Cleaning Guidance

International Journal of Generic Drugs

http://www.locumusa.com International Journal 97 of Generic Drugs e-* [email protected] 0793 694X US/Canada ISSN 0793 7784 Euro ISSN 0793 7822 Pacific Rim

C L E A N I N G D E V E L O P M E N T

Cleaning Proceduresas the FDA See It

‘ … T h e s e g u i d e l i n e s a r e i n t e n d e d f o r B u l k P h a r m a c e u t i c a lC h e m i c a l s b u t a l m o s t e v e r y w o r d a p p l i e s t o D r u g P r o d u c t s …

BPC CLEANING (& Drug Products)

GUIDE TO INSPECTIONSVALIDATION OF CLEANINGPROCESSESI. INTRODUCTION

alidation of cleaning procedureshas generated considerablediscussion since agency

documents, including the InspectionGuide for Bulk PharmaceuticalChemicals and the BiotechnologyInspection Guide, have brieflyaddressed this issue.

CLEANINGPROCEDURES

must be validatedThese Agency documents clearlyestablish the expectation that cleaningprocedures (processes) be validated.This guide is designed to establishinspection consistency and uniformityby discussing practices that havebeen found acceptable (orunacceptable). Simultaneously, onemust recognize that for cleaningvalidation, as with validation of otherprocesses, there may be more thanone way to validate a process.

In the end, the test of any validationprocess is whether scientific datashows that the system consistentlydoes as expected and produces aresult that consistently meetspredetermined specifications.

This guide is intended to coverequipment cleaning for chemicalresidues only.

II. BACKGROUNDFor FDA to require that equipment beclean prior to use is nothing new, the1963 GMP Regulations (Part 133.4)stated as follows "Equipment… shallbe maintained in a clean and orderlymanner…." A very similar section onequipment cleaning (211.67) wasincluded in the 1978 cGMPregulations.

Of course, the main rationale forrequiring clean equipment is toprevent contamination or adulterationof drug products.Historically, FDA investigators havelooked for gross non-sanitation due toinadequate cleaning and maintenanceof equipment and/or poor dust controlsystems.

CLEANINGPROCEDURES

must consistentlymeet the limits set

Also, historically speaking, FDA wasmore concerned about thecontamination of non-penicillin drugproducts with penicillins or the cross-contamination of drug products withpotent steroids or hormones.A number of products have beenrecalled over the past decade due to

V



actual or potential penicillin cross-contamination.PAST HISTORY - 'The Resin Story'One event which increased FDAawareness of the potential for crosscontamination due to inadequateprocedures was the 1988 recall of afinished drug product, CholestyramineResin USP.

The bulk pharmaceutical chemicalused to produce the product hadbecome contaminated with low levelsof intermediates and degradants fromthe production of agriculturalpesticides.The cross-contamination in that caseis believed to have been due to thereuse of recovered solvents. Therecovered solvents had beencontaminated because of a lack ofcontrol over the reuse of solventdrums.

Solvent RecoveryStorage Drumswere used twice

without cleaning limitsDrums that had been used to storerecovered solvents from a pesticideproduction process were later used tostore recovered solvents used for theresin manufacturing process.

The firm did not have adequatecontrols over these solvent drums, didnot do adequate testing of drummedsolvents, and did not have validatedcleaning procedures for the drums.

Some shipments of this pesticidecontaminated bulk pharmaceuticalwere supplied to a second facility at adifferent location for finishing. Thisresulted in the contamination of thebags used in that facility's fluid beddryers with pesticide contamination.

This in turn led to cross contaminationof lots produced at that site, a sitewhere no pesticides were normallyproduced.

FDA instituted an import alert in 1992on a foreign bulk pharmaceuticalmanufacturer which manufacturedpotent steroid products as well asnon-steroidal products using commonequipment.

This firm was a multi-use bulkpharmaceutical facility. FDAconsidered the potential for cross-contamination to be significant and topose a serious health risk to thepublic.

