`
`VOLUME 2: FACIU7Y DESIGN, STERILIZATION AND PROCESSING
`
`and materials are disinfected in stages or encapsulated in previously sterilized gowns, gloves,
`hoods as attempted with personnel for entry. Their level of disinfection and aseptic condition
`should be equal or better than the environment in which they are used. From the core to the
`supporting rooms there is a cascading air overpressure for facilities engaged in filling nontoxic
`(to filling personnel) drug products. In conjunction with the cascading pressure differentials
`from room to room, there must be sufficient air changes per room to ensure nonviable and
`viable particulates are controlled to specified levels. The air is supplied through HEP A filters
`and the air flows, and exchange rates must be justifiable and aligned with regulatory
`requirements. Air flows from .higher graded or classified areas to lower graded areas, that is,
`class 100, class 10,000, class 100,000 depending on requirements (USFDA guidance classifi(cid:173)
`cation cited). Rooms themselves must be composed of smooth, nonshedding materials
`impervious to disinfectants or the exposure to ultraviolet radiation that may be used in the
`disinfection process.
`Using this foot print as a supporting backdrop it is then necessary to superimpose the
`process flow dynamics to the facility. Drug substance in and drug product out. Materials and
`clean and disinfected equipment in and "dirty" equipment and wastes out. If the
`manufacturing process requires support by people necessary personnel must enter and
`leave. This is to be accomplished without compromising clean with "dirty" and within the
`confines of the physical plant and established room classifications. lt represents the challenge
`to maintain the aseptic core and core supporting areas as contamination free. This is the
`complexity of aseptic processing that gives it the high risk status it is known for.
`
`Sanitization and Disinfection
`An integral part of any contamination control program is facility deaning and disinfection
`practices. The program must be su.fficiently robust to maintain the environment in a steady
`state of microbial control on the basis of criticality of operations and classification designation.
`The program must be designed to account for any routine variability to the day-to-day
`function and be flexible enough to respond to the nonroutine perturbations that may cl1allenge
`the program. Trending and interpretation of both viable and nonviable control data will
`provide the evidence to make "mid-course" corrections in a well designed and implemented
`program. Fortunately there are guides available and experience to help determine anticipated
`frequency and concentrations of cleaning and disinfecting solutions to attain validation goals
`and maintain control. Adjustments can be made on the basis of data. Selection of agents should
`be made on the basis of usual, "nonnal," flora and control of biotypes. Application of
`disinfectants is typically done on a manual basis with regiments and schemes available from
`industry benchmarks and vendor information.
`Much discussion has been devoted to the topic of rotation of dfainfectants and the "need"
`to control the adaptation of microbes to the mode of action of a disinfectant. It is now generally
`understood and agreed that the ability of microbes to adapt to the mode of action of a
`disinfectant and gain "resistance" is negligible or nonexistent. However, it is acknowledged
`that by not rotating disinfectants there can be a "selection" process for those microbes not
`susceptible to the mode of action of the chemical of use. It has been shown and is logical that
`naturally resistant microbes can persist and potentially increase in number by not rotating
`disinfectants. This has been demonstrated with in vivo data of endospore formers in the
`presence of quaternary ammonium compounds versus sporicidal agents to which they are
`susceptible.
`
`Environmental Monitoring System
`Evaluating the quality of air and surfaces in a clean room environment should start with a
`well-defined and written program. PDA TR #13 revised (2001) (12) as a starting point wiJJ
`provide a variety of information to help develop a program suitable to your facility and
`aligned with industry practices. Other reference documents such as ISO 14644, clean rooms,
`and associated controlled environments should be considered (13). Methods employed should
`be qualified/validated prior to implementation. The number of sampling sites will vary
`depending on the design of the area and the clean room technology employed (conventional
`fiJliJ1g line, RABS, or isolator). An aseptic core utilizing a conventional filling line would
`
`Regeneron Exhibit 1016.101
`
`
`
`THE FUNDAMENTALS OF AN ENVIRONMENTAL CONTROL PROGRAM
`
`87
`
`present the greatest control risk due to much more operator interaction and intervention. With
`employment of more restrictive technologies such as RABS or isolators, human contact would
`be reduced or virtually eliminated and hence the number of sampling sites could be reduced
`accordingly. The program should address all shifts and include all exposed surfaces of the
`room and equipment and include critical surfaces that come in contact with the product,
`containers, and stoppers. Sample timing, frequency, and location should be carefully selected
`on the basis of their relationship to the operations performed. Recent application of risk-based
`approaches to site selection and decision making has gained acceptance.
