Functional Safety in the Life Science Industries

SEPTEMBER/OCTOBER 2008 PHARMACEUTICAL ENGINEERING 1

Functional Safety

©Copyright ISPE 2008

This articlepresents anoverview offunctional safetywithin the lifescience industrybased oninternationalstandards.

Functional Safety in the Life ScienceIndustries

by David Hatch, Iwan van Beurden, and Eric W. Scharpf

Introduction

For life science companies, the chemicalsafety of the process (plant design andoperation) is as critical as the pharma-cological safety of the products (drug

quality). The use of flammable solvents, corro-sive fluids, toxic gases, and explosive dustspresent significant threats to the safety ofproduction personnel, the local community,surrounding environment, and often expen-sive manufacturing equipment.

Functional safety contributes toward over-all process safety and relies on the correctreaction (both action and speed) of automaticdevices in response to actual or potential dan-gerous conditions, thus preventing hazardousevents or mitigating harmful consequences.

Instrumented protective functions usingelectrical or electronic technologies achieve thisvia sensors to detect process deviations, logicsolvers to evaluate the sensor data, and finalelements to execute the required action toachieve or maintain a safe state. These SafetyInstrumented Systems (SIS) have been widelyused in the general process and pharmaceuti-cal industries for many years, providing pro-tection against deviations in pressure, tem-perature, level and flow, and other criticalprocess parameters.

Correct management of the functional safetyaspects of process plants is now globally recog-nized as the best way to reduce the inherentrisks in hazardous industrial processes. Inter-national standards are driving major end usersin the process industry to adopt the IEC615111

(ANSI/ISA–84.00.012 equivalent) lifecycle ap-proach to safety. The aim of these standards isto make safety a priority throughout the entirelife of any potentially hazardous plant or pro-cess.

Risk ReductionRisk is a measure of the likelihood and conse-quences of a hazardous scenario when a processgoes out of control or is otherwise compromised,leading to a loss of containment with the subse-quent release of material and/or energy.

Companies have moral, legal, and financialresponsibilities to limit the risk their opera-tions pose to employees and members of thepublic to a level that is considered tolerable.

Determining whether a process plant is “safeenough” may sound easy, but in practice itmeans a very well thought out calibration of theAs Low As Reasonably Practicable (ALARP)risk tolerability principle. ALARP has beendocumented by the UK Health and Safety Ex-ecutive in their R2P2 publication,3 which aimsto provide a methodology for defining targetfrequency and severity (i.e., risk) of hazards toa minimum “tolerable” level.

The ALARP principle states that there is alevel of risk that is “intolerable.” Above thislevel, risks cannot be justified on any grounds.Below this intolerable level is the ALARP re-gion where risks can be undertaken only if asuitable benefit can be achieved. In the ALARPregion, risks are only tolerable if risk reductionis impracticable or if the cost of risk reductionis greatly outweighed by the benefit of the riskreduction that is gained. Below the ALARPregion is the “broadly acceptable” region wherethe risks are so low that no consideration ofthem is warranted and detailed work is notneeded to demonstrate ALARP because therisk is negligible. In addition, in this broadlyacceptable region, risk is so low that no riskreduction is likely to be cost-effective, so a cost-benefit analysis of risk reduction is typicallynot undertaken.

In some countries, the law mandates toler-able risk levels whereas in others, such as the

Reprinted from

PHARMACEUTICAL ENGINEERING®

The Official Magazine of ISPE

September/October 2008, Vol. 28 No. 5

2 PHARMACEUTICAL ENGINEERING SEPTEMBER/OCTOBER 2008

Functional Safety


United States, tolerable risk is determined by each companyor organization and must be adopted consistently. Tolerablerisk cannot be applied on a personal preference basis sinceeveryone has their own view on what is tolerable (consideryour own driving style for example).

Once an estimate is made on the likelihood of an unwantedevent occurring, and the potential consequence of that eventis calculated with a commercial value, the decision on whetherto implement further protection measures is often a straight-forward economical one. If the risk reduction is significantand the cost not prohibitive, then clearly the measure shouldgo ahead. Conversely, if a measure is deemed to offer littleimpact on the overall risk reduction and it is prohibitivelyexpensive, it is perfectly valid to consider the associated riskas “tolerable,” assuming no other risk reduction measuresare practical.

Safety StandardsIEC 61511 has been developed as a Process Sector implemen-tation of the international standard IEC 61508:4 which wasprepared as an ‘umbrella’ standard from which industryspecific standards (such as IEC 61511 for the Process Indus-try and IEC 62061 for the Machinery Industry) could bederived.

For end-users in the Life Science process industries (pri-mary and secondary manufacture), IEC 61511 is applicablewith the broader IEC 61508 limited to those who manufac-ture or supply Safety Instrumented System equipment orcomponents. Hereafter we shall refer to IEC 61511 as ‘thestandard,’ which has two key concepts that are fundamentalto its application: the safety lifecycle process, and safetyintegrity levels which define required and achieved func-tional safety performance.

Safety LifecycleSimilar to GAMP,5 a lifecycle approach forms the centralframework that links together the key concepts of the stan-dard and is acknowledged good engineering practice forSafety Instrumented System implementation.

In order to achieve and sustain functional safety through-out the life of a facility (i.e., from initial conceptual design tofinal decommissioning), a number of technical and manage-ment activities must be performed, reviewed, and docu-mented. Similar in many ways to the ISO 9000 qualityprocess, the execution of the functional safety lifecycle ispresented in Figure 1.

This lifecycle can be classified into three distinct groups ofphases:

Analysis – How Much Safety is RequiredAnalysis focuses on identifying hazards and hazardous events,the likelihood that these hazardous events will occur andtheir potential consequences, the availability of layers ofprotection, as well as the need for any Safety InstrumentedFunctions (SIF) and their allocated Safety Integrity Level.The phase concludes with the development of the SafetyRequirement Specification (SRS) to properly define the re-quirements for all Safety Instrumented Functions.

Implementation – How Much Safety can beAchievedThe implementation phases begin with a conceptual design ofeach Safety Instrumented Function based on equipmentselection, architectural voting configuration, and periodictest interval to achieve the risk reduction defined in theSafety Requirement Specification.

These phases include detailed hardware design and build,software configuration, system integration, and testing priorto delivery to site. Implementation also includes advancedplanning for installation, commissioning and validation, aswell as long-term operation and maintenance.

Operation – How to Sustain SafetyThe operation phases are the longest phases of the safetylifecycle and involve the validation of the Safety Instru-mented System and all its Safety Instrumented Functions toconfirm that it functions as per the requirements in theSafety Requirement Specification.

Following successful validation, the system is put intoservice and must be properly operated, maintained, andtested until permanently taken out of service. During thisphase, all modifications must be fully evaluated and docu-mented to ensure they do not compromise safety.

Management of Functional SafetyLike any execution process, functional safety managementrequires careful forward planning which defines the requiredactivities along with the persons, departments, or organiza-tions responsible to carry out these activities. The mainpurpose is to reduce the risk associated with systematicfailures of specification, design, and procedure execution thatFigure 1. Functional safety lifecycle.


Functional Safety


can lead to harmful accidents.Plans should be updated and related activities adjusted as

necessary throughout the safety lifecycle to reflect any non-conformance, changes in scope, technology, or other influ-ences. Regular independent monitoring and objective audit-ing is key to ensure that proper management is provided tosupport the technical execution of the project.

A key element of resource planning is to ensure that,according to the standard, “Persons, departments, or or-ganizations involved in safety lifecycle activities shallbe competent to carry out the activities for which theyare accountable.”

This competence can be demonstrated with qualifications,experience, and qualities appropriate to their duties andshould include:

• training to ensure acquisition of the necessary knowledgeof the field for the tasks that they are required to perform

• adequate knowledge of the hazards and failures of theequipment for which they are responsible

• knowledge and understanding of the working practicesused in the organization for which they work

• an appreciation of their own limitations and constraints,whether of knowledge, experience, facilities, resources,etc., and a willingness to point these out

Internationally recognized accredited schemes are availablewhich formally establish the competency of those engaged inthe practice of safety system application in the process andmanufacturing industries.

Verification and ValidationPharmaceutical validation is defined as “Establishing docu-mented evidence that provides a high degree of assur-ance that a specific process will consistently produce aproduct meeting its pre-determined specifications andquality attributes.” Safety validation is synonymous withthis principle and we could simply replace the word ‘quality’with ‘safety’ to provide a mission statement for functionalsafety.

In common with pharmaceutical compliance, safety com-pliance adopts the proven principles of verification and vali-dation. Very often these terms are misused as they definesimilar, but fundamentally different concepts.

Verification is defined as “demonstrating for eachphase of the safety lifecycle by analysis and/or teststhat, for the specific inputs, the deliverables meet the

objectives and requirements set for the specific phase”and ensures that the final product meets the original design(low-level checking), i.e., you built the product right. This isdone through procedural cross checks such as inspections,reviews, and audits.

Validation is defined as “demonstrating that the safetyinstrumented function(s) and safety instrumentedsystem(s) under consideration after installation meetsin all respects the safety requirements specification”and checks that the product design satisfies or fits theintended usage (high-level checking), i.e., you built the rightproduct. This is done through dynamic testing and otherforms of challenge or trial.

In other words, verification is an ongoing quality assuranceactivity throughout the lifecycle ensuring that the procedureshave been followed and the Safety Instrumented System hasbeen built according to the requirements and design specifica-tions, while validation is a quality control activity at a specificpoint in the lifecycle, which ensures that the Safety Instru-mented System actually meets the user’s needs.

Information and DocumentationRequirements

In common with process validation principles, accurate infor-mation and documentation underpin the implementation ofa successful project and provide ongoing reference materialfor the support of operating processes.

An important aspect of functional safety management is toensure that the necessary information is available, docu-mented, and maintained in order that all phases of the safetylifecycle can be efficiently executed and that the necessaryverification and validation activities can be effectively per-formed.

Process Hazard and Risk AnalysisEach process has its own inherent risk (i.e., potential to causeharm) by virtue of the chemicals handled (flammability andtoxicity), the operating conditions (pressure and tempera-ture), and the inventory (volume or mass), as well as theconstruction (materials), the location of the plant, and theoccupancy (personnel exposure).

If we consider a typical pharmaceutical process, it oftenhandles dangerous materials, but generally does not operateat extreme pressures or temperatures, and in common withother batch processes, has a limited inventory which istypically the reactor capacity – but the volume of storagetanks should not be discounted as these can often be signifi-cant. However, the threat often comes from the manual orsemi-automatic nature of many processes which requiresome operator intervention and thereby exposure to thehazards of the process as well as the potential for humanerror.

Process Hazard Analysis (PHA) is an established activityin all process industries, including life sciences. Commonlyknown as a Hazard and Operability (HAZOP) study,6 a PHAis only the start of the safety journey – identifying what cango wrong – now we must evaluate and address it.

Executive SummaryCompliance with local, national, and international pro-cess safety regulations can be achieved efficiently andeffectively by following established and well provenfunctional safety standards and principles. Best prac-tice for achieving and sustaining functional safety hasclose parallels with pharmaceutical compliance.


Functional Safety


Allocation of Safety Functionsto Protection Layers

Overall risk reduction is achieved by combinations of inde-pendent layers of protection. These layers may take manyforms – mechanical, instrument/electrical, procedural, etc.,and the standard shows typical risk reduction methods inprocess plants in terms of these “layers of protection” whichare presented in Figure 2.

The Safety Instrumented System is one of many potentialmeasures that can be taken at the “Prevention” level, asopposed to “Mitigation” (cure). Once a “tolerable” level of riskhas been established, an analysis of the layers of protectionshould allow a function-by-function comparison of the haz-ardous event frequency.

Assuming this hazardous event frequency is lower thanwhat is considered tolerable, no additional layers of protec-tion will be required. Conversely, if the frequency is higherthan the tolerable level set, then additional independentlayers need to be applied. One of these layers may be a SafetyInstrumented System.

Safety Integrity Levels (SIL) are order of magnitude bandsof risk reduction. There are four levels defined in the stan-dard ranging from SIL1 with the lowest level of risk to SIL4that provides the highest (and rarest) level of risk reduction.These levels are documented in Table A according to the riskreduction that they provide.

For example, to achieve a tolerable risk of one death in1,000,000 years (indicative value for illustrative purposesonly) from a residual (i.e., with all other protection measurescredited) risk of one death in 50,000 years, the Risk Reduction

Factor (RRF) would be 20 (i.e., 1,000,000 ÷ 50,000) which liesin the SIL1 band. Therefore, we would require a SafetyInstrumented Function with this integrity to achieve therequired risk reduction.

Note that the standard suggests that applications whichrequire the use of a single safety instrumented function of SIL4 are rare in the process industry and that they shall beavoided where reasonably practicable by reviewing the pro-cess design to implement more reliable and (wherever pos-sible) inherently-safe non-instrumented protection measures.

There are various methods of determining the SafetyIntegrity Level for a particular hazardous scenario and theseare generally classified into two types:

• Qualitative• Quantitative

Qualitative methods group numerical targets into more broadcategories of risk reduction, while Quantitative methods givespecific numerical targets for risk. Often qualitative methodsare used for quick initial screening with quantitative meth-

Figure 2. Typical protection layers.

Risk Reduction Factor Safety Integrity Level

10000 – 100000 4

1000 – 10000 3

100 – 1000 2

10 – 100 1

Table A. Risk reduction and safety integrity levels.


Functional Safety


High 2 3 4

Moderate 1 2 3

Low – 1 2

Minor Serious Extensive

Hazardous Event Severity Rating

Table B. Risk matrix (EXAMPLE).

Even

tLik

eliho

od

Hazardous Severity Initiating Initiating Protection Layers Residual Tolerable RRF SAFETYEvent Cause Likelihood Risk Risk INTEGRITY

BPCS Alarms Relief Other LEVEL

Reactor 1 death Loss of 0.05/year 0.1 0.5 0.01 1 2.5 x 10-5 1 x 10-6 25 1rupture cooling water (1 in 20 yrs) (1 death in (1 death in

40,000 yrs) 1,000,000 yrs)

Table C. Layer of protection analysis (EXAMPLE).

ods reserved for higher risk scenarios that require moredetailed investigation and evaluation.

Two basic types of qualitative Safety Integrity Levelselection are commonly used in the process industry:

• Safety Matrix• Risk Graph

The Safety Matrix example in Table B considers the severityof a specific hazardous event (X-axis) against the likelihoodthat the hazardous event will occur when all other creditedprotection layers have failed (Y-axis). The intersection ofseverity and likelihood gives a grade of risk reduction (orSafety Integrity Level) that the Safety Instrumented Func-tion (specific to this hazard) must achieve.

The Risk Graph is a development of the safety matrix witha similar severity (Consequence) grading on the Y-axis andthen consideration of three elements (Exposure, Avoidance,and Demand) that make up the likelihood X-axis of thehazardous event. Both safety matrix and risk graph methodsproduce a Safety Integrity Level, not a specific Risk Reduc-tion Factor.

Quantitative methods are more powerful, but requiremore information. They enable the user to perform sensitiv-ity analysis on all the risk reduction measures, allowing themto identify the weaker protection layers, which may requiremore attention.

One of the most common quantitative methods of RiskReduction Factor determination (and therefore Safety Integ-rity Level selection) is Layer of Protection Analysis (LOPA) asshown in Table C.

This method is used to calculate the risk reduction factorfor a specific hazard based on the unmitigated risk of thehazard severity and initial likelihood, which is then reducedby taking credit for appropriate and independent protectionlayers, each with their own probability of failure (lowerprobability of failure means more reliable), to yield a residualrisk which is then compared to the target tolerable risk. Thestrength of this method relies heavily on having adequate,appropriate, and accurate (not necessarily exact) data, pref-

erably from the users own experience.Further guidance on the determination of Safety Integrity

Level is given in the ISA publication Safety Integrity LevelSelection7 as well as part 3 of IEC 61511 and the AIChE CCPSpublication Layer of Protection Analysis.8

Safety Instrumented System SafetyRequirements Specification

The concept of a User Requirement Specification (URS) iswell understood by pharmaceutical companies who follow theprinciples of Good Automated Manufacturing Practice(GAMP). It is a document (or set of documents), which definesclearly, concisely, and unambiguously what the user requiresand is provided to the supplier as the definitive statement ofwhat the system must do. The URS details functional andnon-functional requirements with the emphasis on the re-quirements themselves (i.e., what) and not the method ofimplementing these requirements (i.e., how).

The implementation of Safety Instrumented System issimilarly documented in a Safety Requirement Specification(SRS), which defines the requirements for the Safety Instru-mented Function(s) within the Safety Instrumented System.These requirements must cover the following three key areasof each function:

• Functionality – what it does• Reliability – how well it does it• Performance – how quickly it does it

The UK Health and Safety Executive conducted a survey9 offailures of computer-based systems and the causes of thesefailures are summarized in Figure 3.

The findings show two key issues. First, nearly 60% offailures are already “in” the system before it even arrives onsite. Second, failures can occur at any time within the lifetimeof a system, it is not just problems that occur due tomaloperation or wear and tear during service. Ironically, it isthe initial stage of the lifecycle where one may expect more‘educated’ personnel to be involved that is the weakest.

The Safety Requirement Specification captures what needsto be done to achieve functional safety and is often a contrac-tual document, which is passed from the User to the Supplierwho is responsible for implementing the Safety InstrumentedSystem.

Safety Instrumented System Designand Engineering

A Safety Instrumented System is a system composed ofsensor(s), for example, pressure transmitters, logic solver(s),


Functional Safety


for example, Programmable Logic Controllers (PLC), andfinal element(s), for example, actuated valves, designed insuch a way as to implement Safety Instrumented Functions.

The Safety Instrumented Functions are specified with aparticular Safety Integrity Level in order to achieve a certainrisk reduction for a defined hazardous event. This in turn setsthe requirements for both hardware and software safetyintegrity of the sub elements in the safety loop by means of atarget probability of failure.

We cannot predict exactly what will happen when, but wecan make educated and well informed judgements regardingwhat is likely to happen both in terms of hazardous eventsand protective equipment failures. We consider that theworst-case scenario of a hazardous event occurring at thesame time as the safety equipment is unavailable has aprobability and that the greater this probability, the greaterthe likelihood of harm.

In order to address this challenge, we must aim to reducethe frequency of the hazardous event occurring and reducethe probability of the protective equipment failing. Every-thing will fail; some will fail sooner than others, but even thenmost ‘reliable’ equipment has the potential (however small)to fail when you really need it most.

If we develop Table A into Table D, we see that the higherthe required risk reduction (i.e., the more ‘unsafe’ the process)then we require a Safety Instrumented Function with ahigher safety reliability (i.e., when we place a demand on theSafety Instrumented Function, there is increased confidencethat it will do what is required).

Although the determination of required safety may bedetermined qualitatively (as a Safety Integrity Level) orquantitatively (as a Risk Reduction Factor), the determina-tion of achievable safety can only be determined quantita-tively as a probability of failure. These probability calcula-tions are performed using a variety of techniques, but all arebased on the following three basic considerations:

1. Reliability – the quality of the equipment used within theSafety Instrumented Function in terms of failures

2. Redundancy – the quantity of equipment used within theSafety Instrumented Function in terms of voting anddiversity

3 Repairability – how often and how thoroughly each SafetyInstrumented Function is tested and faults repaired

Therefore, to achieve a Safety Instrumented Function with

the lowest probability of failure, we should use the bestequipment with multiple combinations and test it as often aspractical. This obviously has a commercial impact, as the bestequipment will often come at a higher price and frequenttesting involves significant maintenance costs as well asproduction interruption with the associated loss of revenue.

Very generic sources of equipment failure data have ex-isted for years in the oil and gas industries, but more manu-facturer data is now being collected, assessed, and publishedby independent evaluation companies and bodies. The mostaccurate failure data comes from your own plant records thatreflect actual devices in actual processes under actual main-tenance regimes.

Further guidance on the confirmation of Safety IntegrityLevel and reliability calculations is given in the ISA publica-tion Safety Integrity Systems Verification.10

Safety Instrumented System Installationand Commissioning

Installation and Commissioning of a Safety InstrumentedSystem is similar to a system within a pharmaceuticallyvalidated process.

Installation Qualification (IQ) is defined as “the docu-mented verification that all aspects of a facility, utilityor equipment that can affect product quality adhere toapproved specifications and are correctly installed.”

For Safety Instrumented Systems, we could simply re-place “product quality” with “process safety” and then con-sider the IQ to include aspects such as undamaged delivery tosite, mechanical completion, and cold (power off) loop check-ing.

Operational Qualification (OQ) is defined as “the docu-mented verification that all aspects of a facility, utilityor equipment that can affect product quality operate asintended throughout all anticipated ranges.”

For Safety Instrumented Systems, we could consider theOQ to include aspects such as hot (power on) loop checking toensure that individual sensors can sense/measure correctlyand individual final elements function correctly.

Safety Instrumented SystemSafety Validation

Safety validation of the Safety Instrumented System is vitalto ensure that all the Safety Instrumented Functions performas required to achieve the necessary risk reduction. Thisactivity is also known as Site Acceptance Testing (SAT) or a

Risk SafetyReduction Integrity Probability of Safety Reliability

Factor Level Failure on Demand

10000 – 100000 4 0.0001 to 0.00001 99.99% to 99.999%

1000 – 10000 3 0.001 to 0.0001 99.9% to 99.99%

100 – 1000 2 0.01 to 0.001 99% to 99.9%

10 – 100 1 0.1 to 0.01 90% to 99%

Table D. Safety integrity levels and probability of failure on demand.Figure 3. Failures of computer-based systems.


Functional Safety


Pre-Start-up Acceptance Test (PSAT).Performance Qualification (PQ) is defined as “the docu-

mented verification that all aspects of a facility, utilityor equipment that can affect product quality performas intended meeting predetermined acceptance crite-ria.”

The acceptance criteria for PQ of a Safety InstrumentedSystem would typically include the confirmation of:

• logical relationships between sensors and final elements(e.g., a deviation detected by a specific input initiates aresponse from a related output)

• time of response between initial detection and final action(albeit in initially clean/ideal service on day one)

Safety Instrumented System Operationand Maintenance

Safety validation of the Safety Instrumented System must becompleted before the hazards are introduced and the systemis put into service. This demonstration that the Safety Instru-mented System can reduce risk is only the first step in ajourney that may last for decades to ensure that the risks arereduced to the defined tolerable level while the facility and itsprotected processes and equipment are in operation (andtherefore continue to present a threat of harm to personnel).

Regular maintenance of equipment is vital to ensure thatthe Safety Instrumented Function is available and capable asand when required and this involves routine inspection andcleaning as well as properly executed proof testing.

The purpose of the proof test is to find component failuresthat are otherwise hidden and make any repairs to restorethe Safety Instrumented System to its fully functional state.It is often assumed that if it works properly, it has not failedand a conventional approach is to check to see if the SafetyInstrumented Function operates and the equipment has notfailed. This is only true for the most part since many proof testprocedures do not completely test all of the equipment usedin the Safety Instrumented System.

Regular and effective proof testing is a key element insustaining functional safety and should be considered asearly as possible within the design of each Safety Instru-mented Function.

Safety Instrumented System Modificationand Decommissioning

As with all quality aspects of regulated facilities, propermanagement of change is vital to ensure that safety is notcompromised by uncontrolled or unevaluated modificationsto the physical (hardware) or functional (software) attributesof a Safety Instrumented System.

Since the purpose of a Safety Instrumented Function is toreduce risk, any change to the risk it must reduce or itscapability to reduce that risk will affect the safety it provides.It is important to note that these changes must includedifferences between the performance of equipment estimatedduring the analysis and design phases relative to its actualperformance in the field.

• Hazard severity – If effects are worse than predicted, therisk reduction requirement is greater.

• Hazard likelihood – If more frequent than predicted, therisk reduction requirement is greater.

• Equipment reliability – If equipment fails more frequentlythan assumed, the risk reduction capability is less.

• Equipment redundancy – If equipment redundancy isreduced, the risk reduction capability is less.

• Equipment repairability – If equipment is tested less fre-quently than declared, the risk reduction capability is less.

Any of these changes must be properly evaluated, docu-mented, and implemented to ensure that the functionalsafety protection is not weakened or eliminated.

The multi-product nature of many pharmaceutical facili-ties means that the same equipment may handle a variety ofchemical regimes with process pressure and temperaturesthat change according to the recipes and production phasesused. Therefore, it is essential that the risk reduction re-quirements and capabilities are properly evaluated for eachplant and processes within that plant.

Decommissioning a Safety Instrumented Function or acomplete Safety Instrumented System is an extreme form ofmodification. The key consideration for decommissioning isto assess the effects of removing some or all of the riskreduction and to ensure that other Safety InstrumentedFunction or non-instrumented protection layers are not com-promised or expected to provide a greater level of risk reduc-tion than they are capable of.

