role DelineationPhase 1, Role Delineation involves defining the key knowl-edge areas, skills, and experience that will provide standardsfor the ChPP credential-holder to meet. The PCC and the jobanalysis consultants will develop an interview questionnaireto be used in telephone interviews with a sampling of employ-ers in which targeted pharmaceutical professionals work.This sampling will be quite diverse, covering all aspects of thedrug development process (see Figure 1) including companiesthat manufacture drug products for the industry, as well asthose that supply services and equipment in support of thedrug development process.

An additional survey instrument will be prepared to gatherinput from professional practitioners (ISPE Members andnon-members). These data will determine the knowledge,skills and abilities required for competent performance intheir jobs. The information collected from telephone inter-views and the Internet-based survey will be analyzed bypsychometrichons, thereby creating a link between the knowl-edge/skill sets and the job tasks. This process validates thecredential requirements and serves to establish the type ofassessment required.

Phase 2 - Assessment SpecificationsPhase 2, Assessment Specifications identifies the methodsand criteria that provide the determining factors for confer-ring the credential. There are many types of assessmentcriteria (education, work experience, continuing education,oral interviews, written examination, etc.) which can becombined to define the qualifications of eligibility for assess-ment. Both the eligibility criteria and assessment require-ments will be the work product of the assessment specifica-tions phase. This document and supporting data is then usedin creating the appropriate assessment for the credential.

The professional certification commissioners encourageall industry professional practitioners to participate in thesurveys to achieve a good sampling throughout the industryand provide a statically sound data base.

Figure 1. Drug development process.

The ChPP credential is the first professional certification to be offeredto the global pharmaceutical industry covering research to manufacturing.

Not only can the new certification bring recognition and opportunity to the credentialholders, but it will provide great value to the industry.

Reprinted from

PHARMACEUTICAL ENGINEERING®

The Official Journal of ISPE

July/August 2005, Vol. 25 No. 4

Country Profile - United States

2 PHARMACEUTICAL ENGINEERING JULY/AUGUST 2005

INTRODUCTION

©Copyright ISPE 2005

Market Trendsand Challenging Times

According to the Pharmaceutical Research andManufacturers of America (PhRMA) organi-zation, pharmaceutical products are thecheapest weapon we have in our efforts to

reduce overall medical expenses. The value of newtherapeutics is immeasurable in terms of saving andimproving lives, and as a result, is often underesti-mated. Pharmaceutical therapeutics strengthen oureconomy by improving productivity and reducingworker absenteeism. On the other hand, US prescrip-tion drugs are the most expensive in the world, and theUS spends more money on drugs than the UK, France,Germany, Italy, Spain, and Japan combined. Europe

US Pharmaceutical Industry:Trends, Paradigms, and Challengesby Janice Abel, Director, Pharmaceutical Business, Invensys

spends 60% less per person than the US due to toughprice controls on drugs. US drug prices have risen atabout 10% per year which is more than three to fivetimes inflation, causing public concerns in the US.1 Inaddition, scandals are erupting on an almost dailybasis in the pharmaceutical industry – there are con-cerns about the efficacy and dangers of several best-selling drugs which have led to product withdrawalsand worse. US Government officials are asking fortougher oversight of the drug industry.

This article will provide a factual US market over-view that reviews and summarizes the market trendsand challenges that are being faced by the industry inthe US today. This information on the US pharmaceu-tical manufacturing market is being provided as an

Pharmaceutical Engineering would like to extend a specialthank you to the Chapter Champions who developed the information

for this US Country Profile.

Midwest RegionGreat Lakes Chapter - Chris Roerig, Manager Business Development, VAI Automation, Inc.Midwest Chapter - Sherry Hanafin, Director, Sourcing, Quintiles ConsultingSouth Central Chapter - Eric Unrau, Project Manager, CRB Consulting Engineers, Inc.

Northeast RegionBoston Area Chapter - Patti Charek, Marketing Manager, LinbeckDelaware Valley Chapter - Jon Hofmeister, Principal Engineer, Hofmeister EngineeringNew England Chapter - Jon Savona, Critical Environment Specialist, Accuspec, Inc.New Jersey Chapter - Matt Ferrier, Project Manager, Commissioning Agents, Inc....and a special thanks to Janice Abel for developing the introduction to the US Country Profile.

Southeast RegionCarolina-South Atlantic Chapter - Vince Miller, Senior Control Systems Engineer, Global Automation

PartnersChesapeake Bay Area Chapter - Jason Rifkin, President, Equilibrium Consulting, LLC

West RegionGreater Los Angeles Area Chapter - Michelle M. Gonzalez, Principal Corp. Eng., Amgen, Inc. and

Scott Tiedge, Vice President, Sales and Marketing, Alphabio Inc.Pacific Northwest Chapter - Todd Gill, Construction Manager Consultant, THG Consulting, LLCRocky Mountain Chapter - Mark Lutgen, Project Manager, Amgen, Inc.San Diego Chapter - Deborah Beetson, Project Executive, DPR Construction Inc., Ian Larson, Business

Development Executive, Rudolph & Sletten, Inc., and Phil Boncer, Principal, P. Boncer ConsultingSan Francisco/Bay Area Chapter - Greg Burg, Senior Project Manager, Genentech, Inc.


JULY/AUGUST 2005 PHARMACEUTICAL ENGINEERING 3

INTRODUCTION


introduction to the United States Country Profile,which has been developed jointly by ISPE’s US Chap-ters. For the purpose of this profile, the United Stateswas broken down into four geographical regions, andthe Chapters located in those regions collaborated onthe development of the “regional” profiles. The break-down is as follows:

Midwest RegionGreat Lakes Chapter

(Ohio, Indiana, Illinois, Michigan, Wisconsin, Kentucky)Web site: www.ispe.org/greatlakes

Midwest Chapter(Kansas, Missouri, Iowa, Nebraska, Minnesota)Web site: www.ispe.org/midwest

South Central Chapter(Texas, Oklahoma, Louisiana)Web site: www.ispe.org/southcentral

Northeast RegionBoston Area Chapter

(eastern Massachusetts, Maine, New Hampshire)Web site: www.ispe.org/boston

Delaware Valley Chapter(eastern Pennsylvania, Delaware, southern New Jersey)Web site: www.ispedvc.org

New England Chapter(excluding Boston area)Web site: www.ispe.org/newengland

New Jersey Chapter(New Jersey, New York, northeastern Pennsylvania)Web site: www.ispe.org/newjersey

Southeast RegionCarolina-South Atlantic Chapter

(North Carolina, South Carolina, Georgia, Florida,Alabama, Tennessee)Web site: www.ispe.org/carolina-southatlantic

Chesapeake Bay Area Chapter(Washington, D.C.; Maryland; northern Virginia)Web site: www.ispe.org/chesapeakebayarea

West RegionGreater Los Angeles Area Chapter

(Los Angeles, Orange, Ventura,Riverside Counties)Web site: www.ispe.org/greaterla

Pacific Northwest Chapter(Washington, Oregon)Web site: www.ispe.org/pacificnorthwest

Rocky Mountain Chapter(Colorado, Utah)Web site: www.ispe.org/rockymountain

San Diego Chapter(North to Orange County)Web site: www.ispe.org/sandiego

San Francisco/Bay Area Chapter(northern California)Web site: www.ispe.org/sanfrancisco

Pressures from regulatory agencies, the government,and public pressure to reduce the overall healthcarecosts are impacting the US pharmaceutical market.Even with all these setbacks, the US pharmaceuticalmarket is still an excellent performer when comparedto other industries.

The US pharmaceutical industry grew at a rate thatis estimated to be between 7% and 7.5%, compared toa global growth of approximately 8%. The US is thelargest market with 45% to 46% of the total globalmarket - Table A.

The pharmaceutical manufacturing industry is adynamic industry, one of the best to work in – with verylow turnover rates. Because of the global economicgrowth, demographics of an aging population, and theproduction and demand for newer and more special-ized therapeutics, the industry will continue to flour-ish and remain strong in the US.

Market SegmentsFrom a market perspective, the pharmaceutical manu-facturing market can be broken down into separatesub-segments that include generics, biopharma-ceuticals, contract manufacturers, and large pharma-ceutical companies (medical devices will not be cov-ered in this overview). In 2004, the sales growth ofbiotech and generics were greater than the total mar-ket growth - Figure 1. The largest sub-segment growthover the past year at 17% was in the biotech industry.The generics market segment grew by 10%. Genericsgrew at a slower pace than in previous years; however,

World Audited Market 2004 Sales % Global Sales % Growth(US$B) Year-over-Year

(Constant $)North America 248 47.8 7.8Europe (EU) 144 27.8 5.7Rest of Europe 9 1.8 12.4Japan 58 11.1 1.5Asia (excluding Japan), Africa, and Australia 40 7.7 13.0Latin America 19 3.8 13.4Total IMS Audited* $518 100% 7.1%

Table A. 2004 pharmaceutical sales by region. (Source: IMS Health)

US Pharmaceutical Industry: Trends, Paradigms, and Challenges

Continued on page 4.



INTRODUCTION


sales of generics accounted for 30% of the US sales in2004 of prescription drugs.

Overall, the industry delivered solidly with 31 NewMolecular Entities (NMEs) in 2004 which was up sig-nificantly from 2003. Eighty-two drugs have sales over$1 billion - defined as blockbusters. This was substan-tially more than the preceding year. Typically, newerblockbusters are produced from the new biotechnologydrugs and target specialty markets such as oncology.

The number one company ranked by healthcarerevenue in 2005 and 2004 was Pfizer. The number twocompany in 2004 by US sales revenue wasGlaxoSmithKline, followed by Johnson & Johnson.Also of note is that in the list of top 50 global pharma-ceutical companies, many biotechnology companiesare now included – such as Amgen, Genentech, Serono,and Genzyme.

Out of the top 50 global pharmaceutical companies(by healthcare revenue), 19 have corporate headquar-ters located in the US, 18 have headquarters in Eu-rope, 12 in Japan, and one in the Middle East.

Industry ChallengesThe following discussion will consider some of theindustry sub-segment challenges, trends, and para-digms for pharmaceuticals, biopharmaceuticals, ge-nerics, and contract manufacturers. Issues will focusmostly on research, clinical and manufacturing.

According to PhRMA, for each additional $1 spenton newer pharmaceuticals, $6.17 is saved in totalhealthcare spending. Even new medicines, which costmore, save an average of $111 when compared withother hospital and non-drug costs that are required.

Advances in prescription medicine haveimmense potential for improving lives. Re-search and development environments in the US areconstantly challenged with potential new scientificdiscoveries in the life sciences into real products withtherapeutic benefits for real people.

Expenditures for R&D in the US pharmaceuticalcompanies are shown in Table C. In 2003, the companywith the highest research and development expendi-tures was Pfizer with $7.13 billion - 37.8% more thanin 2002. Pfizer’s expenditure was almost twice that ofany of its major competitors. Sanofi-Aventis rankedsecond with $4.8 billion, followed by J&J with $4.68billion in 2003.

The fastest growing companies, as measured bytheir growth percentage in 2003 are shown in Table D.

Schwarz Pharma AG, a specialty pharmaceuticalcompany, generated the greatest increase in healthcareand consolidated revenue in 2003 with a 55.3% growth,followed closely by biotechnology company Amgen.

Rank Rank Company Revenue2003 2002 in 2003

($/thousands)1 1 Amgen Inc. 8,356,0002 2 Genentech Inc. 3,300,3273 3 Serono SA 2,018,6174 5 Chiron Corp. 1,766,3615 4 Genzyme Corp. 1,713,8716 7 MedImmune Inc. 1,054,3347 11 Gilead Sciences Inc. 867,8648 8 CSL Ltd. 856,2069 9 Cephalon Inc. 714,807

10 6/12 Biogen Idec Inc. 679,18311 10 Celltech Group Plc. 577,32812 13 Millennium Pharmaceuticals Inc. 433,68713 14 Genencor International Inc. 383,16214 21 Acambis Plc. 276,32615 18 Celgene Corp. 271,47516 26 Actelion Ltd. 228,58917 19 Berna Biotech Ltd. 191,22718 16 Nabi Biopharmaceuticals 176,57019 22 InterMune Inc. 154,13820 28 Enzon Pharmaceuticals Inc. 146,40621 24 Ligand Pharmaceuticals Inc. 141,14022 65 Neurocrine Biosciences Inc. 139,07823 36 Cangene Corp. 130,30924 29 Aeterna Laboratories Inc. 118,75625 30 ImClone Systems Inc. 80,830

Table B. The 25 global biotechnology companies ranked byhealthcare revenue.2 (Source: Med Ad News, September2004)

US Pharmaceutical Industry: Trends, Paradigms, and ChallengesContinued from page 3.

Figure 1. Dollar growth 2004 vs. 2003. (Source: IMSHealth)



INTRODUCTION


Research and Development (R&D)Challenges – Products in the Pipeline

To strengthen the pharmaceutical business and sus-tain growth, pharmaceutical companies are investingheavily in research and development. The US leads theworld in investment of biopharmaceutical R&D. TheUS accounted for 85% of the total global spending forR&D, growing at a rate of 7% in 2003.

US pharmaceutical companies spent more than $40billion on National Institutes of Health (NIH), andspent an additional $28 billion on research to developnew and better medicines. European research effortswere recently reduced due to government price con-trols and cost containment measures for therapeuticswith the US being the main beneficiary of this shift inR&D expenditures. As a result of the new controls,

according to the European Commission, theUS pharmaceutical research companies

should gain even more leadership in terms ofgenerating new medicines and will continue to

dominate the world market for prescription drugs.