`It is useful to utilize current approaches emphasizing good science and risk-based
`approaches to contamination control. Application of Quality Risk Management (QRM)
`approaches as delimited in docmuents such as ICH Q9, a11d tools that support such approaches
`can be utilized effectively for snch purposes. Identifying contamination/risk points in your
`process stream and applying a QRM mindset will document the thought and support the
`process for establishing your program. Emphasis should be placed on contamination caused
`by operator interaction/intervention and at intervals in the process where there is the
`intersection of inputs and outputs into and from the core area. These transition points are most
`vulnerable to contamination introduction. These assessment points will be indkative of your
`operation, and there can be variability from line to line or between facilities. Each will exhibit
`its own "fingerprint" for usua I microbial flora, their numbers, and location of recoveries, and is
`influenced by disinfection program, facility layout, work flow, that is, the total control
`program. This fingerprint is developed from trended data. Although some variability of such
`data is normal the level of control should be sufficiently robust and sufficiently detailed to
`recognize trends indicating a perturbation in control and potential risk. Implementation of a
`total control program provides sufficient control to maintain operations within guidance or
`regulatory levels.
`
`Air Monitoring, Nonviable and Viable
`A11y comprehensive environmental conh·ol program should i11clude both nonviable and viable
`monjtoring. There has always been some discussion about the potential interrelatedJ1ess of the
`data generated from each activity; the position microorganisms need carriers such as nonviable
`particles for dispersal has won both skeptics and advocates. On that topic the evidence seems
`to support tl1e premise that you must have particles for microbial transfer but not all particles
`necessarily carry microbes. This premise has also been reflected in the argument supporting
`both 0.5 and 5.0 µm particle monitoring. 1 do not think any one would refute the basic premise
`of microbes "piggy backing" on particles but to say HEPA (high efficiency particulate filters)
`filtered air is a significant source of microbes is questionable. In any case it is generally
`recognized that some level of interrelatedness is apparent and warrants the pursuit of control
`with both approaches, nonviable and viable contributing to the total program. A great injustice
`would be done by looking at the data so prescriptively that tl1e value of the information is lost
`in the argument over the interpretation of the details. Examining the inside out approach to
`environmental control described earlier, core to lesser controlled areas, it is evident that
`monitoring data generated in areas of lesser control and consequently "closer" to an
`uncontrolled state are more difficult to interpret than data generated in more highly controlled
`areas, e.g., class 100 compared to class 100,000 since the frequency of testing tails off sharply as
`we move out from the class 100 area.
`
`Data Interpretation
`Environmental monitoring is for tl1e most part not an exact science but it represents our best
`attempt, given the tools currently available to us, to help characterize an environmental control
`program. Individual data represents a static point at which the data is co!Jected, a snapshot in
`time is the terminology sometimes used. The totality of tl1at data, all the data points collected,
`typifies the level of control during the operation on any given day at any given time. No one
`data point absolutely defines the conditions or level of control of a fill or of the total program. li
`the total control program is designed and implemented appropriately no single data point
`would indicate a process breakdown tmless there was a catastrophic failure. That scenario
`would be the best case situation verifying the program as implemented is working as intended.
`
`Regeneron Exhibit 1016.102
`
`
`
`88
`
`VOLUME 2: FACIU7Y DESIGN, STERJUZAT/ON AND PROCESSING
`
`Single events should typify an isolated incidence that could be assessed through investigation
`and risk assessment to determii,e impact. Since the level of control often varies on the basis of
`the type of technologies employed and operator interaction with that technology, some
`variability in data can be expected. Each data point and the collective array of data should be
`evaluated against the data history for the line, area, or process to determine consistency in
`control levels. Alert and action levels should be established and a mechanism to identify
`deviations and formulate responses should be clearly elaborated. Responses should be science(cid:173)
`based and appropriate to the potential risk presented by the devi.ation. Responses should be
`holistic in approach as well as mitigating the immediate risk posed by any event keeping in
`mind the lag time between the event and the results.