ConclusionsFunctional safety mirrors the principles of process qualityand can be summarized as follows:

• What can go wrong? (HAZOP or PHA)

• How bad can it be? (Risk Analysis)

• What can be done about it? (Safety Integrity LevelSelection and SafetyRequirement Specification)

• How reliable will it be? (Safety Integrity LevelVerification)

• How do I stay safe? (Safety Integrity LevelSustain)

It is vital to know how much safety is actually required andwhat measures are available to achieve it before embarkingon the expensive implementation of Safety InstrumentedSystems, which may actually be unnecessary.

If they are necessary, then Safety Instrumented Functionsmust be appropriately designed, implemented, installed,operated, maintained, and regularly tested in order to opti-mally achieve and continue to achieve the required riskreduction throughout their life.


Functional Safety


Appropriate management must be exercised with compe-tent personnel, accurate data, and proven methods to sup-port the analysis, realization, and sustainment of func-tional safety.

References1. IEC 61511: Functional Safety – Safety Instrumented

Systems for the Process Industry Sector (2003).

2. ANSI/ISA–84.00.01–2004 (IEC 61511–1 Mod): FunctionalSafety: Safety Instrumented Systems for the ProcessIndustry Sector (2004).

3. UK Health and Safety Executive R2P2: Reducing Risks,Protecting People (2001).

4. IEC 61508: Functional Safety of Electrical/Electronic/Programmable Electronic Safety-Related Systems (1998).

5. ISPE GAMP: Good Practice Guide – Validation of ProcessControl Systems (2003).

6. IEC 61882: Hazard and Operability Studies (HAZOPStudies) – Application Guide (2001).

7. ISA Safety Integrity Level Selection – Systematic Meth-ods Including Layer of Protection Analysis (2002).

8. AIChE CCPS: Layer of Protection Analysis – SimplifiedProcess Risk Assessment (2001).

9. UK Health and Safety Executive out of Control: WhyControl Systems go Wrong and How to Prevent Failure(2003).

10. ISA Safety Instrumented Systems Verification – Practi-cal Probabilistic Calculations (2005).

11. AIChE CCPS: Guidelines for Safe Automation of Chemi-cal Processes (1993).

About the AuthorsDavid Hatch is a Partner with Exida andhas more than 20 years of design, commis-sioning, operating, and consulting experi-ence in the life science, energy, and chemicalindustries. This experience covers blue-chipoperating companies, major multinationalengineering contractors, and control andsafety system suppliers. He has a BSc (Hons)

in chemical and process engineering from the University ofStrathclyde (UK), is a Chartered Engineer and Fellow of theInstitution of Chemical Engineers in the UK, and an IChemERegistered Safety Professional. He is a Certified FunctionalSafety Expert, a member of the IChemE Safety and LossPrevention Subject Group, and a specialist in process auto-

mation, hazard analysis, and alarm management; authoringand presenting papers on these key subjects. Hatch is anactive member of ISPE, the Instrument Systems and Auto-mation Society (ISA), and is currently contributing to thedevelopment of an international standard for alarm manage-ment. He can be contacted by telephone: +44-7909-973719 orby email: [email protected].

Exida, Epic House, Eliot Park, Barling Way, Nuneaton,CV10 7RH, United Kingdom.

Iwan van Beurden is the Director of Engi-neering at Exida and previously worked forYokogawa Industrial Safety Systems inApeldoorn, the Netherlands. He worked bothin the R&D and in the Operations Depart-ment as a safety assessment specialist. Hewas involved with the implementation ofIEC 61508 within Yokogawa Industrial Safety

Systems dealing with SIS reliability calculations and thesafety education of engineers. He is a Certified FunctionalSafety Expert, a member of the Instrument Systems andAutomation Society (ISA), and has published various papersand magazine articles. He holds a Master of Science degreefrom Eindhoven University of Technology in Eindhoven, theNetherlands, where he majored in reliability engineeringand graduated cum laude. He can be contacted by telephone:+1-215-453-1720 or by email: [email protected].

Exida, 64 N. Main St., Sellersville, Pennsylvania 18960,USA.

Dr Eric W. Scharpf is a Partner with Exidaand is based out of Port Chalmers, NewZealand. He has more than 10 years of pro-fessional experience in the chemical process-ing industry. Scharpf is recognized as anexpert in chemical process efficiency analy-sis and improvement. He developed severalof the cryogenic gas processing and separa-

tion techniques currently used for hydrogen, synthesis gas,and power generation. He was most recently a lead processchemical engineer with a Fortune 200 gas and chemicalprocessing company. Principal work responsibilities haveincluded engineering research and development, project ex-ecution, safety and risk analysis, and operations/manufac-turing efficiency and reliability improvement in the bulk gasand specialty chemicals industry. He has authored more than10 US and international patents as well as several journalarticles and conference publications. Currently, he teachesprocess optimization and energy management at the Univer-sity of Otago in Dunedin, New Zealand. He has a BChE fromthe University of Delaware and a PhD in chemical engineer-ing from Princeton University, both in the USA. He can becontacted by telephone: +64-3-472-7707 or by email:[email protected].

Exida, 278 Blueskin Road RD1, Port Chalmers 9081, NewZealand.


Process Technology


This articlediscusses thefuture of APIManufacturingand the roleISPE can play indriving industryinnovation.

Innovations in Process Technology forManufacture of APIs and BPCs

by John Nichols

Introduction

This article presents a glimpse at someof the transformational changes im-pacting the technology and processesfor manufacturing Active Pharmaceu-

tical Ingredients (APIs). My main problem inwriting this article was to convey in words theexcitement I feel at seeing these changes start-ing to take place.

The ISPE API Community of Practice hasformed a Process Technology Subcommitteewith a mission to drive innovations in processtechnology for manufacture of APIs and BulkPharmaceutical Chemicals (BPCs) by the iden-tification of subject matter experts, interactionand partnership with related groups; by sup-port and input to the education program; thepromotion/marketing of new innovations inprocess technology; and input to industry guide-lines and best practice documents. I am Co-Chair of this subcommittee and together wehave recruited thought leaders from industryand academia to help define this direction anddrive this change.

The API COP Process Technology Subcom-mittee sponsored a conference in Copenhagenin April of 2008 on Innovations in API/BPCManufacture. This article summarizes thevarious presentations which gave us a glimpseof this future in the following areas:

• Efficient Processing Implementation (e.g.,Continuous)

• New Technology/Equipment• ISPE/ASTM Enabling Activities

Change is Needed and is ComingFor many years, APIs/BPCs manufacturinghas generally existed on batch unit operationswith little difference in principle from thosefound in 3000 to 4000BC. However, things arestarting to change.

If we are looking for the point at which thisstarted, we would most probably identify thatas 2003 with the FDA’s cGMPs for the 21stCentury (announced originally August 2002)and PAT initiative as the obvious critical initia-tors. One of their objectives was to encourageearly adoption of new technological advancesby the industry.9 Emer Cooke as Head of Sector,Inspections EMEA confirmed11 Europe’s simi-lar view. The timing of these initiatives alsocombined with an external business environ-ment of cost/price pressures,10 social pressuresfor “greener” processes, and improved operatorhealth and safety which has accentuated theopportunity. Companies have been hesitant tochange until the new way is seen as “Industry”practice - a “Catch 22” situation. The regulatorswere to that point “seen” as the inhibitors of“change,” but here they were saying the oppo-site. Often quoted is a statement by JanetWoodcock of the FDA that the desired state is:“A maximally efficient, agile, flexible pharma-ceutical manufacturing sector that reliably pro-duces high-quality drug products without ex-tensive regulatory oversight,” along with vari-ous comments on the poor efficiency of thepharmaceutical industry compared to others.Benchmarking the pharmaceutical industryagainst others suggests that there is huge po-tential to increase manufacturing performancewith lower new investment.18,19

Most recently we have seen the ProductQuality Lifecycle Implementation (PQLI) ini-tiative4 bringing together a Quality by Design(QbD) approach based on scientific understand-ing, the perfect basis for developing a processand technology targeted to the needs of themolecular transformations.

A Vision of the FutureProcesses may be designed to be fully ‘continu-ous,’ i.e., material continually flows through an

Reprinted from





Process Technology


integrated system of unit operations without a break. ‘Hy-brid’ operation also is possible with some continuous opera-tions as part of a batch process.

Process Intensification is the use of understanding offundamental science (e.g., heat transfer, mass transfer, effectof gravity) to design highly efficient cost effective smallequipment, e.g., compact heat exchangers/microreactors.

The concept of targeted processes is simple: understandthe required transformations and what drives them in thedesired direction, take a holistic view and select the bestpathway, and design the process and technology to best meetthose objectives.7 However, to have that understanding, thereneeds to be a different approach to product and processdevelopment, a different approach to the use of PAT, andstrong interaction and cohesion between the developmentchemists and the engineers. These are points developed inthe PQLI philosophy, noted by many of the speakers atCopenhagen, and well presented in References 12 and 16.

In Reference 2, Roger Benson discusses Lean Manufactur-ing as an alternative to “Offshoring,” an issue of primecurrent interest. Some of the particularly relevant principlesof lean manufacture are listed below.5 We can use these as abasis for looking at alternative technologies and means ofprocessing.

• The customer’s needs – price, quality, delivery• Simplicity – ease of control• Waste – elimination• Flow – integrated synchronized operations• Postponement – delay activities to the last possible time• Time – reduce time to produce• Variation reduction• Thinking small – specify smallest machine and build in

increments.• Knowledge – build and distribute• To these is added “agility”

The following discussion looks at how targeted production,continuous operation, hybrid batch/continuous, or processintensified operation can help achieve these objectives:

• The operating cost is reduced (see reduced waste below)and new facilities are smaller and have lower capital costgiving reduced product cost - Figure 1.

• Steady state operation in continuous manufacture andtargeted production giving less by-products eases control,improves quality, and reduces variation.

• Reduced inventory in continuous or intensified manufac-ture gives improved delivery, more agility, and reducedwaste.

• Intensified Equipment can be designed to give agility.8

• Steady state gives reduced raw materials (higher yield),utilities and environmental emissions, reduced waste.Reference 15 gives a methodology to factor these intodecision making.

• Targeted production gives less by-product waste.• Continuous or intensified facilities are typically smaller.• In order to have a targeted design, we need scientific

understanding of the transformations, improved knowl-edge. This article does not cover knowledge handling, butreaders could refer to Reference 17 for a general discussion.

• Better control and better scientific understanding willlead to less “bad” batches, better quality.

• The scale of continuous or intensified processes is suchthat that the laboratory proof of concept plant, kilo labora-tory, pilot plant, and clinical trials plant can become oneand the same, the only difference is the duration for whichthey are run. Increase in production can then simply andflexibly be made by adding a second line. This enablesearly production, a critical feature for many pharmaceuti-cal compounds.13

In addition, techniques such as continuous processing andprocess intensification by their nature enable the use ofprocesses that would be impossible or difficult to control in abatch plant.3 From the safety and health point of view, theseimproved techniques can reduce inventory, improve integra-tion, thereby reducing handling, and give better control.

While not applicable to every process due to the nature ofthe transformations necessary, it has been seen that to adoptthese techniques can be very beneficial in a substantialpercentage of cases.

Innovations in API/BPC Manufacture –Copenhagen Conference 2008

The ISPE Copenhagen Conference covered a wide range ofissues pertinent to current and future innovation in API/BPCmanufacture. The Conference showcased the significant op-portunities for improvement in operating efficiency beingimplemented in different areas, including:

• demonstrating where Genzyme, Novartis, and Pfizer arein implementing these new technologies

• highlighting new, available equipment technology• featuring the work being done by ISPE and ASTM to

facilitate this new technology implementation

Efficient Processing ImplementationUnderstanding the Factors Involved –Huw Thomas, Foster WheelerThomas covered the business case for innovation, particu-

Figure 1. Economics of continuous plants. (Courtesy of Foster Wheeler)


Process Technology


larly continuous processing, based on work already com-pleted – Figure 1. He discussed the implications on processdevelopment, plant and process design, qualification, andoperation.6

Aspects to consider are:

• How to chase the money• What is the meaning of scale and scale-up for a continuous

process• Targeted equipment selection - Figure 2• Hybrid implementation• The challenges of small scale equipment• Plant control and the need for buffering• The implications on environmental and safety issues,

qualification, and plant culture.

Continuous Processing, An Overview ofOpportunities and Challenges –Paul Sharratt, Manchester UniversityContinuous processing promises more robust processing andsmaller, more efficient plants for both primary and secondarymanufacturing. Sharratt’s presentation discussed:

• The nature of the benefits of continuous processing – bothtechnical and commercial

• The changes in skill base necessary to access continuousprocessing

• The challenges that continuous processing presents toorganization and management - Figure 3

• Examples of individual continuous technologies and wholecontinuous process.

Challenges in the Validation of ContinuousProcessing – Peter McDonnell, GenzymeThe pharmaceutical industry has been slow to adopt continu-ous processing methods in comparison to other process indus-tries. Recent FDA led initiatives have opened the door toexamining ways of reducing manufacturing costs and in-creasing process understanding through implementation ofQuality by Design principles.

Continuous processing with appropriate controls, offersthe possibility of exercising exquisite levels of control andmanaging variability of inputs to give consistently highquality products - Figure 4. McDonnell summarized Genzyme’sexperience in the validation of their continuous manufactur-ing process for an API and how they have confronted unfamil-iar challenges.

Continuous Manufacturing; The Ultra Lean Wayof Manufacturing – Walter Bisson, NovartisTraditional batch manufacturing has been very successfullyoptimized and offers in the future limited opportunities todrive significant efficiency gains. Transformation from batchto continuous, end to end processing will unlock significantefficiency gains. Together with academia, a blue sky visionhas been formulated by Novartis, which Bisson described anddiscussed. He presented a view of the more distant futurewith a totally integrated end to end process conducted to meetultra lean and high quality attributes.

API Process Technology, Improving theProductivity of the Manufacturing Process –Sarah Mancini, PfizerUse of new technology is being driven by strong cost pressuresin the industry. Manufacturing processes need to achievevery high productivity, i.e., having both low cash and process-ing costs. While past cost improvement efforts have utilizednew, more efficient chemical routes to reduce cost, applica-tion of new technology opens up the possibility to directlyaddress processing cost. This will change process develop-ment from adjusting the process to function in existingequipment, to selecting equipment to meet processing de-mands. The resulting processes will have improved processcontrol, maximized yields, minimized unwanted reactions,increased throughput, and decreased waste.

From new chemical entities to marketed products, effortsare underway to implement new technologies into batchprocesses and to implement fully continuous processes. Pfizeris implementing new technology as part of their processintensification efforts from implementing platform technol-ogy into batch processes to fully continuous process imple-mentation. Mancini’s presentation covered how processesare selected for technology enhancements and provided anoverview of the key technologies that Pfizer is taking forward

Figure 2. Equipment selection – reactor technology. (Courtesy ofFoster Wheeler)

Figure 3. The innovation process.


Process Technology


for implementation into its manufacturing processes. Thiswas followed by a presentation of the facilities that areavailable within Pfizer for new technology development -Figure 5, and the challenges they are encountering as theymove forward with the implementation of new technologies.

New Technology/EquipmentNew Technologies for Safe and ContainedPowder Handling in the Bulk PharmaceuticalIndustry – Frederic Dietrich, Dietrich EngineeringDietrich’s presentation introduced various requirements forand aspects of operating an API plant in regard to powderhandling, and presented various technologies and examplesof technologies, which can be implemented in order to opti-mize a production facility and decrease the operating costs.

The presentation elaborated on a number of points includ-ing:

• comparison of traditional operation with new powderhandling technologies

• criteria for selecting proper technologies respecting thevarious aspects of containment, safety, and quality controlrequirement.

• impact of technologies on the productivity of a plant andhow to decrease the process downtime (faster productchanges for multipurpose operation)

• improvement of processes from batch to semi-continuousprocesses

• technology to handle powders as if they were fluids• design for multipurpose operations in order to adapt

quickly to the market/process requirements. The design ofproduction units should allow fast changes of the processby using mobile and modular technologies.

The presentation introduced various ways to answer to thesechallenges and show how the selection of the appropriateconcept for the powder handling in an API plant can have alarge impact on the productivity of a plant and can lead tohigh cost saving when selected properly.

Batch Reactor Innovations – Jean-MarieEslinger, De DietrichGlass-lined equipment have many well known advantages inAPI manufacturing because the glass is chemically resistant,non contaminating, and easy to clean thanks to the smooth-ness of the finish.

On the other hand, glass-lining manufacturing technologyimplies restrictions in the equipment design, mainly mini-mum radius of curvature, that create dead zones and reten-tion areas that can be problematic in terms of cleaning andcleanability and poor mixing.

New technologies and products were presented, answer-ing the needs of mixing, cleanliness, and cleaning validationin the field of glass-lined equipment, for example:

• new generation of reactors with baffles integrated onto thewall that delete dead zones that are difficult to rinse, and

free up all the nozzles for process or cleaning devicesinstallation

• glass-lined batch reactors that are no more under-baffledand less efficient compared to metallic reactors

• new range of bottom outlet flush valves equipped withcleaning additional port and retractable spray device

• new design of valve seats, avoiding cross contamination• integration of sight glasses directly into the equipment,

without gasket• new type of nozzle flanges, deleting dead zones of tradi-

tional nozzles• glass-lined stainless steel, for a better outside cleanliness

(no painting but polishing)

Continuous Crystallization using OscillatoryBaffled Crystallizer – Xiong-Wei Ni, Nitech/Herriott Watt.In his presentation, Ni reviewed the fundamental theories insolution crystallization and outlined the challenges in indus-trial batch crystallization. Since no complete theory is avail-able to model nucleation/crystallization, their behaviors canonly be anticipated by experimentation; accurate processmeasurements are essential in understanding crystalliza-tion processes. A number of techniques have been imple-mented to monitor crystallization/process variables at labo-ratory scale. [e.g., optical turbidometric UVvis probe formetastable zone width; X-Ray Diffraction (XRD) for poly-morph; Fourier Transform Infrared spectroscopy (FTIR) forsupersaturation; Ultra Sound Spectroscopy (USS) for crystalsize; Focused Beam Reflectance Measurement (FBRM) foronline chord length of crystals and crystal size distribution;differential scanning calorimetry for phase transition; Par-ticle Vision Measurement (PVM) for online crystal shape,Particle Image Velocimetry (PIV) for local velocity]

These techniques/measurements have promoted signifi-cant advances in understanding lab scale batch crystalliza-tion and have assisted in designing better crystallizationprocesses. However these advances have not been matchedby an equivalent increase in our ability and understanding in

Figure 4. Design space and alerts/alarms.


Process Technology


scaling up STR, the workhorse of industrial crystallization.The simple example is that the LINEAR cooling profile hasbeen identified as one of the key operational parameters insolution crystallization, while it is fairly easy to achieve suchprofiles in small STR, but is still a near impossible mission inindustrial scale batch STR with huge volume, inherited non-uniform mixing, and significant velocity/temperature gradi-ents.

He introduced the novel continuous crystallization tech-nology, the Continuous Oscillatory Baffled Crystallizer(COBC), that offers near plug flow conditions under laminarflows with near perfect controllable radial mixing and excel-lent heat transfer coefficient. He noted that it providescontrollable LINEAR cooling profiles that cannot be achievedin large batch operations, allowing well-defined continuouscrystallization.

He then used real relevant specialty chemical and phar-maceutical examples of industrial crystallization in COBC todemonstrate the significant benefits obtained, e.g., consis-tent crystal morphology and size; better filterability; signifi-cantly reduced crystallization time, space usage, utility, andenergy consumption.

Innovative Blending and Drying Technology –Jean-Francois Demeyre, TriaprocessDemeyre presented an example of a new innovative piece ofequipment, the Triaxe, for mixing and granulation processes.The Triaxe is a completely spherical mixer with a novelgyrating and rotating mixing blades, originally developed in1998 for achieving homogeneity in jam.

Demeyre introduced the history of this mixer to date,general theory of mixing, and details of the machine and itspeculiarities.

He then presented the mixing results in mixing viscousfluids, powders, and in granulation.

Results showed:

• It compares favorably against other equipment types, e.g.,ribbon mixer.

• It is a very polyvalent system (interesting if you want tomix some products with changing properties).

• Ability to realize recipes that are not possible in otherequipment.

• Very short mixing time to achieve homogeneity.• Low power consumption.• Capable of effective granulation.• Size, distribution of size, shape, and mechanical strength

can be altered by altering operating conditions.

ISPE/ASTM ActivitiesISPE Activities, the API Community of Practice– John Nichols, ISPEThe presentation began with an explanation of how ISPEactivities associated with the API Community of Practice(COP) fit into the current industry strategic environment,and parallel European industry initiatives for improvementin efficiency/process intensification, in particular how youcan help make a paradigm shift in the API industry’s appli-cation of new technologies, details of the new COP ProcessTechnology Subgroup, its current work and its aims, and howinput to the ASTM standards has been enabled.

The main messages of the presentation were:

• There is a real opportunity for API manufacture to reduceits costs, and ensure its future.

Figure 5. Technology development facilities.


Process Technology


Prof. Prabir Basu Pharma Engineering, Purdue UnivBritest, NIPTE

Carol Bye Senior Dir. Head API PfizerQuality Operations

Nigel Fletcher API Design Foster Wheeler UKProf. Brian Glennon Scale-up of Chemical Dublin

and BioprocessesBenny Goosen Chemist, Process Janssen

DevelopmentRuth Hardy Technical Support Pilot Pfizer

PlantDr. Barry Johnson Equipment Development Alfa LavalRick McCabe Proj. Manager Innovative Pfizer

ProjectsDr. Peter McDonnell Implementing Cont. Genzyme

ProcessingStan Newberger Pilot Plant, API Design CE & ICJohn Nichols Cont. Fine Chem.

Operation, API Design Prof. Åke C Rasmuson Crystallization in Pharma KTH/Royal Institute

Processes of TechnologyDr. Raf Reintjens Coordinator for PI DSMProf. Paul Sharratt Prof of Sustainable Manchester Univ

Processing and BritestProf. Andrzej Stankiewicz Chair EFCE PI Working Delft TU

PartyProf. Bernhardt Trout Dir. Novartis-MIT Cont. MIT

Man. CentreProf. Xiong Wei Ni OFR Science and Herriott Watt Univ/

Development Nitech

Table A. Membership of the API Process Technology Subcommittee.

• To do this, multidiscipline interaction between develop-ment science, engineering, and quality disciplines is es-sential.

• There are lessons available to be learned from similarindustries.

• We also need to note that some of the old paradigms nolonger apply, circumstances have changed. For example:“Production Rates: Plants having a capacity of greaterthan 10 × 106 lb/yr (approx. 5000 T/yr) are usually continu-ous, whereas plants having a capacity of less than 1 × 106

lb/yr are normally batch types,” etc.1 For the pharmaceu-tical and specialty chemical industries, this is no longertrue.

• There is a lot of activity currently to make these improve-ments and various industry bodies are involved.

• Based on sound scientific, engineering, and business prin-ciples, Product Quality Lifecycle Implementation (PQLI)provides a technical framework for the implementation ofQuality by Design (QbD). PQLI helps craft a pragmaticapproach to implementing Q8 and Q8R, Q9, and Q10,using a risk-based approach to the lifecycle of a product –from regulatory submission, to end of life migration.Uniquely, PQLI involves worldwide regulators in thedevelopment and implementation of this critical thinking.

• The ISPE API COP is a good place to get the necessarysupport and action.

• The new Process Technology Group is specifically dedi-cated to making these paradigm shifts. Support it. It canhelp you. Only by working together will we get mostsynergy.

• We should not be afraid of making a change.

Developing a Consensus Standard for theApplication of Continuous ProcessingPharmaceutical Industry (9192) – Trevor Page,Niro Pharma Systems/ASTMPage introduced his reasons for being involved in this initia-tive as: a passion for the concepts of lean thinking and a beliefthat applying continuous processing to pharmaceutical pro-duction has the potential to generate huge benefits.

He explained the background to ASTM, why it is relevantto the pharmaceutical industry, what is E55, what the newstandard on continuous processing addresses, how ISPE hasbeen involved in this, and the next steps for the ASTMstandard.

Important messages were:

• ASTM has a proven track record of developing and deliv-ering common sense standards.

• The standard is being encouraged and supported by theFDA, it meets the objectives of the “cGMPs for the 21st

Century Initiative.”• The FDA believes it should not be developing its own

government specific standard, but should be participatingwith voluntary consensus standards bodies.

• E55 has now been in operation for nearly four years and itsrole is broader than PAT. It includes various interlinked

pharmaceutical standards.• Continuous Processing is defined as:

- Materials are fed into the system at the same time asproduct is removed from the system.

- Material condition is a function of its position withinthe process as it flows from inlet to outlet.

- The quantity of product produced is a function of theduration for which the process is operated and thethroughput rate of the process.

• The standard is addressing key aspects (that may inhibitinnovation implementation) for example:- Process Understanding (how to achieve necessary level

of understanding, the desired state, residence timedistribution, process dynamics/mixing, process con-trol)

- Material Traceability (e.g., lot identification, batchdefinition, composition)

- Cleaning (e.g., degradation, microbiological)• ISPE is involved in providing comment via the API Com-

munity of Practice.• New draft of standard 9192 expected Summer 2008. Aim

to get 60 percent positive vote by December 2008.