Companies are supporting maturing pipelines withclinical programs in the areas of oncology, neuro-science, and diabetes and metabolic disease. Abbottdelivered seven drug candidates from discovery todevelopment in 2003 and has doubled the number ofquality drug candidates generated by its discoveryoperations compared with just a few years ago. Accord-ing to Med Ad News, Abbott’s goal is to get ninecandidates into clinical trials each year between 2003and 2007. New Drug Approvals (NDAs) started to leveloff in the beginning of this decade; however, NDAsramped up in 2004 to 92, and 2005 appears to beanother winning year with 31 NDAs from Januarythrough 17 May 2005.

The biggest challenge faced by companies in re-search is in getting effective therapeutics to market asquickly as possible, scaling the processes up rapidly,and reducing R&D costs. Streamlining R&D processesas well as manufacturing transfer and scale-up prac-tices will increase productivity. New innovative toolshave the potential to reduce development costs sub-stantially.

New tools are being developed for researchers tobetter evaluate new drug molecules as potential candi-



Rank Rank Company Healthcare2003 2002 R&D

20031 1 Pfizer Inc. 7,131,000,0002 16/5 Sanofi-Aventis 4,797,560,0003 3 Johnson & Johnson 4,684,000,0004 2 GlaxoSmithKline Plc. 4,560,773,1005 6 Novartis 3,756,000,0006 7 Roche 3,542,440,9107 4 AstraZeneca Plc. 3,451,000,0008 8 Merck & Co. 3,178,100,0009 10 Eli Lilly and Co. 2,350,200,000

10 9 Bristol-Myers Squibb Co. 2,279,000,00011 11 Wyeth 2,093,533,00012 12 Abbott Laboratories 1,733,472,00013 17 Amgen Inc. 1,655,400,00014 13 Schering-Plough Corp. 1,469,000,00015 14 Bayer 1,413,243,50016 15 Boehringer Ingelheim GmbH 1,330,644,00017 18 Takeda Chemical Industries Ltd. 1,127,902,69118 19 Schering AG 1,045,506,00019 20 Sankyo Co. 799,115,37020 43 Allergan Inc. 763,500,00021 21 Genentech Inc. 721,970,00022 25 Fujisawa Pharmaceutical Co 678,603,02323 22 Akzo Nobel NV 640,429,00024 23 Yamanouchi Pharmaceutical Co. 637,670,47525 24 Novo Nordisk AS 637,234,043

Table C. Top 25 companies ranked by healthcare R&Dexpenditure. (Source: Med Ad News, September 2004)

Rank Rank Company % Growth2003 2002 in 2003

1 --- Schwarz Pharma AG 55.3%2 2 Amgen Inc. 51.3%3 16 Pfizer Inc. 39.6%4 15 Chiron Corp. 38.4%5 12 Serono SA 31.3%6 6 Teva Pharmaceutical Industries Ltd. 30.1%7 22 Genzyme Corp. 28.9%8 3 Genentech Inc. 27.7%9 5 Allergan Inc. 26.7%

10 --- Mylan Laboratories Inc./ 20.8%King Pharmaceuticals Inc.

11 8 Procter & Gamble Co. 20.6%12 1 Forest Laboratories Inc. 19.3%13 --- Watson Pharmaceuticals Inc. 19.2%14 37 Novartis 19.1%15 41 Bristol-Myers Squibb Co. 15.4%16 13 Johnson & Johnson 15.3%17 47 Eli Lilly and Co. 13.6%18 20 Alcon Inc. 13.2%19 22 Abbott Laboratories 11.3%20 21 Bausch & Lomb Inc. 11.2%21 18 Baxter International Inc. 9.9%22 33 Wyeth 8.7%23 25 Eisai Co. 8.0%

Table D. Fastest growing companies: growth in healthcarerevenue. (Source: Med Ad News, September 2004)



INTRODUCTION


dates. For example, early toxicity identification andelimination methods will increase efficiencies, andtherefore, reduce costs so that energies can be focusedon developing more promising biologics (non-toxic).Tools that eliminate toxic compounds and drug safetycan be used early in the research process to improvepatient safety while reducing costs associated withunproductive initiatives. Previously, animal modelsand other laboratory techniques were used, but todaythe focus is on computer-based predictive models us-ing biomarkers (quantitative measures of biologicaleffects that provide informative links between thedrug’s action and clinical effectiveness), knowledgemanagement, and applying other new technologies toevaluate the efficacy.

Pharmaceutical Challenges –Increase Efficiencies and Reduce CostsTraditional pharmaceutical companies need to con-solidate resources and implement global standards

and practices worldwide to improve productiv-ity and efficiency. Larger companies are in theprocess of integrating functions that were previouslyseparate, such as pharmaceutical research and devel-opment, and manufacturing operations into a singleglobal structure with common goals and priorities.

The cost to bring a drug to market is estimated to bewell more than $800 million to more than $1 billionwith variations occurring from indication to indicationand company to company. One fact is obvious, thenumbers are very high. If costs keep rising, the possi-bilities for better-targeted medicines aimed at smallerpatient groups will not be readily investigated becauseinvestors will not fund indications that will not beprofitable. Traditional pharmaceutical companies willcontinue to invest in ways to operate more efficientlyand compete on a global basis. Mergers and acquisi-tions will continue, but partnerships will be even moreprevalent.

Generic Challenges – Competition,Biogenerics, and Reducing Costs

Under the Hatch-Waxman law – which is the DrugPrice Competition and Patent Term Restoration Act of1984 – generic drugs are expedited by allowing theproducts to be approved without clinical trials, basedon the safety and effectiveness data developed by theoriginal innovator. In the US, the first company to filestatus for generic versions of drugs that have gone off-patent, has a 180-day period of exclusivity, meaningthat only one generic firm, the first to file, can manu-facture the product for the first six months after thepatent expires. After this period, the drug price gener-ally falls by as much as 80%. The firm must makeprofits immediately. In the US, the added challengeresults from high labor costs compared to other partsof the world, making it difficult to compete globallyafter the six month exclusivity period. The challenge isto manufacture products as efficiently as possiblewhile maintaining the highest possible quality.

The generic market share is still expanding quiterapidly at a rate of 14% annually. However, the compe-tition for generics is increasing and speed to market isbecoming more critical than ever. The first company tolaunch a generic version of the product tends to cap-ture from 60% to 80% of the market. Globally, govern-ments, health management companies, and insurancecompanies are encouraging the use of generics to cuthealthcare costs. The generic market will continue togrow and help to reduce prescription drug costs in theUS.

Figure 3. The generics share of the US prescription marketcontinues to increase. (Source: PhRMA, New Medicines,New Hope)


Figure 2. New Drug Approvals (NDAs). (Source PhRMAand Drugs.com)



INTRODUCTION


Biotechnology Challenges – ProductDevelopment, Clinical Trials,Biogenerics, and Regulatory

RequirementsThirteen of the 67 blockbuster drugs with annual salesmore than $1 billion are biopharmaceuticals.2 Biotechproducts accounted for approximately 27% of the ac-tive research and development pipeline and 10% ofglobal sales in 2004. The US biotech industry is grow-ing at a rapid pace – 17% for 2003 to 2004 – becauseresearch and development efforts attributed to strongUS patent protection laws.

According to Cutting Edge Information, a businessinformation company, conservative estimates predictthat biogenerics will command more than $12 billion ofthe drug market by 2010.

A recent survey described in Biopharm Interna-tional asked biopharmaceutical companies why thecost of development of biologics was so expensive. Thetop three answers were overcoming the technical chal-lenges in product development, costs of clinical trials,and regulatory requirements. The survey also askedthis same group how they would reduce costs whilemaintaining safety, quality, and efficacy of the newbiopharmaceutical medicines. Two prominent strate-gies ranked highest: 1. using ‘best practice skills tostreamline product development to reduce time andcosts,’ and 2. developing ‘new process technologies thatincrease productivity.’ Thus, utilizing know-how tolearn from past experience and utilizing innovativetechnologies and techniques seem to be the overallstrategy for increasing bottom-line profits.

One major challenge to the biogeneric growth in thebiotechnology industry is the outcome of the currentdebate for establishing the equivalency of biogenerics.Should a law be enacted that is similar to the Hatch-Waxman law for the generic manufacturing of biologicsupon expiration of a patent? A new law would establishthe criteria for bioequivalence and could affect theindustry immensely since many of the biologics beingsold today are already off-patent or will be in the nearfuture. Proponents believe that biogenerics will reducethe cost of a biological drug, and opponents believe thatit is more difficult if not impossible to establishbioequivalence for biogenetics.

According to Arthur Levinson, Ph.D., CEO,Genentech, “Genentech does not believe that the tech-

nology currently exists to prove a genericbiotechnology product safe and effective out-

side of the new drug application and biologicslicense application process. Unlike traditional

generics, we believe that differing cell lines and

manufacturing processes mean that different manu-facturers will make different protein products that arenot substitutable.”

In the meantime, Europe has already started manu-facturing and selling biogenerics.

Contract Pharma Provider Challenges -Staffing, Commoditization, and

Reducing CostsPharmaceutical companies will continue to rely onboth research, clinical, and manufacturing contractorsenabling them to focus resources and investments oncores specialties. It is now possible to outsource virtu-ally every function involved in bringing a drug tomarket and pharmaceutical companies are starting tolook at outsourcing as a long-term strategic measurerather than a temporary measure. As manufacturersfocus on the process, the manufacturing contract busi-ness may slow a little, but research and clinical trialoutsourcing will increase in an effort to reduce costs.Some of the challenges the contract provider facesinclude reducing costs, meeting GMPs as well as otherglobal standards for quality and regulatory compli-ance, quick scale-up of processes, staffing, andcommoditization. In the US, contract manufacturingis a $30 billion business. Pharmaceutical contractmanufacturing is estimated to grow to $48 billion by2008.

Previously, the pharmaceutical industry utilizedContract Resource Organizations (CROs) to get addi-tional manpower when needed; however, contract ser-



Figure 4. Top reasons for outsourcing. (Source: ContractPharma, First Annual Outsourcing Survey 2005)



INTRODUCTION


vices are becoming part of traditional pharmaceuticalcompanies’ strategy today. In a recent survey con-ducted by Contract Pharma, 68% of the pharmaceuti-cal/biopharmaceutical outsourcers would describe theirrelationships with the contract manufacturer as apartnership with 54% stating that the reason foroutsourcing was strategic. The number one reason foroutsourcing was to focus on their core competencies(strategic), while the second reason cited was due to atemporary lack of capacity (tactical). When determin-ing the outsource manufacturer, the number one rea-son cited was GMPs. This was followed by technologyand third was (lowest) cost as the reason for choosingto outsource. Still, pharmaceutical industry customersare putting substantial pressure on contract providersto reduce their costs. Outsourcing has grown substan-tially over the past decade and has helped drug compa-nies cut capital investment while preventing capacitybottlenecks, enabling the industry to focus on corecompetencies.

The US FDA, Regulatory Challenges,Government, and Legal Challenges

Patent ChallengesThese are very challenging times for the US pharma-ceutical industry with an outcome that will affect us allglobally. The generics business will continue to growand subsequently drive down prescription drug costs.Generic companies continue to debate pharmaceuticalpatents, and some branded pharmaceutical compa-nies face a substantial amount of patent expirations.The industry continues to pour money into researchand development, but only a few companies have beensuccessful in filling their pipelines with innovativedrugs and bringing new chemical entities that couldturn into successful drugs to market. Even innovativebiotechnology products will be impacted by patentexpirations if the government establishes bioequiva-lence for generic biotechnology – which may happen inthe next year or two.

Internationally, the US is working to strengthenthe intellectual property protection of medicines. TheWorld Trade Organization’s (WTO) Trade-Relatedaspects of Intellectual Property (TRIPS) agreementestablished minimum international obligations forthe patent protection of medicines. Additional workwill continue in this area.

When ranking policy actions that would be the mostbeneficial in providing innovation incentives for intel-lectual property, two-thirds of executives polled byErnst & Young say the most beneficial policy is appro-priate market exclusivity for branded products.

Some of the patent challenges include theinterpretation of the Food and Drug Adminis-tration exemption for studies “reasonably related” to afuture drug application, and many of these issues arebeing debated as we write. The current US patentsystem provides the incentive necessary to continue thedevelopment of innovative and valuable research bygranting the inventor exclusive rights to control the newproduct or technology. Patent concerns and legal de-bates will continue.

Food, Drug, and Cosmetic ActThe US Country Profile would not be complete withouta discussion on the US Food and Drug Administration(FDA). FDA regulations are known around the worldbecause any product that makes a medical claim andis shipped into the US must comply with the applicableCode of Federal Regulations (CFR). Since almost halfof the therapeutics sold in the world are sold to theUnited States consumer, meeting US regulations be-comes a priority for anyone interested in selling prod-ucts to the US market. Additionally today, there is alsoan important initiative being undertaken to harmo-nize global regulations. The US FDA is workingwith other countries to accomplish this harmonization– which may take several years.

The FDA began when Congress enacted the FederalFood, Drug, and Cosmetic Act in 1937, which requiredcompanies to prove the safety of new drugs beforeputting them on the market. Cosmetics and therapeu-tic devices also were added to this Act, which has beenupdated to improve consumer protection. Over theyears, Congress has continued to give the FDA newresponsibilities. As a result, the FDA has an enormousrange of responsibilities that include:

• Reviewing labeling, animal and human testing toassure product safety and efficacy of new products.The FDA tracks how they are manufactured andresponds to reports of problems or newly identifiedrisks.