`Aseptic processing is based on the separation of the filling environment from potential
`contamination and any inputs to the process stream including operators. With the technologies
`available, and considering the conglomeration of life on this planet, this is essentially done in a
`somewhat arduous and cascading manner. The least areas of control are furthest from but
`connect to the greatest areas of control. The areas of greatest control are the class 100/class
`10,000 core area, using USFDA termmology, which are the cleanest and the most free of
`contamination and furthest from the natural uncontrolled world. Through that connection and
`by devising increasing levels of environmental control outside to inside, from natural
`uncontroUed areas to highly controlled core area, maintenance of a class 100 area can be
`accomplished. Much attention is given to the control of that class "IOO area and understandably
`so, but it is often overlooked or underestimated that inputs in your process stream including
`operators must traverse the lesser controlled areas to perform the end ftu1ctions of filling and
`stoppering of the product. What happens upstream will decidedly have an impact on
`operations downstream. By sequentially cascading areas, from nature outside the physical
`plant to unclassified area inside the plant to various levels of classification, we achieve what is
`regarded as a suitable level of control of the area to accomplish aseptic processing. It is not
`considered the best of situations by regulators for product manufacture, but when the nature
`of the drug product does not permit sterilization by other accepted technologies, aseptic
`processing is permitted for mamtfacture of sterile products.
`When something happens that impacts our control it should be detected. If it is not
`detected it may indicate we are not "measuring/monitoring" the appropriate indicators of
`change for the area. Since major control parameters can be more readily measured, it is usually
`the subtle events, those not directly measured or detectable, that impact control. That
`underscores the importance of looking for trends. Often our concentrated efforts are at the
`immediate point of focus and do not consider the control of areas and operations cascading
`from areas of lesser control, and as a result we struggle with appropriate corrective actions.
`Having said that, it is not reasonable that we monitor less critically controlled areas with the
`same rigor that we monitor class 100 areas but in an effort to get meaningful data from less
`controlled areas we must have greater assurance that what we do monitor from those areas
`adds value to the program. In many instances decisions of when and where and even possibly
`who we monitor in lesser controlled areas are not scientificaUy sound and not as well
`conceived as what we do for class 100 areas. Of course there is more prescriptive guidance on
`what to do in the class 100 area and frankly in some instances we have adopted such guidance
`to lesser controlled areas without thinking through our actions. Here is where the QRM
`process and a risk-based assessment can greatly enhance the value of our monitoring activity
`and make our control programs more robust. .In numerous instances, using a QRM approach to
`improve control in our lesser controlled areas has decreased costs through reduced monitoring
`and manpower expenditures and at the same time increased its value because of acquisition of
`more meaningful data.
`Jf data generated has greater capability to indicate perturbations in microbial control we
`should be able to respond in a more deliberate and effective manner. To do so this requires a
`more in depth review of data and a more frequent review of data. Not only is a response
`required at the point of deviation but one should also look at the potential impact upstream
`and downstream of that event. This expanded look and response could be proactive as well as
`reactive even if fttrther contamination is not detected. We tend to be too focused in our reviews
`and responses on specific events and as a result subtle irregularities were not seen until they
`
`Regeneron Exhibit 1016.103
`
`
`
`THE FUNDAMENTALS OF AN ENVIHONMENTAL CONTROL PROGRAM
`
`89
`
`became events and were widespread and problematic. This has been exacerbated due to the
`delay in getting data from monitoring to detect them they must grow. Reaction time is the
`cornerstone of control.