The Way ForwardIn looking at the way forward, we see that we are at the stageof early adopters taking action to improve the position, but itis going to need proof of these concepts, general publicity ofthe benefits, and sharing of experience before the movement


Process Technology


becomes more widespread. Knowledge is nothing if it isnot shared.

Corporate culture can be one of the biggest inhibitors toinnovation and we need to overcome this. Innovation isfostered by flexibility, commitment, communication, andcreativity.14 We have a great opportunity at the moment, weneed to work together, and should not be afraid to grasp theopportunity.

One enabler for this, a vehicle for this communication isthe ISPE API COP and in particular, its Process TechnologySubcommittee.

The API Process Technology Subcommittee is currentlyworking on a roadmap identifying all the main issues inhib-iting the introduction of these new technologies, plus someenabling tools and tabling a path forward. The roadmap alsowill include an appendix of potential technologies. In addi-tion, the team will maintain its ongoing input to the ASTMContinuous Processing Standard development. The groupalso has future plans to generate a white paper covering inmore details the particular areas of compliance/quality as-surance and control associated with continuous processing.

Future educational offerings will build on this work at USand EU conferences in 2008 and 2009, and there are longterm plans to expand the API Baseline Guide with sections onprocess intensification/continuous processing and small vol-ume/high potency manufacture.

This collaboration between industry and top-level aca-demics highlights the role ISPE can play in advancing manu-facturing technology, and building these strategic partner-ships.

For further information, visit ISPE’s Web site underCommunities of Practice (www.ISPE.org).

References1. Douglas, James M, Conceptual Design of Chemical Pro-

cesses. New York, McGraw-Hill. 1988, p. 108. Under theheading, “Guidelines for Selecting Batch Processes.”

2. Benson, Roger S., ““Offshoring” Life Science Production isnot the Only Answer,” Pharmaceutical Engineering, July/August 2007, Vol. 27, No. 4, p. 22.

3. Vasen, Daniela, “Taming a Powerful Couple,” TCE, July2008, pp. 32-34.

4. PQLI various e.g., “ISPE Leading PQLI Industry Initia-tive,” Pharmaceutical Engineering, September/October2007, Vol. 27, No. 5, and ISPE Web site: www.ISPE.org.

5. Bicheno, John, “The New Lean Toolbox Towards Fast,Flexible Flow,” PICSIE Books 2004.

6. Thomas, Huw, “Coming of Age,” TCE, July 2008, p. 38-40.7. Davison, Sushma, “The Whole Picture,” TCE, July 2008,

pp. 42-43.8. Patel, Mayank, Ashe, Robert, “Marrying Efficiency with

Flexibility,” TCE, July 2008, pp. 35-37.9. Phillips, Joseph X., “Update on FDA Pharmaceutical

cGMPs for the 21st Century – A Risk based approach.”Pharmaceutical Engineering, May/June 2006, Vol. 26,No. 3, pp. 108-110.

10. Hall, Stephen and Stoker, Andy, “APIs: Current Trends,”

Pharmaceutical Engineering, May/June 2003, Vol. 23,No. 3, pp. 112-124.

11. Cooke, Emer, “Pharmaceutical Engineering InterviewsEmer Cooke” Pharmaceutical Engineering, January/Feb-ruary 2006, Vol. 26, No. 1, pp. 44-50.

12. Goeffroy, Jean-Marie, “Use PAT to Gain a FundamentalUnderstanding of Your Process and then Move Forwardto Knowledge Development,” Pharmaceutical Engineer-ing, January/February 2004, Vol. 24, No. 1, pp. 14-22.

13. Hall, Donald R., “Plants on Demand,” PharmaceuticalEngineering, May/June 2004, Vol. 24, No. 3, pp. 24-34.

14. Odum, Jeff, “Technology Transfer as a Strategic Tool:Bridging the Valley of Death,” Pharmaceutical Engineer-ing, September/October 2007, Vol. 27, No. 5, pp. 18-26.

15. Rogers, Joseph E.L., “Merging Environmental, Healthand Safety Costs into a Decision-Making Model,” Phar-maceutical Engineering, March/April 2002, Vol. 22, No. 2,pp. 56-64.

16. Davis, John R., and Wasynczuk, John “The Four Steps ofPAT Implementation,” Pharmaceutical Engineering,January/February 2005, Vol. 25, No. 1, pp. 10-22.

17. Zhao, C., Jain, A., Hailemarian, L., Suresh, P., Akkisetty,P., Joglekar, G., Venkatasubramanian, V., Reklaitis, G.V.,Morris, K., Basu, P., “Toward Intelligent Decision Sup-port for Pharmaceutical Product Development,” Journalof Pharmaceutical Innovation, September/October 2006,Vol. 1, No. 1-2, pp. 23-35.

18. Benson, Roger S., McCabe, Jim D.J., “From Good Manu-facturing Practice to Good Manufacturing Performance,”Pharmaceutical Engineering, July/Aug 2004, Vol. 24, No.4, pp. 26-34.

19. Benson, Roger S., “From World Class Research to WorldClass Manufacturing: The Challenges,” PharmaceuticalEngineering, September/October 2005, Vol. 25, No .5, pp.100-108.

About the AuthorJohn Nichols has more than 35 years ofexperience in the chemical engineering field,and is an expert in the areas of high contain-ment and process design having facilitatedand coordinated multi-national teams forbiochemical, fine chemical, and pharmaceu-tical API projects. He recently retired fromhis role as Director of Pharmaceutical Tech-

nology at Foster Wheeler after more than 25 years. In thisrole, he was responsible for global technology coordinationand consulting to API facilities for specific clients and projects,as well as being a Director of Britest. Previously, Nichols wasa Director of Carlisle Consulting and Engineering Directorfor Extract Technology Ltd. a manufacturer of a wide varietyof containment equipment. He is also former Manager ofPharmaceutical Engineering for Foster Wheeler’s ReadingOffice, where he was responsible for the staff involved in thedesign of bulk pharmaceuticals, biochemicals, and secondaryfinishing facilities. Prior to that, he had 10 years experiencein design, commissioning, and plant management of various


Process Technology


batch and continuous fine chemical facilities. Nichols isa Chartered Chemical Engineer and a Fellow of the Institu-tion of Chemical Engineers. He has a BSc in chemical engi-neering from Imperial College, London, and a Diploma inmanagement from Thames Polytechnic, London. He is amember of ISPE’s International Board of Directors, Co-Chairman of the API COP Subcommittee on Process Technol-ogy, and a member of the EFCE Working Party on ProcessIntensification. He can be contacted by email: [email protected].


OSD Revision Baseline® Guide Executive Summary


This executivesummaryprovides anoverview of thesecond editionof the ISPEBaseline® Guide:Oral SolidDosage Forms.The revision ofthe Guide wasprompted by anumber ofdevelopmentswithin theindustryrequiring theguidance to berealigned andrefreshed.

Visitwww.ISPE.orgfor additionalinformationabout this Guideand othertechnicaldocumentsavailable fromISPE.

ISPE Baseline® Pharmaceutical Engineering Guidefor New and Renovated Facilities Volume 2:Oral Solid Dosage Forms (Revision) –Executive Summary

1 Introduction1.1 BackgroundThe design, construction, commissioning, op-erations, and qualification of Oral Solid Dos-age (OSD) facilities are significant challengesfor manufacturers, engineering professionals,and equipment suppliers. These facilitiesshould meet Current Good Manufacturing Prac-tice (cGMP) regulations while complying withall other governing codes, laws, and regula-tions.

The cost of bringing these facilities on-linehas been rising, in many cases due to inconsis-tent interpretation of regulatory requirements.ISPE and engineering design and constructionprofessionals from the pharmaceutical indus-try have entered into a partnership with theFood and Drug Administration (FDA) and Eu-ropean Agency for Evaluation of MedicinalProducts (EMEA) to jointly develop a commonunderstanding and interpretation of GMP re-quirements for production facilities. This Guideis intended to offer a consistent framework forinterpretation, while still allowing a flexibleand innovative approach to facility design, con-struction, commissioning, and qualification.This approach will allow manufacturers to bet-ter serve their customers by helping to reduceproduction cost and maintenance, and improveproduct quality, thus make more funds avail-able for the discovery of new and innovativemedicines. Additionally, this Guide will pro-vide an overview of potential new technologieswhich are being applied selectively in the in-dustry. The reader should consider this Guideas a tool to use in conjunction with designguides that are already available in the indus-try.

This is the revision to the ISPE Baseline®

Guide, originally published in February 1998

for new and renovated OSD facilities. Asidefrom focusing on compliance with the currentFDA cGMP regulatory expectations, whereapplicable this Guide also discusses the differ-ences between the US, European, and Japa-nese practices. Some major changes to the Guideinclude:

• Addition of the new chapter on ‘risk manage-ment and risk assessment.’ This chapterdiscusses, at a high level, approaches relat-ing to assessment and management of riskswithin the lifecycle of the facility design andconstruction project

• expanded definitions and discussion of prod-uct risk of contamination relating to the‘levels of protection’ concept

• the chapter on ‘product and processing con-siderations’ has expanded to include emerg-ing processing technologies, containment,MES, PAT, and cost considerations

• introduction of area definitions and zoningconcepts with architectural detail

• updated the current philosophy of commis-sioning and qualification

• the appendices provide an overview of therisk assessment process as it pertains tonon-GMP subjects such as HSE and con-trolled substances

• the facility cost section has been moved tothe appendix, which will include informa-tion on lifecycle cost, and it will be refer-enced by appropriate chapters

• demonstrate and introduce methods to im-prove or maintain process

• deeper focus on quality and cost aspects

1.2 Scope of this GuideThis Guide may be used by industry for theplanning, design, construction, and qualifica-

Reprinted from







tion of new OSD facilities. It is neither a standard nor acommissioning design guide. It is not intended to supersedegoverning laws or regulations which apply to facilities of thistype, nor is it intended to apply to existing facilities whichmay fall short of the baseline described. The use of thisdocument for new facilities and major renovations of existingfacilities is at the discretion of the facility manufacturer oroperator and the current status of codes and standards mayaffect the appropriateness of certain design solutions.

The Guide covers pharmaceutical facilities for the manu-facture of oral solid dosage forms including tablets, capsules,and powder. It may also be applied to clinical supply facilities.It is not intended to address the manufacture of dietarysupplements, excipients, sterile products, topicals, oral liq-uids, or aerosols. Guidance on facility requirements for themanufacture of excipients can be found in the Baseline®

Guide, Volume 1 – Active Pharmaceutical Ingredients (Sec-ond Edition), while sterile products, topicals, and oral liquidsand aerosols are being addressed in the Baseline® GuideVolume 3 – Sterile Manufacturing Facilities. Wherever ap-plicable, references are provided to existing Baseline® Guidesfor further detail on specific systems or operations.

The Guide is intended primarily for facilities meeting theregulatory requirements to supply the United States (US)market, and follows US standards and references. Whereapplicable, pertinent European and other non-US standardsare referenced in the appendices.

The concepts proposed constitute a Guide or base pointfrom which to proceed. ISPE and FDA have agreed that thisdocument is an acceptable guide to achieving regulatorycompliance. Other applicable codes, standards, and govern-ing laws still apply. These are mentioned for completeness,and for where they may impact upon GMP design issues.

1.3 Key Features of this GuideThe following key concepts are a basis for this Guide:

• proper application of facility design and procedures toassist with GMP compliance

• risk-based approach• non-GMP technology and its impact upon facility design

and costs• contamination risk as assessed by the manufacturer• design conditions versus operating range• good engineering practice• enhanced documentation

Proper Balance of Facility Design and Procedures: it isnot necessary to address each GMP issue only by design; oneor more of the following may be applied:

• procedural control• facility layout• containment and barrier technologies

This will allow the flexibility to design for appropriate levelsof protection or containment, while avoiding costly designsthat result in no significant improvement in quality, efficacyof the drug product, or protection of personnel. For example,based upon an assessment of contamination risk within atableting room, one or more of the following may be appliedto prevent contamination:

• airlocks• multiple pressurization levels• one-way personnel flow• special gowning procedures• special cleaning procedures

The risk-based approach involves using innovative manufac-turing science and technology to assess, mitigate, and controlthe potential hazard in a manufacturing process that affectthe quality of the drug product. As an example, utilizingstatistical data analysis in conjunction with Process Analyti-cal Technology (PAT) for continuous process monitoring andcontrol will lead to higher quality product. Sharing such riskmitigation strategies with the FDA may be beneficial. (SeeSection 17, reference 1, ICH Q9 Quality Management Sys-tem.)

Why a Revision?

This Guide has been revised both to reflect changes inregulations and industry practice, and refers to guid-ances such as ICH Q8 and ICH Q9.

It supports taking a risk-based approach and a newchapter on ‘risk management and risk assessment’ hasbeen added to discuss, at a high level, approachesrelating to assessment and management of risks withinthe lifecycle of the facility design and constructionproject.

The appendices now provide an overview of the riskassessment process as it pertains to non-GMP sub-jects, such as HSE and controlled substances.

The facility cost section includes information onlifecycle cost.

The chapter on ‘product and processing consider-ations’ has expanded to include emerging processingtechnologies, containment, MES, PAT, and cost con-siderations.

The Guide takes into account that several new ISPEBaseline® Guides have been developed and othersrevised. It relates closely to both the new Risk-BasedManufacture of Active Pharmaceutical Products (Risk-MaPP) Baseline® Guide and the Good EngineeringPractice (GEP) Good Practice Guide.

The Guide aims to define product protection require-ments based on product risk factors and the discussionof product risk of contamination relating to the ‘levelsof protection’ concept has been expanded.

New sections have been added to discuss areadefinitions and zoning concepts with architecturaldetail.




Non-GMP Technologies: some facility design require-ments arise from decisions made to address non-GMP issuesor preferences of the manufacturer, such as operator safety orstrategic operating decisions. These non-GMP driven tech-nologies often affect facility design features aimed at achiev-ing GMP compliance and are discussed in Chapter 10 of thisGuide.

With proper planning, both GMP and non-GMP risk as-sessments may be completed in parallel so that key driversfor capital investment are included in the project scope.

The level of protection required is based upon the risk ofcontamination as assessed by the manufacturer, and assess-ment criteria include:

• The duration of product exposure

• The product mix and product changeover (productchangeover is the frequency of change of product processedin a room or in a piece of equipment)

• The characteristics of those products, such as potency ortoxicity

• Human activities performed during the manufacturingprocess

• Facility design and performance factors• Environment in which the plant is located

Operating Conditions are based upon product acceptancecriteria, while design set points and conditions are targetvalues for the engineering designer to achieve. For example,a blending room may have a setpoint of 40% RH and a designrange of 30 to 50% RH (Relative Humidity), but the product

Oral Solid Dosage Forms Baseline® Guide Revision: Table of Contents

1 Introduction1.1 Background1.2 Scope of This Guide1.3 Key Features of This Guide1.4 How to Use This Guide

2 Concepts and Regulatory Philosophy2.1 Introduction2.2 Regulatory Philosophy2.3 Science-Based Quality Risk

Management2.4 Risk Management Approach2.5 Principles of Risk Management

for OSD Facilities2.6 Quality Risk Management Process

for OSD Facilities

3 Product Protection3.1 Potential Product Protection

Requirements3.2 Design Concepts for Product

Protection3.3 Process Control Concepts

4 Product and Processing4.1 Introduction4.2 Preliminary Design Considerations4.3 Impact of Powder Physical and

Chemical Properties on EquipmentDesign

4.4 General Equipment Considerations4.5 Overview of Isolation and

Containment Technology4.6 Architectural and Facility

Interface Considerations4.7 Material Handling4.8 Process Unit Operations4.9 Materials Handling and Storage4.10 Process Utilities4.11 Supporting Unit Operations4.12 Regulatory Requirements4.13 Emerging Technologies

5 Architectural5.1 Introduction5.2 High Level Considerations5.3 Functional Areas5.4 Layout5.5 Surface Finishes and Materials of

Construction5.6 Other Considerations

6 Process Support and Utilities6.1 Introduction6.2 Approach6.3 OSD Process Support Systems

and Utilities6.4 Supporting Management Systems

7 HVAC7.1 Introduction7.2 HVAC Description7.3 Environmental Standards and GMP7.4 Design Considerations7.5 Non-GMP Considerations7.6 Room Air Systems7.7 Commissioning and Qualification

Consideration7.8 Lifecycle Cost Considerations

8 Electrical8.1 Introduction8.2 Power Distribution Overview8.3 Electrically Classified Areas8.4 Identification and Labeling8.5 Lighting8.6 Wiring Methods8.7 Motors and Drives8.8 Grounding8.9 Preventive Maintenance

9 Control and Instrumentation9.1 Introduction9.2 Monitoring and Control9.3 Validation

10 Other Considerations10.1 Introduction and Non-cGMP Risk

Considerations10.2 Strategies to Control Employee

Exposures to Hazardous Materials10.3 Preventing Physical Injury and

Life Safety10.4 Hazardous Operations10.5 Environmental10.6 Community Response and

Emergency Preparedness10.7 Controlled Substances

11 Revision of the Commissioning andQualification Baseline Guide11.1 Introduction11.2 Principles and Concepts of ASTM

2500

12 Appendix A: Cost Factors In OSDManufacturing12.1 Introduction12.2 Lifecycle Costing12.3 Cost Drivers

13 Appendix B: Summary of Quality RiskManagement Process

14 Appendix C: Risk Management Tools14.1 Introduction14.2 Example Risk Assessment for an

OSD Facility Upgrade

15 Appendix D: Other Considerations15.1 General Regulatory Approach by

Region15.2 Japanese Laws and Regulations

16 Glossary and Acronyms16.1 Glossary16.2 Acronyms and Abbreviations

17 References




Volume 5 is being revised to align with the new ASTM E2500,which focuses on risk-based approach for the verificationprocess, ‘a new umbrella term that encompasses commission-ing and qualification’.

Four Appendices have been added to provide more detailon some relevant aspects related to OSD facilities:

• Appendix A contains discussion and guidance on thelifecycle cost analysis relating to optimizing the design ofthe facility and selection of equipment and system.

• Appendix B contains an summary of the quality riskmanagement process for an OSD facility.

• Appendix C contains an example of risk assessment for anOSD facility.

• Appendix D contains HSE Regulatory and ConsensusStandards, Codes, and Guidelines References from theUnited States, the EU, Canada, and Japan.

2 Concepts and Regulatory Philosophy2.1 IntroductionThis chapter is intended to provide guidance on regulatoryphilosophy as applied to the design of Oral Solid Dosageprocess systems and facilities. Understanding and interpre-tation of regulatory requirements and guidance is key, as theengineering solutions adopted will affect both the initial costand operating costs throughout the life of the facility.

GEP can help to ensure that the products meet the re-quired standards of quality and purity.

2.2 Regulatory PhilosophyRegulation of the pharmaceutical industry is conducted bynational and international agencies, for example, the USFood and Drug Administration (FDA), the European Medi-cines Agency (EMEA), and the Ministry of Health, Labour,and Welfare (MHLW) in Japan. A list of international andnational regulatory agencies can be found on the FDA homepage: http://www.fda.gov/oia/agencies.htm.

These agencies are empowered by legislation, such as theUS Food, Drugs, and Cosmetics Act (FD&C Act), and the EU

Table A. Facility designation.

Type Facility Processing Methodology

1. Dedicated - Facility manufactures products from a singleActive Pharmaceutical Ingredient (API)

- Same API all the time- The equipment and facility are dedicated- Dedicated Operation

2. Multi-Use - Facility manufactures products from differentAPIs, but they use dedicated equipment

- Facility has a mix of APIs- Dedicated equipment in a multi-use facility- Concurrent Operation

3. Multi-Purpose - Facility manufactures products from differentAPIs, but specific equipment is not dedicated toany product

- Different products/changeover within the facility- Multi-purpose facility and multi-purpose

equipment- Both concurrent and campaign operation

in that room may be unaffected by humidity in the range of 20to 70% (validated product acceptance criteria). The accept-able operating range for the room is therefore 20% to 70%, not30% to 50%. Additionally, non-product requirements, such ashuman comfort, are also criteria for the design. This isdiscussed further in Chapter 2 of this Guide.

Good Engineering Practice (GEP) is defined as engi-neering practices that are applied throughout the businessto provide organization and control, balance risk and cost,and ultimately deliver appropriate and effective solutions.

The term GEP is used to describe an engineering manage-ment system that is expected in a pharmaceutical enterprisebut not mandated by GxP regulations. GEP recognizes thatall systems in a facility undergo some form of commissioning,which include inspection, testing, and documentation basedon agreed protocols, while direct impact systems requireenhanced documentation which include an enhanced designreview, and Quality Assurance inspection and approval thatare appropriate and acceptable to regulators. GEP capital-izes upon this by suggesting that manufacturers engage allstakeholders (engineers, managers, operators, Quality As-surance experts, and others) very early in the planning,design, construction, and commissioning and qualificationphases to ensure that systems are documented only once.This is discussed further in Chapter 6 and Chapter 11 of thisGuide.

A key element of good engineering practice is to developappropriate documentation throughout the life of the projectto ensure that the equipment and facility is fit for its intendeduse. The documentation should be reviewed, approved byappropriate subject matter experts, updated in a timelyfashion, and stored in a secured location for retrieval.

1.4 How to Use this GuideThis Guide for Oral Solid Dosage Forms facility design isorganized in a format similar to that followed by other ISPEBaseline® Guides.

Chapter 2 ‘Concept and Regulatory Philosophy’ provides ageneral overview of regulatory expectations for OSD facili-ties, and identifies key design requirements for regulatorycompliance. Chapter 3 ‘Risk Management’ discusses a highlevel approach relating to assessment, control, communica-tion, and review of risks to the quality of the product withinthe lifecycle of the facility design and construction project.These two chapters provide the regulatory framework for thedesign model to follow.

Subsequent chapters provide guidance on engineeringsystems such as HVAC, material and equipment selection,and installation. Chapter 10 ‘Other Considerations’ focuseson implementing risk assessment and risk managementstrategies in the design for compliance with appropriateHealth, Safety, and Environment (HSE) regulations. Chap-ter 11 ‘Commissioning and Qualification’ provides the ap-proach for commissioning the facility to properly qualify thedirect components of an OSD facility. This discussion pro-vided in this chapter is in alignment with the Baseline® GuideVolume 5 ‘Commissioning and Qualification.’ At this writing,




Directive 2003/94/EC. The legislation requires the pharma-ceutical industry to follow current Good Manufacturing Prac-tice (cGMP), e.g.:

The FD&C Act, Chapter V, subchapter A, section 501states: “A drug ...... shall be deemed to be adulterated —(B) if ...... the methods used in, or the facilities or controlsused for, its manufacture, processing, packing, or hold-ing do not conform to or are not operated or adminis-tered in conformity with current good manufacturingpractice to assure that such drug meets the requirementsof this Act as to safety and has the identity and strength,and meets the quality and purity characteristics, whichit purports or is represented to possess; ...”

Paragraph 1 of the EC directive states: “All medicinal prod-ucts for human use manufactured or imported into the (Euro-pean) Community are to be manufactured in accordance withthe principles and guidelines of good manufacturing practice.”

CGMP is set out in US Code of Federal Regulations Title21, referenced as 21CFR210 and 21CFR211. The Center forDrug Evaluation and Research (CDER) also issues guidancedocuments, which represent the FDA’s current thinking on aparticular subject.

In Europe, cGMP is set out in EUDRALEX Volume 4 –Medicinal Products for Human and Veterinary Use: GoodManufacturing Practice. Chapter 3 covers premises andequipment.

In Japan, GMP is required for manufacturing and market-ing approval under Article 14.2, paragraph 4 of the Pharma-ceutical Affairs Law. GMP requirements for OSD facilitydesign are in MHLW Ordinance ‘Regulations for Buildingsand Facilities for Pharmacies, etc.’

OSD facilities may be designed and constructed to operateunder different production scenarios - Table A.

The products manufactured in OSD facilities may vary indosage form and degree of product hazard. The equipmentused to produce OSD forms can range from open-type manualprocessing to enclosed, highly automated processing. Thiscan lead to numerous possible facility layouts, but all aregoverned by the same basic GMP requirements for the de-sign, construction, and validation of OSD facilities and equip-ment:

• should be of suitable size, construction, and layout to allowall required manufacturing operations; personnel, prod-uct, and equipment movement; and permit effective clean-ing and maintenance

• should be designed with adequate space and orderly flowto prevent product mix-ups and product cross-contamina-tion

• should provide protection of the product from chemical,physical, microbiological, and all environmental contami-nation

• should be designed and operated with facilities for breaks,toilets, hand washing, and garment changing, provided asappropriate for product protection

• should include specific precautions to ensure that hazard-ous materials do not present an unacceptable level ofproduct cross-contamination risk or a risk to personnel orthe environment

• elements of the facility and equipment which are criticalto product quality should be qualified

These general principles represent the regulatory philosophythat drives the basic requirements within OSD facilities toensure product quality and operator safety during the manu-facturing process.