The firm had only recently started acleaning validation program at thetime of the inspection and it wasconsidered inadequate by FDA.

One of the reasons it was consideredinadequate was that the firm was onlylooking for evidence of the absence ofthe previous compound. The firm hadevidence, from TLC tests on the rinsewater, of the presence of residues ofreaction byproducts and degradantsfrom the previous process.

III. GENERAL REQUIREMENTSFDA expects firms to have writtenprocedures (SOP's) detailing thecleaning processes used for variouspieces of equipment.If firms have one cleaning process forcleaning between different batches ofthe same product and use a differentprocess for cleaning between productchanges, we expect the writtenprocedures to address these differentscenario.

Minor CleaningSame Product

Major CleaningDifferent Product

Similarly, if firms have one processfor removing water soluble residuesand another process for non-watersolubleresidues, the written procedure should



address both scenarios and make itclear when a given procedure is to befollowed.

Cleaning SOPWater Soluble Residues

Cleaning SOPNon-Soluble Residues

Bulk pharmaceutical firms may decideto dedicate certain equipment forcertain chemical manufacturingprocess steps that produce tarry orgummy residues that are difficult toremove from the equipment.

Fluid bed dryer bags are anotherexample of equipment that is difficultto clean and is often dedicated to aspecific product. Any residues fromthe cleaning process itself(detergents, solvents, etc.) also haveto be removed from the equipment.

Cleaning LimitsFor 'Active Residues'

& LimitsFor 'Cleaning Agents'

What Do The FDA Expect FDA expects firms to have writtengeneral procedures on how cleaningprocesses will be validated. FDA expects the general validationprocedures to address who isresponsible for performing andapproving the validation study, theacceptance criteria, and whenrevalidation will be required. FDA expects firms to preparespecific written validation protocols inadvance for the studies to beperformed on each manufacturingsystem or piece of equipment whichshould address such issues assampling procedures, and analyticalmethods to be used including thesensitivity of those methods.

FDA expects firms to conduct thevalidation studies in accordance withthe protocols and to document theresults of studies. FDA expects a final validationreport which is approved bymanagement and which stateswhether or not the cleaning process isvalid.

The data should support a conclusionthat residues have been reduced to an"acceptable level."IV. EVALUATION OF CLEANINGVALIDATIONThe first step is to focus on theobjective of the validation process,and we have seen that somecompanies have failed to developsuch objectives.It is not unusual to see manufacturersuse extensive sampling and testingprograms following the cleaningprocess without ever really evaluatingthe effectiveness of the steps used toclean the equipment.

Several questions need to beaddressed when evaluating thecleaning process. For example:Q. At what point does a piece ofequipment or system become clean?Q. Does it have to be scrubbed byhand?Q. What is accomplished by handscrubbing rather than just a solventwash?Q. How variable are manual cleaningprocesses from batch to batch andproduct to product?

The answers to these questions areobviously important to the inspectionand evaluation of the cleaning processsince one must determine the overalleffectiveness of the process.

Answers to these questions may alsoidentify steps that can be eliminatedfor more effective measures and resultin resource savings for the company.



Determine the number of cleaningprocesses for each piece ofequipment.Ideally, a piece of equipment orsystem will have one process forcleaning, however this will depend onthe products being produced andwhether the cleanup occurs betweenbatches of the same product (as in alarge campaign) or between batchesof different products.

Minor Cleaningof Same ProductDoes not Require

Validation or LimitsWhen the cleaning process is usedonly between batches of the sameproduct (or different lots of the sameintermediate in a bulk process) thefirm need only meet a criteria of,"visibly clean" for the equipment.

Such between batch cleaningprocesses do not require validation('Minor Clean').1. Equipment DesignExamine the design of equipment,particularly in those large systems thatmay employ semi-automatic or fullyautomatic clean-in-place (CIP)systems since they representsignificant concern.