`With our current capabilities even for the most controlled environments it is broadly
`recognized that interpretation of data in the vast amount of cases can not be directly
`extrapolated to product risk. To deliver value in most instances the entire control picture must
`be formatted and analyzed to get an understanding of what if anything is going on
`mic.robiologically, knowledge versus information. This necessitates looking at data in other
`ways then the daily format associated with the collection of information. Trends, departure
`from routine or expected data, can best be recognized by grouping information into related
`categories and looking for cha nges. Multiple approaches to grouping information can be used
`to assist with change identification. Groupings focused by room, by classified area, and by
`microbial characteristics are but a few. Microbial characteristics st1ch as Gram reaction, Genus
`and species identification where and when appropriate, and numbers isolated in CFUs are
`common approaches. The transition between classified areas, airlocks, and pass-throt1ghs are
`excel.lent areas where potential issues can be identified as they ar-e developing. Changes in any
`one of these parameters may have significance in the overall microbiological maintenance/
`control of the facility. Frequency of review is paramott11t to identification of trends, events, or
`perturbations in the control program. Generally, the more data the better the capability of
`analysis but more is only useful ttnless it is relevant and if it is analyzed. Data coUection for
`data's sake is not productive. U you have information then you must do something with it and
`that it adds value to the program or it is not worth collecting. U it does not add value to your
`control program then do not pmsue it.
`Where or when does environmental monitoring not add value? There is a movement of
`late to monitor areas remote from the classified areas. The data generated may be taken at
`long/infreqttent intervals. Such data adds little knowledge to the maintenance of the control
`process and is likely to expe11d significant manpower and cost to generate the data. A well(cid:173)
`devised and implemented facility cleaning and control program as described earlier will
`contribute more to maintenance of facility control than environmental monitoring. If microbial
`monitoring is to be done in support of unclassified area control, the application of microbial
`acceptance criteria to cleaning and sanitization validation would be appropriate. Routine
`monitoring at long intervals is not scientifically sound and adds little value. The best approach
`is to incorporate a risk-based approach to decision making when devising a total program. Use
`of tools like HACCP has been shown to be effective in such applications.
`There are multiple reguJatory or guidance recommendations for periodicity of data
`collection. Recommended frequency of data collection is generally higher in more controlled
`areas and less in lesser controlled areas. However, since facility equipment and personnel
`control is basicaUy from the outside in, class 100,000 -+ 10,000-+ 100 or grade D > C > B > A,
`whatever scheme you choose to follow, it might be advisable to monitor or review data with
`greater frequency in the outer areas to give a more dynamic picture and information that can
`be the key to response and subsequent control. What appears outside by design generally
`works its way to the inside. Actually frequencies of data collection and review should be
`flexible based on the operations they are intended to control and the data recovered for
`analysis. To maintain control in a cascading system, it is most effective to put an emphasis on
`the upstream process to mitigate issues before they get to your critical areas. This is a position
`that is now being stressed in guidance information but is not always heeded by QA/QC. units.
`Regulatory requirements or gu idance recommendations should be regarded as mfoimums and
`adjusted to fit yom circumstances and total control program.
`
`How to Handle Excursions
`Like any deviation to requirements an investigation into the cause is expected. The extent of
`the investigation should be commensurate with the event and its proximity to the core A/8
`classified area or potential impact on the process. Investigations should have some consistency
`in approach or scope again based on criticality. Identified trends, alert level excursions, and
`action level excursions can be treated differently but must be defined by procedure. A rationale
`based on science and risk should support the position. Investigation should be systematic in
`
`Regeneron Exhibit 1016.104
`
`
`
`90
`
`VOLUME 2: FACIL/7Y DESIGN, STERJLIZAT/ON AND PROCESSING
`
`approach and of sufficient scope to determine potential impact both upstream and
`downstream of the event and between recently manufactured batches. All investigations
`having the potential for batch implication should be closed and reviewed by the quality unit
`prior to batch release.