OSD facilities have common issues across the differentunit operations and product types encountered. Dust con-tainment often presents a design challenge for OSD forms,where highly hazardous active ingredients are becomingmore common. This Guide is intended to help establishconsistent and minimum parameters for facility design, whichaddress these concerns and meet GMP requirements.

This chapter outlines guidance on the following concepts,which are further developed in Chapters 3 through 12:

• risk-based approach• product protection requirements• design concepts for product protection• process control concepts• commissioning, qualification, and validation

Each manufacturer should define the level of control, protec-tion, and validation appropriate to each manufacturing op-eration, based upon a sound understanding of the productand process critical parameters. They should determine therisk of product contamination to the product mix within eachmanufacturing area.

When existing facility renovations or modifications aremade or manufacturing procedures are changed, the natureof the changes should be evaluated in advance to determinehow they may affect patient safety and product quality.Appropriate change control procedures should be followed inmaking any change to qualified equipment, systems or vali-dated processes. In addition, depending upon the extent ofthe change and its potential to affect patient safety andproduct quality, the governing regulatory agency may re-quire notification or prior-approval of a change before it isimplemented. Understanding the potential impact of a pro-posed change and the corresponding regulatory require-ments is critical to maintaining facilities, equipment, andprocesses in a qualified and validated state.

3 Product ProtectionThe ISPE Baseline® Guide: Risk-MaPP describes a scientificrisk-based approach for the prevention of cross-contamina-tion that will, on a case-by-case basis, determine the need fordedicated or segregated facilities in the manufacture ofpharmaceutical products. For further information see section2.4.4 of this Guide.




4 Product and ProcessingThis chapter presents a guide to process and equipmentchoices available that meet the demands for OSD forms innew or renovated pharmaceutical manufacturing, formula-tion development, and pilot plant operations. Aspects ofcommonly used technologies, unit operations, and associatedequipment that are pertinent to product quality and facilitydesign are reviewed. In addition, material characteristicsand properties, material handling, cleaning, and mainte-nance are addressed. The application of PAT and Manufac-turing Execution Systems (MESs) is considered.

Decision trees are used to guide the reader through theoptions to a baseline operation and equipment training selec-tion for a variety of process requirements and material types.Differences between European, Japanese, and US practicesare discussed. For further information, see the ISPE Baseline®

Guides on Active Pharmaceutical Ingredients (API) andPackaging, Labeling, and Warehousing Facilities.

5 ArchitecturalThis chapter provides a guide for requirements to be consid-ered for OSD facility design and construction regardingbuilding programs, building layouts, space definition, de-tails, and finish materials. These aspects of the facility aredeveloped in the context of cGMP, risk, and process require-ments to establish baseline guidance and parameters.

6 Process Support and UtilitiesThis chapter discusses the categorization of mechanical sys-tems used in OSD pharmaceutical manufacturing. There aretwo broad categories:

1. process support systems2. utility systems

This chapter discusses the categorization of mechanical sys-tems using a risk-based approach methodology. This pro-vides a robust methodology for the identification of processsupport and utility systems. It also provides a method for thedetermination of system applications and the selection ofresulting commissioning and qualification strategies. Ex-amples for each type of system are discussed.

Other considerations which impact system design, includ-ing user requirements, engineering design elements, andstart-up requirements also are addressed. The discussionalso focuses on GEP for utility system design. Approaches formultiple-use requirement scenarios also are addressed.

A listing of typical OSD utility systems is provided withgeneral descriptions and requirements for each system.

7 HVACThis chapter focuses on the GMP requirements for HVACsystems. It provides guidance on the design of HVAC sys-tems based upon clearly defined user requirements (e.g.,level of product protection, product and process require-ments, and architectural design). Non-GMP requirements,such as operator protection, expectations on monitoring,

energy efficiency, and safety requirements also are ad-dressed.

HVAC systems can help mitigate risks to both the productand the people who work around it. Their contributionsinclude: GMP risks and non-GMP risks.

HVAC designers should understand cGMP regulatoryrequirements and also be familiar with industrial HVAC asdefined in various documents by the:

• American Society of Heating, Refrigeration and Air-Con-ditioning Engineers (ASHRAE)

• American Conference of Governmental Industrial Hy-gienists (ACGIH)

Knowledge of all local construction codes, the National FireProtection Association (NFPA) standards, environmentalregulations, and Occupational Safety and Health Adminis-tration (OSHA) regulations is also assumed. The design andinstallation of the HVAC system should comply with theseand all applicable building, safety, hygiene, and environmen-tal regulations.

8 ElectricalFrom the electrical perspective, the most important issues tobe addressed in an OSD facility include:

• cleanability of all exposed electrical equipment• flush lighting should be used wherever possible• all conduits and raceways should be hidden (not exposed)

in the production areas

Although the degree may differ based upon the level ofprotection required, the cleanability of the exposed electricalequipment is the primary concern of electrical systems in anOSD facility.

Electrical power distribution systems do not directlyaffect the quality of Oral Solid Dosage (OSD) pharmaceuti-cal drugs and hence are not critical systems and are notsubject to regulatory oversight and validation requirements.The ‘equipment,’ which produces and controls the pharma-ceutical processes and provides clean and controlled envi-ronments for the manufacturing areas for OSD pharmaceu-tical drugs, requires a source of electric power and anelectrical power distribution system that are reliable andmaintainable.

A properly designed electrical distribution system shouldprovide reliable electricity to the OSD pharmaceutical equip-ment. This Guide provides design and maintenance criteriato assist in the proper design of an electrical system to providereliable electrical service.

9 Control and InstrumentationThis chapter considers Control and Instrumentation (C&I)systems for OSD manufacturing facilities and focuses on:

• those facility and environment controls which affect pa-tient safety and product quality




• the major topics that drive decisions regarding the designand set-up of Process Control Systems (PCSs)

The objective is to provide design guidance, which results incost-effective system designs, capable of being qualified.

C&I systems are used in many facility systems. They maybe deemed to affect patient safety and product quality if theycontrol, monitor, or record a CPP or directly affect a CQA.Components of C&I systems may also be considered criticalif they come into direct physical contact with the product.

The functions described may be combined within a singleC&I system, or be performed by several independent sys-tems.

Specific design advice has been given where possible, butit is stressed that different operational preferences andpriorities will influence the preferred solution.

The designer also should consider other relevant designcriteria, such as safety, reliability, and design for mainte-nance.

Process Control Systems encompass a wide variety ofsystems such as, PLC, SCADA, DCS, and MES. For furtherinformation see the GAMP Good Practice Guides on Valida-tion of Process Control Systems (VPCS) and ManufacturingExecution Systems (MES) (see Section 17, reference 13).

Control systems may be complex and validation strategiesshould be based on a defined risk-based approach. Testingstrategy should be based on a predefined estimate of the levelof risk, providing traceability and information allowing criti-cal parameters to be determined.

Process Analytical Technology (PAT): the FDA pub-lished a PAT framework Guidance for Industry in 2004 (seeSection 17, reference 28). The purpose is to foster innovationand efficiency in the pharmaceutical industry. The basis ofPAT is process understanding and the application of appro-priate control strategy for the critical process parameters toassure the quality of in-process material and final drugproducts. A robust control system is a key component tosupport a PAT system implementation.

Statistical Process Control (SPC) may be considered as aprecursor to PAT.

A control system should assist with:

• collection of process data (60% of an SPC implementation)• transfer of process data to standard statistical tools• eventually implement monitoring and later on control

charts in a batch context

At the time of publication, PAT often is only in the inceptionphase and the objective is to provide an approach whichfosters PAT implementation. The ability to connect PATmeasurement devices with process control systems and totransfer related data to statistical tools is required beforefiling for PAT. This is considered to be an important early stepin the process.

10 Other ConsiderationsThis chapter provides an overview of HSE and controlled

substances considerations, which should be considered dur-ing a comprehensive risk assessment (described in Chapter 3of this Guide). Non-cGMP risks that should be considered aresummarized and an overview of the basic technical andprocedural approaches that may be used to mitigate theserisks is provided. In addition to information provided in thischapter, project teams should be aware of local requirementsand facility policies.

This chapter is intended for use as a guide to the types ofnon-cGMP information that should be gathered as part of anorganization’s risk assessment and mitigation process.

See Chapter 15 of this Guide for a comparison of safetyregulations for the US, Canada, and the EU. A summary ofJapanese requirements is provided separately.

11 Revision of the Commissioning andQualification Baseline® Guide

In the Final Report Pharmaceutical CGMPs For the 21stCentury – A Risk-Based Approach, FDA described how theirobjectives included the following:

• Encourage the early adoption of new technological ad-vances by the pharmaceutical industry

• Facilitate industry application of modern quality manage-ment techniques, including implementation of qualitymanagement systems approaches, to all aspects of phar-maceutical production and quality assurance

• Encourage implementation of risk-based approaches thatfocus both industry and Agency attention on critical areas

A revision of the Baseline® Guide Volume 5 is underway thatseeks to support these objectives by describing an efficientand effective design, installation, and verification processthat focuses on safeguarding product quality and publichealth. It is intended that risk management should underpinthe specification, design, and verification process, and shouldbe applied appropriately at each stage.

As part of this revision, Baseline® Guide Volume 5 is beingaligned with ASTM standard E2500, which describes a risk-based and science-based approach to the specification, de-sign, and verification of manufacturing systems and equip-ment that have the potential to affect product quality andpatient safety.

12 Appendix A: Cost Factors InOSD Manufacturing

How much does it cost to build a new pharmaceutical manu-facturing facility? The answer to this important questionshould be carefully considered. It will drive the mathematicsof profit projections and has the potential to terminate aproject at an early stage if incorrect.

Project funding decisions are usually based on the firstcost alone and value engineering efforts to reduce first costare a common element in project planning. Most projectplanners are familiar with the concept of Lifecycle Costingbut are generally not tasked with or experienced in doing it.

During the beginning phases of a project, the cost involved




in owning and operating a new pharmaceutical manufactur-ing facility throughout its useful life should be considered.This cost has significant long term ramifications. The needfor radical changes may not become apparent until they aretoo costly to implement.

Lifecycle Costing offers an opportunity for a much moreinformed decision making process and, if properly imple-mented, can deliver significant savings. During the selectionof options that deliver the required functionality, LifecycleCosting considers both initial costs and future operatingcosts.

13 Appendix B: Summary ofQuality Risk Management Process

This chapter provides a summary, for both GMP and non-GMP elements, of the quality risk management process.

14 Appendix C: Risk Management ToolsThis chapter provides an example showing a systematicapproach to risk management for the design (e.g., new,retrofit), qualification, and maintenance of an OSD facility.

Several qualitative and quantitative risk managementtools are available. Tools that are routinely used are listedbelow. Additional references include ICH Q9 ‘Quality RiskManagement, and the ISPE Baseline® Guide on Risk-MaPP(in draft form at the time of this writing). Selection of aspecific tool will depend on the level of rigor of the data andthe criticality of the risk assessment (e.g., higher risks topatient safety may require a more in-depth risk analysistool.)

• Failure Mode Effects Analysis (FMEA; EN ISO 9001:2000)• Failure Mode, Effects and Criticality Analysis (FMECA)• Fault Tree Analysis (FTA)• Hazard Analysis and Critical Control Points (HACCP)• Hazard Operability Analysis (HAZOP)• Preliminary Hazard Analysis (PHA)• Risk ranking and filtering• Supporting statistical tools• Basic risk management facilitation methods (flowcharts,

check sheets etc.)

15 Appendix D: Other ConsiderationsThis chapter provides information on HSE regulatory andconsensus standards, codes, and guidelines for the US, theEU, Canada, and Japan.

16 Glossary and AcronymsA glossary of pharmaceutical industry terminology relevantto this Guide.

17 ReferencesA list of publications referenced by the Guide which providesfurther reading on the topic of this Guide.


Industry Interview


In thisinterview,Norman Winskilland KelvinCooper discusshow Pfizermanages theimportantinterfacebetweendevelopmentandmanufacturing.They also giveinsight intoPfizer’s strategicthinking onQbD, leanapproaches,outsourcing,newmanufacturingtechnologies,and greenchemistry.

by Rochelle Runas, ISPE Technical Writer

Norman Winskillis the Vice Presi-dent and TeamLeader of GlobalTechnology, PfizerGlobal Manufac-turing (PGM). Hisorganization in-cludes Pfizer Glo-bal Engineering,which has engi-neering responsi-bility for the whole

of Pfizer; Right First Time, PGM’s continuousimprovement program; and Global Manufac-turing Services (GMS), the manufacturing sci-ence and technology group of PGM. The mainresponsibility of GMS is the co-development(with Worldwide Pharmaceutical Sciences),scale up and launch of new products, processimprovement and process optimization, andthe development and implementation of newmanufacturing technologies.

Winskill joined Pfizer Sandwich in 1975 asa biochemist in the International Process De-velopment Group. In the next eight years, heheld a number of positions in Manufacturing,Development, and PGRD before joining Inter-national Manufacturing in New York in 1984.By 1996, he was the group’s Executive Direc-tor, Operations and Technology.

Winskill has a BSc in chemistry and a PhDin microbial chemistry from the University ofManchester, UK. Winskill is a Fellow of theRoyal Society of Chemistry.

Kelvin Cooper isthe Senior VicePresident of World-wide Pharmaceu-tical Sciences(PharmSci). Coo-per led this organi-zation since mid-2000 through themergers and acqui-sitions of Warner-Lambert and Phar-macia.

Cooper joined Pfizer in 1981 as a medicinalchemist in the Sandwich Discovery Laborato-ries, where he worked in several different projectareas, including anti-fungals, anti-bacterials,respiratory, urology, and exploratory medici-nal chemistry. In 1991, Cooper transferred tothe Discovery Laboratories in Groton, Con-necticut to lead the inflammation, allergy, res-piratory, and immunology medicinal chemistryprograms and in 1994 became the MedicinalChemistry Director of the cancer, inflamma-tion, immunology, allergy, respiratory, and in-fectious disease programs. In 1998, Cooper leda new technology-based group encompassinglarge-scale synthesis, general pharmacology,drug metabolism technology development, anddrug metabolism candidate support.

Cooper started his scientific career as ajunior technician and analytical chemist in thedrug metabolism department at Wellcome Re-search, Beckenham, England. He earned abachelor’s degree in chemistry, an honors de-gree in chemistry from Kingston University,

PHARMACEUTICAL ENGINEERING InterviewsNorman Winskill, Vice President, GlobalTechnology, Pfizer Global Manufacturing andKelvin Cooper, Senior Vice President,Pfizer Worldwide Pharmaceutical Sciences

Reprinted from





Industry Interview


London, and his PhD from NottinghamUniversity, where he conducted re-search in organic chemistry.

Cooper is a member of the RoyalSociety of Chemistry, the AmericanChemical Society, and a Fellow of theAmerican Association for the Advance-ment of Science. Cooper is also a boardmember on the University of Connecti-cut Foundation.

Q The partnership between prod-uct development and manufac-

turing is an especially important one.How does Pfizer manage that impor-tant interface?

Winskill: The interface betweenthese two key functions is one uponwhich Pfizer has worked for manyyears. We call this “co-development”and we manage it jointly between PfizerGlobal Manufacturing (PGM) andWorldwide Pharmaceutical Science(WWPS). Though we’ve been practic-ing co-development for years, it be-came more rigorous and systematicwhen we acquired Warner-Lambert(2001) and then Pharmacia (2003) –forcing us to institutionalize and for-malize our co-development procedures.

Our co-development process incor-porates all of the elements of Qualityby Design, lean manufacturing, greenchemistry, and many other critical ini-tiatives within Pfizer. What is inter-esting is how Pfizer manages this, withone core team comprised of colleaguesfrom all relevant disciplines in bothPGM and WWPS. We use a joint gover-nance/joint procedures approach be-tween the two organizations, built onmutual responsibility and mutual re-spect. We believe this is a competitiveadvantage. Others in our industry talkabout doing this, but when we discussthis in detail with our peers in the field,it is evident that Pfizer is going beyondthe kind of co-development happeningin most other pharmaceutical compa-nies right now.

Cooper: The structured co-develop-ment relationship emerged some sixyears ago, and was developed from anevolving relationship that had its rootssome 15 to 20 years ago when we had

more of a sharp interface between whathad historically been considered to betwo discreet functions. Just recently,the latest evolution in co-developmentadjusted the critical interface againand now we have a single team ap-proach to technology development —through co-development. The result isa very smooth transition from R&Dthrough commercial with no realhandoffs and no sharp interfaces.

Q Pfizer has been an active partici-pant in the FDA’s QbD pilot pro-

gram with the first approval in thatprogram. How does QbD fit into thefuture of pharmaceutical developmentand manufacturing?

Winskill: It’s true that Pfizer hasbeen very active in the area of Qualityby Design. In fact, we had two out of theoriginal nine applications in the pilotprogram, including the first approvalin the program. We learned a greatdeal about our readiness and expecta-tions on the part of the FDA from thispilot, and Pfizer shared its experiencesextensively at AAPS, PhRMA, PDA,and ISPE meetings and other industryforums. Our second pilot program, dem-onstrating continuous improvementwith lessons learned from the first,went even smoother. There is still a lotmore to learn.

The industry is under enormouspressures, including cost control and

product development. Trying to developincreasingly complicated technologieswith fewer resources than were avail-able in the past is a challenge. Withinour co-development process, we startedexploring how best to do more withless, and out of these efforts came ourRight First Time (RFT) program, ourterminology for Six Sigma. This evolvedinto QbD, something we were alreadypartially doing when we learned of theFDA’s intentions. We were quick toembrace the pilot program because ofour long established co-developmentprocess. A QbD submission is now thestandard for all new drug products inthe Pfizer R&D pipeline. The applica-tion of QbD at Pfizer is a scientificallygrounded, prospective, and risk ori-ented approach to development and webelieve it will improve the understand-ing of our products and the processesthat produce them and lead to increasedinnovation.

QbD has great potential benefits tothe industry, and the FDA deservesenormous credit for its bold steps inthis area and its willingness to shareperspectives and feedback. Throughthe efforts of Dr. Janet Woodcock,Helen Winkle, Moheb Nasr, JoeFamulare, and others at the FDA,QbD is evolving in ways where theindustry is able to learn how to imple-ment the concept on a practical basis.It helps pharmaceutical companiesfocus on more critical, promising prod-

Norman Winskill and Kelvin Cooper, representing the strong partnership between Pfizermanufacturing and development.


Industry Interview


ucts and process areas. This will onlybecome more important as pressureson the industry remain or increaseover time. We greatly appreciate thatEU and Japan regulators are activelyengaged in the building and content ofthis program, people like Emer Cooke,Jean-Louis Robert, Jacques Morenas,and Yukio Hiyama. And we are hope-ful that industry and regulators fromemerging markets will participate inthe near future.

Cooper: Our participation in theFDA QbD pilots was extremely usefulfor Pfizer, and demonstrated the po-tential to achieve much greater controland understanding about the process,which in turn ensures quality and en-ables greater flexibility.

The insights gained during thevarenicline pilot, to give one example,enabled us to respond quickly and ex-pertly when commercial demands farexceeded expectation. Because our re-sponse capability was already withinour design space and our quality con-trol parameters, we were confident wecould increase volume very quickly andchange things around as needed tomeet the market demand. Measuringit another way, that quick responsetranslated to somewhere in the area ofa million patients getting on varenicline– without QbD and continuous improve-ment built on QbD (i.e., the old way)we’d have been stocked out within prob-ably a month.

This benchmark collaboration be-tween the FDA and industry has beenvery useful in taking the industry for-ward; it’s my understanding that theseQbD pilot programs are being used astemplates for other pharmaceuticalcompanies.

Q Are you familiar with ISPE’sProduct Quality Lifecycle Initia-

tive (PQLI)? What role do you see thisinitiative playing in the future of phar-maceutical development and manufac-turing?

Winskill: Yes, PQLI is an impor-tant program providing practical, con-sistent advice to companies trying toimplement ICH guidance. Companies

interpret guidance in different ways,leading to inconsistencies, confusion,and delays. PQLI aims to eliminatethis confusion, thus leading to morerapid adoption.

Pfizer has been very active from thebeginning with ISPE’s PQLI initiativeand our support is consistent with ourdesire to have global industry and regu-latory discussions on best practices forimplementing concepts of ICH Q8, Q9,and Q10. ISPE’s strong reputation as aglobal educational and professional de-velopment organization makes thegroup ideally suited to run the PQLIprogram, and we think this initiativewill play an important role in manufac-turing and development in the future.

PQLI is a limited duration program,slated to run about five years. Hope-fully, the guidelines it is helping to getimplemented will become part of theindustry’s daily modus operandi. Ifsuccessful, the ICH guidelines thatunderpin PQLI will be used more regu-larly and consistently in both develop-ment and manufacturing. PQLI alsohas been very successful assuring aglobal perspective on QbD.

Cooper: We’ve been very active inPQLI from the beginning — Pfizer col-leagues serve on a number of teams,including quality, design space, criti-cality, legacy products, and controlstrategy, as well as on the steeringcommittee. It’s an important initiativethat to a great degree is made possibleby the support and collaboration ofboth the industry and regulators.

Q With cost pressures and the trendtoward more outsourcing, how

will Pfizer manage supply chain integ-rity and assure high product quality?

Winskill: We get asked this ques-tion a lot these days. The answer is“with care,” as we always have done.The factors that affect Pfizer’s decisionto outsource include: sourcing flexibil-ity, competitiveness, need for specialtechnology, cost control, and site dives-titures. However, whether we produceinternally or outsource, a secure sup-ply chain is paramount in protectingthe patients who use Pfizer products.

There are several considerations inorder to qualify a supplier for the in-tended purpose. Once a working rela-tionship is established, many ongoingissues need to be addressed, includingquality of product and services, prac-tices, safety, regulatory compliance, re-ports, and metrics. Special consider-ations when working with suppliers inemerging markets include, first and fore-most, product integrity and safety. Anypotential supplier is evaluated on itsability to produce material in a mannerthat is fully compliant with all regula-tory procedures. Pfizer has taken stepsto educate, evaluate, and essentially,enforce appropriate quality standardswith suppliers in emerging markets.

We partner with companies in lowcost locations as needed. If there areissues, we will try to resolve them toour satisfaction. If necessary, we willwalk away. It’s important that we holdexternal members of our supply net-work to Pfizer’s stringent standards.This is somewhat different from somecompanies who may still purchase fromlow cost locations, often from a broker,without much knowledge of thesupplier’s manufacturing conditions.

Cooper: Quality is number one hereand dominates our processes both in-ternal and external. We have a very,very strong internal quality systemthat includes extensive oversight ofthe supply chain whether it is Pfizerowned or not. Every partner that weconsider working with goes throughrigorous quality, scientific, financial,EHS, work practice, and related auditsand must meet and work within Pfizerstandards.

We’re actually helping some exter-nal suppliers get into QbD via a sort ofmentoring process. Suppliers that em-brace QbD should have a competitiveadvantage in our industry in the future.

Q What are the new manufactur-ing technologies that will change

the way pharmaceutical manufactur-ing is done from today?

Winskill: I’d prefer to answer thequestion in a broader context: ‘Howwill pharmaceutical manufacturing


Industry Interview


differ in the future?’ I say this becausethere is more than just a technologydifference involved; it is really amindset change. In the past, our indus-try was largely compliance based. Weinternally benchmarked ourselves yearon year, and considered that goodenough. The industry was not preparedto push the envelope to find a betterway of managing manufacturing andregulatory issues.

This is all changing rapidly with anumber of factors leading to a signifi-cant increase in cost pressures. Almostevery company in the industry has aneed to be more competitive. The in-dustry can learn a lot from other indus-tries where cost has been an issue formany years. For example, companieslike Pfizer are already looking side-ways to see how cost-sensitive indus-tries such as food or chemicals, forexample, view production issues. True,technology is a part of it, but otherfactors are important.

Today, Pfizer is benchmarking otherindustries, looking at best practicesfrom a variety of fields. We are alreadywell known for our PAT program; thefact is that many of our PAT applica-tions were “borrowed” from other in-dustries. As the industry’s mindsetchanges, it will free up pharmaceuticalcompanies to more boldly explore noveltechnologies and approaches like PAT,continuous processing, biocatalysis,and advanced process control systems.In the immediate future, the industrywill be leaner, more agile, and cer-tainly greener. The past mindset wasto hide behind the regulators… ‘Theywon’t allow us to do this, so we won’task.’ Today, the industry is open toother ideas for moving forward.

Cooper: We are seeing a big trendaway from just tablets and capsules asa way to present medicines and towardbiologics in different delivery systems,more devices, and the potential to getinto combination therapies. In addi-tion, changes in technology will be en-abled as a result of our investments inQbD. Understanding the design spacein which you can operate makes itpossible to start to build continuousmanufacturing which allows you to

completely change the way you man-age manufacturing, and to do so withabsolute confidence in quality and yourability to respond efficiently and effec-tively to market demand. The environ-mental benefits also will be signifi-cant; imagine the plant of the future assmaller and ‘greener,’ with fewer emis-sions, better energy utilization, lesswaste, and a correspondingly smallerfootprint.