• Ensuring the safety of marketed products by in-specting domestic and foreign manufacturers.

• Monitoring for risks using a scientific approach toensure that regulatory decisions are sound andassess risk. The FDA initiates corrective actionsand legal actions for dealing with problems.

Risk-Based FocusThe FDA uses science-based, efficient risk manage-ment in the following regulatory activities so that theAgency’s limited resources can provide the most health




INTRODUCTION



promotion and protection at the least cost for thepublic:

• improve health through better information• enable consumers to make smarter decisions by

getting them better information to weigh the ben-efits and risks of FDA-regulated products

• improve patient and consumer safety• seek continuous improvements in patient and con-

sumer safety by reducing risks associated withFDA-regulated products

• protect the USA from terrorism• strengthen the FDA’s capability to identify, prepare

for, and respond to terrorist threats and incidents• develop more effective regulation through a stron-

ger workforce• ensure a world-class professional workforce, effec-

tive and efficient operations, and adequate resourcesto accomplish the Agency’s mission

• establish science-based policies and standards• issue final “Aseptic Processing Guidance”• issue Process Analytical Technology (PAT) guid-

ance for the pharmaceutical industry• conduct comparability studies• facilitate integrated quality systems orientation• facilitate international cooperation (ICH, European

Agencies)• protect the public health• 21 CFR Part 11 and Predicate Rules

Risk AssessmentRisk assessment is being used by the FDA and indus-try to identify, assess, prioritize, mitigate, and monitorthe likelihood of the occurrence and impact of risks toproduct quality, patient safety, and record integrity.This approach is similar to the FDA’s approach to therisk-based model that is being used to determineinspection sites (see risk-based model).

cGMPS for the 21st Century InitiativeThe cGMPs for the 21st Century is an FDA initiativethat was started in 2002 intended to modernize FDA’sregulation of pharmaceutical quality for veterinaryand human drugs and select human biological prod-ucts such as vaccines.

Objectives of the 21st Century Initiative• encourage the early adoption of new tech-

nological advances in the pharmaceuticalindustry

• facilitate industry application of modern qualitymanagement techniques (quality systems) to allaspects of production and QC

• encourage implementation of risk-based approachesthat focus both industry and Agency attention oncritical issues

• ensure that regulatory review, compliance, andinspection policies are based on state-of-the-artpharmaceutical science

• enhance the consistency and coordination of FDA’sdrug quality regulatory program

• provide possible guiding principles for the restruc-turing of the FDA

As part of the FDA’s 21st Century Initiative, one of therecent changes made is the FDA’s acceptance of usingnew technologies and the drive to increase efficienciesas part of the PAT Initiative. The final PAT Guidelinewas issued by the Agency in September 2004. ThisGuideline is particularly exciting. Many companieshave implemented or are implementing PAT typeprojects or solutions – some even prior to the FDAGuideline.

Process Analytical Technology (PAT)FDA’s focus on PAT is encouraging industry to investin new technologies and improve product developmentand manufacturing efficiencies in terms of using moreautomation, advanced controls and optimization, newon-line sensors, and other exciting new technologies.Some of these technologies are already being used inother industries, but due to regulatory constraints,many of these techniques and technologies were notapplied in the pharmaceutical industry. PAT shouldhave a major impact on the industry in terms ofinnovative methodologies and technologies being usedto increase efficiencies in the development and manu-facturing areas. As the industry begins to see andmeasure the results of these initiatives deploymentefforts for PAT solutions will increase dramatically.PAT should enable industry to reduce time to market,reduce the regulatory burden, increase efficiencies,decrease costs, and fundamentally change the wayindustry measures and tracks business performance.

Risk Ranking ModelAs regulators and consumers intensify their scrutinyand global competition accelerates, pharmaceuticalcompanies are being forced to adopt ever-more exact-ing measures to evaluate and manage risks. According




INTRODUCTION



to ARC’s John Blanchard, “The FDA’s new risk-basedmodel has arrived, giving the industry additional in-centive to develop enhanced process understanding,improve process control, and implement more exten-sive risk mitigation techniques.” The new model takesinto account potential product risks, potential process-ing risks, as well as risks the facility might pose. Theresults of the model are expected to be used to deter-mine 50% of the sites to be inspected in each district in2005.

Bioterrorism ActThe events of 11 September 2001 reinforced the needto enhance the security of the United States. Congressresponded by passing the Public Health Security andBioterrorism Preparedness and Response Act (theBioterrorism Act), which President Bush signed intolaw 12 June 2002.

The FDA and Customs and Border Protection (CBP)Compliance Policy Guide (CPG) describes the strategyfor maintaining an uninterrupted flow of food and drugimports while improving their safety in accordancewith the Public Health Security and BioterrorismPreparedness and Response Act of 2002.The purposeof the Bioterrorism Act is to protect the health andsafety from an intended or actual terrorist attack onthe nation’s food and drug supply. Provisions of theBioterrorism Act include: national preparedness forbioterrorism and other public health emergencies,enhancing controls on dangerous biological agents andtoxins, protecting safety and security of the food anddrug supply, drinking water security and safety, andadditional provisions.

The Act requests that the FDA receive aprior notification of all human and animalfood, drinks, and dietary supplements imported to theUS. It addresses threat assessments; technologies andprocedures for securing food processing and manufac-turing facilities and modes of transportation; responseand notification procedures; and risk communicationsto the public. In order to protect the drug supply, theact requires yearly registration of all imported drugs.In addition to being registered yearly with a statementprovided at the time of importation, all imported drugsmust include the following information: that the ‘ar-ticle is intended to be processed into a drug, biologicalproduct, device, food, food additive, color additive, ordietary supplement that will be exported under thePublic Health Safety Act.’ The statement also requiresname and place of business, US agent, name of im-porter, name of person who imports. Additional re-quirements include a certificate of analysis to identifythe article, and information on the level, potency,identity, strength, quality, and purity of the drug.

Radiofrequency Identification (RFID)ChallengesRFID is a state-of-the-art technology that uses elec-tronic tags on product packaging to allow manufactur-ers and distributors to more precisely keep track ofdrug products as they move through the supply chain.It is similar to the technology used for tollbooth andfuel purchasing passes.

Tracking and traceability will become more impor-tant as counterfeit drugs become more widespread,and the apprehension of counterfeit drugs will becomemore of an issue in this industry as additional cases arereported. There were 22 cases of counterfeit drugsreported in 2003 in the US - Figure 5. The industry willramp up security and tracking systems rapidly in thenext few years. Bioterrorism scares also have indi-cated a need for additional regulatory and company-wide security measures being implemented.

The FDA published a Compliance Policy Guide(CPG) on RFID in an effort to improve the safety andsecurity of the nation’s drug supply through the use ofRFID technology.

The goal of this CPG is to facilitate the performanceof RFID studies and allow industry to gain experiencewith the use of RFID. The FDA believes that use ofRFID technology is critical to ensuring the long-termsafety and integrity of the US drug supply. This Guideis part of the FDA’s commitment to promote the use ofRFID by the US drug supply chain by 2007.

Figure 5. FDA open investigations for counterfeiting.(Source: Combating Counterfeit Drugs - A Report of theUS FDA, February 2004)



INTRODUCTION



Re-Importation ChallengesPublic pressure to control US pharmaceutical costshas recently focused the US government on re-impor-tation proposals to reduce the cost of prescriptiondrugs. The products may have been manufactured inthe US or elsewhere; however, the current law does notallow for re-importation due to realistic health andsafety issues. Some of the issues revolve around risksto the patient in terms of adulteration, contamination,or even counterfeit (or efficacy). Additionally, prescrip-tion drugs currently purchased over the Internet canpose the same threats. Although it appears to bepolitically appealing to amend our current laws, fed-eral laws on drug imports and reimports reflect welldocumented concerns about the safety of importeddrugs.

ConclusionThese are very challenging times for the US pharma-ceutical industry. The complex environment is beingdriven by increased global market competition andconsumer pressures to reduce healthcare costs, highlabor costs, regulatory changes, and an increased needfor life-saving and life-altering therapeutics. Pharma-ceutical companies must maximize their portfolios’effectiveness while at the same time streamline opera-tions.

The pharmaceutical industry will continue to lookfor ways to increase efficiencies and improve the bot-tom line. There does not seem to be any one action thatwould lead to overall healthcare cost reductions. Dis-cussions on how to lower costs have revolved aroundimproving the use of innovative technologies (see PAT),sharing know-how and knowledge bases, reducing theregulatory burden, and developing new therapeutics.Productivity enhancers include everything from in-creasing manufacturing performance to having bettertraining for the workforce. Some state governmentsare trying to educate the local workforce to enhanceproductivity. For example, the state of North Carolinahas established an ambitious biomanufacturing train-ing program and the state of Massachusetts also hasfunded biotechnology training programs, as have sev-eral other states in the US.

Thoughts that will continue to preoccupy the indus-try include questions on what the US pharmaceuticalmarket will be like in the future. The only thing that

that we can write with certainty is that the USpharmaceutical industry will continue to be

challenged over the next decade and beyond.

References1. www.PHRMA.org., New Medicines, New Hope.

2. IMS Health, IMS National Sales Perspectives,2005.

3. PhRMA, Pharmaceutical Industry Profile 2005.

4. Med Ad News, “Top 50 Companies”, September2004.

5. www.fda.gov

6. www.Forbes.com, “Drugmaker 1Q Results DifferWildly”, 19 April 2005.

7. www.Forbes.com, “Supreme Court Reviews DrugPatents”, 21 April 2005.

8. www.cuttingedge.com

9. Biopharmaceutical International, Developing theFuture, Big Questions, Multiple Answers, April2005.

10. Pharmaceutical Formulations and Quality, 2004.

11. Contract Pharma, “First Annual Outsourcing Sur-vey”, May 2005.

12. Blanchard, John, “Ready or Not, FDA’s Risk Rank-ing Model will Determine cGMP Inspections”, ARCAdvisory Group Report, 2005.

13. Combating Counterfeit Drugs; A Report of the USFood and Drug Administration, February 2004.

14. http://www.biopharminternational.com/biopharm/article/articleDetail.jsp?id=157252.



INTRODUCTION


About the US Food and Drug Administration (FDA)

Organization

The FDA’s top official, LesterM. Crawford, DVM PhD, was

appointed Commissioner of theFDA on 14 February 2005. Com-missioner Crawford’s job is to en-sure that the Agency carries outits mission of protecting and ad-vancing the public health.

The US FDA is responsible forprotecting consumers by oversee-ing the safety and efficacy of allproducts with medical claims be-fore they are sold in the US. TheFDA regulates complex and sophis-ticated drugs, medical products, andother consumable products whileprotecting consumers. The FDA’smission is quoted below:

“FDA is responsible for protectingthe public health by assuring thesafety, efficacy, and security of hu-man and veterinary drugs, biologi-cal products, medical devices, USAfood supply, cosmetics, and prod-ucts that emit radiation. FDA isalso responsible for advancing thepublic health by helping to speedinnovations that make medicinesand food more effective, safer, andmore affordable; and helping thepublic get the accurate, science-based information they need to usemedicines and food to improve theirhealth.” (www.fda.gov)

At the heart of all FDA’s regulatoryactivities is a judgment aboutwhether a new product’s benefits tousers will outweigh its risks. TheFDA uses science-based, efficientrisk management to allow theAgency to protect the public at thelowest cost. While no regulatedproduct is totally risk-free, it isimportant that the FDA evaluateeach product individually and con-sider the risk against the potentialbenefit — especially for productsused to treat serious, life-threaten-ing conditions. In addition to pre-scription therapeutics, the FDA is

responsible for regulating food, foodadditives, non-prescription phar-maceutical products (anything thatmakes a medical claim or is used inthe prevention of a disease),biologics (vaccines, blood products,biotechnology products and genetherapy products), and medical de-vices (thermometers to pacemak-ers and dialysis machines).

Additionally, the FDA regulatesdrugs and devices used for ani-mals, and veterinary medical de-vices for both pets and animalsthat produce food. Before manu-facturers can market animal drugs(including drugs used in animalfeeds), all of the regulated prod-ucts must gain FDA approval byproviding proof of their safety andeffectiveness. The FDA monitorscosmetic products to be sure thatthey are safe and properly labeled.The Agency also oversees productlabeling for food, drugs, and medi-cal devices used by health profes-sionals to ensure that the productshave the information needed forproper use.

Most of the FDA’s budget goestoward paying its highly skilledand internationally respected workforce. The FDA employs some 9,000science and public health profes-sionals — including biologists,chemists, physicians, biomedicalengineers, pharmacologists, vet-erinarians, toxicologists, and spe-cialists in public health educationand communication.

StructureThe FDA is a US Agency within theDepartment of Health and HumanServices and is structured into eightcenters/offices that include:

Center for Biologics Evalua-tion and Research (CBER) -CBER regulates biological prod-ucts for disease prevention andtreatment that are inherently morecomplex than chemically synthe-

sized pharmaceuticals in-cluding:

• blood and blood products(plasma), blood-derived pro-teins including clotting factorsfor hemophilia, tests used toscreen blood donors and devicesused to make blood products

• vaccines and allergenic prod-ucts

• protein-based drugs, such asmonoclonal antibodies andcytokines that stimulate the im-mune system to fight cancer,and enzyme therapies that stopheart attacks

CBER also plays an integral rolein several initiatives for protect-ing the US against bioterrorism.

Center for Drug Evaluationand Research (CDER) - CDERpromotes and protects the publichealth by assuring that all pre-scription and over-the-counterdrugs are safe and effective. CDERevaluates all new drugs before theyare sold, and serves as a consumerwatchdog for drugs on the marketto be sure they continue to meetthe highest standards.