For example, sanitary type pipingwithout ball valves should be used.When such non-sanitary ball valvesare used, as is common in the bulkdrug industry, the cleaning process ismore difficult.

When such systems are identified, itis important that operators performingcleaning operations be aware ofproblems and have special training incleaning these systems and valves.

Determine whether the cleaningoperators have knowledge of thesesystems and the level of training andexperience in cleaning these systems.

Also check the written and validatedcleaning process to determine if thesesystems have been properly identifiedand validated.

LIST ALLComplex Cleaning

andValidate

In larger systems, such as thoseemploying long transfer lines orpiping, check the flow charts andpiping diagrams for the identificationof valves and written cleaningprocedures.

Piping and valves should be taggedand easily identifiable by the operatorperforming the cleaning function.Sometimes, inadequately identifiedvalves, both on prints and physically,have led to incorrect cleaningpractices.

LIST maximumTime Limits Allowable

between Process Endand Start of Cleaning

Always check for the presence of anoften critical element in thedocumentation of the cleaningprocesses; identifying and controllingthe length of time between the end ofprocessing and each cleaning step.

This is especially important fortopicals, suspensions, and bulkdrug operations.

In such operations, the drying ofresidues will directly affect theefficiency of a cleaning process.

Whether or not CIP systems are usedfor cleaning of processing equipment,microbiological aspects of equipmentcleaning should be considered.

This consists largely of preventivemeasures rather than removal of



contamination once it has occurred.

There should be some evidence thatroutine cleaning and storage ofequipment does not allow microbialproliferation.For example, equipment should bedried before storage, and under nocircumstances should stagnant waterbe allowed to remain in equipmentsubsequent to cleaning operations.

Protect'Cleaned Equipment'

from subsequentmicrobial contamination

Subsequent to the cleaning process,equipment may be subjected tosterilization or sanitization procedureswhere such equipment is used forsterile processing, or for non-sterileprocessing where the products maysupport microbial growth.

While such sterilization or sanitizationprocedures are beyond the scope ofthis guide, it is important to note thatcontrol of the bioburden throughadequate cleaning and storage ofequipment is important to ensure thatsubsequent sterilization or sanitizationprocedures achieve the necessaryassurance of sterility.This is also particularly important fromthe standpoint of the control ofpyrogens in sterile processing sinceequipment sterilization processes maynot be adequate to achieve significantinactivation or removal of pyrogens.2. Cleaning Process WrittenProcedure and DocumentationExamine the detail and specificity ofthe procedure for the (cleaning)process being validated, and theamount of documentation required.

We have seen general SOPs, whileothers use a batch record or log sheetsystem that requires some type ofspecific documentation for performingeach step.

Depending upon the complexity of thesystem and cleaning process and theability and training of operators, theamount of documentation necessaryfor executing various cleaning steps orprocedures will vary.

When more complex cleaningprocedures are required, it isimportant to document the criticalcleaning steps (for example certainbulk drug synthesis processes).

In this regard, specific documentationon the equipment itself which includesinformation about who cleaned it andwhen is valuable.However, for relatively simple cleaningoperations, the mere documentationthat the overall cleaning process wasperformed might be sufficient.

Other factors such as history ofcleaning, residue levels found aftercleaning, and variability of test resultsmay also dictate the amount ofdocumentation required.

'Clean Equipment'to a specificresidue limit

For example, when variable residuelevels are detected following cleaning,particularly for a process that isbelieved to be acceptable, one mustestablish the effectiveness of theprocess and operator performance.

Appropriate evaluations must bemade and when operator performanceis deemed a problem, more extensivedocumentation (guidance) and trainingmay be required.3. Analytical MethodsDetermine the specificity andsensitivity of the analytical methodused to detect residuals orcontaminants. With advances inanalytical technology, residues fromthe manufacturing and cleaningprocesses can be detected at very lowlevels.