Q Six-sigma, Lean, and similar ap-proaches have been strongly en-

dorsed across the industry. How arethey managed through the develop-ment to manufacturing continuum?

Winskill: Pfizer’s co-developmentprocess is an ideal framework to as-sure our lean approaches and continu-ous improvement efforts can be carriedout across a product’s lifecycle, start-ing at the pre-filing stage while thecompound is still under development.This builds a foundation for lifecyclefeedback and continuous improvementthat offers great benefits.

Several years ago we formally in-corporated Right First Time (RFT) intothe co-development process and cre-ated a jointly staffed RFT ProgramOffice to incorporate Six Sigma andlean concepts into new products. Thisestablished a sound scientific basis forfuture continuous improvementthroughout the lifecycle of a product.The boundaries between our WWPSand PGM organizations continue toblur, allowing us to more effectivelymanage company initiatives from de-velopment through manufacturing.

Cooper: RFT can be looked at as astructured risk assessment programthat looks at the entire development/manufacturing process and devisesstrategies to minimize and mitigaterisks. In PharmSci, we have investedheavily in a science of scale approachthat aligns very well with RFT, therebyinforming and bolstering a lean ap-proach to understanding and applyingthe science involved.

Q How is Pfizer building a cultureof continuous improvement? Can

you provide some specific examples?

Winskill: When it comes to impacton culture, the most significant thingwe have done is to recognize that tochange the culture we had to get every-one in the organization to speak thesame language, to understand the toolsand how to use them. So training playsa huge part in culture change at Pfizer.

We have three levels of training incontinuous improvement: Yellow,Green, and Black Belt. Our goal was tohave 100% of PGM colleagues trainedin Yellow Belt by 2008; 5% in GreenBelt, and 1% in Black Belt. We are ontarget to reach this objective. WithinR&D, more than half of the colleagueswill be formally trained by year-end.

Of course, it’s not enough to haveour colleagues onboard solely throughtraining. We stress that continuousimprovement is everyone’s responsi-bility, not just a select few. This inclu-sive spirit has resulted in thousands ofsuccessful projects (on average we com-plete more than one Green Belt orBlack Belt project and countless Yel-low Belt projects every day of the year)and the business benefits from the re-sulting savings.

There is broad-based managementsupport within Pfizer for continuousimprovement, evidenced by our rewardand recognition program. We rewardcolleagues who use continuous im-provement tools that make a real dif-ference in how our business operates.And when quality or safety investiga-tions are conducted, we expect col-leagues to use the RFT tools and tech-niques.

About a year ago, Pfizer Inc. lookedat PGM’s program and decided to adoptcontinuous improvement as a company-wide strategy. They appointed JohnScott as the Vice President in charge,the person who ran PGM’s RFT initia-tive originally. Each part of Pfizer isnow doing continuous improvement,though slightly differently to accom-modate different business needs.

Cooper: Currently, we have nearly100 continuous improvement programsgoing on at all levels across PharmSci.Yellow belt colleagues are encouraged


Industry Interview


to apply that training to a specificproject to demonstrate a good returnon investment — such as cycle timereduction, eliminating several waste-ful steps, all which contribute to driv-ing the cost of the process down — andthey are rewarded for doing so. Theresults have been impressive. The ulti-mate goal is not just to save time oreliminate steps, but for colleagues tobe fully engaged in what they knowbest and to understand and evaluatethe overall processes, and as appropri-ate, to suggest those changes that willachieve an optimum balance of effec-tiveness and efficiency.

One example that comes to mind isa project at our (Chesterfield, MO)biologics pilot plant to reduce cycletimes and increase throughput for themanufacture of monoclonal antibodies(mAB)—the active ingredient in somebiotherepeutics. This Black Belt projectwill increase our internal manufactur-ing capacity and save an estimatedannual cost of $8 million due to de-creased outsourcing expenditures.

Q Can you tell us about the GreenChemistry program at Pfizer and

why Pfizer decided to implement sucha program?

Winskill: Pfizer has been interestedin Green Chemistry for a long time.About six years ago, we formally de-clared that it was the morally andsocially responsible thing to preserveresources to the best of the company’sability. I can say confidently that we allfeel that green initiatives are the rightthing to do for the planet and the life onit. What is interesting is that initiallythe program was about minimizingresources. As we started to work on thedetails of Green Chemistry, Pfizer sawthat there were important economicbenefits as well. That fact alone will behelpful in getting Green Chemistrymore widely adopted throughout theindustry.

Cooper: Green chemistry is the de-sign of efficient chemical processes thatreduce energy use and reduce or elimi-nate the use of hazardous substances,reduce the use of solvents. Minimizing

our environmental impact is key toPfizer’s mission of creating a healthierworld. We have always committed tothis concept and in 2006, we decided tofurther elevate our commitment orga-nizationally. Pfizer appointed a greenchemistry leader, Peter Dunn, who inturn created a cross divisional steeringcommittee that integrates GreenChemistry into our research, develop-ment, and manufacturing organiza-tions; is active in community and edu-cational initiatives, including greenchemistry education at high schoolsand colleges, and has established greenchemistry awards within Pfizer.

In addition, we have received recog-nition for our environmental initia-tives and accomplishments from orga-nizations, including the U.S. Environ-mental Protection Agency. In 2002,Pfizer took first place in the EPA’sPresidential Green Chemistry Chal-lenge. In 2003, we were awarded theCrystal Faraday Award, and in 2006,we received the Astra Zeneca Awardfor Excellence in Green Chemistry.

Career Paths in Developmentand Manufacturing

Q What led you into or sparkedyour interest in a career in phar-

maceutical manufacturing?

Winskill: What attracted me toPfizer was the science. I was doing myPhD in the United Kingdom when I sawan ad for a company I had never heardof. I applied and joined Pfizer for a jobco-located at their manufacturing sitein Sandwich. That was 33 years ago.

Though I was involved in R&D, I gotto work very closely with the manufac-turing team. I spent eight years thereand took various jobs shuffling backand forth between manufacturing andR&D. What finally drew me to staywith manufacturing was the realiza-tion that the manufacturing environ-ment had very unique challenges,which were equally stimulating asthose in research. But the big advan-tage was that the time frames weremuch smaller. I could see a tangibleimpact of my work whereas in the fieldof research you can work all your lifeand not work on a product that makes

it to the marketplace.

Cooper: I never set out to be in phar-maceutical sciences or manufacturing;it kind of evolved with my career. Istarted with Pfizer as a discovery chem-ist trying to identify new medicines. AsI attained leadership positions in thatrole, I got interested in the develop-ment phases for a medicine, which thengot me interested in pharmaceuticalsciences.

At the time of the Pfizer Warner-Lambert merger, I was asked if I wouldlike to create the Pharmaceutical Sci-ences organization within Pfizer. I havebeen with Pfizer 27 years and in mycurrent role for almost eight years. Ihave a PhD in organic chemistry, andstarted my career at Wellcome.

Q What kind of training and experi-ence best prepared you for your

current position?

Winskill: My experience in the R&Denvironment definitely helps me tobetter evaluate and understand sci-ence and technology. Also, I have hadthe opportunity to work on AnimalHealth, Chemicals, and ConsumerHealth products within Pfizer – allvery cost-sensitive areas of the busi-ness. This is proving invaluable con-sidering the direction in which the phar-maceutical business is now headed.

Cooper: What most prepared me formy current position were my chemis-try education, essential in a chemistry-based organization such as PharmSci,and the training and development thatI was privy to at Pfizer. For example,just learning the discovery and drugdevelopment business over the yearshas been an amazing experience.

Q What significant changes haveyou seen in manufacturing and

what changes do you anticipate in thenext few years?

Winskill: In 33 years, I’ve seen somekey shifts in manufacturing. I remem-ber a period of expansion when costpressures were minimal and the em-phasis was essentially on supply and


Industry Interview


quality. This has changed and the in-dustry is more competitive now. Weneed to get much closer to our custom-ers and to be more agile in our re-sponses. I envision this trend continu-ing, and in fact, increasing. Anotherfeature of pharmaceutical manufactur-ing that is changing is the movementaway from traditional products – liketablets, capsules, and small moleculeAPIs. In five years time, the industrywill manage more complex technolo-gies, more diversity of delivery sys-tems, and more biological products.

Cooper: The changes I’ve seen inmy 35 years in this industry have beenabsolutely incredible in terms of thescience and technology now available.We are now gaining a good fundamen-tal understanding of the science of for-mulation, a huge advance that bringsus all the way back to science of scaleand QbD again. Some of the science

changes have been equally dramatic...going from simple tablets and capsulesto much more complex medicines, andnow all the way to the potential forstem cell therapies. The science ofmanufacturing stem cells under therigorous quality standards Pfizer main-tains is certainly a big challenge tothink through.

Q What are some of the challengesyou are currently dealing with in

manufacturing?

Winskill: There are a great deal ofchallenges in this field today, includ-ing cost pressure, excess capacity, lossof exclusivity, pipeline issues, the exit-ing of some sites, and maintainingmorale and motivation in the midst ofit all. How to balance all this is a majorissue. What will separate Pfizer fromthe rest is how we handle it.

Cooper: The business and financialchallenges we are experiencing are simi-lar to those being experienced acrossthe industry. As demand for more com-plex medicines drives the cost of manu-facturing up, there is pressure to cutcosts to remain competitive. New kindsof therapies are emerging that presenttechnical challenges; iRNA is just oneexample. We can design molecules thatinterfere with RNA, the protein synthe-sis machinery of the cell, but we cannotdeliver these innovative new therapiesusing any of the conventional deliverysystems at the present time. What willdistinguish Pfizer is its pursuit of suchinnovative therapies, and as it has dem-onstrated throughout its history, whereothers see challenges, Pfizer envisionssolutions.

PE would like to thank Jim Spavins forhis contributions to the development ofthis interview.


PAT Compliance


This articledescribes aPAT-approachto the coatingprocessperformed in aperforated pan.The main topicsinclude processunderstanding,Quality byDesign, andprocessregulation andcontrol.

How to Make a Perforated Pan PAT-Compliant

by Davide Manca, Caterina Funaro, Fiorenzo Cembali,Giusi Mondelli, Giorgio Tarozzi, and PierAntonio Ragazzini

Introduction

Process Analytical Technology or PAT,is a set of systems for the analysis andcontrol of manufacturing processesbased on the measurement of critical

quality parameters and the performance at-tributes of raw materials and in-process prod-ucts to assure an acceptable end-product qual-ity.2 The PAT initiative promotes the processknowledge and understanding to integrate newmanufacturing technologies into pharmaceu-tical production.5 As reported by the FDA (2004)“the term analytical in PAT is viewed broadlyto include chemical, physical, microbiological,mathematical, and risk analysis conducted inan integrated manner.” Several current andnew tools enable scientific, risk managed phar-maceutical development, manufacture, andquality assurance. Quoting from the FDA:“these tools, when used within a system canprovide effective and efficient means for ac-quiring information to facilitate process under-standing, develop risk-mitigation strategies,achieve continuous improvement, and shareinformation and knowledge. A desired goal of

the PAT framework is to design and developprocesses that can consistently ensure a pre-defined quality at the end of the manufacturingprocess.” As a result, it is possible to reduceproduction cycle times by using on-, in-, and/orat-line measurements and controls. It also ispossible to prevent rejects, scrap, and re-pro-cessing, while achieving the so-called real timerelease. Process automation allows improvedoperator safety, while reducing human error.

This article focuses on a perforated pan andon the effort played by the authors to move bothequipment manufacturing and process opera-tion from well-established and conventionalknowledge toward the PAT-approach. Thismindset change is not chosen, adopted, andcovered because it is fashionable, but because itrepresents a challenge to uphold the currenttechnological gap with competitors, while in-creasing the understanding of the process. Thislast point is of paramount importance for pro-cess management because working with a PATcompliant approach to process operation allowsreplacing the static and rigid approach utilizedby conventional recipes to embrace an advancedprocess control that through measurement, pro-cess knowledge, and regulation of operatingconditions permits achieving the real time re-lease.

The PAT initiative attempts to move phar-maceutical manufacturing from a recipe drivenapproach toward a process control methodology.The process should not be managed according toa pre-established recipe, but instead a “measureand control” system should follow an optimalprocess trajectory. By doing so, the end-pointproblem would be upset. On one hand, the recipeassigns a predefined time bucket for every pro-duction stage of the pharmaceutical process; onthe other hand, the “measure and control” sys-tem follows a completely different approach based

Figure 1. Frontal view ofa perforated pan.

Reprinted from





PAT Compliance


Nominal size [l] 200

Min/Max pan capacity [l] 50-200

Pan diameter [mm] 1330

Pan engine power [kW] 4

Maximum air flowrate [m3/h] 4000

Table A. Technical data of the perforated pan.

Total number of tablets 285,714

Droplet mean volume [m3] 6.37063E-15

Mean flight time [s] 0.013

Total number of sprayed droplets 3.139409E+12

Number of droplets sprayed in 1 s 872,057,988

Total number of droplets sprayed on a tablet 10,987,931

Number of droplets sprayed on a tablet in 1 s 3,052

Table B. Mean operating data of a pan coater based on 100 kgtablets load.

on product quality achievement. Once the process is thor-oughly understood, the equipment is designed according toproduction demands, and a dedicated control system is setup,it is eventually possible to make the process follow an optimalprocess trajectory. By doing so, the end-point problem isautomatically solved and production time is optimized againstthe external disturbances that make the process deviate fromthe design conditions.

Before addressing the model and control of the coatingprocess, it is necessary to understand in-depth how a perfo-rated pan works and what are the most important param-eters that influence its operation.

Understanding the Perforated PanA perforated pan coater consists of a horizontal cylindricaldrum that rotates around its axis - Figure 1. The tablets areloaded into the pan and the basket containing the tablets isperforated with holes smaller in dimension than the tablet inorder to avoid any loss of cores.

Once the pan is loaded, the basket begins to rotate and aflow of hot, dry air passes through the perforated basket andthe rolling bed of tablets. The metal basket comprises somebaffles that increase the mixing of the tablets within therolling bed. A number of nozzles spray a liquid solution ofpolymer and solvent (usually water) to the surface of thetablets. The air flowrate dries the wet tablets and an increas-ing polymer thickness covers the cores. Once the optimalcoating thickness and the residual moisture content meet theprocess specifications, the coating process stops and thetablets are unloaded.

Several parameters condition the coating process. Threedistinct categories can be identified: the design specification,the static process conditions, and the dynamic process vari-ables. Some examples for each category are:

• design specifications: drum diameter, drum depth, num-ber and geometry of the mixing baffles, number, pitch, anddistance of the holes, geometry of air inlet and outlet ductsand their relative positions in respect to the drum

• static process conditions: number of nozzles, pan load (i.e.,amount of tablets loaded), polymer/solvent ratio

• dynamic process variables: inlet air flowrate, inlet airtemperature and moisture, pan speed, solution spray rate,nozzle patterns and operating pressure, nozzle angles anddistance from the tablet bed (the dynamic angle of repose[i.e., the angle formed by the tablet bed in respect to ahorizontal plane when the pan is rotating] changes withthe pan speed).

The geometrical dimensions and layout of the pan coater areof paramount importance in the scale-up of the process units.A deep understanding of the physical phenomena involved intablet coating can be of real help in the design and scale-upof industrial units.

Static process conditions are adjustable parameters thatmust be set before the coating starts. A deep understandingof their action may significantly influence the quality andyield of the final product. Usually, it is up to the coatingsupervisor to assign these parameters as input data for theprocess recipe. Once again, a shared and validated knowledgeof the impact of these parameters on the product quality helpsin configuring the pan coater, while avoiding any misinter-pretations by the operator.

The dynamic process variables allow the regulating andcontrolling of the coating process according to the productquality specifications. Instead of blindly implementing apredefined and static recipe, it is possible to configure acontrol system capable of driving the process through anoptimal operative trajectory. The optimal product quality canbe accomplished by continuously changing the operatingconditions, while complying with the process constraints.The implementation of a control system permits attainingthe product specifications, independently of the externaldisturbances that invalidate any predefined recipes.

Some Words about the Coating ProcessTable A reports the technical and geometrical data of anindustrial perforated pan.

We assume a partial load of the pan in terms of 100 kg ofplacebo tablets. The tablets are round, have a diameter of11 mm, and each weigh 350 mg. The apparent tablet density(bulk density) is 800 kg/m3. Consequently, the tablets take up avolume of 125 l, which is about 63% of the nominal pan size.

The mean coating weight increment is 4% in respect to theuncoated tablets. Since we have 100 kg of uncoated tablets,the solid coating weighs 4 kg. The sprayed suspension ofpolymer and excipients in water is 20% on a weight basis.Consequently, the liquid solution weighs 20 kg (16 kg of waterand 4 kg of solids). The spraying period takes about one hour,and according to the pan speed (usually 6 to 10 rpm), thesprayed flowrate is 250 to 350 ml/min.

Figure 2 shows the spray guns. Usually, for a 100 kg batch,there are four nozzles, which are aligned with the drum axis.The bore diameter for each nozzle is 1.2 mm, and the atomi-zation air pressure is 1.5 to 2.0 bar. The distance between the


PAT Compliance


nozzle and the rolling bed is about 200 mm according to thepan load and this gives rise to a spray breadth of about 160mm. The nozzle manufacturer performed a number of experi-ments to measure the spray pattern as well as the meandiameter and velocity of the droplets.

Typical droplet diameters are 18 to 28 mm and dropletvelocities are 7 to 24 m/s according to the emission angle of thedroplet in respect to the nozzle axis. Based on these meandata, it is straightforward to compute the average operatingvalues reported in Table B.

The droplet mean volume (in the order of picoliters) andthe number of sprayed droplets (teradrops, i.e., thousands ofbillions of drops) are quite impressive values. It also isamazing the average number of droplets that hit a tablet inonly one second (some thousands). Actually, the coatingprocess is even more complex.

Spray Pattern and Droplet DryingAs per the aforementioned, a nozzle sprays an endless num-ber of droplets through its bore. Two streams of compressedair separately feed the spray gun. The former atomizes thesprayed liquid; the latter moulds the spray and is calledpattern air. Usually, the flowrate and pressure of the at-omization air are larger than those of the pattern air. Thepattern air drives the sprayed droplets toward the rollingtablets and avoids any losses on the pan walls. As reported inTable B, the flight time of a droplet is quite short. During itsflight, the droplet dries due to evaporation. While it flies in analmost straight line (gravity does not play a significant roledue to the high initial velocity and short flight distance), thedroplet crosses a hot and parallel airflow. The hot air flowrateplays a dual role: it partially dries the flying droplets and thesprayed tablets within the rolling bed. Its velocity is signifi-cantly lower than the sprayed droplets. While the dropletdecreases its water content, the exsiccating air increases itsmoisture content. The droplet does not evaporate completely,but a certain amount of water is still present when it im-pinges the tablet surface. This water content is of paramountimportance for the coating process. Actually, if the droplethas excessive water content then the tablet gets soaked andthis may compromise its integrity and that of the active

principle. Conversely, if the droplet is too dry then it does notstick on the tablet surface and the coating turns into dust thatis lost through the perforated cylinder. Consequently, thereis an optimal operating range for the hot air flowrate and itsinlet moisture content.14 Due to the spray breadth and thedifferent flight times (as a function of the launch angle inrespect to the normal bore direction), it is possible to adjustthe pattern air pressure to either narrow or widen the spraypattern in order to achieve a convenient final distribution ofthe droplet diameters (i.e., droplet moisture).

The Moisture BalanceWith reference to the perforated pan, it is possible to write anoverall dynamic balance for the process water. The mainreason for this water mass balance is sketching out theprocess moisture content. Actually, the main task of a perfo-rated pan is producing tablets with optimal coating thicknessand uniformity, and optimal moisture content. Usually, thecoating thickness and uniformity are measured at-line by anoperator who, at predefined time intervals, weighs a certainnumber of tablets withdrawn from the rotating basket, andvisually checks the surface for irregularities. The moisturecontent of the coated tablets is more difficult and expensiveto measure (in terms of required time). It is usually per-formed afterward, in a laboratory, by measuring the moisturefraction of a number of tablets whose coating has beenscratched out in advance. These tablets are withdrawn fromthe pan at predefined time-intervals. Consequently, the dy-namics of the tablet moisture content may only be deter-mined at the end of the coating process. This measure is notdesirable from an on-line point of view. It is used only to verifya posteriori the product specifications. Conversely, if thetablet coating moisture was measured on-line and in realtime, it could be used to identify the end-point condition.

Figure 3 shows the experimental set-up, which helps toconclude the moisture content of the tablets. If one measuresthe inlet and outlet water factors, it is possible to write arather simple dynamic mass balance for the water crossingthe perforated pan. We have to evaluate four quantities.

Figure 2. Spray guns applied on the perforated pan.

Figure 3. Simplified experimental setup for measuring the moisturecontent in the perforated pan.


PAT Compliance


Two are relative to the moisture content of the inlet andoutlet air flowrates (ρinxinQin and ρoutxoutQout). The third quan-w w a w w a

tity is the inlet water flowrate sprayed by the nozzles on thetablet bed (ωsprayWspray). The fourth is the moisture content ofw sol

the atomization and pattern air flowrates to the nozzles.

pandMwThe overall water mass balance is: ________ =

dt

ρinxinQin – ρoutxoutQout + ωsprayWspray + ρnozzlexnozzle (Qpattern + Qatom)w w a w w a w sol w w a a

where M pan is the total water amount inside the perforatedw

pan tabpan. More specifically, we can assume that Mw = Mw , i.e.,the water amount inside the pan moistens the tablets.

This hypothesis is reasonable since the vapor condensatesneither on the rotating basket nor on the outlet metallic duct.Moreover, the spray guns control the pattern in order to avoidany droplets spraying on the pan walls. Therefore, it issufficient to add two moisture probes to the inlet and outlethot air ducts since the inlet sprayed solution is measuredeither volumetrically (peristaltic pump) or massively (dedi-cated flowmeter). The water fraction in the sprayed solutionis constant as the polymer solution is continuously stirred.The same can be said for the compressed air fed to the nozzles.Finally, the remaining overall water mass allows the estima-tion of the mean moisture content of the tablets. This esti-mated measure is almost instantaneous as the moistureprobes and the flowmeter give a quick response. Therefore, itis possible to speak of a soft-sensor determined according tothe PAT directives. The mean moisture content of the tabletbed is of paramount importance since it allows the identifica-tion of the end-point condition of the coating process. It alsocan be used for on-line control of the coating process. Let ussuppose that a parametric design of experiments has identi-fied an optimal exsiccation trajectory, which is a compromisebetween tablets that are either too wet or too dry (to preserveboth mechanical integrity of tablets and active principledegradation.)11,15 The on-line controller, based on the tabletmoisture soft sensor, will regulate the operation of the coat-ing process (e.g., inlet air flowrate and temperature, sprayedflowrate, pan speed) to follow the process optimal trajectory.It is worth mentioning that once the spraying is over thecoating process proceeds with the exsiccation stage where thehot air stream reduces further the tablet moisture to thedesired value. In this case, as well, we have an optimaltrajectory problem where the end-point time is unknown andvaries from batch to batch along with the spray features andhistory. In reality, very often, the coating process is managedin a more conventional way where the recipe dictates itsaprioristic rules. Actually, the exsiccation period is assigneda priori and is kept constant, but this approach does notmatch the PAT methodology that supports the trajectorytracking and end-point achievement.

Tablets TrajectoryThe homogeneous mixing of tablets during the coating pro-cess is a prerequisite for good coating uniformity. This fea-

ture is significantly influenced by pan speed, spray rate, inletair temperature and flowrate, tablet shape, and baffle geom-etry.1,3,8,10 Therefore, understanding the tablet motion in aperforated pan may lead to an improved coating process.

As reported, several attempts were made to understandand record particle motion using a variety of particle trackingtechniques, such as video imaging, particle imagingvelocimetry, positron emission, and photometry.9,12,13 Leavershowed that the surface time (the time that the tablet spendson the surface of the bed) is directly related to the pan speedand load.4 In contrast, the circulation time distribution isonly affected by the pan design and mainly by the bafflespresence/absence. Wilson and Crossman showed that coatinguniformity increases with pan speed.16

According to Mehta, the optimization of pan coating opera-tions may involve more than 20 design and process variables.6

This set of variables would require a too demanding experi-mental effort. This is one of the reasons for simulating themotion of particles in the pan and to develop models tosimulate the coating dynamics. Turton and coauthors dis-cussed extensively the tablet kinematics in a pan-coatingdevice,7,9,12 using an imaging technique to sample in real-timethe tablet motion. A few variables give an overall picture of thetablet kinematics: the time spent by a tablet in the bulk of thebed also called circulation time (tcirc), the time spent on thesurface of the bed (tsurf), the projected area exposed to the spray(Aproj), and the surface velocity (νsurf). Actually, the tablet surfsand rolls on the bed surface for short times, while it travels andshuffles in the bed bulk for longer times. When the tablet floatson the bed surface, it receives the droplets sprayed by thenozzles. Since the tablet rolls and tumbles, the area exposed tothe jet stream keeps on changing in a random way.