Center for Devices and Radio-logical Health (CDHR) - CDHRensures that new medical devices(pacemakers, contact lenses, hear-ing aids, etc.) are safe and effectivebefore they are marketed.

Center for Food Safety andApplied Nutrition (CFSAN) -CFSAN has one of the Agency’sbiggest jobs: it is responsible forthe safety of 80% of all food con-sumed in the US except for meat,poultry, and some egg products,which are regulated by the USDepartment of Agriculture.

Center for Veterinary Medi-cine (CVM) - CVM affects mil-



INTRODUCTION


lions of consumers by helping toassure that animal feed productsare safe. CVM also evaluates thesafety and effectiveness of drugsused to treat pets and livestock.

National Center for Toxicologi-cal Research (NCTR) - The mis-sion of NCTR is to conduct peer-reviewed scientific research thatsupports and anticipates the FDA’scurrent and future regulatoryneeds. This involves fundamentaland applied research specificallydesigned to define biologicalmechanisms of action underlyingthe toxicity of products regulatedby the FDA. This research is aimedat understanding critical biologi-cal events in the expression of tox-icity and at developing methods toimprove assessment of human ex-posure, susceptibility, and risk.

Office of the Commissioner(OC) - OC is responsible for theefficient and effective implemen-tation of the FDA mission. It ismade up of several componentsand offices (Good Clinical Practice(GCP), Office of International Pro-grams, Office of Orphan ProductsDevelopment, etc.)

Office of Regulatory Affairs(ORA) - This is the lead office forall field activities of the FDA. ORAregulates the business establish-ments that annually produce,warehouse, import, and transportconsumer goods. Highly trainedstaff ensures the implementationof the FDA’s high public healthstandards such as GxPs.

The FDA’s Web site is a great re-source offering a wealth of infor-mation about all of the programs

and product areas. Foradditional information

about the FDA, visitwww.fda.gov.

About the FDAUS Country Profile Regional Articles

Table of Contents

Midwest RegionMidwest Overview ............................................................ 14

Issues and Opportunities in Contract Manufacturing ............. 15

Trends in the Animal Health Industry .................................. 17

Northeast RegionNortheast Overview .......................................................... 19

History of Pharmaceutical and Biopharmaceutical Companiesin New England .............................................................. 20

Southeast RegionMid-Atlantic and Southeast United States - Pharmaceutical andBiotechnology Manufacturing Growth ................................ 21

Interview with Dr. Leslie Alexandre, President and CEO,North Carolina Biotechnology Center ................................. 24

West RegionWest Overview................................................................. 26

The Life Sciences Industry in Washington State ................... 27

The Life Sciences Industry in the Rocky Mountain Region ..... 28

A Brief History of Biotechnology in Northern California .......... 30

San Diego Regional Profile ................................................. 32



MIDWEST REGION


Midwest Overview

The Midwest Region encompasses the ISPE GreatLakes (Michigan, Illinois, Indiana, Ohio, Wiscon-

sin), Midwest (Kansas, Nebraska, North Dakota, SouthDakota, Minnesota, Iowa, Missouri), and South Cen-tral (Oklahoma, Texas, Louisiana) Chapters. The com-bined ISPE membership in this region is 2247, orapproximately 20% of the total ISPE membership inthe 15 state geography.

The companies with manufacturing capacity repre-sented in this region include: Abbott, Alcon, Allergan,Baxter, Bayer, Ben Venue Labs, Bioport, Boston Sci-entific, Cambrex, Cardinal Health, Centocor, ColgateOral Pharmaceuticals, Cook Group, Depuy, Dow, DSMPharmaceutical, DPT Laboratories, Eli Lilly, Ethicon,Fort Dodge, Gerber, Guidant, Hospira, King Pharma-ceutical, KV Pharmaceutical, Mallinckrodt, MeadJohnson, Medtronic, 3M, Nestle, Novartis, Perrigo,Pfizer, Protein Design Labs, Retractable Technolo-gies, Roche Diagnostics, Ross Products, Roxane Labs,Serologics, Shaklee Technica, Solvay Pharmaceutical,

Sovereign Pharmaceuticals, Stryker, TanoxBiosystems, Vetmedica, Zimmer, and ZRBBehring.

The breakdown of firms by category is as follows:

Visit the following Web sites to read moreon the Midwest Region:

Biotechnology Industry ReviewInterview with Tom Kowalski, President, TexasHealthcare and Bioscience Institute (THBI)by Eric Unrau, Project Manager, CRB Consulting Engrs, Inc.www.ispe.org/countryprofile

JETT Consortiumwww.ispe.org/countryprofile

States Compete to Capture Bioscience Marketwww.ispe.org/countryprofile

Overview of the ISPE Great Lakes Chapterwww.ispe.org/greatlakes or www.ispe.org/countryprofile

Overview of the ISPE Midwest Chapterwww.ispe.org/midwest or www.ispe.org/countryprofile

Overview of the ISPE South Central Chapterwww.ispe.org/southcentral or www.ispe.org/countryprofile



MIDWEST REGION


Overview

T he market for ContractManufacturing Organiza-tions (CMOs) is estimated

at $15 billion per year with anannual growth rate of 20%. Why somany opportunities and growth incontract manufacturing? Cur-rently, the pharmaceutical indus-try outsources 30% of R&D and40% of all manufacturing activi-ties. Contract manufacturingranges from simple packaging ac-tivities to the application of lead-ing edge technologies - Figure 1.

Out of this $15 billion market,the largest percentage of contractmanufacturing occurs in pharma-ceutical chemicals, followed bysolid dosage formulations. Pack-aging and biologics round out themarket, but at significantlysmaller capacities versus the otherareas.

At face value, the concept ofcontract manufacturing is straightforward. However, typical relation-ships between the contract manu-facturers and their customers(pharmaceutical companies for ex-ample) can be complex and multi-dimensional, taking years to de-velop into a full partnership insome cases. This is also a marketthat contains multiple segments,as shown in Figure 1, that are influx as market pressures changecontinuously.

The opportunities for contractmanufacturers in outsourcingcome from four major areas:

• relative economics –cost savings throughleveraging resources

Issues and Opportunities in ContractManufacturingby Eric Unrau, Project Manager, CRB Consulting Engineers, Inc.

• technology and skill availabil-ity – delivering value throughspecialized services

• start-up operations – creatinga virtual company

• increasing pressure to increaseefficiency and returns – marketexpectations

For DPT Labs, headquartered inSan Antonio, Texas, the approachto the market needed to providedifferentiation from other poten-tial low cost competitor CMOs.DPT has focused on differentia-tion through providing specializedskills and products their competi-tors are unable to match. This cre-ates more opportunities for part-nership with their clients and re-duces the chances that DPT will beviewed as a commodity – a diffi-cult place to compete where oftenthe only negotiating tool is price.

One area that is growing in theCMO business is the area of engi-neering support. Utilizing an in-house project engineering groupcan provide added value to clientswhen manufacturing challengesarise. For ex-ample, DPThas developedexpertise inthis group toassist clients inmanaging andl a u n c h i n gnovel pro-cesses. As theywork with anumber of theirc u s t o m e r s ,they are find-ing that manyinnovations arecoming from

small companies that are in a modeof expansion. Some of these com-panies may not have the money orfinancing needed to open their ownmanufacturing facility, so theyturn to CMOs for their increasedmanufacturing needs. Thesesmaller companies’ innovationsoften bring technology require-ments that are outside the realmof the typical CMO operation,which is where DPT has been ableto offer advantages to clients withtheir specialized services. Thesespecialized services have evenreached beyond engineering andinto R&D. In addition to the engi-neering and R&D support, on themanufacturing side Speed-to-Mar-ket is key. Many companies haveseasonal products (such as allergymedicines that are offered overthe counter) where they need as-sistance quickly in bringing theirformulations and manufacturingonline fast. The ability of CMOs torespond quickly and provide thatseasonal boost to manufacturingbrings affordable capacity to theirclients.

Figure 1. A breakdown of the annual $15 billion CMO market.Concludes on page 16.



MIDWEST REGION


As a major part of DPT’s busi-ness, they consider engineeringservices to be a “Focal Point ofSuccess” in supporting their cli-ents’ multiple innovating deliveryplatforms, product scale-up issues,and integration with product de-velopment teams. This broad ap-proach provides the users of CMOsmore flexibility in their operationsand planning.

CMO ChallengesWhat are some challenges forCMOs? In certain areas of themarket today, such as solid dosagepackaging for example, traditionalmanufacturers have excess capac-ity in-house creating pressurewithin these companies to keeptheir own plants running and re-duce outsourcing. Economic pres-sures developing in other globalregions, such as Asia – especiallyIndia, are creating a global com-petitive market.

In certain market segments,such as biologics and API, uncer-tain market conditions abound.Many CMOs are rethinking thestrategies employed in these ar-eas to increase their success. Thefailure of a multi-client businessmodel and weakness in capacityavailability has provided difficul-ties. Other issues including prod-uct failures, pipeline uncertainty,and the potential overhaul of theFDA approval process createshurdles to success for CMOs inthese markets. These areas alsocan add additional costs to CMOsin the regulatory demand for in-creased clinical and preclinicaltesting. When working with spe-cial or unique compounds, CMOs’risk goes up as they increase thecost of development, equipmentand facilities requirements to sup-port novel processes and complexformulations, as well as increasedraw material costs.

Issues and Opportunities in Contract ManufacturingContinued from page 15.

For DPT, they focus on long-term relationships with their cli-ents. They believe this approachprovides them an improved chancefor long-term growth and success.The longer term relationships alsocan assist both companies inweathering the changes that canoccur in the marketplace.

In general, there continue tobe other pressures facing the over-all pharmaceutical market thatcan hold back growth in the CMOindustry. These include contin-ued pressure to hold downhealthcare costs through drug re-importation, insurance companypressure to support the lowest costtherapy, and in some cases, dic-tated formulary pricing models.One area many companies arelooking at currently is the occur-rence of what seems to be a sig-nificant amount of blockbusterdrugs falling out of favor with theFDA and the public. The concernis how this may affect CMOs’ pipe-lines down the road.

CMO OpportunitiesSo what is coming down the pipefor CMOs? Some of CMOs’ tradi-tional advantages continue to betheir ability to provide speed tomarket and manufacturing flex-ibility as well as their supply chainexpertise in inventory manage-ment. The ability for fast decisionsto take advantage of current mar-ket opportunities allows them toget increased savings for their cli-ents, especially on the purchasing(raw materials) side of the busi-ness. These savings provide an-other advantage to their clients –freeing up assets and cash for morestrategic investments.

Sterile product manufacturingis one area that required moreevaluation for CMOs. This marketcan be broken into two main tiers:a larger segment where manufac-

turers are looking to pro-vide millions of units tothe market, and a smaller seg-ment where more specializedmanufacturing will producesmaller runs on products. The lat-ter is where CMOs have found aneed they can fill. Typically, com-panies looking for shorter andsmaller runs of products haveneeded additional technical/engi-neering expertise in addition tothe facility to manufacture smallerruns. CMOs who are equippedproperly can offer both of these toprovide added value.

One final area of potentialgrowth is the area of supply chainsupport for their clients. CMOsare looking at how they can workwith their customers in the areasof information infrastructure,managing costs and capital fromraw materials through distribu-tion, and creating the partnershipto support immediate inventoryreplenishment models. Close col-laboration and communication isneeded to make these endeavorswork.

ConclusionBeing successful means workingclosely and successfully with yourclients. Customer service is a bigmetric, and one that is tied closelyto information exchange with cli-ents. By focusing on deliveringvalue at multiple levels and cre-ating the right relationship withtheir clients, the added value ofspecialized services have madesome CMOs successful at whatthey do.

AcknowledgmentSpecial thanks to Mark Fite, VPOperations of DPT Labs and hispresentation at the ISPE SouthCentral Chapter Education Dayin San Antonio, Texas on 7 April2005.



MIDWEST REGION


Trends in the Animal Health Industryby George Heidgerken, President and Damian Gerstner, Manager,Engineering, Boehringer Ingelheim Vetmedica, Inc.

2001 2002 2003 2004 2005* Growth*

North American Sales 4.138 4.180 4.475 4.700 5.010 2.0%

Global Sales 11.050 11.330 12.545 13.710 15.070 1.7%

(billion US $) (* = Projected)

Table A. North American real growth rate vs. the global real growth rate.(Source: Wood Mackenzie, LLC)

The animal health industryprovides medications, vac-cines, anti-infectives, para-

siticides, and medicated feed addi-tives for a variety of species, in-cluding livestock (cattle, pigs, poul-try, sheep) and companion ani-mals (dogs, cats, horses). The $4.7billion North American animalhealth industry is projected to havea real growth rate slightly higherthan the global real growth rate -Table A.

Recent US sales growth hasbeen driven by parasite treatmentswithin the companion animal seg-ment. As the parasite treatmentmarket has matured, there hasbeen a notable shift toward medi-cations that improve the quality oflife for companion animals, andthus, improve the quality of life fortheir owners.