`
`Process Simulation-Media Fills
`By definition media fills are designed to mimic the manufacturing process. They also represent
`an opporttrnity to collect additional data and knowledge about aseptic control mechanisms in
`your process. During media fills we are qualifying operators and stressing our processing
`environment by running our processes to their defined limits both ttpper and lower, both
`mechanically and by simulating in a condensed format the extremes of anticipated operating
`conditions. By doing added environmental and personnel monitoring during these exercises
`we can determine when and where risks are elevated due to the added interactions of
`personnel with the equipment. Not only do these activities test the robustness of your aseptic
`process, but the information obtained can give additional insight into where and when
`microbial risk arises with activities. From such data modification of location or timing of EM
`can be enhanced. Recall, however, that media fills are to represent the normal activities during
`the process a11d are not meant to justify practices that pose an unnecessary contamination risk.
`During routine aseptic manufacturing conditions you would want to do the minimum amount
`of monitoring since the activity in and of itself introduces a level of risk. Media fills give the
`latitude to explore that arena and apply leanings.
`
`REFERENCES
`1. Federal Standard Airborne Particulate Cleanliness Classes in Cleanrooms and Clean Zones, Fed. Std.
`No. 209E. ANSI/ ASQ American Society for Quality Control 611 East Wisconsin Avenue, Milwaukee,
`Wisconsin, 1992:53202.
`2. Clean Rooms and Work Stations for the Microbially Controlled Environment. National Aeronautics
`a.nd Space Administration, Washington, DC, 1967.
`3. FDA's Summary of Judge Wolin's Interpretation of GMP Issues Contained in the Court's Ruling in
`USA vs Oarr Laboratories. 1993. Available at: http://www.gmplst.com/barrsu.tn.hlm.
`4. Pharmaceutical cGMPs for the 21st Century: A Risk Based Approach. A science and risk based
`approach to product quality regulation incorporati.ng an integrated quality systems approach. 2002.
`A vai.lable at: http://www.fda.gov/ oc/ gujdance/ gm p. html
`5. International Conference on Harmonisation of Technical Requirements for Registration of
`Pharmaceuticals for Human Use. ICH Tripartite Guideline, ICH Q8, Q9, QlO. Geneva, Switzerland.
`6. Friedman RL, Mal1oney SC. Risk factors in aseptic processing. Am Pharm Rev 2003; 6(2):44 45,96.
`7. Hussong D. Environmental monitoring for aseptic processing. Am Pharm Rev 2003; 6(2):20 24.
`8. Guideline on Sterile Drug Products Produced by Aseptic Processing. FDA/CDE.R/Office of
`Compliance, 1987. (Reprinted in 1997).
`9. Guidance for Industry, Sterile Drug Products Produced for Aseptic Processing Current Good
`Manufacturing Practice. FDA/ CDER/ CBER/ Office of Regulatory Affairs/ Divisi.on of Drug
`Information, 2004.
`10. Eudralex, The Rules Governing Medicinal Products in the European Union. Volume 4. Annex 1.,
`Manufactu,ring of Sterile Medicinal Products. European Commission, 2008.
`11 . Akers J, Moore C. Microbiological Monitoring of Pharmaceutical Cleanrooms: The Need for
`Pragmatisn1. A2C2 Magazine, January 1998.
`12. Fundamentals of an Environmental Monitoring Pwgram. PDA J Pharrn Sci Technol. 55 Technical
`Report 13, 2001
`1:\. Microbiologic·al Evaluation of Clean Rooms and Other Controlled Environments < 1116> . United
`States Pharmacopeia USP 26 National Formulary 21, The United States Pharmacopeial Convention,
`RockvUle, MD, 2003.
`14. Cleanroom and Associated Controlled Environments, ISO 14644 Parts 1,2. International Orgal1ization
`for Standardization, Geneva, Switzerland . 1999, 2000.
`
`Regeneron Exhibit 1016.105
`
`
`
`INTRODUCTION
`The most challenging pharmaceutical water applications are typically associated with
`parenteral products since, by their nature, parenterals are uniquely able to access bodily
`flujds and tissues. In this aqueous environment, water is uniquely suited to serve in a
`multitude of roles associated with drug development, testing, manufacture, ru1d delivery.