With reference to the circulation time (tcirc), the experi-ments showed that:

• There is a decrease in tcirc as the pan speed increases (sincethe tablet residence time inside the bed decreases whenthe pan rotates more rapidly).

• Under the same operating conditions (i.e., pan speed andload), larger tablets are characterized by lower tcirc.

• tcirc increases with the pan load (the more the tablets, thehigher the residence time inside the bed. This is due to thefact that, when the pan load is higher, the bed is thickerand the tablets have to cross a larger volume).

• The baffles promote the tablet motion and reduce the deadzones. As a consequence, the coating uniformity is pro-moted by baffles and tcirc is slightly reduced.

With reference to the surface time (tsurf), the experimentsshowed that:

• tsurf decreases when the pan speed increases• tsurf decreases when the pan load increases• The influence of the tablet dimensions on tsurf is negligible.

With reference to the projected surface area (Aproj), the experi-ments showed that:


PAT Compliance


• Aproj decreases when the pan speed increases• Aproj increases with the dimensions of the tablet

Finally, with reference to the tablet surface velocity, Sandadishowed that its horizontal component has a null mean valuedue to the random collisions among tablets.12 Conversely, thevertical component of the surface velocity (ν surf) has a gaussian-Y

like distribution, whose mean value is positive (in the down-hill direction). Turton and coworkers (2004) observed thatthis component increases with the pan load. This is due to thedirect proportionality between the pan load and the wall-tablet friction (the friction term increases the dynamic angleof repose of the rolling tablets).4,7 Consequently, the higherslope of the bed surface makes the tablets move at a highervelocity. This point also explains the reduction of tsurf whenthe pan load increases. On the other hand, ν

surf is ratherY

independent of the tablet dimensions.Mueller and Kleinebudde (2007) showed that the periph-

eral pan speed is not a valid scale-up criterion for designpurposes.7 Actually, the most important parameter is thesurface time, which measures the amount of sprayed solutionreceived, while the tablet moves inside the pan. As previouslyshown, there is a direct dependency between tsurf and ν

surf.Y

Mueller and Kleinebudde showed that, under constant panloads and peripheral pan speeds, the tablet surface velocity isnot constant, but may change significantly.7 They proposed ascale-up parameter (ξpan) that allows for the determination ofthe surface velocity when passing from lab-scale to pilot-scaleand even to an industrial-scale pan:

d2νpan2

= ξpan 3 ______ νpan1√ d1

where d is the diameter of the basket, and ξpan may bedetermined by measuring the main process variables.

All these considerations delineate the complex world ofpan coating: a non-linear sum of effects, conditions, andparameters that is necessary to account for and to under-stand deeply the process sensitivity before addressing theregulation and control themes.

Control FrameworkThis section is devoted to the analysis of the regulation andcontrol of the coating process inside a perforated pan. Beforeaddressing individually the control topics, we would like toset clearly the boundary line between regulation and control.

The term regulation means the modification and adjust-ment of a process variable by means of another processvariable. This adjustment is based on a recipe, which isassigned a priori. An example follows: when the pan speedincreases from 6 to 8 rpm then the spray rate is increasedfrom 300 to 350 ml/min. This regulation is performed inde-pendently of the real process conditions that should accountfor the disturbances either measurable or immeasurable. Asfar as the regulation is concerned, we do not assign anysetpoints, we do not define any controlled variables, weneither setup nor configure any control loops, and we do noteven define any pairings among manipulated and controlled

variables. Actually, we have only some rules that are dictatedby the recipe. When these rules are implemented, the regu-lator adjusts the manipulated variables to the new recipesettings (this also is known as actuator device).

On the other hand, the classical term for control refers tothe pairing existing between manipulated and controlledvariables. The operator assigns a setpoint value to the con-trolled variable and the controller adjusts the manipulatedvariable accordingly in order to avoid possible offsets pro-duced either by external disturbances or by process variabil-ity.

Process Regulation in a Perforated PanWith reference to a typical pharmaceutical recipe, we can saythat the coating process is usually split into a few timebuckets where the manipulated variables are kept constant.These variables change value at the end of every time bucket.Consequently, we have the typical step-like dynamics thatcontributes to process discontinuities. An example is given bythe pan speed that, according to the recipe, is usually ad-justed three times by predefined increments. Instead of step-changing the process variables, we suggest adopting asmoother approach to process regulation. The step functionthat describes the dynamics of a manipulated variable can besubstituted by a continuous interpolating function, which isa sounder approach to process stability. A cubic spline inter-polation is a good compromise between simplicity and flex-ibility without compromising the process robustness. Topreserve the same impact of the manipulated variable on theprocess dynamics, it is possible to assume the conservation ofits integral action.

With reference to the perforated pan, the main manipu-lated variables that are usually adjusted at predefined timeintervals are the pan speed (ωpan), the sprayed flowrate(Qspray), the inlet hot air temperature (T

in), and seldom thea

angle of the spray gun in respect to the bed tablets (ϕspray). Toacutely understand the regulation and control framework, itis necessary to give a short explanation of the role played bythese manipulated variables. There is a direct proportional-ity between the pan speed and the degree of mixing requiredof the tablets within the bed. We have already reported thatan increase in the pan speed means a decrease of time spentby the tablet inside the bed, and an increase of the number ofsprayed droplets that collide with the tablet surface in thetime unit. At early stages, when the tablet is still uncoated,the pan speed should not be excessive to avoid the tabletcrumbling due to heavy mechanical friction. The coating,besides preserving the active principle from external agents,also enhances the mechanical strength of the tablet. There-fore, it is then possible to increase progressively the panspeed and the degree of mixing within the bed. This allows theincrease of the spray flowrate while reducing the total pro-cessing time. Therefore, to support the increment of sprayedflowrate and coating moisture, it is recommended to increasethe inlet air temperature. The dynamic angle of the tablet bedis a function of the pan speed. A clever operator adjusts thespray gun angle to keep the longitudinal spray axis perpen-


PAT Compliance


Table C. Pairing between the manipulated and controlled variablesof the control loops.

Controlled variable Manipulated variable

Loop #1 Bulk temperature of the tablets Inlet air temperature

Loop #2 Moisture content of the tablets Inlet air flowrate

dicular to the tablet bed. According to these considerations, itis then possible to define the following hierarchical depen-dencies: Qspray (ωpan(t)), T

in (ωpan(t)), and possibly ϕspray (ωpan(t)).a

Conversely, the inlet air flowrate is kept constant to avoidthe tablets sticking to the pan walls, due to the resistanceexerted by the gas flowing through the perforated basket ofthe pan. Primarily, its value depends on the total tablet loadand only secondarily on the sprayed flowrate.

Process Control in a Perforated PanFrom the aforementioned arguments, it might look as nomore manipulated variables are available for control pur-poses. This is not true. In fact, it is possible to set-up a controlloop for the bulk temperature of the tablet bed where the mostsuitable manipulated variable is the inlet air temperature.The reason for choosing the bed temperature as a controlledvariable is the coating quality and integrity. Actually, if thebed temperature is too low then the tablets are too wet (evensoaked) and they stick together. This phenomenon also isknown as twinning. Conversely, if the bed temperature is toohigh, besides the danger of compromising the active prin-ciple, the coating uniformity decreases. For too high tablettemperatures, common phenomena are the so-called orange-peel effect as well as a manifest roughness of the coating (withchinks and polymer detaching).

When this control loop is implemented, the process opera-tor assigns an optimal bulk temperature belonging to theinterval between the lower and upper bounds that define theattainment of the process quality. It is quite easy to measurethe bulk temperature of the tablet bed by means of either athermocouple or a PT-100 probe. It is worth mentioning that,due to the size of the pan, the temperature probe does notaffect significantly the bed flow and the mixing degree of thetablets.

To make the control framework more complex and flexible,it also is possible to use the inferred measure of the moisturecontent of the tablets. In this case, a possible manipulatedvariable is the inlet air flowrate. Formerly, we wrote that theprocess operator prefers to keep the inlet air flowrate constant.This happens because pharmaceutical processes are usuallyunder-controlled and a number of process variables are keptconstant to reduce both the degrees of freedom and complexityof the process. This is particularly true when the process isoperated manually. Conversely, a robust control system allowsthe increase of the operative elements, while making theprocess more flexible and responsive. As a matter of fact, a newcontrol loop, based on the mean moisture content of the tablets(controlled variable) and the inlet air flowrate (manipulatedvariable) may be implemented and coupled to the previous one(see Table C for more details on the variables pairing).

ConclusionsThis article discussed the PAT approach to the coating pro-cess carried out in a perforated pan. In respect to the conven-tional process operation, which is based on the recipe para-digm, the PAT approach is a significant change to this well-established practice.

One of the main achievements is process understanding.This is the basis for further innovative procedures such asonline process regulation and control. The optimal trajectoryproblem, disturbance rejection, and constrained processingare distinctive features of PAT. Actually, they are quitechallenging for the pharmaceutical world, which is restrainedby the inflexibility of the recipe paradigm. At the same time,the proposed PAT features open pharmaceutical manufac-tures to new frontiers and high quality control.

References1. Faroongsang, D., Peck, G.E., “The Role of Liquid Water

Uptake by an Insoluble Tablet Containing a Disintegrant,”Drug Development and Industrial Pharmacy, 20, 1777-1794, (1994).

2. FDA, “Guidance for Industry PAT: A Framework forInnovative Pharmaceutical Development, Manufactur-ing, and Quality Assurance,” CDER & FDA, www.fda.gov/cder/OPS/PAT.htm, (2004).

3. Fourman, G.L., Hines, C.W., Hritsko, R.S., “Assessing theUniformity on Aqueous Film Coatings Applied to Com-pressed Tablets,” Pharmaceutical Technology, 19, 70-76,(1995).

4. Leaver, T.M., Shannon, H.D., Rowe, R.C., “A Photomet-ric Analysis of Tablet Movement in a Side-Vented Perfo-rated Drum (Accela-Cota),” Journal of Pharmacy andPharmacology, 37, 17-21, (1985).

5. Lopes J.A., Costa, P.F., Alves, T.P., Menezes, J.C.,“Chemometrics in Bioprocess Engineering: Process Ana-lytical Technology (PAT) Applications,” Chemometricsand Intelligent Laboratory Systems, 74, 269-275, (2004).

6. Mehta, A.M., “Scale up Considerations in the Fluid BedProcess for Controlled Release Products,” Pharmaceuti-cal Technology, 12, 46-52, (1988).

7. Mueller, R., Kleinebudde P., “Prediction of Tablet Veloc-ity in Pan Coaters for Scale Up,” Powder Technology, 173,51-58, (2007).

8. Okutgen E., Jordan, M., Hogan, J.E., Aulton, M.E., “Ef-fects of Tablet Core Dimensional Instability on the Gen-eration of Internal Stresses within Film Coats. Part III:Exposure to Temperatures and Relative Humidities whichMimic the Film Coating Process,” Drug Development andIndustrial Pharmacy, 17, 2005-2016, (1991).

9. Pandey P., Song, Y., Kayihan, F., Turton, R., “Simulationof Particle Movement in a Pan Coating Device usingDiscreet Element Modeling and its Comparison withVideo-Imaging Experiments,” Powder Technology, 161,79-88, (2006).

10. Poukavoos N., Peck, G.E., “The Effect of Swelling Charac-teristics of Superdisintegrants on the Aqueous CoatingSolution Penetration into the Tablet Matrix during the


PAT Compliance


Film Coating Process,” Pharmaceutical Research, 10, 9,1363-1370, (1993).

11. Ruotsalainen M., Heinämäki, J., Guo, H., Laitinen, N.,Yliruusi, J., “A Novel Technique for Imaging Film Coat-ing Defects in the Film-Core Interface and Surface ofCoated Tablets,” European Journal of Pharmaceutics andBiopharmaceutics, 56, 381-388, (2003).

12. Sandadi S., Pandey, P., Turton, R., “In Situ, Near Real-Time Acquisition of Particle Motion in Rotating PanCoating Equipment using Imaging Techniques,” Chemi-cal Engineering Science, 59, 5807-5817, (2004).

13. Tobiska S., Kleinebudde, P., “A Simple Method for Evalu-ating the Mixing Efficiency of a New Type of Pan Coater,”International Journal of Pharmaceutics, 224, 141-149,(2001).

14. Tobiska S., Kleinebudde, P., “Coating Uniformity andCoating Efficiency in a Bohle Lab-Coater using OvalTablets,” European Journal of Pharmaceutics andBiopharmaceutics, 56, 3–9, (2003).

15. Twitchell A.M., Hogan, J.E., Aulton, M.E., “The Behaviourof Film Coating Droplets on the Impingement onto Un-coated and Coated Tablet,” S.T.P. Pharma Sciences, 5,190-195, (1995).

16. Wilson K., Crossman, E., “The Influence of Tablet Shapeand Pan Speed on Intra-Tablet Film Coating Unifor-mity,” Drug Development and Industrial Pharmacy, 23,12, 1239-1243, (1997).

About the AuthorsDavide Manca is Associate Professor of“Conceptual Design” and “Dynamics andControl of Chemical Processes” at Politecnicodi Milano. He obtained a PhD in chemicalengineering in 1994. In the last decade, heworked on a number of research projects onpharmaceutical and manufacturing pro-cesses, e.g., microencapsulation of active prin-

ciples, high energy mechanochemical activation, planningand scheduling of batch processes, and process analyticaltechnologies. He can be contacted by email: [email protected].

Politecnico di Milano, P.zza Leonardo da Vinci, 32, 20133Milano, Italy.

Caterina Funaro received her degree inpharmaceutical chemistry and technology atthe University of Bologna in 1999. From1999 to 2000, she was employed in GS Coat-ing System in Bologna as laboratory tech-nologist. In 2001, Funaro joined IMA S.p.A.in Bologna where she presently is in chargeof process laboratory management with the

following responsibilities: technical support to marketingand commercial department, pre-sales trials (granulation,coating, tabletting), after-sales assistance to customers, co-operation with marketing and commercial department, R&Din cooperation with raw materials suppliers and academia,

support to IMA technical department for development of newtechnologies, drawing up of technical data for scientific pub-lications and promotional papers, training of customers, IMAsales managers, IMA technicians, attendance to pharmaceu-tical meetings for paper presentations and personal updat-ing. She can be contacted by email: [email protected]

Fiorenzo Cembali started his work experi-ence in 1987 at GS Coating System as tech-nical assistance in coating machines field.He has been employed at IMA S.p.A. since2000 in Bologna as mechanical engineer andprocess laboratory technician. He can be con-tacted by email: [email protected]

Giusi Mondelli received her degree in phar-maceutical chemistry and technology at Uni-versity of Bologna in July 2006. From Sep-tember 2006 to December 2006, she wastrained in the laboratory R&D for processdevelopment of oral solid dosage forms inIMA S.p.A. In January 2007, she startedworking in IMA S.p.A. as a laboratory tech-

nologist with the following responsibilities: technologicalsupport to pre-sales trials (granulation, coating, tabletting),after-sales assistance to customers for granulation equip-ments start-up and technology transfer. She can be contactedby email: [email protected].

Giorgio Tarozzi graduated in electronicengineering at the University of Bologna inthe academic year 1997/1998 with an experi-mental research activity on Real Time Oper-ating System (RTOS) and Fieldbus RemoteControl Architecture. In 1999, he joined IMAS.p.A. in Ozzano dell’Emilia (Bologna) in theR&D Corporate Department to develop pack-

aging systems for the pharmaceuticals and tea and coffeebusiness. From 2004 to 2007, he was the “Special Technolo-gies” specialist in “Innovation, Research & New Technolo-gies” IMA Department. In January 2008, Tarozzi joined theSolid Dose Division Electrical Department with the respon-sibility of coating systems. He can be contacted by email:[email protected].

PierAntonio Ragazzini received hisMaster’s degree in electrical engineering atthe Politecnico di Torino in 1979. In 1980, hejoined a company dealing with cryptographyin order to develop machines protecting com-munications against “eavesdroppers.” Therehe designed both hardware and software forthe first fully digital ciphering machine. In

1986, he joined IMA S.p.A. responsible for the electricaldepartment. In 1992, he went to the corporate IMA Depart-ment for the strategic development of innovative machines asthe coordinator of all the electrical departments with the aim


PAT Compliance


of creating synergies. In addition, he was responsible for theinnovations in automation. At present, he is responsible forthe research and new technologies, a corporate departmentdealing with automation, but also looking for emerging tech-nologies to apply to new machines in order to improve theirperformance and the quality of final product. He is coordina-tor of the PAT team in IMA. He can be contacted by email:[email protected].

IMA S.p.A., via Tosarelli, 184, 40055 Castenaso (BO),Italy.


Process Parameters


by John Eisenbrey, Phyllis Huang, Michael Soulen,and Margaret Wheatley

Introduction

Traditional chemotherapy relies on asystemic dosage of a highly toxic agentwith hopes that a small proportion ofthe drug will be delivered to the actual

site of need. This approach to cancer treat-ments has led to the well known debilitatingside effects commonly associated with chemo-therapy. Additionally, this approach leads to arelatively low therapeutic level at the actualtumor site, greatly reducing the potency of thedrug. While this approach has had success withsome forms of cancer, it has been woefullyunsuccessful in the treatment of liver cancer.In 2006, roughly 1.3 million new cases of pri-mary liver cancer were reported worldwidewith a five year survival rate of 7%.1 Two of thesalient problems facing treatment of liver can-cer are the low early detection rate and theinability to gauge tumor size,2 both of whichcould be solved using a delivery platform ca-pable of improving imaging contrast as well asdelivering drug. Additionally, many other formsof more treatable cancer metastasize to theliver and become fatal.

Doxorubicin (Dox) is one of the few chemo-therapeutics that has shown clinical efficacyagainst liver cancer.3 However, its use bringswith it severe side effects, including cardio-toxicity and congestive heart failure.4,5 Thiseffectiveness coupled with its systemic sideeffects makes it an ideal candidate for a tar-geted drug delivery platform. The proposedplatform currently being developed in our labuses ultrasound sensitive carriers injected in-travenously, while ultrasound is being applied

to a particular treatment area with the goal toprovide a systemic dosage of carrier, but alocalized delivery of chemotherapeutic drug.

Ultrasound has many advantages as animaging modality because it is relatively inex-pensive, portable, real time, uses non-ionizingradiation, and is readily available.6 Ultrasounduses the transmission and reflection of gener-ated pressure waves to create an image. De-pending on insonation frequency and acousticpressure, it can be focused on areas smallerthan one cm2 at any depth in the body.7 Despitethe technical advances within ultrasound,8 themodality continues to have trouble distinguish-ing between healthy and diseased tissue, mak-ing cancer detection difficult.

These shortcomings in detection have givenrise to development of ultrasound ContrastAgents (CA), which increase image contrast byupwards of 20 dB.9 CA consist of a coremicrobubble (generally a non toxic, relativelyinert gas such as air, sulfur hexafluoride ofperfluorocarbon) stabilized with an outer shell.A CA must be intravenously injectable, stableenough to last the duration of the imagingsession, non-toxic, non-carcinogenic, andsmaller than 8 µm in order to freely pass throughthe capillary beds during circulation. Theseagents have been successfully synthesized frompolymers, proteins, surfactants, and lipids.10 Ithas been shown that ultrasound imaging hasbecome sensitive enough to detect a single CAin circulation.11 A method for producing hollow,stable, echogenic microbubbles from both polylactic and poly lactic-co-glycolic acids has beendeveloped in our lab, giving ultrasound en-

Doxorubicin Loaded Contrast Agentsfor Ultrasound Triggered DrugDelivery: Importance of ProcessParameters

This articlediscussesresearchcontributing tothe developmentof a drug deliveryplatform that hasthe potential togreatly reducesystemic toxicitycommonlyassociated withchemotherapy,while increasingthe concentrationof the drug at atumor site.

First presented byJohn Eisenbrey atthe 2007 ISPEDelaware ValleyStudent PosterCompetition, thisresearch went on towin the GraduateLevel award at theISPE InternationalStudent PosterCompetition in LasVegas, NV later thatyear.

Reprinted from





Process Parameters


hancement of more than 20 dB both in vitro and in vivo.12,13

These agents show a tight size distribution with a mean sizeof 1.23 +/- 0.62 µm. It also has been shown that these agentsare completely biodegradable after use and break down dueto hydrolysis.14

The development of CA for targeted drug delivery hasprogressed along two paths: the loading of drug onto andwithin the CA, and the addition of tumor specific ligands tothe surface of the CA in order to provide a higher bindingaffinity of the agent to the area of need. While the majority ofresearch in these areas has remained separate, the twoapproaches are expected to merge, eventually creating a CAwith a high affinity to solid tumors that also can releasechemotherapeutics when triggered with ultrasound.

Several labs have had success with adhesion of tumorspecific ligands to the surface to CA. Poly-lactic-acid CA havebeen developed with the αvβ3 integrin specific Arg-Gly-Asp(RGD) ligand on the surface. These agents were shown todemonstrate a significantly greater affinity for cancer cellsexpressing the αvβ3 integrin than non-targeted CA.15 Othergroups have shown that targeting αv-Integrins is a viablemeans for assessing angiogenesis using ultrasound.16 Re-search such as this reinforces the importance of the CA drugdelivery platform for use for both diagnosis and therapy.

Additionally, several labs have shown the potential fordrug-loaded CA. Huang and MacDonald have shown that itis possible to encapsulate drugs within acoustically activeliposomes.17 Using calcein as a model, the group was able toachieve an encapsulation efficiency of roughly 14% withoutsacrificing ultrasound enhancement of the agent. In vitro,the agents showed a 60% release when insonated andalmost no release when stirred in solution.17 Pong et al. haveshown that it is possible to use ultrasound to mediateleakage from large, acoustically sensitive phospholipidvesicles, again triggering high release amounts in vitro withlow frequency ultrasound.18 High-Intensity Focused Ultra-sound (HIFU) has been used as a source of hyperthermiatogether with injected thermosensitive liposomes to en-hance delivery and efficacy of doxorubicin in tumors.19 It hasbeen shown that HIFU can be used to induce tumor ablationthrough heating.20 HIFU also has been used in conjunctionwith magnetic resonance imaging which is used for in situtarget definition and identification of nearby healthy tissuethat is to be spared.21 In addition to the use of ultrasound todisrupt drug-loaded agents and for ablation, the use of theplatform also may be beneficial due to the well documentedeffects of ultrasound with traditional systemic drug deliv-ery. Ultrasound has been shown to lead to increased efficacyat the area of insonation through increased drug retentionat the tumor site. Several studies have been done showingthat systemic administration of doxorubicin combined withlow intensity ultrasound (0.25-2 W.cm2) leads to increaseddrug efficacy.22,23 These therapeutic benefits of ultrasoundwith chemotherapy become even more significant in thepresence of CA. Higher intensity ultrasound can be used torupture CA in circulation by inducing inertial or collapsecavitation.24 This cavitation has been shown to result in

higher membrane permeability, increasing drug uptake atthe site of insonation.25 Clearly, many advantages existusing a drug delivery platform of targeted, drug-loaded CAcompared to traditional chemotherapy.

Several methods of loading Dox in PLA CA have beendeveloped within our lab. Drug has successfully been loadedboth on the surface and within the shell of these agentswithout sacrificing the acoustic performance of the originalagent.26 The successful synthesis of these agents has shownthe feasibility of a drug delivery system using CA and re-sulted in several optimization studies. This paper will exam-ine the optimization of one particular method, dry surfaceadsorption, in which Dox is applied to the surface of the CAafter synthesis. Effects of temperature and adsorption timewill be examined. Both in vivo and in vitro results areincluded showing the method’s feasibility for both ultrasoundenhancement and targeted drug delivery.

MaterialsPoly-lactic-acid (PLA) (MW = 83 kDa) was purchased fromLakeshore Biomaterials. Poly (vinyl alcohol), 88% mole hy-drolyzed, with a MW of 25 kDa, Dox, and camphor werepurchased from Sigma-Aldrich. Ammonium carbonate waspurchased from J.T. Baker. All other chemicals were reagentgrade from Fisher Scientific.

MethodsMicrobubble PreparationPLA CA were prepared using a double emulsion method (W/O)/W developed in our lab.12 Briefly, 0.5 g of PLA combinedwith 0.05 g of camphor were dissolved in 10 ml of methylenechloride. One ml of ammonium carbonate (4% w/v) was addedand the mixture sonicated at 110 Watts for 30 seconds atthree seconds on and one second off using a Misonix Inc.sonicator probe. After sonication, the resulting (W/O) emul-sion was poured into 50 ml of 4°C, 5% polyvinyl alcohol withina 600 ml beaker and homogenized for five minutes at 9000rpm. CA were then collected by centrifugation and washedwith hexane. After hexane evaporation, the capsules wereflash frozen and lyophilized for 48 hours. During this process,ammonium carbonate and camphor sublime out of the cap-sule, leaving a void in their place that later fills with air afterbeing exposed to atmospheric pressure. All dry samples arethen stored in a desicator at -20°C until used.