Within the livestock segment,the industry has been migratingaway from products that treat dis-eased animals toward biologicalvaccines that prevent disease. Thismove has been in response to con-sumer concerns about misuse ofantibiotics and other pharmaceu-tical treatments, and the potentialimpact on food residues and anti-biotic resistance. Animal health companies in-teract with and are regulated bythe same regulatory agencies ashuman health companies. In ad-dition, biological manufacturerswithin the animal health indus-try also are regulated by the USDepartment of Agriculture(USDA). Although USDA regula-

tions and manufacturinginspections assure prod-

uct quality, the USDA doesnot have formal require-

ments for Good Manufactur-

ing Practices (GMPs). However,the USDA is increasingly collabo-rating with the US FDA tostrengthen regulatory require-ments for veterinary biologicalmanufacturers. With extremely few exceptions,the animal health industry GMPsare in lock step with human healthGMPs. However, because animalhealth products typically do nothave the profitability that humanhealth products have, the cost ofregulatory compliance becomes asignificant factor in determiningwhether to bring new products tomarket or not. Ultimately, con-sumers will decide how much ad-ditional regulatory oversight theycan afford for their pets.

Bovine Spongiform Encephal-opathy (BSE or Mad Cow Dis-ease) has made a huge impact onthe animal health industry. Notonly has BSE impact on biologicalmanufacturing been similar to theeffects on human healthbiologicals, but BSE impact onthe food chain also causes disrup-tive factors. One outbreak of BSEcan upset the economics of beefproduction instantly, which maymake the use of normal produc-tion practices unprofitable. Oneresult of financial losses due to anoutbreak of BSE may be that pro-ducers no longer can afford to pur-chase cattle vaccines and medica-tions. Cattle integrators will riskthe chance of disease rather than

incur the cost of the vaccines. Fromthis author’s perspective, the USneeds to address the problem ofBSE from a health safety and sci-entific viewpoint, rather than fo-cusing on political solutions suchas trade barriers. By understand-ing how BSE propagates and ulti-mately affects human health, wecan focus on solutions that ad-dress the root cause.

Business consolidation affectsthe animal health industry in sev-eral ways. Food companies con-tinue to consolidate regularly, be-coming more vertically integratedalong the meat production supplychain from the farm to the dinnertable. This reduces the number ofpotential customers purchasinganimal health products and in-creases the dependency/risk thatanimal health companies face withthose customers that remain. Con-solidation within the humanhealth industry also greatly af-fects the animal health industrybecause animal health companiesare often associated with, but usu-ally a very small segment of, ahuman health company. Operat-ing philosophies and capital allo-cated to animal health versus otherbusiness areas like human drugscan change drastically with achange in ownership. Finally, ani-mal health distributors are con-solidating often as they face thechoice of reducing their costs orwatch their customers leave to

Concludes on page 18.



MIDWEST REGION


Trends in the Animal Health IndustryContinued from page 17.

purchase directly from animalhealth companies.

Although the US animal healthmarket is growing at a slow rate,its size and business environmentstill provides the greatest oppor-tunity for animal health manufac-turers. US patent regulations givemanufacturers the incentives theyneed to research and develop inno-vative products that improve meatproduction efficiency and improvethe quality of life for companionanimals and their owners.

Within this friendly, but verycompetitive US market place, thefollowing critical success factorswill determine which animalhealth companies are ultimatelywinners in the next decade:

Quality FirstManufacturers that continuallyimprove quality can stay ahead ofthe regulatory curve and enjoyaccess to more markets.

Move FasterManufacturers must fig-ure out how to innovate faster tobring products to market morequickly, without sacrificing qual-ity.

Listen to CustomersConsumers are becoming more so-phisticated in their tastes and pref-erences. If customers want moreorganic, chemical-free meat prod-ucts or if they want their pet tohave a high quality of life, thenmanufacturers must innovate toprovide those solutions.

ISPE Affiliates and Chapters

ISPE Affiliates and Chapters vary in size, number, andnature of activities. Although each has a distinctcharacter, all of them:

• promote and support educational programs designedto enhance professional performance

• foster and encourage favorable relations between itsMembers and related professionals

• collect and disseminate information for its Members

• establish pharmaceutical manufacturing as a profes-sion, and educate and promote the valuable role ofthis profession within the industry

If you are already an ISPE member but do not belong toa local Affiliate or Chapter, you may join one by visitingthe ISPE Web site www.ispe.org and indicating yourchoice on your Member Profile.



NORTHEAST REGION


The Northeast Region of the United States asrepresented by ISPE is comprised of the follow-ing 10 northeastern states: Connecticut, Dela-

ware, Maine, Massachusetts, New Hampshire, NewJersey, New York, Pennsylvania, Rhode Island, andVermont. This region is supported by ISPE’s NewJersey, Delaware Valley, New England, and BostonArea Chapters.

The Northeast Region of the United States is cur-rently and has historically been the strongest geo-graphical region in the world with regard to the phar-maceutical industry. Of the top 25 pharmaceuticalcompanies globally, 80% have headquarters or majorfacilities in the Northeast Region. These companiesgenerate 87% of global pharmaceutical sales and 86%of all research and development spending and they sell82% of the top pharmaceutical products in the world.

Companies with headquarters in the region includePfizer, Merck, Johnson & Johnson, Bristol-MyersSquibb, Wyeth, and Schering-Plough. Global compa-nies with North American headquarters in the regioninclude GlaxoSmithKline, Aventis, AstraZeneca,Novartis, Roche, Boehringer Ingelheim, Eisai, NovoNordisk, and Teva. All of these companies have majorresearch, development, and manufacturing facilitiesin the area as well. This is quite an array of companiesto be concentrated in one geographical area makingthe Northeast Region a very important place in thepharmaceutical world.

While some of the manufacturing operations ofthese companies have moved from the region to thesouth and west, the area remains strong in its concen-tration of headquarters, administrative, and researchand development campuses. Many large R&D facili-ties have been constructed on new and existing sites inthe region in the past 10 years.

Bastions of the pharmaceutical industry are lo-cated primarily in the Philadelphia and New York

Northeast Overviewmetropolitan areas in a region stretching from north-ern Delaware to Connecticut. Like elsewhere in thecountry, the biotechnology industry is growing in thenortheast with concentrations in the Boston and south-ern New England areas, northern New Jersey, and thegreater Philadelphia metropolitan area.

Because of the concentration of pharmaceuticalbusiness in the region, there is also a high concentra-tion of top quality support companies. Some of thelargest construction companies in the world are lo-cated in the northeast and build for the pharmaceuti-cal industry. Supporting them are architectural, engi-neering, and consulting firms that design, commis-sion, and validate the facilities that they construct.Equipment manufacturers and vendors who make andsell everything from air handling equipment to scien-tific and manufacturing apparatus support this indus-try as well, and a plethora of satellite companiesproviding the industry with consumable supplies andother consulting and resource services are locatedhere.

The region has a strong constituent of educationalinstitutions including colleges, universities, and techni-cal schools that feed professional talent into these alliedindustries. Their supporting strengths are generally inthe sciences and engineering fields, but there are alsostrong medical schools and technical trade schools inthis region as well. Historically, many of these institu-tions have sprung up around the industrial centers thatthey support. However, growth trends in biotechnologyhave shown that this industry has tended to germinatearound its supporting scholastic institutions.

This region with its historical technical strength inthe pharmaceutical industry has exported its prod-ucts, businesses, and expertise throughout the countryand the world. The northeastern United States hasbeen and will continue to be the leading region in theworld of pharmaceutical sciences.

Visit the following Web sites to read moreon the Northeast Region:

Overview of the ISPE Boston AreaChapterwww.ispe.org/boston

Overview of the ISPE DelawareValley Chapterwww.ispedvc.org

Overview of the ISPE New EnglandChapterwww.ispe.org/newengland

Overview of the ISPE New JerseyChapterwww.ispe.org/newjersey

or www.ispe.org/countryprofile



NORTHEAST REGION


History of Pharmaceutical andBiopharmaceutical Companies In New Englandby Hank Moes, Consulting Director, Arion Water Inc. andJames Blackwell, Consultant, BioProcess TechnologyConsultants

Pharmaceutical Company Locations

New England encompasses the states of Connecti-cut, Massachusetts, Maine, New Hampshire, and

Vermont. The larger pharmaceutical companies are con-centrated in Connecticut, including for example suchcompanies as Bayer, Bristol-Myers Squibb, BoehringerIngleheim, and Pfizer. These are all well-establishedcompanies whose facilities were built many years ago.

AstraZeneca, Novartis, and Bristol-Myers Squibbare the major pharmaceutical companies in Massa-chusetts. AstraZeneca has been around a long timesince it was originally the Swedish company, AstraPharmaceutical. Novartis is a relative newcomer.

Biopharmaceutical Company LocationsThere are biopharmaceutical companies in almost everyNew England state, but the majority are located in thegreater Boston area, which includes Worcester. Secondplace goes to Connecticut and Rhode Island. New Hamp-shire and Maine also have a number of companies.

The University and Medical FactorThe greater Boston area has one of the highest concen-trations of biotechnology companies in the world. Thereason, of course, is the presence of universities thatspecialize in the sciences as well as world-renownedmedical research facilities and hospitals. Among uni-versities, Harvard University and the MassachusettsInstitute of Technology (MIT) are renowned, but North-eastern University, Tufts University, the University ofMassachusetts, Boston College, Boston University,Brandeis University, and Worcester Polytechnic Insti-tute (WPI) contribute to this vast knowledge base.

The Biotechnology HistoryThe term “biotechnology” was probably coined in the1970s, but biological processes in pharmaceutical com-panies were prevalent long before that. The majority ofthe biopharmaceutical companies in the Boston areawere formed in the 1980s with additional companies inthe 1990s. The major companies with capitalization ofmore than one billion are Abbott Laboratories’ Biore-search Center, Biogen-Idec, Wyeth Biopharma,Genzyme Corporation, Amgen (in Rhode Island),Serono, and Lonza Biologics (in New Hampshire). Abrief descripion of these companies is located at

www.ispe.org/countryprofile. This listing does notinclude Novartis and AstraZeneca.

The total number of biopharmaceutical companiesin the area is more than 300. A listing and details of325 companies can be found on the Web atwww.massbio.org/directory/companies.

Biotechnology “Flavor”New England is host to biotechnology firms of all sizesand flavors. Smaller firms are focused on leveragingnew technologies, including stem cell research andgene therapy to develop products that meet unmetmedical needs. These firms benefit from the closeaccess to talent, research, venture capital, and newventure expertise, such as intellectual property lawfirms. Medium-sized firms typically have at least oneproduct and have most of their production capacity inthe state in which they reside. The larger firms typi-cally have multiple products and operations aroundthe world. The region is competing with other regionsin the US and the rest of world to attract and maintainbiotech manufacturing. Its attractions are its highlyskilled workforce and proximity to research and devel-opment. The remainder of the discussion will focus onthese larger firms.

Amgen is known for its oncology, immunology, andnephrology products. In Rhode Island, Amgen pro-duces Enbrel for rheumatoid arthritis along with otherproducts; Enbrel is a direct competitor to Abbott’sHumira product. Abbott Laboratories is a multi-na-tional, diversified healthcare company and producesHumira at its Worcester, Massachusetts productionfacility. Humira is a fully humanized monoclonal an-tibody for the treatment of rheumatoid arthritis.Biogen-Idec is known for its oncology and immunologyproducts, most notably Avonex, Amevive, and Rituxan.While carving a niche in enzyme replacement therapy,most notably Cerezyme, Genzyme has been a diversi-fied company for some time and has products fordiagnostics, surgery and orthopedics, autologous celltherapy, and renal care, among others. Lonza Biologicsis a CMO for biologics with small to large-scale cellculture manufacturing capabilities. Lonza has produc-tion agreements with many industry leaders. WyethBiopharma manufactures ReFacto AntihemophilicFactor, and BeneFix Coagulation Factor IX.



SOUTHEAST REGION


Mid-Atlantic and Southeast United States

Pharmaceutical and BiotechnologyManufacturing Growthby Alan Jones, CRB Consulting Engineers and Builders andJason Rifkin, President, Equilibrium Consulting, LLC

The Carolina South-Atlantic and Chesapeake BayArea Chapters geographic region has been attractive to pharmaceutical and biotechnology

manufacturers due to the low cost of labor, low taxes,good climate, access to research scientists, and a quali-fied labor pool.

MarylandMaryland’s pharmaceutical industry is predominatelyfocused on R&D and early stage product developmentalthough manufacturing facilities are on the rise. Hu-man Genome Sciences and MedImmune recently builtcorporate campuses that include pilot and scale-upmanufacturing. Contract manufacturing organizationskeep growing to meet the demands of the industry.These firms include Cambrex BioSciences, ChesapeakeBiological Laboratories, and UPM Pharmaceuticals.

MedImmune plans to continue this trend with amanufacturing expansion in Frederick, Maryland. Therecent approval of $5.6 billion for Project BioShieldalso will likely lead to an increasing number of compa-nies relocating to the Washington, DC area. One ex-ample of this is BioPort’s recent entrance into Mary-land where they plan to expand their Anthrax Vaccineproduction capabilities.

VirginiaEli Lilly & Co. is building an insulin fill/finish manufac-turing facility on 120-acres in Manassas. The location isprojected to employ more than 700 employees with anaverage annual salary of $44,000. Pat McGarrah, Gen-eral Manager of the Lilly Prince William site, says ahighly skilled workforce and high quality of life were theprimary deciding factors for the location.

North CarolinaIn North Carolina, the Research Triangle Park (RTP)near Raleigh-Durham has a large concentration ofmanufacturers located within 100 miles of RTP. From1990 to 2000, more than 42 new companies have

established facilities in RTP. New construc-tion and expansion has totaled more than five

million square feet. Some of the larger pharma-ceutical manufacturers in the RTP area include

BiogenIdec, Diosynth Biotechnology, and Eisai Phar-maceuticals - Figure 1.