If levels of contamination or residualare not detected, it does not mean thatthere is no residual contaminantpresent after cleaning. It only meansthat levels of contaminant greater thanthe sensitivity or detection limit of theanalytical method are not present inthe sample.

Firms should challenge the analyticalmethod in combination with thesampling method(s) used to show thatcontaminants can be recovered fromthe equipment surface and at whatlevel, i.e. 50% recovery, 90%, etc.

This is necessary before anyconclusions can be made based onthe sample results. A negative testmay also be the result of poorsampling technique (see below).4. SamplingThere are two general types ofsampling that have been foundacceptable. The most desirable is thedirect method of sampling the surfaceof the equipment. Another method isthe use of rinse solutions.

Surface Cleaningcan be checked bySurface Swabs

or

Rinse Samplesa. Direct Surface Sampling - Determinethe type of sampling material usedand its impact on the test data sincethe sampling material may interferewith the test. For example, theadhesive used in swabs has beenfound to interfere with the analysis ofsamples.Therefore, early in the validationprogram, it is important to assure thatthe sampling medium and solvent(used for extraction from the medium)are satisfactory and can be readilyused.

Advantages of direct sampling arethat areas hardest to clean and whichare reasonably accessible can beevaluated, leading to establishing alevel of contamination or residue pergiven surface area.

Additionally, residues that are "driedout" or are insoluble can be sampledby physical removal.b. RINSE SAMPLES

Two advantages of using rinsesamples are that a larger surface areamay be sampled, and inaccessiblesystems or ones that cannot beroutinely disassembled can besampled and evaluated.A disadvantage of rinse samples isthat the residue or contaminant maynot be soluble or may be physicallyoccluded in the equipment.An analogy that can be used is the"dirty pot." In the evaluation ofcleaning of a dirty pot, particularly withdried out residue, one does not look atthe rinse water to see that it is clean;one looks at the pot.

Check the Babynot the

Bath WaterCheck to see that a directmeasurement of the residue orcontaminant has been made for therinse water when it is used to validatethe cleaning process.For example, it is not acceptable tosimply test rinse water for waterquality (does it meet the compendiatests) rather than test it for potentialcontaminates.c. Routine Production In-ProcessControlMonitoring - Indirect testing, such asconductivity testing, may be of somevalue for routine monitoring once acleaning process has been validated.



This would be particularly true for thebulk drug substance manufacturerwhere reactors and centrifuges andpiping between such large equipmentcan be sampled only using rinsesolution samples.Any indirect test method must havebeen shown to correlate with thecondition of the equipment.During validation, the firm shoulddocument that testing the uncleanedequipment gives a not acceptableresult for the indirect test.V. ESTABLISHMENT OF LIMITSFDA does not intend to setacceptance specifications or methodsfor determining whether a cleaningprocess is validated. It is impracticalfor FDA to do so due to the widevariation in equipment and productsused throughout the bulk and finisheddosage form industries.

The firm's rationale for the residuelimits established should be logicalbased on the manufacturer'sknowledge of the materials involvedand be practical, achievable, andverifiable.

It is important to define the sensitivityof the analytical methods in order toset reasonable limits.

Some limits that have beenmentioned by industry representativesin the literature or in presentationsinclude analytical detection levelssuch as 10 PPM, biological activitylevels such as 1/1000 of the normaltherapeutic dose, and organolepticlevels such as no visible residue.

Check the manner in which limits areestablished. Unlike finishedpharmaceuticals where the chemicalidentity of residuals are known (i.e.,from actives, inactives, detergents)bulk processes may have partialreactants and unwanted by-productswhich may never have beenchemically identified.

In establishing residual limits, it maynot be adequate to focus only on theprincipal reactant since other chemicalvariations may be more difficult toremove.

There are circumstances where TLCscreening, in addition to chemicalanalyses, may be needed. In a bulkprocess, particularly for very potentchemicals such as some steroids, theissue of by-products needs to beconsidered if equipment is notdedicated.