Drug LoadingSeveral methods for loading Dox onto the CA have beendeveloped in our lab. For this study, we focus on adsorption,were Dox is adsorbed to the surface of pre- fabricated CA. Anelectrostatic attraction between the drug and polymer resultsin a strong adhesion between the two. One hundred mg of dryagent was added to 3 ml of 3% Dox (w/v). The sample was thenshaken end over end for varying time periods at varyingtemperatures. Drug adsorption was done at 20°C and 4°Cwith loading times of five and 30 minutes, one, two, and 24hours. After addition of Dox to the shell surface, the CA wascentrifuged and washed three times to remove any excess


Process Parameters


drug that had not adsorbed to the surface. After centrifuga-tion, samples were flash frozen and lyophilized for 48 hours.All samples were then stored in a desicator at 20°C to avoiddegradation through hydrolysis.

Particle SizingParticle size was measured on a Malvern Nano ZS ParticleSizer in order to characterize the mean size and distributionof the agent. Roughly one mg of dry sample was suspended in1 ml of deionized water. Each sample was measured threetimes and the results averaged.

Scanning Electron Microscopy (SEM) ImagingDrug loaded agent was imaged using an environmental scan-ning electron microscope. Freeze dried agent was sputtercoated with platinum for 60 seconds prior to imaging. Allimages were taken at a magnification of 6000x with an accel-erating voltage of 3.0 kV. All electron microscopy was done atthe Drexel University Materials Characterization Facility.

Acoustic TestingUltrasound enhancement, stability, size distribution overtime, and drug release were all measured in vitro to deter-mine the agent’s feasibility as a platform for drug delivery.Acoustic enhancement and stability under ultrasonic condi-tions were both measured in vitro to determine microbubble

performance as a CA as well as sensitivity to ultrasound.These properties were characterized using the setup shownin Figure 1.

A 12.7 mm diameter, 5 MHz transducer with a focal lengthof 50.8 mm, 6 dB bandwidth of 91% and a pulse length of 1.2mm was placed in a water tank filled with 37°C, 18.6 MΩ-cmdeionized water. These acoustic parameters are all wellwithin current ultrasound imaging standards.7 The ultra-sound transducer was focused through an acoustically trans-parent window in the sample holder onto the sample volume.A pulser-receiver generated an acoustic wave with a pulserepetition frequency of 100 Hz. The reflected signal from theCA was detected by the transducer and amplified 40 dBbefore being read by an oscilloscope. Data was then acquiredand values calculated using Lab View 7 Express and storedon a CPU.

Ultrasound Contrast EnhancementBackscattering enhancing was measured as a function of CAdosage in order to gauge both ultrasound contrast enhance-ment as well as the agent’s sensitivity to ultrasound. Thesample holder shown in Figure 1 was filled with 50 ml of PBSat 37°C. Three mg of dry agent was suspended in 800 µl ofPBS. Samples were pipetted into the holder in incrementaldosages from 0-16 µg/ml. The agent was allowed to mix for 10seconds before readings were taken to ensure homogeneoussample volume. Sixty readings of each individual samplewere taken using LabView. All final values were based on anaverage of three readings from three individual samples.

Physical Effects of Insonation on CAIn vitro insonation experiments were all performed in theacoustic setup shown in Figure 1 in order to assess the agent’sbehavior in the sound field, and hence its suitability as aplatform for triggered drug delivery. Ten mg of sample wereadded to 50 ml of 37°C PBS within the sample holder. Theagent was then insonated using the 5 MHz transducer at 0.68MPa pressure. At varying time intervals, one ml of mixture

Figure 1. Acoustic setup used to test ultrasound enhancement,stability, and in vitro drug release under ultrasonic conditions.

Figure 2. Ultrasound enhancement as a function of CA dosage foragent loaded for varying time intervals at 20°C. Beyond 30minutes of loading at room temperature, the agent does notprovide enough enhancement to be used as a CA.

Figure 3. Ultrasound enhancement as a function of CA dosage foragent loaded at varying time intervals at 4°C. After 24 hours ofloading at 4°C, the agent still provides enough enhancement tobe used as a CA.


Process Parameters


was taken using reverse pippeting. CA size distributionswere then measured using the particle sizer. Controls wereperformed using the same setup and time scale, but withoutinsonation.

In Vivo ImagingIn vivo imaging was performed in New Zealand rabbits with2 to 3 cm2 VX2 tumors implanted within the liver in order todetermine ultrasound enhancement and the agent’s ability topenetrate within a vascular tumor model. Seventy mg ofagent suspended in 510 ml of physiological saline wereinjected intravenously through an auricular line and thetumor was imaged continuously both before and after injec-tion. Doppler ultrasound was used to image the solid tumorat a frequency of 5 MHz at a mechanical index of 1.0 for a totalof 20 minutes. All images were saved and digitized for laterprocessing.

Statistical AnalysisA one-way ANOVA was used to determine statistical signifi-cance and was performed using Prism 3.0. Statistical signifi-cance was determined using α = 0.05. A Newman-Keuls testwas performed as a post test to determine significant vari-ance between groups.

Results and DiscussionOptimization of Drug Adsorption ParametersLonger loading times result in higher encapsulation effi-ciency, but jeopardize the CA’s sensitivity to ultrasound.Thus, after drug loading, ultrasound enhancement of the CAwas measured to ensure that the agent was both still usableas a CA and still sensitive enough to be used as an ultrasoundsensitive carrier.

Figure 2 shows the ultrasound enhancement of the agentsloaded for various times at 20°C, while Figure 3 shows thesame loading times at 4°C.

Figures 2 and 3 show the importance of loading tempera-ture on the acoustic properties of the CA. Enhancement foreach loading time was found to be significantly higher whenloaded at 4°C compared to 20°C (α = 0.05).

Drug loading at 4°C for 24 hours produces a drug loadedCA that provides more ultrasound enhancement than anagent loaded for one hour at 20°C. High enhancement is animportant characteristic of the CA in order for the agent toprovide good contrast enhancement during the scan, and alsofor triggering drug release.

As mentioned previously, longer loading times are neededto improve encapsulation efficiency. Encapsulation efficiencywas not statistically different between the two loading tem-peratures, with both methods giving an efficiency of 62 +/- 8%after 24 hours. Thus, the optimal loading technique withinour lab has shown to be at 4°C over 24 hours. SEM image ofthese agents is shown in Figure 4. The agents show smoothsurface morphology, a mean size of 1.3 µm and a relativelytight size distribution. Drug loading of the CA could furtherbe optimized by performing adsorptions at lower tempera-tures at mediums with freezing points below zero.

Physical Effects of Insonation on CA In VitroAdditionally, the size distribution of the drug loaded CA wasexamined over the course of insonation as described in themethods section. Figure 5 shows the change in size distribu-tion of the CA over time with and without insonation.

Figure 5 shows the manner in which the CA populationchanges when in solution over time. Notice over the course of15 minutes with no insonation, the agent swells slightly,becoming 10% larger. However, when ultrasound is appliedto the sample, the agent population decreases drastically. CAhave been shown to rupture when insonated and others losetheir gas cores, causing them to shrivel.27 These findings areconfirmed by our results. After 15 minutes of insonation, wesee a population size reduction of almost 200% with the finalpopulation having a mean size of roughly 300 nm. Thesedifferences were found to be statistically significant with afinal size small enough for particles to exit through the

Figure 5. Size distributions of the CA population over time areshown for samples under insonation, and samples under noinsonation while circulating in the sample holder described inFigure 1.

Figure 4. SEM image of CA population at 6000x. (Size bar =5μm, Accelerating voltage = 3 kV).


Process Parameters


vasculature within the tumor and de-liver their contents on the cellular level.

Results from preliminary studiesshow great promise for a drug loadedCA to be injected intravenously andwith localized drug delivery initiatedthrough focused ultrasound. We haveseen that our agent gets excellent pen-etration into a solid tumor, and afterinsonation becomes small enough toescape out of the vasculature and ini-tiate change on the cellular level. Thus,the platform can provide a primaryrelease of drug within the solid tumoras well as a carry drug out of the vascu-lature and to the cancer cells afterinsonation.

In Vivo Imaging and TumorPenetrationIn vivo experiments were conductedshowing the agent’s potential to bothprovide enhancement during ultra-sound scans, and penetrate into thevasculature of a solid tumor. Dopplerultrasound was used to image the solidtumor at a frequency of 5 MHz at amechanical index of 1.0. This corre-sponds to pressures that have beenshown to cause destruction of polymershelled CA,27 a potential mode of trig-gering drug release. Figure 6 shows Doppler images of thetumor pre and post injection of the CA.

Results from imaging studies with our agent show that theCA penetrates well into the solid tumor and the in vitrostudies indicate that focused ultrasound can be used to causedestruction of the microbubbles in an isolated region. Fordrug release studies in vivo, the ultrasound beam can befocused entirely within the solid tumor (1-2 cm2), initiatingdestruction of the CA primarily within the tumor.

ConclusionsA drug delivery platform has been developed in conjunctionwith an ultrasound contrast agent, and parameters havebeen identified which do not destroy the echogenicity of theCA, which can be administered systemically. This type ofplatform can greatly reduce systemic toxicity commonly asso-ciated with chemotherapy, while increasing the concentra-tion of the drug at the tumor site. The process parameters ofadsorbing drug to the surface of the CA have been optimizedwith a loading time of 24 hours at 4°C resulting in anencapsulation efficiency of 62 +/- 8%, with loss of roughly only25% in echogenicity. These agents respond well to ultra-sound, show smooth surface morphology, and a tight sizedistribution. Preliminary drug delivery studies show theplatform is detectable by ultrasound in vivo and providesgood tumor penetration. In vitro, the size distribution of the

agent under ultrasound was shown to become small enoughto escape out of the vasculature which would cause contact ofthe drug with the cancer at the cellular level. This work,combined with research on the addition of tumor specificligands to the surface of CA may ultimately result in a drugdelivery platform in which drug carriers with a higher affin-ity for solid tumors carry drug through the body and releasetheir contents at a desired location when triggered withfocused ultrasound.

References1. American Cancer Society., “Detailed Guide: Liver Can-

cer,” www.cancer.org, 2007.2. Yang, B. et al., “Prospective Study of Early Detection for

Primary Liver Cancer,” Research and Clinical Oncology,Vol. 123, No. 6, 1997, pp.357-360.

3. Kemeny, N., Sugarbaker, P.H., “Treatment of MetastaticCancer: Treatment of Metastatic Cancer to the Liver,” In:De Vita VT, Hellman S and Roseberg SA, eds. Cancer,1989.

4. Van-Hoff, L., “Risk Factors for Doxorubicin-induced Con-gestive Heart Failure,” Annals of Internal Medicine, Vol.5, No. 91, 1979, pp. 710-717.

5. Singal, P., Ilskovic, N., “Doxorubicin-induced Cardiomy-opathy,” New England Journal of Medicine, Vol. 339,1998, pp. 900-905.

Figure 6. Power Doppler image of an implanted VX2 tumor within the liver of a 2-3 kgNew Zealand Rabbit, a) pre injection, b) post injection. The upper left oval shows the solidtumor within the liver. Notice that after injection, CA penetrates well into the tumor aswell as in the tumor feeding vessels surrounding it. Both Doppler and grey scale imagesare enhanced after injection, with the tumor mass showing much greater detail. 179 ×154 mm (72 × 72 DPI).


Process Parameters


6. Lewin, P., “Quo Vadis Medical Ultrasound,” Ultrasonics,Vol. 42, 2004, pp. 1-7.

7. Szabo, T., “Diagnostic Ultrasound Imaging: Inside Out,”New York: Elsevier Academic Press, 2004.

8. Kudo, M., “New Sonographic Techniques for the Diagno-sis and Treatment of Hepatocellular Carcinoma,”Hepatology Research, Vol. 37, Suppl. 2, 2007, pp.S139-S199.

9. Forsberg, F., et al., “Clinical Applications of UltrasoundContrast Agents,” Ultrasonics, Vol. 35, 1998, pp. 695-701.

10. Correas, J.M., et al., “Ultrasound Contrast Agents: Prop-erties, Principles of Action, Tolerance and Artifacts,”European Radiology, Vol. 11, 2001, pp. 1316-1328.

11. Klibanov, A., et al., “Detection of Individual Microbubblesof Ultrasound Contrast Agents,” Investigative Radiology,Vol. 39, No. 3, 2004, pp. 187-195.

12. El-Sherif, D., Wheatley, M., “Development of a NovelMethod for the Synthesis of a Polymer Ultrasound Con-trast Agent,” Journal of Biomedical Materials, 66A, 2003,pp. 247-357.

13. Wheatley, M., et al., “Comparison of in Vitro and in VivoAcoustic Response of a novel 50:50 PLGA Contrast Agent,”Ultrasonics, Vol. 44, 2006, pp. 360-367.

14. Narayan, P., Wheatley, M., “Preparation and Character-ization of Hollow Microcapsules for Use as UltrasoundContrast Agents,” Polymer Engineering and Science, Vol.11, No. 39, 1999, pp. 2242-2255.

15. Lathia, J.D., Leodore, L., Wheatley, M., “Polymeric Con-trast Agents with Targeting Potential,” Ultrasonics, Vol.42, 2004, pp. 763-768.

16. Leong-Poi, H., et al., “Noninvasive assessment of Angio-genesis by Ultrasound and Microbubbles Targeted toAlpha-v-Integrins,” Circulation, Vol. 107, 2002, pp. 455-460.

17. Huang, S., MacDonald, R.C., “Acoustically Active Lipo-somes for Drug Encapsulation and Ultrasound TriggeredRelease,” Biochimica, et Biophysica Acta, Vol. 1665, 2004,pp. 134-141.

18. Pong, M., et al., “In Vitro Ultrasound-Mediated Leakagefrom Phospholipid Vesicles,” Ultrasonics, Vol. 45, 2006,pp. 133-145.

19. Drom, S., et al., “Pulsed-High Intensity Focused Ultra-sound and Low Temperature-Sensitive Liposomes forEnhanced Targeted Drug Delivery and Antitumor Ef-fect,” Clin Cancer Res, Vol. 9, No. 13, 2007, pp. 2722-2727.

20. Chapelon, J., et al., “In Vivo Effects of High-IntensityUltrasound on prostatic Adenocarcinoma DunningR33271,” Cancer Research, Vol. 52, 1992, pp. 6353-6357.

21. Moonen, C.T.W., “Spatio-Temporal Control of Gene Ex-pression and Cancer Treatment Using Magnetic Reso-nance-Imaging- Guided Focused Ultrasound,” Clin Can-cer Res, Vol. 12, No. 13, 2007, pp. 3482-3489.

22. Harrison, G., Balcer-Kubiczek, E., Eddy, H., “Potential ofChemotherapy by Low-Level Ultrasound,” InternationalJournal of Radiation, Vol. 59, 1987, pp.1453-1466.

23. Umemura, S., et al., “Studies of Sonodynamic CancerTherapy,” Engineering in Medicine and Biology Society,

No. 14, 1992, pp. 345-355.24. Hussenii, G.A., et al., “The Role of Cavitation in Acousti-

cally Activated Drug Delivery,” J Control Release, Vol. 2,No. 107, 2005, pp. 253-261.

25. Unger, E., et al., “Local Drug and Gene Delivery throughMicrobubbles,” Progress in Cardiovascular Diseases, Vol.1, No. 44, 2001, pp. 45-54.

26. Mualem-Burnstein, O., Eisenbrey, J., Wheatley, M., “Doxo-rubicin Loaded Polymeric Contrast Agents: Comparisonof Shell Material and Drug Loading Methods,”Biomaterials, Submitted January 2008.

27. Bloch, S.H., et al., “Optical Observation of Lipid- andPolymer-Shelled Ultrasound Microbubble ContrastAgents,” Applied Physics Letters, Vol. 84, No. 4 2004, pp.631-633.

AcknowledgmentsIn vivo imaging studies were performed at The University ofPennsylvania with help from Susan Schultz. Electron Mi-croscopy was done at Drexel University’s Materials Charac-terization Institute with the help of Kelleny Oum. Fundingwas provided through NIH grants HL 52901 and CA 52823,and The Coulter Foundation.

About the AuthorsJohn Eisenbrey did his undergraduate workat the University of Delaware where he earnedbachelors degrees in mechanical engineer-ing and management. He is currently a PhDCandidate at Drexel University in Biomedi-cal Engineering focusing on ultrasound sen-sitive drug delivery vehicles. Academicachievements include winning the 2007 ISPE

Annual Meeting Graduate Student Poster Competition, the2005 Charles B. Evans award for undergraduate research,and graduating magna cum laude from the University ofDelaware. He is currently a member of ISPE, ASME, and TauBeta Pi Engineering Honor Society. He can be contacted byemail: [email protected].

Drexel University, School of Biomedical Engineering, Sci-ence and Health Systems, 3141 Chestnut St., Philadelphia,Pennsylvania 19104, USA.

Phyllis Huang is currently studying bio-logical sciences as an undergraduate at DrexelUniversity. In the biomedical engineeringdepartment, she is assisting in research forultrasound drug delivery, such asmicrobubble fabrication and polymer test-ing. Her academic achievements include at-taining Dean’s List and becoming a STAR

Scholar at Drexel University. She is currently a member ofTri-Beta Biological Honors Society and AMSA. She can becontacted by email: [email protected].



Process Parameters


Michael Soulen is Professor of Radiologyand Surgery at the University of Pennsylva-nia, where he directs the Interventional On-cology program, which utilizes minimally-invasive image-guided therapies to treat solidtumors in the liver, kidney, lung, and muscu-loskeletal system. His research concentrateson animal and clinical trials of novel image-

guided tumor therapies. He can be contacted by email:[email protected].

Hospital of the University of Pennsylvania, Division ofInterventional Radiology, 1 Silverstein, 3400 Spruce St.,Philadelphia, Pennsylvania 19104, USA.

Margaret Wheatley is the John M ReidProfessor of Biomedical Engineering at DrexelUniversity. She obtained her undergraduatetraining at Oxford University, United King-dom, graduate training at University ofToronto, Canada, and postdoctoral training atMIT, Cambridge, Massachusetts. One cur-rent research interest includes development

of ultrasound contrast agent for diagnosis and therapy. Par-ticular focus areas are drug delivery, tumor specific target-ing, and contrast agents in the nano scale. Other researchareas include design and construction of smart constructs forspinal cord repair, and development of drug-eluting tissuejoining devices for small vessel anatomizes. Drexel has ac-knowledged her contribution by award of the Drexel Univer-sity Research Achievement Award. She is an elected fellow ofAIMBE. She can be contacted by email: [email protected].



ISPE Update


API COP Forms Subcommittee to Drive Innovation inProcess Technologyby Rochelle Runas, ISPE Technical Writer

Taking a lead role in the future of Active PharmaceuticalIngredient (API) manufacturing, the API Communityof Practice (COP) has formed a Process Technology

Subcommittee.“We are seeing a period of change catalyzed by the FDA’s

GMPs for the 21st Century and Industry initiatives to reducethe cost of drugs,” says the Subcommittee’s draft missionstatement. “This is generating many new concepts based onthe strength of scientific understanding, for example, QbD,risk-based approach, PAT application, PQLI, etc. leading to apotential revolution in the way that we approach the processdevelopment, technology utilization, and facility design.”

“There is a real opportunity for API manufacture to reduceits costs and ensure its future,” said John Nichols, Co-Chairof the new Subcommittee, at the ISPE Copenhagen Confer-ence in April. “To do this, multidiscipline interaction betweendevelopment science, engineering, and quality disciplines isessential.”

The API COP Process Technology Subcommittee’s missionis to drive innovations in process technology for manufactureof APIs and BPCs by the identification of subject matterexperts, interaction and partnership with related groups, bysupport and input to the education program, the promotion/marketing of new innovations in process technology, and inputto industry guidelines and best practice documents.

“We also need to note that some of the old paradigms nolonger apply, circumstances have changed,” Nichols said. Forexample: “Production Rates: Plants having a capacity ofgreater than 10 × 106 lb/yr (approx. 5000 T/yr) are usuallycontinuous, whereas plants having a capacity of less than 1× 106 lb/yr are normally batch types,” etc.1 For the pharmaceu-tical industry and specialty chemical industries, this is nolonger true.”

Nichols said currently there is a lot of activity both in theUS and Europe toward making such paradigm shifts. Forfurther details about these activities, see “Innovations inProcess Technology for Manufacture of APIs and BPCs” byNichols (pgs. 24 – 34). The article summarizes presentationsfrom the ISPE Copenhagen Conference, giving a glimpse intothe future of efficient processing implementation (e.g., con-tinuous), new technology/equipment, and ISPE/ASTM en-abling activities.

In addition to sponsoring the ISPE Copenhagen Confer-ence seminar on Innovations in API/BPC Manufacture, theSubcommittee has recruited thought leaders from academia,API manufacturers, and equipment and services suppliers tofurther help define the group’s direction and drive change.

The Subcommittee is currently working on an appendix ofpotential technologies; a roadmap identifying main issues

inhibiting the introduction of new technologies; and is main-taining its ongoing input to the ASTM Continuous ProcessingStandard development. The Subcommittee also has futureplans to generate a white paper covering in more detail theparticular areas of compliance/quality assurance and controlassociated with continuous processing. Future educationalofferings will build on this work at US and EU conferences,and there are long term plans to expand the API BaselineGuide with sections on process intensification/continuousprocessing and small volume/high potency manufacture, saidNichols.

To join the Subcommittee and for further information onthe group, visit the API Community of Practice site atwww.ISPE.org.

The API COP is one of 17 ISPE Communities of Practice(COPs). ISPE’s COPs enable like-minded professionals toconnect through an interactive online community. Throughprofessional networking and peer collaboration, ISPE’s COPsproduce discipline-specific content that deepens members’knowledge and expertise, increases quality and continuousimprovement in the industry, and help to achieve ISPE’s corepurpose of leading global innovation. Current COPs are:

• Active Pharmaceutical Ingredients (API)• Biotechnology• Commissioning and Qualification (C&Q)• Containment• Critical Utilities (CU)• Disposables• Engineering Standards Benchmarking• Good Automated Manufacturing Practice (GAMP)• Good Control Laboratory Practices (GCLP)• Heating, Ventilation, and Air Conditioning (HVAC)• Investigational Products (IP)• Packaging• Process Analytical Technology (PAT)• Process/Product Development (PPD)• Project Management (PM)• Sterile Products Processing (SPP)• Sustainable Facilities

Joining an ISPE COP is easy, free, and open to both Membersand nonmembers. Visit www.ISPE.org under the People andGroups heading for further information.

Reference1. Douglas, James M, Conceptual Design of Chemical Pro-cesses. New York, McGraw-Hill. 1988, p.108. Under theheading, “Guidelines for Selecting Batch Processes.”

Reprinted from





ISPE Update


Annual Meeting to Help Navigate Horizon of IndustryChange

The industry compass continues topoint toward a horizon of changethat includes more science- and

risk-based approaches, lean and flex-ible manufacturing, and complex tech-nologies. To help navigate this hori-zon, the educational sessions plannedfor the 2008 ISPE Annual Meeting (26– 29 October, Boca Raton, Florida, USA)have been tailored to address theseand other emerging trends and prac-tices.

The following are highlights of what toexpect at the Keynote Session and inseven educational tracks:

Keynote SessionThis year’s Annual Meeting KeynoteSpeakers are:

• Hans Rosling, MD, PhDProfessor of International Health,Karolinska Institute and Directorof Gapminder Foundation,Stockholm, Sweden

• Janet Woodcock, MDDirector, Center for Drug Evalua-tion and Research, US FDA

• Patrick Y. Yang, PhDExecutive Vice President, ProductOperations, Genentech, USA

Regulatory TrackMore than a dozen regulators havebeen invited to speak and are confirm-ing now. This track is for anyone whodeals with regulatory compliance orquality issues and focuses on the fol-lowing topics:

• Product Quality Lifecycle Imple-mentation (PQLI) and a practicalapproach to global implementationof ICH documents

• Lessons learned from use of ProcessAnalytical Technology (PAT) andhow to apply them to your own fast-track PAT program

• Risk-MaPP and the use of risk-basedtechniques to manage cross contami-nation – key regulatory agenciespresent

• Financial, legal, technical, and regu-latory issues facing biosimilars

• Regulatory compliance and qualityissues as related to quality labora-tory facilities and an update on themuch anticipated Baseline® Guidefor Quality Laboratory Facilitiesscheduled for publication duringfirst quarter 2009

Innovation TrackThe industry faces many challengesthat require companies to develop prod-ucts faster and at less cost. Manufac-turing is tasked to reduce costs andbecome more lean and flexible, whilemeeting compliance in an increasingglobal environment. This track exam-ines how today’s leaders are meetingsome of these challenges, including:

• Continuous and high throughputprocesses for everything from APIsto biologics and innovative processesto increase the downstream effi-ciency in biotechnology

• Implementing PQLI, Quality byDesign (QbD), and other initiativesto reduce time and cost for develop-ing new products

• Vaccines, from pre-pandemic vac-cines to current manufacturing chal-lenges

• How nanotechnology can be appliedand developed

• Changes in regulatory thoughts onGMPs

• A special session probes the idea ofinnovation and how it can be/is be-ing taught and practiced in and outof our industry

ManufacturingOperations Track

This session will address real concerns,including:

• Higher costs requiring increasedefficiencies

• Decisions on internal productionversus outsourcing opportunities

• Ways to be more productive andfaster using ideas from other indus-tries

• Trouble-shooting your aseptic pro-cess

Engineering Design TrackThis track presents a unique opportu-nity to see and hear in a few hours whatthe experts have devoted many yearsdeveloping and improving through ap-plication of innovation and response todemands of a dynamic industry. Del-egates will have a chance to reviewdeveloping technologies and techniquesand practical solutions, including:

• Working details of systems in thetransfer technologies field, optionsin the selection of equipment, andanalysis of competing systems

• Facilities retrofit design processes,available choices between retrofit-ting and new construction, and con-struction logistics planning

• Project presentations by the Cat-egory Winners of the Facility of theYear Awards (FOYA) program for2008

• Energy reduction techniques forcritical utility applications and com-parison with new, more efficientwater and pure steam technologies

• Identification and description of newand developing equipment construc-tion standards and current trendsin their generation

• Function and importance of kilo labsand pilot plants, in addition to costand scheduling of related projects

InvestigationalProducts Track

Education sessions are offered fromtwo perspectives of critical importancefor today’s clinical supply chain profes-sional: operations and management.