Visit the following Web sitesto read more

on the Southeast Region:

Balancing Cost and OperatingParadigms are Key Facility Issues forBurgeoning Life Science Companiesin Today' Marketby Ken Berkman, Executive VicePresident, Scheer Partners, Inc. andJason Rifkin, President, EquilibriumConsulting, LLCwww.ispe.org/countryprofile

Overview of the ISPE Carolina-SouthAtlantic Chapterwww.ispe.org/carolina-southatlanticor www.ispe.org/countryprofile

Overview of the ISPE ChesapeakeBay Area Chapterwww.ispe.org/chesapeakebayareaor www.ispe.org/countryprofile


Merck and Co. is building a new 272,000 square footvaccine manufacturing facility to be located in Durham.Talecris Biotherapeutics launched its new worldwidetherapeutic proteins business on 1 April 2005 afteracquiring the contributed assets of Bayer’s plasmabusiness, including the state-of-the-art production fa-

Figure 1. Eisai Pharmaceuticals.



SOUTHEAST REGION


Pharmaceutical and Biotechnology Manufacturing GrowthContinued from page 21.

cility in Clayton - Figure 2.Wyeth Vaccines, located in Sanford, also has seen

phenomenal growth in both its workforce and sitelayout - Figure 3. The Sanford site is the exclusiveproducer of HibTITER, a conjugated vaccine that hasbeen effective against haemophilus influenza b (Hib)and is licensed for use in more than 60 countriesworldwide. The site has more than doubled in size tomore than 1,100 employees since 2001 creating not onlyjobs for the surrounding communities, but also openingunlimited professional development for current staff.

Other large manufacturers in North Carolina include:Novo Nordisk Biochem, Novo Nordisk PharmaceuticalIndustries, Inc., Banner Pharmacaps, Baxter HealthcareIV Systems, GlaxoSmithKline, EON Pharma, PurduePharmaceuticals, Merck (two Facilities), AKZO Nobel,and Cardinal Health Sterile Technologies.

Several contract manufacturing firms have recentlyexpanded to meet improved customer demand. DSMPharmaceuticals in Greenville formulates pharmaceu-tical products into various dosage forms and recentlyexpanded its fill/finish operations - Figure 4. DSM

Pharmaceutical Products focuses on the devel-opment, scale-up, and custom manufacture ofpharmaceuticals. It acts as a customer-oriented andreliable partner supplying the pharmaceutical industrywith a sophisticated range of chemical intermediates,biotech-based drugs, and dosage formulations such asorals, topicals, and steriles. They recently expandedtheir fill/finish operations to include sterile productmanufacturing capacity to 410,000 square feet withmore than 3,700 square feet of lyophilization capacityand innovative aseptic liquid filling suites.

South CarolinaMartek Bioscience, Inc., located in Kingstree, recentlyexpanded. They are “an innovator in the research anddevelopment of products derived from microalgae.”Martek has developed and patented two fermentablestrains of microalgae which produce oils rich indocosahexaenoic acid, DHA. In a similar manner,another patented process was developed for a fungusthat produces an oil rich in arachidonic acid, ARA.Both DHA and ARA are found in breast milk and areimportant nutrients in infant development. Thus, thetwo oils are used in infant formulas, while the DHA-rich oil also can be used in supplements and functionalfoods for older children and adults. Martek also makesand sells a series of proprietary and nonproprietaryfluorescent markers. These products have applica-tions in drug discovery (high-throughput screening),DNA microarray detection and flow cytometry.”Martek’s key markets include: infant formula, nutri-tional supplements and functional foods, life science,and drug discovery.

Other key manufacturers in South Carolina in-clude: Bausch and Lomb (Figure 5), Roche Carolina,Pfizer Capsugel, GlaxoSmithKline, Holopack Interna-tional, Perrigo Company, Leiner Health Products, Phar-maceutical Associates, and Irix Pharmaceuticals.

GeorgiaGeorgia has a robust manufacturing base with severalrenovations and expansions to the Northeast of At-lanta. Elan Holdings, Inc. (Figure 6) and Merial areboth located in Gainesville while Merial also hasmanufacturing facilities in Athens. Solvay Pharma-ceuticals’ North American Headquarters is also inGeorgia. Merck has had a manufacturing facility inAlbany for many years. Augusta is home to Monsantoand UCB Bioproducts manufacturing operations.Mikart, a recognized leader in providing formulationdevelopment, contract manufacturing, and packagingservices to the pharmaceutical industry, has state-of-the-art facilities in Atlanta.

Figure 2. Talecris Biotherapeutics.

Figure 3. Wyeth Vaccines.

Figure 4. DSM Pharmaceuticals.



SOUTHEAST REGION


FloridaFlorida has rapidly grown its pharmaceutical manu-facturing business over the last 15 years and they nowhave more than 4,000 employees involved in pharma/bio manufacturing. A majority of the companies areclustered in the South Florida area.

Among biotechnology companies in South Floridais Goodwin Biotechnology, Inc. (GBI). GBI is a Con-tract Manufacturing Organization (CMO) specializingin the production of biologics for toxicology studies andPhase I, II, and III Clinical Trials. GBI has been inbusiness as a CMO longer than any existing competi-tor and has historically specialized in mammalian cellculture. Recently, GBI was acquired by Wallace Phar-maceuticals Pvt. Ltd., an India-based multi-specialtypharmaceutical company. GBI has received a signifi-cant capital infusion in connection with this transac-tion, and is expanding into 200L mammalian stirredtanks and microbial fermentation. GBI’s business isgrowing rapidly, and the Wallace organization is eagerto participate both in the growth of biotechnology inthe US in general and specifically, South Florida’sbiotech growth as a result of the Scripps transaction.

A recent accomplishment for the state was thecommitment by North American Biologics Inc. (NABI)to build a state of the art vaccine production facility forStaphVAX, an investigational vaccine to prevent lifethreatening S. aureus infections. This will be the firstcommercial vaccine facility in the state. The facility islocated in Boca Raton.

Andrx Corporation is located in South Florida andper their Web site, Andrx “develops, manufactures,and commercializes generic versions of controlled-release, niche, and immediate-release pharmaceuticalproducts, including oral contraceptives; and distrib-utes pharmaceuticals, primarily generics, which havebeen commercialized by others, as well as their own,

primarily to independent pharmacies, phar-macy chains, pharmacy buying groups, and

physicians’ offices.”IVAX Pharmaceuticals, headquartered in Mi-

ami, has experienced considerable growth the last

few years - Figure 7.IVAX was recently rec-ognized as a leading USgeneric company movingup to a #5 ranking basedon dispensing 84 millionprescriptions in 2004.IVAX ranked #1 in termsof absolute total pre-scription growth with14.8 million prescrip-tions. For the total phar-maceutical industry (brand and generic), IVAX nowranks as the 11th largest pharmaceutical company inthe US, and the fastest growing among the Top 20Pharmaceutical firms.

Aphton Corporation in Miami currently has severaldrugs in Phase II and III Clinical Trails. They alsohave alliances with Aventis Pasteur, GlaxoSmith Kline,and Schering Plough.

TennesseeThe Eastern part of Tennessee has two large pharma-ceutical manufacturing facilities in GlaxoSmithKlineand King Pharmaceuticals. King is headquartered inBristol.

It is evident that the Southeast and Mid-Atlantic re-gions will continue to compete for new manufacturingfacilities due to the large capital investments in aca-demic research facilities and worker training programscurrently being implemented to introduce current GoodManufacturing Practices to the existing labor pool.

We hope you have time to visit our region in thefuture and that this article enlightens your knowledgeof the pharmaceutical and biotechnology manufactur-ing industry in the Mid-Atlantic and Southeast Regionof the United States.

Figure 5. Bausch and Lomb.

Figure 7. IVAX Pharmaceuticals.

Figure 6. Elan Holdings, Inc.

Pharmaceutical and Biotechnology Manufacturing Growth



SOUTHEAST REGION


Interview with Dr. Leslie Alexandreconducted by Jeffery N. Odum, Principal, NCBioSource

Dr. Leslie Alexandre is Presi-dent and Chief Executive Of-ficer of the North Carolina

Biotechnology Center, a private,non-profit corporation establishedby the North Carolina General As-sembly in 1984. The mission of theBiotechnology Center is to providelong-term economic and societalbenefits to North Carolina by sup-porting biotechnology research,business, and education statewide.

Before joining the Biotechnol-ogy Center in August 2002, Dr.Alexandre was assistant directorfor industrial relations in the Of-fice of Technology and IndustrialRelations of the National CancerInstitute (NCI) in Bethesda, Mary-land. She was responsible for build-ing relationships with the privatesector on behalf of NCI and facili-tating the development of scien-tific collaborations with industryto accelerate the progress of can-cer research.

Looking at the biotech industryin the southeast, where do yousee the greatest changes overthe past five years?

I do not think of biotechnology asan industry per se, but rather as acollection of scientific tools andtechnologies that use living cellsand their molecules to make prod-ucts and solve problems in manydifferent industries. Over the pastseveral years, the southeast haswitnessed tremendous growth inthe importance of biotechnologyas an engine for economic develop-ment at the state and local level,as well as regionally. Five yearsago, the southeast was not viewedas a major force in terms of attract-ing investment or intellectual“capital” for biotechnology compa-

nies. There also was not broadrecognition of the large and grow-ing concentration of manufactur-ing and support companies withinthe region. Today, that has changeddramatically as major biotechnol-ogy companies such as BiogenIdechave located in the southeast, andmany “home grown” companieshave not only become viable, buthave successfully launched prod-ucts to the marketplace. In addi-tion, we are seeing broad diversifi-cation in the application of bio-technology to multiple industriesin the southeast – perhaps more sothan in any other region. In NorthCarolina, for example, not only doyou find lots of biopharmaceuticalcompanies, you also find manycompanies focused on industrial,agriculture, forestry, and marinebiotechnology applications.

Now looking ahead to the nextfive years, what trends do youforesee for biotechnology com-panies in the southeast?

I believe that the growth of bio-technology companies in the re-gion will be even more dramaticover the next five years. Helping tofuel this growth will be a concertedpush by states to transition theireconomies from those dependenton brawn to those dependent onbrains, related efforts by state andlocal economic developers to stemthe transfer of jobs overseas, thecontinued commercial successes of“home grown” southeastern-basedcompanies, and the exciting ad-vances in the sciences coming outof the region’s major universities.In addition, I believe you will see alarge increase in the number ofpharmaceutical and biotechnologycompanies that choose to manu-

facture in the southeast. With theirbusiness friendly environments,low cost of doing business and sea-soned manufacturing workforce –not to mention low cost of livingand high quality of life – south-eastern states are well positionedto attract biotechnology and otherlife science companies that wish todo at least some of their productmanufacturing in the US.

What issues do you see asbeing potential roadblocks tocontinued growth?

While the future of biotechnologyin the southeast looks extremelybright, there are some critical is-sues that must be addressed. Oneis expanding access to investmentcapital throughout the region. Un-like the more abundant supplies ofinvestment capital found in the SanFrancisco Bay area and the metro-Boston area, there are currentlyonly a few large pools of venturecapital within our region. Part ofour challenge stems from the factthat the southeast is still relativelyunknown with respect to biotech-nology. We have not yet createdany big, household name compa-nies, such as an Amgen or aGenentech. Those companies and afew others in California, Massa-chusetts, and a handful of otherstates provided phenomenal re-turns to their early investors andmotivated substantial additionalbiotechnology investing in their re-gions.

Expectation management is an-other big challenge economic de-velopers in the southeast will face,particularly in those states thatcurrently lack a large concentra-tion or “cluster” of biotechnologycompanies. I see some such states



SOUTHEAST REGION


committing enormous sums of tax-payer dollars to initiatives aimedat stimulating biotechnology re-search and commercialization withthe ultimate goal of creating largenumbers of high paying biotech-nology jobs in the near future. Whatwe have learned in North Carolinais that it takes decades, not years,to create a sustainable biotechnol-ogy cluster and it takes the in-volvement and full cooperation ofa huge constellation of partner or-ganizations in the public and pri-vate sector. Research cannot berushed, nor can the process ofmoving a new therapeutic agentfrom the bench to the bedside. Ifear that substantial taxpayer dol-lars will be wasted in those statesthat have come most recently tobiotechnology as a form of eco-nomic development and lack thecritical ingredients required forsuccess. In all likelihood, thosewill be the states that can leastafford wasted expenditures.

More than 20 years ago, NorthCarolina was fortunate to haveextraordinarily visionary leaderswho recognized that the new sci-ence of biotechnology would oneday yield tremendous commercialopportunity. To help catalyze thedevelopment of biotechnology, theGeneral Assembly proceeded tocreate and fund every year sincethen, the North Carolina Biotech-nology Center. Today, we are thethird leading state for biotechnol-ogy with nearly 200 biotechnologyand related life science companiesemploying nearly 40,000 workersin clean, safe high paying jobs.

Do you see challenges on theregulatory front that the indus-

try must address in or-der for the growth to

continue?Protection of the consumer

will always be a top priority.

But there also must be a focus onthe business needs of the compa-nies that manufacture humantherapeutic products. Answeringthe question of how to ensure bothsafety and compliance in an af-fordable and timely manner willbe critical to patients and compa-nies alike.