The objective of the inspection is toensure that the basis for any limits isscientifically justifiable.

VI. OTHER ISSUESa. Placebo ProductIn order to evaluate and validatecleaning processes somemanufacturers have processed aplacebo batch in the equipment underessentially the same operatingparameters used for processingproduct.A sample of the placebo batch is thentested for residual contamination.

Testing aPlacebo Batch

may only beSupportive Data

However, we have documentedseveral significant issues that need tobe addressed when using placeboproduct to validate cleaningprocesses.One cannot assure that thecontaminate will be uniformlydistributed throughout the system.

For example, if the discharge valve orchute of a blender are contaminated,the contaminant would probably notbe uniformly dispersed in the placebo;it would most likely be concentrated inthe initial discharge portion of thebatch.



Additionally, if the contaminant orresidue is of a larger particle size, itmay not be uniformly dispersed in theplacebo.

Some firms have made theassumption that a residualcontaminant would be worn off theequipment surface uniformly; this isalso an invalid conclusion.Finally, the analytical power may begreatly reduced by dilution of thecontaminate. Because of suchproblems, rinse and/or swab samplesshould be used in conjunction with theplacebo method.

b. DetergentIf a detergent or soap is used forcleaning, determine and consider thedifficulty that may arise whenattempting to test for residues.

A common problem associated withdetergent use is its composition. Manydetergent suppliers will not providespecific composition, which makes itdifficult for the user to evaluateresidues.

As with product residues, it isimportant and it is expected that themanufacturer evaluate the efficiencyof the cleaning process for theremoval of residues.

However, unlike product residues, it isexpected that no (or for ultra sensitiveanalytical test methods - very low)detergent levels remain after cleaning.

Detergents are not part of themanufacturing process and are onlyadded to facilitate cleaning during thecleaning process.

c. Test Until CleanExamine and evaluate the level oftesting and the retest results sincetesting until clean is a concept utilizedby some manufacturers.

They test, resample, and retestequipment or systems until an"acceptable" residue level is attained.

For the system or equipment with avalidated cleaning process, thispractice of resampling should not beutilized and is acceptable only in rarecases.

Constant retesting and resamplingcan show that the cleaning process isnot validated since these retestsactually document the presence ofunacceptable residue andcontaminants from an ineffectivecleaning process.

Repeat Testing to'Clean Compliance'

is NOT the

Route to Go

REFERENCES

1) J. Rodehamel, "Cleaning and Maintenance,"Pgs 82-87, University of Wisconsin's ControlProcedures in Drug Production Seminar, July 17-22, 1966, William Blockstein, Editor, Published bythe University of Wisconsin, L.O.C.#66-64234.

2) J.A. Constance, "Why Some Dust ControlExhaust Systems Don't Work," Pharm. Eng.,January-February, 24-26 (1983).

3) S.W. Harder, "The Validation of CleaningProcedures," Pharm. Technol. 8 (5), 29-34 (1984)

4) W.J. Mead, "Maintenance: Its Interrelationshipwith Drug Quality," Pharm. Eng. 7(3), 29-33(1987).

5) J.A. Smith, "A Modified Swabbing Technique forValidation of Detergent Residues in Clean-in-PlaceSystems," Pharm. Technol. 16(1), 60-66 (1992).

6) Fourman, G.L. and Mullen, M.V., "DeterminingCleaning Validation Acceptance Limits forPharmaceutical Manufacturing Operations,"Pharm. Technol. 17(4), 54-60 (1993).

7) McCormick, P.Y. and Cullen, L.F., inPharmaceutical Process Validation, 2nd Ed.,edited by I.R. Berry and R.A. Nash, 319-349(1993)

Note:This document is reference material for investigators and other FDA personnel. The document does not bind

FDA, and does no confer any rights, privileges, benefits, or immunities for or on any person(s).

Documents

Cleaning Guidance