`Because of their unparalleled access to critical areas of the body, and compottnded by their
`irretrievable nature, parenterals must meet extremely stringent requirements both in the United
`States and around the globe. Injectables necessitate the use of water that is chemically and
`microbiologically pure, to exacting standards, both from a practical perspective and based on
`regulatory dictates to avoid patient risk and to ensure product and treatment efficacy. As a
`product, excipient, cleaning agent, solvent, etc., water is used in significant quantities and, in
`many cases, it is the single largest volumetric commodity associated with any finished product.
`This chapter will discuss the uniqueness of parenteral water applications including their
`current regulatory requirements. Discussion will focus on injectable risks, compendia! limits
`for chemical purity, viable and nonviable microbial contamination, and added substances.
`Approved water treatment methods vary significantly throughout the world in spite of
`harmonization efforts and pose a significant hurdle for global firms wishing to reduce costs,
`consolidate manufacture, and standardize operations. As part of the discussion associated
`with approved methods of producing parenteral waters, common equipment types, basic
`~-ystem designs, operational challenges, and delivery / utiliz.ation issues will be reviewed. Of
`paramount concern is the ability to design, install, operate, and maintain a system that will
`consistently produce suitable quality water. Sanitization, testing, and monitoring are a few of
`the other key items that will also be addressed.
`
`WATER GRADES
`There is a considerable number of water grades used for pharmaceutical applications with
`varying regulatory requirements. The most significant of these are tabularized with their
`primary criteria in Table 1. These various grades are best characterized by their use, falling
`primarily into two groups, within which both bulk and packaged waters are defined. [t should
`be noted that these two primary groups are identified by their role relative to parenteral
`products, such that one is specifically designated for parenteral use while the other is
`designated not for parenteral use. Hence, it will be most practical to begin the discussion of
`water types and their application by reviewing the two primary types of bulk water, followed
`by discussion of each individual packaged grade including the requirements that make each
`unique and specialized.
`
`BULK PHARMACEUTICAL WATERS
`Bulk waters are those waters produced by pharmaceutical manufacturers for use in or during
`production of their products and ust1ally within their facility, while packaged waters are waters
`typically produced for incorporation into limited sized containers, most often one liter or less,
`and sold as a finished product for use in a multitude of applications ranging from
`extemporaneous compounding to laboratory testing. Packaged waters are most often sterilized
`to ensure that any residual microbial contamination does not multiply out of control, resulting
`in a compromised product or injured patient and owing to a general avoidance of preservatives.
`Sterility is not required for bulk water, including Water for Injection (WFJ); however,
`responsibility is placed on the drng manufacturer to ensure the safety and efficacy of their
`products. This reliance is mo11Hored by regulatory bodies tasked with protection of the public
`health. The lack of a sterility requirement for bulk water is not based on a lack of concern but
`instead on the recognition that WFl in bulk form will often require additional processing, and
`
`Regeneron Exhibit 1016.106
`
`
`
`Table 1 Water Grades with Monograph Designation of Applicable Standard (Water for Hemodialysis Has Been Omitted)
`
`Water for
`njection
`
`Water Highly
`Purifieda
`
`Bacteriostatic
`Water for
`njectionb
`
`Sterile water
`for nhalationb
`
`Sterile water
`for njectionh
`
`Sterile water
`for rrigationb
`
`Bacterial endotoxins < 0 25 EU/ml
`< 85 >
`Total organic carbon Meets the
`requirements
`< 643>
`Water conductivity
`Meets the
`requirements
`< 645>
`NA
`Packaging and
`storage
`
`Labeling
`
`NA
`
`Antimicrobial agents NA
`< 51 > & < 341 >