As the industry becomes moreglobal, we must also look for waysto expand the concept of harmoni-zation. More products manufac-tured in the southeast and through-out the United States are beingtargeted for global markets. Weneed to focus on ways to improvecommunication within the regula-tory framework under which wemust operate. If we continue tohave a fragmented regulatory land-scape, it will be especially difficultfor smaller companies to succeed.Professional organizations such asISPE will play a critical role in thisarea. Being able to look at the sub-stance of issues and propose solu-tions that are focused on the needsof the industry is something that iscritical. Having worked on CapitolHill, I have seen first hand howimportant it is to have strong pro-fessional organizations become in-volved in providing information andhelping simplify issues to makethem clear and straight-forward tolawmakers and their staffs. Thatway, legislative and regulatory is-sues can be resolved much easier.

If you had to pick a few “hotbuttons” for the southeast tofocus on in the coming years,what would they be?

Again, access to investment capi-tal is critical. Without plentifulventure capital, we will not be ableto keep pace with other regions.We also must continue to grow thetalent pool of entrepreneurs andexecutive leadership and find new

Interview with Dr. Leslie Alexandre

venues for nurturing young com-panies in supportive environmentsuntil they are ready to stand inde-pendently. That may requiregreater involvement by the uni-versities and companies fromwhich new technologies are spin-ning out. And finally, as I saidearlier, the region needs more“home runs.” More home-grownsuccess stories that become visiblewill do great things to fuel thegrowth engine of biotechnology inthe southeast.

• • •

For more information, please con-tact the Center at Tel: 1-919/541-9366.



WEST REGION


The West Region of the UnitedStates as represented by ISPE

is comprised of the 11 western moststates of: Arizona, California, Colo-rado, Idaho, Montana, Nevada,New Mexico, Oregon, Utah, Wyo-ming, and Washington.

This region is supported byISPE’s Rocky Mountain, PacificNorthwest, and three CaliforniaChapters including the San Fran-cisco/Bay Area, Greater Los Ange-les Area, and San Diego Chapters.

The Western Region hosts theheadquarters of two of the largestbiotech/pharmaceutical companiesin the world: Amgen andGenentech. It hosts a myriad ofothers including, but not limitedto: Abbott, Abgenix, Allergan, Alza,Baxter, Bayer, B Braun, Berlex,BiogenIdec, Biomarin, Biosite,Cardinal Health, Cephalon,Chiron, Gambro BCT, Gen-Probe,Icos, Medtronic, Nektar, Novartis,NPS Pharmaceuticals, QLT,

West OverviewRoche, Sandoz, Watson Laborato-ries, and Zymogenetics.

Two of the major draws the westhas for the pharmaceutical indus-try are university collaboration andthe quantity and quality of localsupport companies. Bioscienceparks have been erected by manyuniversities in the region and theseuniversities are working with lo-cal companies to develop the prod-ucts and technologies of the fu-ture. The region produces a largequantity of PhD level academi-cians and attracts even more withits climate, salary scale, and qual-ity of life. In addition, local sup-port companies make the regionvery attractive for manufacturers.Venture capital, experienced andhighly skilled construction com-panies, equipment and productmanufacturers, validation and QA/QC companies, certified inspectors,and others compliment the welleducated talent base and draw

employers and employeesalike.

The pharmaceutical industry inCalifornia is concentrated in SanFrancisco, Los Angeles, and SanDiego. All of these locales boastsignificant bioscience parks andcenters. The Northwest, RockyMountain, and Southwest areasare up-and-coming pharmaceuti-cal centers where university col-laboration, an educated talentbase, and tax incentives offer acompelling draw for both existingresidents and potential companies.

California alone is home to morethan 4,000 bioscience companiesthat employ more than 150,000people. Academic institutions inCalifornia like USC, Scripps,Stanford, Cal Poly Pomona, CalState Fullerton, and the Universi-ties of California at Berkeley, LosAngeles (UCLA), Irvine (UCI), andSan Diego (UCSD) play a vital rolein the development of new tech-nologies and bioscience parks inaffiliation with their local phar-maceutical communities.

The Pacific Northwest has ex-perienced a 20% increase in phar-maceutical employment over thelast 10 years with more than 30,000people now working in the indus-try. Nearly two-thirds of thesepeople are employed at companiesof less than 50 employees. Theseinclude many significant start-upcompanies with technologies thathave been developed in collabora-tion with local institutions like theUniversity of Washington, Wash-ington State University, FredHutchinson Cancer Research Cen-ter, University of Oregon, and Or-egon State University.

The Rocky Mountain area iscomprised of Utah and Coloradoand has more than 800 registeredcompanies and 30,000 employeesin the pharmaceutical industry.The University of Colorado,

Visit the following Web sites to read moreon the West Region:

Amgenby Michelle M. Gonzalez, Principal Corp. Eng., Amgen, Inc.www.ispe.org/countryprofile

Overview of the ISPE Greater Los Angeles Area Chapterwww.ispe.org/greaterla or www.ispe.org/countryprofile

Overview of the ISPE Pacific Northwest Chapterwww.ispe.org/pacificnorthwest or www.ispe.org/countryprofile

Overview of the ISPE Rocky Mountain Chapterwww.ispe.org/rockymountain or www.ispe.org/countryprofile

Overview of the ISPE San Diego Chapterwww.ispe.org/sandiego or www.ispe.org/countryprofile

Overview of the ISPE San Francisco/Bay Area Chapterwww.ispe.org/sanfrancisco or www.ispe.org/countryprofile




WEST REGION


The Life Sciences Industry inWashington Stateby Kevin Brettmann, Director of Life Sciences, JE DunnConstruction and Pam Love, Director of Mktg. and MemberServices, Washington Biotechnology and Biomedical Association

The boom you hear from the Pacific Northwestisn’t just the sonic boom of a Boeing jet any more.Businesses based on new technology and upscale

taste - Microsoft, Amazon.com, Starbucks, andNordstrom - have infused the region with wealth andentrepreneurial energy.

Although yet a small portion of the overall stateeconomy, biotechnology is among the most dynamicgrowth sectors in the state, experiencing a 20 per-cent increase in employment in the last 10 years,more than the state’s overall job growth during thesame period. More than 20,000 people are employedin biotechnology here, nearly two-thirds of them insmaller companies of less than 50 employees. Morethan three-fourths of Washington biotechnologyfirms are not publicly traded, but the state is hometo some larger companies. Amgen’s significant R&Dunit (formerly Immunex), is in Seattle; ICOS em-ploys nearly 700 in Bothell, and Seattle also is hometo Dendreon, Corixa, and Rosetta Inpharmatics, adivision of Merck.

There are several important reasons for theindustry’s growth in Washington State: excellent re-search institutions, access to capital, a high quality oflife, and several business initiatives that have createda friendly climate for start-ups. Industry leaders areunanimous in crediting the University of Washington,Washington State University, and the Fred HutchinsonCancer Research Center, as the technology foundationfor Washington’s biotechnology industry. More thanone-half of biotechnology firms in the state are eitherfounded on technologies developed at these institu-tions or have ongoing collaborative relationships as animportant component of their business operation.Battelle Pacific Northwest National Laboratory inRichland is also an important research partner withmany of the state’s biotechnology companies.

Several institutions are in place to encourage en-trepreneurship: the Washington Technology Center

(WTC) is the state-funded enterprise thatsupports commercially promising research

and technology development of direct benefitto the economic vitality of Washington State.

WTC provides grants to professors to encourage

them to team up with entrepreneurs. A private eco-nomic development group, the Alliance of Angels,nurtures the growth of technology-based businessesin Washington State and improves the interactionsamong angel investors and emerging local technologycompanies seeking funding. The Washington Bio-technology and Biomedical Association (WBBA) is atthe center of networking events and political advo-cacy for the industry.

Washington has no corporate or personal incometax. Capital investments for qualified high technologyfirms (including biotechnology and medical devices)are exempt from the state sales tax. Firms that aremanufacturing in the state that have repair and re-placement costs have those costs exempted as well.Washington has a tax credit for businesses that locatein distressed areas although the biotechnology hot-beds of the state, King and Snohomish counties, do notqualify as distressed.

A major expense for most businesses is the Businessand Occupancy (B&O) tax on gross receipts. However,qualified high technology firms, including those in thebiosciences, receive a credit against B&O taxes (up to $2million per year) for their research and developmentexpenditures. Another organization, the WashingtonBiotechnology Foundation (WBF), provides the tools toteachers to foster awareness of biotechnology in theclassroom. WBF provides resources and training forteachers, including workshops, seminars, biotechnol-ogy lessons, and opportunities to meet and learn frombiotechnology professionals.

Washington is rich in education and training op-portunities for people seeking careers in all levels ofbiotechnology. Washington State University in Pull-man is one of nine campuses across the countrychosen for a National Institutes of Health graduatetraining program in the science and applications ofprotein chemistry. In addition to UW’s Department ofMolecular Biotechnology, which attracts studentsinternationally, UW is a leading center for training invirtual reality software development with many ex-citing implications for biotechnology.

Many of the state’s community colleges and four-year colleges have programs that prepare students for




WEST REGION


The life sciences industry isthriving in the Rocky Moun-tain Region. Manufacturing

companies, research and develop-ment enterprises, research facili-ties (institutes, universities, hos-pitals, and laboratories), and sup-porting companies have estab-lished themselves from the Cana-dian border to the Mexican borderalong the Rocky Mountains. Themajority of these life sciences enti-ties are clustered in the greaterDenver area in Colorado and inthe greater Salt Lake City area inUtah. In those two states, thereare more than 800 registered lifescience companies in all. Estimatesfrom 2003 show these two statescontain nearly 30,000 workers inthe life sciences industry. Indus-try sectors in the Rocky Mountainregion include: aerospace, animalbiotechnology, bioinformatics, bio-technology, pharmaceutical, medi-cal device, diagnostics, plant andagricultural biotechnology, nano-technology, photonics, and nutra-ceuticals. Fueling these sectors isa strong research base in the areawith numerous institutes, univer-sities, and research centers thatannually receive more than $300million for life science research.Helping drive the expansion of theindustry to the area are the manyquality of life factors, favorabletax environment, and a high con-centration of ‘High Tech’ workersin the area.

Colorado’sLife Sciences Industry

Colorado’s state government hasidentified the life sciences indus-try as a key to its future economicsuccess and is in the process ofmaking the state an even moreattractive location for companies.Currently, the state has more than

The Life Sciences Industry in theRocky Mountain Region

600 registered companies, employ-ing nearly 17,000 workers withthe majority of these in medicaldevice, biotechnology, and phar-maceutical sectors. There are sev-eral representatives from estab-lished companies, includingAmgen, Novartis, Roche, BaxaCorporation, Medtronic, COBECardiovascular, Gambro BCT, andCargill. There are numerous uni-versities, research and develop-ment enterprises, and institutesfueling the life sciences industry.These include the University ofColorado, Eleanor Roosevelt In-stitute, and the Division of Vector-Borne Infectious Diseases. Colo-rado also is home to the Fitz-simmons “Life Science City.” This$4.3 billion ‘square mile of life sci-ences,’ located in the Denver metroarea, is the largest medical-relatedredevelopment effort in the coun-try. The investment in this“biopark” is a major step forwardfor Colorado and the life sciencesindustry and represents a majorhub for the industry in the RockyMountain region and the UnitedStates.

In addition to this positive-lifescience environment, Coloradoboasts one of the highest concen-trations of ‘high tech’ workers inthe US, a state cabinet technologyposition, and a top five rating for‘tax-friendly’ states. These factorsalong with the investments in in-frastructure mentioned above aswell as other factors such as qual-ity of life make it an attractivelocation for the life sciences indus-try.

Colorado’s current challenge toexpand the industry includes thelack of notoriety in the industryover other areas such as Bostonand California, and the fact thatthe industry hasn’t reached a ‘criti-

cal mass’ of companies. However,with all the positive attributesfound in the state, it is just a mat-ter of time until it becomes a na-tional hub for the industry.

Utah’sLife Sciences Industry

Utah has been a growing player inthe life sciences industry since theearly 1980s when the Universityof Utah successfully implanted anartificial heart into a human. Priorto that milestone, the medical de-vice market in Utah had explodedwith the invention of disposablemedical devices by newly foundedDeseret Pharmaceuticals in 1956.For the state of Utah, this growthin the medical device industry wasbolstered with the development ofBallard Medical Products a fewyears later and the Sorenson Fam-ily of medical science companies.

Since this time, Utah has be-gun to build a reputation as a medi-cal products and biotechnologycenter with up and coming biotechcompanies spinning off from theUniversity of Utah and Utah StateUniversity making national news-paper headlines. Currently, Utahis home to more than 150 life sci-ence companies primarily in themedical device and biotechnologysectors. Utah has major biomedi-cal devices and supply companies,two world-class biopharmaceuticalcompanies, a national drug deliv-ery company, and a worldwideleader in biological products. Withindustry leaders that includeCephalon, Myriad Genetics, AbbottCritical Care Systems, NPS Phar-maceuticals, and Watson Labora-tories, as well as associated part-nerships with most of the leadingpharmaceutical companies, Utahis positioning itself to continue sig-nificant expansion in the bio-



WEST REGION


sciences market.Utah’s state universities have

major genetic medicine andbiotech/agricultural resources, aswell as significant research capa-bilities in bioinformatics and sci-entific imaging. The education sys-tem in Utah is committed to facili-tate the growth of biotechnologythroughout the state. The Univer-sity of Utah, Brigham Young Uni-versity, and Utah State Univer-sity are all active in research anddevelopment in the bioscience are-na as well as facilitating the trans-fer of university developed tech-nologies to the marketplace. TheHuntsman Cancer Institute in col-laboration with the University ofUtah has discovered more gene-related diseases than any otheruniversity. This adds to the poten-tial of university spin-offs in ge-netic medicine.