ISPE Update


Annual Meeting...Continued from page 82.

Sessions in the operations track in-clude business process management;supply pooling through protocol inter-pretation; distribution; managementof controlled substances; and compara-tor strategy. Sessions in the manage-ment track review business processmanagement; developing and engag-ing talent; successful partnerships;delivering supplies to emerging mar-kets; and applying six sigma conceptsto study supply planning. In the gen-eral session, recognized experts willtalk about the future of clinical sup-plies; removing use-by dates from clini-cal trial materials in Europe; and clini-cal research perspective on handling/managing clinical supplies.

Project Management TrackIn today’s troubled economic times,the demands to deliver projects ahead

of schedule and under budget forceproject managers to adopt innovativemethodologies. While classic tools andprograms will never be abandoned,the industry is demanding an out-of-the-box mindset, and this is most of-ten affected by analyzing lessonslearned from both previous successand failure.

This interactive track will analyzecurrent, real world, applied innova-tions in project delivery and take ad-vantage of insights from industry ex-perts discussing how current and fu-ture trends can be used to transformcurrent projects and help shape futureproject success. There will be discus-sions on successful project executionon a wide variety of project types; man-agement of risk and innovation; thedifferences between projects of differ-ent size and scope; and the human side

of managing projects.

Efficient and EffectiveCompliance Track

Attendees of this track will have theopportunity to learn from the earlyadopters of the new approaches to com-pliance. The program covers a broadarray of key topics, such as projectmanagement challenges; best practiceapproaches to Computer System Vali-dation (CSV); and incorporation of newapproaches to compliance in facilities,utilities, and manufacturing systems.Case studies provide insight on howcompanies are refining project ap-proaches through cooperative efforts ofquality, IT, and engineering. Discus-sions will include:

• ASTM E2500 – Practical examplesof a science- and risk-based approach

Concludes on page 4.

ISPE and CCPIE Work Together to Bring LearningOpportunities to China

As part of a pioneering relationshipbetween an American not-for-profit

organization and the Chinese govern-ment, the China Center for Pharma-ceutical International Exchange(CCPIE) has invited ISPE to collabo-rate with them in bringing a first-of-its-kind learning opportunity to Chi-nese pharmaceutical professionals.

The 2008 ISPE Conference in Chinawill take place 11-12 November 2008in conjunction with the 13th ChinaInternational Pharmaceutical Indus-try Exhibition (12-14 November) at theChina International Exhibition Cen-ter in Beijing, China.

The CCPIE acts as the conduit be-tween professional organizations andthe SFDA, China’s pharmaceuticalregulatory agency. The collaborationbetween ISPE and CCPIE is as pivotalas it is historic, as the CCPIE alsotrains members of the SFDA.

The keynote session on the first dayof the conference will provide delegateswith insights on China’s GMP regula-tions, perspectives of the China StateFood and Drug Administration (SFDA),and the US Food and Drug Adminis-tration (FDA).

The second day of the conferenceconsists of the following three paralleltracks and their respective track lead-ers:

• GAMP® 5 – Sion Wyn, Director,Conformity Ltd.

• Biotechnology – Ron Branning, VPCorporate QA, Gilead Sciences Inc.

• Validation – Steve Wisniewski, Se-nior Associate, Director of Compli-ance, IPS

Many other notable speakers will takethe podium during the two-day confer-ence, including:

• Robert Best, President and CEO,ISPE

• A representative from SFDA• A representative from US FDA• Zhao Yajun, Director, CCPIE• Paul D’Eramo, Executive Director,

Quality and Compliance Worldwide,Johnson & Johnson

• Professor Zheng Qiang, Director,Centre for Pharmaceutical Informa-tion and Engineering Research, Pe-king University

• Robert Tribe, ISPE Asia-PacificRegulatory Affairs Advisor

To review the conference agenda,speakers’ biographies, and to down-load a registration form, please visitwww.ISPE.org.cn or email: [email protected].


ISPE Update


COMING SOON:Downloadable Glossary ofPharmaceutical and BiotechnologyTerms

In fall 2008, ISPE will introduce a portable PDF version of its acclaimedGlossary of Pharmaceutical and Biotechnology Terminology for ISPE Mem-

bers only. The Glossary contains more than 5,800 abbreviations, acronyms,and terms focusing on specific areas in biotechnology and pharmaceuticalscience and manufacturing, such as computer technology, manufacturingprocesses, water treatment, welding, metallurgy, HVAC, medicine, biology,chemistry, and many other subjects. This tool is designed to help industryprofessionals with daily job functions, understand terminology used in ISPE’stechnical documents and education seminars, and standardize industryterms around the globe.

for specification, design, and verifi-cation of facilities and equipment

• Process validation – Hear from con-firmed regulatory speaker RickFriedman, US FDA

• Computer System Validation (CSV)– Benchmarking your current prac-tices and solving your real ques-tions

• Case studies for transitioning to thenew verification models

• Best practice case studies in CSV• Project information management –

Lessons learned and recommenda-tions from case study projects

• C&Q Baseline® Guide updates –Review status of Volume 12, a sci-ence- and risk-based approach toverification, and contribute to con-tent for the revision of the originalCommissioning and QualificationBaseline Guide (5.1)

Annual Meeting...Continued from page 3.

Knowledge Briefs to Offer Snapshots of Industry Issues,Processes, and Technologies

ISPE has established the Knowledge Brief publication pro-gram to provide access to general information on issues,

processes, and technologies impacting the contemporary phar-maceutical industry. Presented in levels of Basic, Intermedi-ate, and Advanced, a Knowledge Brief is intended to providean overview written in such a manner that the non-technicalreader can understand the context, substance, and relevanceof the subject.

Each Knowledge Brief will include links to technicaldocuments, Pharmaceutical Engineering articles, Communi-ties of Practice, seminars, and the like to provide readersaccess to more specific and detailed information on thesubject.

Knowledge Briefs are free to ISPE Members and $5 US/€ 3 to nonmembers and are available for immediate down-load from the ISPE Web site:

Biotechnology BasicsLevel: BasicAdapted from the ISPE Training Course on Biotech Basics,this Knowledge Brief provides basic concepts explaining thescience of biotechnology and how science and process arecombined to lead to the manufacture of a human therapeuticproduct.

Commissioning and Qualification ofBiopharmaceutical FacilitiesLevel: IntermediateThis Knowledge Brief summarizes the considerations in-volved in the commissioning and qualification of abiopharmaceutical manufacturing facility. The informationcontained in this Knowledge Brief was extracted from theISPE Baseline® Guide: Biopharmaceutical Facilities.

Quality by Designby John Berridge, PhDLevel: IntermediateThis Knowledge Brief provides and explains the basic ele-ments of Quality by Design (QbD).

For information on submitting an article to the KnowledgeBrief program, please contact Rochelle Runas, ISPE Techni-cal Writer, by email: [email protected]. Authorship recogni-tion will be provided to all individuals whose KnowledgeBrief is published. Publications emanating from COPs ortechnical documents will reflect authorship by those groups,rather than by individual. Prospective authors are encour-aged to review currently available Knowledge Briefs to betterunderstand the type of document being sought.


Classified Advertising


Architects, Engineers – Constructors

CRB Consulting Engineers, 7410 N.W.Tiffany Springs Pkwy., Suite 100, KansasCity, MO 64153. (816) 880-9800. See ourad in this issue.

EI Associates, 8 Ridgedale Ave., CedarKnolls, NJ 07927. (973) 775-7777. Seeour ad in this issue.

IPS – Integrated Project Services, 2001Joshua Rd., Lafayette Hill, PA 19444.(610) 828-4090. See our ad in this issue.

Parsons, 150 Federal St., Boston, MA02110. (617)-946-9400. See our ad inthis issue.

Bioreactors/Fermenters

Cleanroom Products/Services

AES Clean Technology, 422 Stump Rd.,Montgomeryville, PA 18936. (215) 393-6810. See our ad in this issue.

Dagard USA Corp., 1251 Avenue of theAmericas, 34th Floor, New York, NY10020. (212) 583-4928. See our ad in thisissue.

Consultants

Employment Search Firms

Jim Crumpley & Associates, 1200 E.Woodhurst Dr., Bldg. B-400, Springfield,MO 65804. (417) 882-7555. See our ad inthis issue.

Filtration Products

MKS Instruments, 5330 Sterling Dr.,Boulder, CO 80301. (800) 345-1967. Seeour ad in this issue.

Siemens Water Technologies, 10Technology Dr., Lowell, MA 01851. (978)934-9349. See our ad in this issue.

Instrumentation

Hach Ultra Analytics, 5600 LindberghDr., Loveland, CO 80539. (970) 663-1377. See our ad in this issue.

Label Removal Equipment

Hurst Corp., Box 737, Devon, PA 19333.(610) 687-2404. See our ad in this issue.

Passivation andContract Cleaning Services

Active Chemical Corp., 4520 Old LincolnHwy., Oakford, PA 19053. (215) 676-1111. See our ad in this issue.

Passivation andContract Cleaning Services

Astro Pak Corp., 270 E. Baker St., Suite100, Costa Mesa, CA 92626. (800) 743-5444. See our ad in this issue.

Cal-Chem Corp., 2102 Merced Ave., SouthEl Monte, CA 91733. (800) 444-6786.See our ad in this issue.

Spray Dryers

GEA Niro Pharma Systems, 9165Rumsey Rd., Columbia, MD 21045. Seeour ad in this issue.

Heinen Drying Inc., 1504 Grundy’s Ln.,Bristol, PA 19007. (215) 788-8196. Seeour ad in this issue.

Sterile Products Manufacturing

Tanks/Vessels

Eagle Stainless, 816 Nina Way,Warminster, PA 18974. (215) 957-9333.See our ad in this issue.

Lee Industries, PO Box 688, Philipsburg,PA 16866. (814) 342-0470. See our ad inthis issue.

Used Machinery

Validation Services

Commissioning Agents, Inc., 1515 N.Girls School Rd., Indianapolis, IN 46214.(317) 710-1530. See our ad in this issue.

ProPharma Group, 10975 Benson Dr.,Suite 330, Overland Park, KS 66210;5235 Westview Dr., Suite 100, Frederick,MD 21703. (888) 242-0559. See our ad inthis issue.

Valves

Gemu GmbH & Co., Fritz-Mueller-Str. 6-8, D-74653 Ingelfingen, Germany. +497940123-0. See our ad in this issue.

Water Treatment

Siemens Water Technologies, 10Technology Dr., Lowell, MA 01851. (978)934-9349. See our ad in this issue.

Veolia Water Solutions & Technologies,Global Headquarters, L’Aquarène – 1,place Montgolfier, 94417 Saint-MauriceCedex, France, www.pharma.veoliawaterst.com, Email: [email protected]. See our ad in this issue.

Reprinted from





Global Regulatory News


South AmericaArgentinaThe Argentinian National Medicines,Food and Medical Technology Admin-istration (ANMAT) has published newmandatory regulations (AnmatDisposición 2372/2008) for GMP in-spectors to follow when gatheringmanufacturing process information.These regulations aid in standardizingthe procedure for GMP investigators.There are two parts to the new regula-tion: the first covers inspections andthe second offers the classification ofGMP compliance deficiencies. The regu-lations cover: human resources, theplant and its general condition, watersystem, storage of raw materials, pack-aging materials, finished products, re-calls, production documentation, batchregistration procedure, and correctivemeasures for company deficiencies inthese areas.1

BrazilANVISA requires that companies in-volved in the manufacture, importa-tion, exportation, repackaging, stor-age, dispatch and distribution of APIssubmit the details of their products tothe regulatory authority so that thesedetails are recorded in the nationalAPI databank. In cases when APIs arenot directly obtained from the manu-facturer, the ANVISA asks companiesto submit detailed information aboutthe suppliers and vendors in order totrack for each specific API from theorigin to the distribution chain. It isthe responsibility of companies to keepthis required information up to date.Therefore, where companies fail to pro-vide detailed information about theirAPIs, ANVISA stated their APIs willno longer be marketed. Pharmaceuti-cal companies will only need to submitdetails of imported APIs (i.e., without adistributor). Companies only involvedin transporting APIs and compound-ing pharmacies are not meant to followthis new regulation.2

Asia/PacificAustraliaAs of 1 July 2008, the TherapeuticGoods Administration (TGA) requiresthat the manufacturer details and the

formulation of Over-The-Counter(OTC) medicines in Australia be madeavailable to sponsors. The regulatoryagency explained that not doing thiswill prevent sponsors from fulfillingtheir obligations under the Therapeu-tic Goods Act 1989. The manufacturerdetails and the formulation of OTCmedicines will be disclosed to sponsorsthrough the three TGA’s online ser-vices (the Australian Register of Thera-peutic Goods, the Electronic ListingFacility, and the OTC Medicines Elec-tronic Lodgement system). The TGAsays the purpose of this regulation is toensure the safety, quality, and efficacyof the OTC drugs, and to be able to fullyinvestigate their adverse events.3

IndiaIn an effort to improve the workingrelationships between the EuropeanDirectorate for the Quality of Medi-cines and Healthcare (EDQM), the In-dian Health Ministry and the IndianPharmacopoeia Commission (IPC),Susanne Keitel, a Director in EDQM,said they would be in favor of an appli-cation from India for observer status atthe European Pharmacopoeia Commis-sion. The Indian officials said the IPCalso would be keen on exchanging sci-entific expertise with their Europeancounterparts. In the future, collabora-tion between Europe and India espe-cially in joint inspections of manufac-turing sites and scientific cooperationis expected to increase.4

EuropeIn July 2008, the European Commis-sion DG Entreprise published a sum-mary of the public consultation on theprovisions for certification of qualityand non-clinical data for small andmedium-sized enterprises (SMEs), pur-suant to Article 18 of Regulation (EC)No 1394/2007. The summary outcomeof the public consultation paper is nowavailable.5

In addition, the Commission Regu-lation (EC) No 542/2008 of 16 June2008 amended Annexes I and II toCouncil Regulation (EEC) No. 2377/90laying down a Community procedure6

for the establishment of maximum resi-due limits of veterinary medicinal prod-

ucts in foodstuffs of animal origin, asregards cyfluthrin and lectin extractedfrom red kidney beans.

In the course of the public consulta-tion, the Commission has received 100stakeholder responses from industry(pharmaceutical industry, wholesalers,pharmacies, companies producing au-thenticity features, etc.), 12 from civilsociety (patient groups, consumer as-sociations, citizens), and three fromHealth Insurances and Health CareCenters. These responses are now avail-able.7

The European Commission pub-lished an article on their Web site thatsummarized the opinion on the combatof counterfeit medicines in Europe.Most of the agencies agree to take ur-gent and decisive actions, but the chal-lenge is to implement new stricter en-forcement regulations without leadingto a whole restructure of the alreadyexisting legal systems.8 The EuropeanFederation of Pharmaceutical Indus-tries and Associations said in its re-sponse that it supports the use of overtand covert authentication features (e.g.,holograms, color shifting inks,taggants), but the technology choiceshould be specific to each manufac-turer.9 In addition, it wants these mea-sures to be extended to all prescriptiondrugs rather than implemented on arisk-based basis. In order to effectivelycombat the medicines counterfeit, theEGA agrees with the commission’s pro-posals to make all participants in thedistribution chain comply with exist-ing pharmaceutical legislation (Direc-tive 2001/83/EC, as amended).

The EMEA has published a reporton the progress of the annual, “sam-pling and testing programs for cen-trally authorized medicines.” Samplingand testing programs run by EDQM,the national laboratories and medi-cines authorities of EU, and EEA Mem-ber States are used to detect qualityissues that need to be addressed, in theinterest of protecting public and ani-mal health. This report covering theperiod 1998-2007 shows the statisticaldata on medicines tested, the resultsobtained, the changes introduced tothe programs over the years, and em-phasizes involvement and cooperation

Reprinted from







of partner organizations across Eu-rope.1 0

A productive collaboration at theTransatlantic Economic Council (TEC)between the European Commission,the European Medicines Agency, andthe U.S. Food and Drug Administra-tion resulted with strengthening oftransatlantic cooperation on medicinesregulation.11 The Commission/EMEAand the FDA will pilot joint inspectionsof companies manufacturing pharma-ceuticals in the U.S. and in the EU andof companies manufacturing activepharmaceutical ingredients in thirdcountries; pilot the exchange of inspec-tion schedules results, and informa-tion on inspected manufacturing sitesin order to attain more GMP inspectioncoverage collectively and to better iden-tify manufacturing sites producing ac-tive pharmaceutical ingredients inthird countries. Moreover, a collabora-tion to determine to what extent dedi-cated production facilities are neces-sary for certain pharmaceuticals tak-ing into account a risk-based approachwill be set up. The EMEA and the FDAalso announced successes in their trans-atlantic work on biomarker develop-ment and validation for various prod-uct development purposes. Both par-ties will continue to work on this initia-tive with further biomarker develop-ment and validation.

The following meetings were heldduring the period covered by this up-date:

• The Committee for Orphan Medici-nal Products (COMP) held its nine-tieth plenary meeting on 13-14 May2008.12

• The Committee on Herbal Medici-nal Products (HMPC) held its 24thmeeting at the EMEA on 2-3 July2008.13

• The Committee for Orphan Medici-nal Products (COMP) held its ninety-second plenary meeting on 8-9 July2008.14

• The Paediatric Committee (PDCO)celebrated its first anniversary atits 2-4 July 2008 meeting.15

• The Committee for Medicinal Prod-ucts for Human Use (CHMP) heldits June plenary meeting from 23-26 June 2008.16

EMEA provides a Questions and An-swer on Active Substance, EuropeanPharmacopeia, Impurities, Manufac-ture of the medicinal products, Specifictypes of product, Stability, Water andData submission topics.17 EMEA pro-vides a Questions and Answer on theStability – Calculation of ExpirationDates topics.18

The following guidance documents arenoteworthy:

• Concept Paper on Developmentof a Guideline on Setting Speci-fications for Related Impuritiesin Antibiotics.EMEA has published on their Website the current thinking for “Set-ting Specifications for Related Im-purities in Antibiotics.”19

• Explanatory note for clarifica-tion of the scope of the VICHGuideline 17 on Stability Test-ing of Biotechnological/Biologi-cal Veterinary Medicinal Prod-uctsEMEA has published an explana-tory note clarifying the VICH Guide-line 17 relating to the “Stability Test-ing of Biotechnological/BiologicalVeterinary Medicinal Products.”20

• ICH Topic Q10 Note for Guid-ance on Pharmaceutical Qual-ity SystemThe International Conference onHarmonisation (ICH) Q10 docu-ment21 on Pharmaceutical QualitySystem was adopted (Step 4) at theICH Steering Committee meetingin June 2008. This describes a modelfor an effective pharmaceutical qual-ity system that is based on Interna-tional Standards Organisation (ISO)quality concepts, and includes ap-plicable Good Manufacturing Prac-tice (GMP) regulations.

• ICH Topic Q4B Annex 5EMEA released this note on the

evaluation and recommendation ofpharmacopoeial texts for use in theICH regions on “Disintegration TestGeneral Chapter.”22

• ICH Topic Q4B Annex 4CThis EMEA note is for evaluationand recommendation of pharmaco-poeial texts for use in the ICH re-gions on tests for microbiologicalexamination of non-sterile products:“Acceptance criteria for pharmaceu-tical preparations and substancesfor pharmaceutical use.”23

• ICH Topic Q4B Annex 4BThis note follows the Annex 4A andis related to the evaluation and rec-ommendation of pharmacopoeialtexts for use in the ICH regions ontest for “Microbiological examina-tion of non-sterile products: testsfor specified micro-organisms.”24

• ICH Topic Q4B Annex 4AThis note from the EMEA is for“Evaluation and recommendationof pharmacopoeial texts for use inthe ICH regions on microbiologicalexamination of non-sterile products:micrological enumeration tests.”25

• ICH Topic Q4B Annex 3This note from EMEA is for “Evalu-ation and recommendation of phar-macopoeial texts for use in the ICHregions on test for particulate con-tamination: sub-visible particlesgeneral chapter.”26

• ICH Topic Q4B Annex 2This note from EMEA is for “Evalu-ation and recommendation of phar-macopoeial texts for use in the ICHregions on test for extractable vol-ume of parenteral preparations gen-eral chapter.”27

• USThe new “Guidance for IndustrycGMP for Phase 1 InvestigationalDrugs” 28 has been published in July2008. This guidance mentions thecurrent thinking of the Food andDrug Administration (FDA) on theGood Manufacturing Practice in themanufacture of most Investigational




New Drugs (INDs) used in phase I ofclinical trials. This guidance doesnot establish enforceable responsi-bilities but should be viewed as rec-ommendations.

References1. Anmat Disposición 2372/2008,

Boletín Oficial de la República Ar-gentina, No 31.395, 30 April 2008referenced in RAJ Pharma May2008.

2. Anvisa Resolution RDC No 30,Diário Oficial da União, 16 May2008 referenced in RAJ PharmaJuly 2008.

3. TGA advisory, 20 June 2008, http://ebshelp.tga.gov.au/docs/Notice-Ingredient_manufacturer_confidentiality-Mar_08.rtf referenced in RAJPharma August 2008.

4. EDQM press release, 30 April 2008,www.edqm.eu/medias/fichiers/Meeting_EDQM_Indian_Authorities_30_April_2008.pdf Informa UK Ltd2008 referenced in RAJ PharmaJuly 2008.

5. http://ec.europa.eu/enterprise/pharmaceuticals/advtherapies/pub l i c _ consu l ta t i on_cer t i f /summary_outcome_2008-07-07.pdf.

6. http://ec.europa.eu/enterprise/p h a r m a c e u t i c a l s / e u d r a l e x /vol5_en.htm#mrl.

7. http://ec.europa.eu/enterprise/pharmaceuticals/pharmacos/docs/d o c 2 0 0 8 / 2 0 0 8 _ 0 6 /pharm63_summary-record_2007-12-10.pdf.

8. European Commission, Summaryof Responses to the Public Consul-tation Document, 19 June 2008,http://ec.europa.eu/enterprise/pharmaceuticals/counterf_par_trade/conterfeit_doc/2008_06_19_summary_of_responses.pdf.

9. EFPIA response, 9 May 2008, http://ec.europa.eu/enterprise/pharma-ceuticals/counterf_par_trade/d o c _ p u b l _ c o n s u l t _ 2 0 0 8 0 3 /114_efpia.pdf.

10. http://www.emea.europa.eu/In-spections/docs/38643408en.pdf.

11. http://www.fda.gov/oia/TEC.htm.

12. http://www.emea.europa.eu/pdfs/human/comp/26044108en.pdf.

13. http://www.emea.europa.eu/pdfs/human/hmpc/36935908en.pdf.

14. http://www.emea.europa.eu/pdfs/human/comp/36313208en.pdf.

15. http://www.emea.europa.eu/pdfs/human/pdco/35532908en.pdf.

16. http://www.emea.europa.eu/pdfs/human/press/pr/32726508en.pdf.

17. http://www.emea.europa.eu/In-spections/qwp/list.htm.

18. http://www.emea.europa.eu/In-spections/qwp/q21.htm.

19. http://www.emea.europa.eu/pdfs/human/ich/21473207en.pdf.

20. http://www.emea.europa.eu/pdfs/human/qwp/13635108en.pdf.

21. http://www.emea.europa.eu/pdfs/vet/iwp/31977108en.pdf.





26. http://www.emea.europa.eu/pdfs/human/ich/56117607enfin.pdf.

27. http://www.emea.europa.eu/pdfs/human/ich/55940907enfin.pdf.

28. Food and Drug Administration,Center for Drug Evaluation andResearch (CDER), July 2008,CGMP.

This information was provided by FrankSayala, MRPharmS, PhD, Pharmaceu-tical Research Associates (UK).

Documents

Functional Safety in the Life Science Industries