`
`Particulate matter
`< 788>
`pH < 791 >
`
`Sterility < 71 >
`
`NA
`
`NA
`
`NA
`
`Other requirements NA
`
`Microbial guidance
`< 1231>
`
`:::; 10 CFU/
`1()() ml
`
`< 025 U/mL
`
`< 0 5 EU/ml
`
`<0 5 EU/ml
`
`< 025 EU/ml
`
`S 025 EU/ml
`
`Meets 2 2 44
`
`Meets the
`requirements
`NA
`
`NA
`
`NA
`
`Clear and
`colorless
`NA
`
`NA
`
`Specific
`Requirements
`per EP
`:::; 10 CFU/
`100 ml
`
`NA
`
`NA
`
`Specific
`requirements
`per monograph
`Specific
`requirements
`per monograph
`Meets
`effectiveness
`and Content
`Meets the
`requirements
`Specific
`requirements
`per monograph
`Meets the
`requirements
`Specific
`requirements
`per monograph
`Meets the
`requirements
`
`NA
`
`Meets the
`requirements
`Specific
`requirements
`per monograph
`Specific
`requirements
`per monograph
`NA
`
`NA
`
`NA
`
`Meets the
`requirements
`Specific
`requirements
`per monograph
`Meets the
`requirements
`
`NA
`
`NA
`
`Specific
`requirements
`per monograph
`Specific
`requirements
`per monograph
`NA
`
`Meets the
`requirements
`Specific
`requirements
`per monograph
`Meets the
`requirements
`Specific
`requirements
`per monograph
`Meets the
`requirements
`
`NA
`
`Meets the
`requirements
`Specific
`requirements
`per monograph
`Specffic
`requirements
`per monograph
`NA
`
`NA
`
`NA
`
`Meets the
`requirements
`Specific
`requirements
`per monograph
`Meets the
`requirements
`
`~Column is only applicable to the European Pharmacopoeia 6 3 (USP designations are not applicable)
`bProducts produced from bulk Water for njection (WF )
`
`Purffied
`Water
`
`NA
`
`Meets the
`requirements
`Meets the
`requirements
`NA
`
`NA
`
`NA
`
`NA
`
`NA
`
`NA
`
`NA
`
`:$100 CFU/ml
`
`Sterile Purified
`Water
`
`NA
`
`NA
`
`Meets the
`requirements
`Specific
`requirements
`per monograph
`Specific
`requirements
`per monograph
`NA
`
`NA
`
`NA
`
`Meets the
`requirements
`Specffic
`Requirements
`per monograph
`Meets the
`requirements
`
`::tJ
`(1) cc
`(1)
`:::s
`(1) a :::s
`m
`>< :::r
`C"
`;:;:
`
`"' N
`
`;§
`r--
`~
`
`(::J
`
`!',:,
`~
`~ ....
`~
`0
`-~
`u,
`ni :x,
`r"'-
`~
`:::j
`c:,
`<:
`~ (::J
`1l c:,
`rn
`~
`
`c-,
`
`
`
`WATER SYSTEMS FOR PARENTERAL FACILITIES
`
`93
`
`sterility on such a large scale is difficult, if not impossible, to achieve and/or prove. This
`situation may be compotmded by testu1g, often in uncontrolled environments. Hence, it is
`frequently more practical to control final sterility using one or more alternative methods such
`as portion sized terminal sterilization or sterilization via filtration, all of which can be assured
`at a higher level of reliability than the bulk counterpart.
`Sterilization reliability is primarily based on statistical methods as absolute testing for
`sterility is effectively impossible with the technology cturently available. As a result, sterility
`can be achieved but cannot be practically measured on a direct quantitative basis without
`compromising the sterility itself or destruction of the product. This is complicated additionally
`by the nature of microorganisms as living entities and tl1eir lifecycle tendencies including their
`reproductive processes, their resilience, their potentially unique and individual response to
`stressed conditions, and their ability to colonize, just to name a few of the traits that improve
`their survivability and make sterilization quite challenging.
`Discussion of microorganisms and biofilm will occur furtller on in tllis chapter to provide
`a basic understanding of those challenges associated with biologic control of water used in,
`and for, pharmace1.ttical product manufacture, as weU as the testing required to ensure the
`suitability of water intended for use in parenteral products. Traditional cu.ltivative pour-plate
`methods that are typically employed result in lengthy delays before data is available, causing
`slower than optimal reaction to failures or anomalies and implementation of less than efficient
`quarantine practices to ensure safety and efficacy prior to product release. System configu(cid:173)
`ration can affect the ability to obtain valid cultivative data compotmding an already difficult
`situation.
`Microbiologists familiar with pharmaceutical water system design, operation, sanitiza(cid:173)
`tion, and, most importantly, the flora likely to exist in a