Utah’s overall culture providesa unique asset to biotech and alsomay be attractive to companieslooking at moving there. The al-truistic culture in Utah encour-ages many to participate in cancerresearch, genetic testing, and as-sociated genealogical disclosure.This culture combined with thequality of life; educational excel-lence, and long-term commitmentof the high-tech workforce helpsfacilitate the process in seeing anew drug through the subsequentyears of clinical trials. Local gov-ernment does not play a huge partin the development of Utah’s lifesciences industry; however, revo-lutionary efforts by Utah’s Sena-tor Orrin Hatch to make the FDAmore efficient coupled with his ef-forts to fund research for pediatricAIDS have contributed to thisstates’ technical growth.

The Life Sciences Industry in the Rocky Mountain Region

Attracting venture capital fund-ing is an ongoing challenge forUtah. Utah based biotechnologycompanies have had to rely prima-rily on business partnerships withhigher profile companies to receivefunding. Utah will need a strongerfinancial and legal infrastructure,including legal services, venturecapital firms, and investmentbanks that specialize in biotech-nology - all of which are key to thesuccessful development of any re-lated high tech industry. The lastchallenge left to overcome is thesuccessful marketing of Utah as astrong option for life science com-panies. Although Utah companiesdo a good job of marketing, theworld does not know that they arethere, and Utah is still not seen asa primary place to invest in bio-technology.



WEST REGION


A university researcher and aventure capitalist formedthe world’s first biotechnol-

ogy company, locating it in 3,000square feet of industrial space inSouth San Francisco. Genentech,founded in 1976 by University ofCalifornia, San Francisco biochem-ist Herb Boyer and RobertSwanson, was one of the first ofmany successful Bay Area biotech-nology companies to come.

Genentech followed the directentrepreneurial lead of pioneer-ing Northern California compa-nies like Cetus Corp, a biologicalengineering company – the firstever – founded in Berkeley in 1971,and Palo Alto-based Syntex Cor-poration, founded in 1964. Syntex,led by Alejandro Zafferoni (later afounder of a number of biotechcompanies), was the first pharma-ceutical company to form sinceWorld War II.

Much of the work in bio-technology’s early years was doneby a handful of brilliant investiga-tors and scientists at world-classresearch institutes – StanfordUniversity, UCSF, and UC Berke-ley. Charged by the federal gov-ernment in the early 1970s to fighta war on cancer, researchers guidedby their interest in pushing thefrontiers of genetic engineeringbuilt more than science. Theyfounded an industry that revolvesaround the Bay Area.

The FoundationsThe research institutions createdthe technology, tools, and intellec-tual climate necessary to build anew industry. The Bay Area in the1960s was percolating innovationand discovery in many quarters,

A Brief History of Biotechnology inNorthern Californiaby BayBio

including science. Since the mid-1950s, Stanford University, UCSF,and UC Berkeley had producedNobel Laureates in chemistry,physics, and biology. In the early1970s, that tradition of rigorousscientific innovation continued asUC Berkeley biochemist BruceAmes devised a method, now astandard for researchers, to detectgenetic mutations in bacteria.Stanford University’s LeonardHerzenberg developed rapid cell-sorters, speeding up research.

In 1973, Boyer, Stanford ge-neticist Stanley Cohen, andStanford biochemist Paul Bergisolated many of the genetic engi-neering methods used. Berg linkedtwo genes from different virusestogether, and Cohen and Boyerdemonstrated that cloning DNAwas possible. Cohen and Boyerthen proved that they could splicegenes into bacteria, using mi-crobes to churn out human hor-mones, growth factors, and othermedically important chemicals. Injust a few short months, thesecomplementary discoveries be-came the intellectual basis of bio-technology.

With the new genetic toolscame new fear. Researcherssought to understand the dangersand possibilities that genetic en-gineering offered. Unsure of thedangers and the regulatory cli-mate they might face, Berg andothers in 1974 called for a world-wide discussion on issues of genesplicing and safety. In 1975 atAsilomar Conference Center inPacific Grove, 140 prominent re-searchers and academicians de-bated their opinions about gene-splicing. Within a year, the Na-

tional Institutes of Health wouldissue guidelines based upon theconference’s recommendations.Swanson and Boyer then foundedGenentech, which would win therace to produce the first humanproteins using biotech techniques,including insulin (FDA approvedin 1982) and the human growthhormone (FDA approved in 1985).Other discoveries were leading tonew companies: UCSF’s WilliamRutter and two university re-searchers co-founded Chiron Cor-poration to find vaccines for hepa-titis B. In 1979, the J. DavidGladstone Cardiovascular Insti-tute began research at UCSF.

As science pushed forward,questions regarding patentablescience would dominate the 1980s.“If companies and universitiespoured millions into researchwithout patent protection, inno-vation would be dampened,” re-called Dr. George Rathmann,founder of Amgen, and Chairmanof Sunnyvale-based Hyseq at arecent Bay Area Bioscience Cen-ter forum, Bay BioNEST. In 1980,the Supreme Court agreed thatlife forms could be patented. Pat-ents in hand, biotech companieswere ready to face Wall Street.Genentech became the first ofmany companies to go public, gen-erating $35 million in its initialpublic offering. A rash of addi-tional public offerings followed ascompanies attempted to duplicateGenentech’s success. Cetus fol-lowed in 1981 with a $107 millionIPO. Scios, Amgen, Chiron, Xoma,and others would follow in thenext five years.



WEST REGION


Into the 1980sDuring its first decade, biotech inNorthern California experiencedtremendous growth. From 1964through 1977, 84 companies werefounded. By 1987, the Bay Area’sbiotechnology industry had grownto 112 companies, supporting19,400 jobs and total sales of $2billion. Throughout the remain-der of the 1980s, that growth con-tinued, and between 1987-1990,81 new companies formed. In 1989,Stanford opened the $100 millionBeckman center for medicine andmolecular biology, headed by NobelLaureate Paul Berg.

An early focus of genetic engi-neering was agriculture. Andagain, Northern California was,and is, at the forefront of agricul-tural biotech, and can lay claim toownership of the first agriculturalbiotech, International Plant Re-search Institute. In 1986, UC Davisfounded its program for biotech-nology in agriculture. Early inbiotech’s history, researchers fo-cused their efforts on helping foodcrops resist frost and pests andBay Area researchers respondedas UC Berkeley professor StevenLindow and Advanced Genetic Sci-ences created and developed ice-inhibiting bacteria, the first re-lease of a bio-engineered organ-ism in 1987. Calgene worked ontesting of a genetically engineeredtomato, later approved in 1994.Groundbreaking agricultural en-gineering work, begun in North-ern California, continues in ear-nest.

Beyond business success, uni-versity researchers and biotech-nology institutions are working tocure diseases. Early on, research

that led to biotech had ananti-cancer goal, but came

to include therapies fornearly every disease. The rela-

tively new disease, AIDS was nodifferent, and research began totry and understand the diseaseimmediately. Northern Californiaresearch led the way as UCSF re-searchers isolated the HIV virusin 1983. Chiron cloned the virus in1984, and in 1986, blood screeningand diagnostic tests were madeavailable worldwide.

The regulatory climate in the1980s was evolving as well. Thefederal government and industryworked to coordinate oversight ofgenetic engineering. In the early1990s, an effort by the Clintonadministration to reform thenation’s healthcare system alsoimpacted biotech’s prospects bythreatening to intervene in phar-maceutical pricing, and thus re-duce the chances of recouping de-velopment costs. Because the in-dustry is new, government over-sight and regulatory frameworkshave shifted frequently.

From the 1990s toTomorrow

Silicon Valley fueled innovation toaccelerate the pace of discovery.An entire industry of tool-makersand testing-equipment manufac-turers developed as the need fornew instruments pushed technol-ogy, led by research done atStanford University, AppliedBiosystems’ early DNA synthesiz-ers, and Bio-Rad Laboratories. Ce-tus Corp. developed PCR, a chainreaction allowing researchers togenerate billions of gene sequencecopies in only hours netting KaryMullis the Nobel Prize in 1993.Later, tool companies would taketechniques learned in the develop-ment of silicon chips for high techand apply them toward biotech.The new tools would be necessaryfor the next big revolution, and thenext decade of discovery. The

A Brief History of Biotechnology in Northern California

much-publicized Human GenomeProject, begun in 1990, led to anew basis of drug research anddevelopment, from viral replica-tion of proteins to gene-based dis-covery and treatment of diseases.Again, Northern California re-search was at the forefront, asLawrence Livermore and Law-rence Berkeley Labs headed upthe Western Facility in WalnutCreek, playing a key role in thegenome. Northern California alsobuilt on its rich legacy of world-leading research as UCSF wouldherald two more Nobel Prizes, onefor Michael Bishop, current chan-cellor.

With the maturation of a num-ber of early biotech companies,many invested heavily in produc-tion facilities. Chiron, Bio-Rad,Genentech, Genencor, and Bayerin Berkeley built or expanded re-search and production facilities inthe Bay Area. And, just over 20years after Genentech’s founding,the company began marketing thefirst genetically engineered mono-clonal antibodies aimed at fight-ing cancer, Rituxan.

Bay Area research universitiesare currently gearing up for thenext wave of research. UC Berke-ley unveiled its life sciences initia-tive. UCSF is expanding into itsMission Bay Campus, which willallow the university to performmore research projects for the NIH,and is building QB3, a joint re-search program between UC Ber-keley, UCSF, and UC Santa Cruz.Stanford University has estab-lished BioX at the Clark Center, amulti-disciplinary program to as-sist in the discovery and under-standing of science and medicine,and UC Berkeley is undergoing asimilar transformation, expandingits college of engineering.



WEST REGION


San Diego Regional Profile

The history of biotechnology in San Diego began withtwo pioneers, Howard Birndorf and Ivor Royston,

and the founding of Hybritech in the late 1970s. Ini-tially researchers at UCSD, their use of monoclonalantibodies revolutionized the tedious process of diag-nosing disease. They, along with the others atHybritech, ended up founding many of the biotechcompanies in San Diego. The two are commonly knownas the fathers of San Diego’s biotechnology industry.

Today, the biotechnology industry in the San Diegoarea is all-encompassing and includes pharmaceuti-cals, medical devices, biotechnology, bio informatics,bio-argriculture, and environmental biotechnology.

San Diego is one of the top bioscience centers in thecountry for NIH funding and is regularly listed as oneof the big nine. The area boasts roughly 625 biosciencecompanies and employs approximately 23,000 people.As a region, these numbers compete with many statesand even some countries.

La Jolla and the nearby Sorrento Valley host promi-nent bioscience parks with the who’s who of bioscienceresidents. Pfizer, Novartis, Gen-Probe, Skyepharma,La Jolla Pharmaceuticals, Cardinal Health, Invitrogen,Diversa, Merck, Elan, Tannox, and Amlyin all call thearea home. San Diego is aggressive in courting large

pharmaceuticals as well. An example of this isthe addition of pharmaceutical giantBiogenIdec in the northern San Diego area.

Biogen Idec was created by the merger of Biogen,Inc. of Massachusetts, and Idec Pharmaceuticals, Inc.of San Diego in November 2003. BiogenIdec, nowheadquartered in Cambridge, Massachusetts, is thethird largest biotechnology company in the world. Thecompany employs 4,000 people worldwide. It cur-rently has approximately 20 products either in clinicaltrials or on the market. Biogen Idec maintains a largeR&D facility in San Diego.

Genentech has recently come into the San Diegomarketplace with the purchase of a 90,000 liter biotechmanufacturing facility in Oceanside, California thathad formerly belonged to BiogenIdec. This $380 mil-lion facility is the largest manufacturing commitmentto the San Diego area.

The San Diego area has many draws for biosciencecompanies. Among these are a highly educated talentbase, local university support, tax incentives, plusincredible weather. In addition, San Diego has a largenumber of support companies, including a number ofcontractors and vendors with extensive pharmaceuti-cal experience and expertise.

The Life Sciences Industry in Washington StateContinued from page 27.

careers in the biotechnology industry. In Ellensurg, Central WashingtonUniversity is home to the oldest biotechnology-training programs on thewest coast.

In addition to the educational opportunities, the quality of life in Wash-ington helps attract talent. The natural attractions include three nationalparks, both the Cascade and Olympic mountain ranges within 90 minutesof Seattle, miles of public beaches, and nearly every form of outdoorrecreation one can imagine. Seattle itself is home to professional sportsteams, a ballet, opera, and dozens of professional theater companies.

Wherever you go among biotechnology professionals in Washington,you find connections. More than a dozen biotechnology firms trace theirroots to the University of Washington, including Cell Therapeutics, Corixa,and ZymoGenetics.

At the confluence of Microsoft and a bioscience cluster, genomics andbioinformatics firms are bound to be born and flourish. Geospiza,Rosetta Inpharmatics (a division of Merck), VizX Labs, and the Insti-tute for Systems Biology, founded by Dr. Leroy Hood, are some of these.On the device side, strong ultrasound research at the UW has resultedin nationally-known device companies such as Philips UltrasoundNorth America (formerly ATL), SonoSite, and Siemens Medical Solu-tions Ultrasound.

West OverviewContinued from page 26.

Eleanor Roosevelt Institute, Divi-sion of Vector-Borne InfectiousDiseases, University of Utah,Brigham Young University, andUtah State are the major educa-tion centers that feed the pharma-ceutical industry in the region.Colorado has the largest medical-related redevelopment effort in thecountry with a $4.3 billion “squaremile of life sciences.” Utah’s strong-hold was the early emergence ofthe medical device industry begin-ning in the early 1950s and thesuccessful implant of an artificialheart by the University of Utah in1980.