You are on page 1of 6

B

I O

PR

O C E S S TECHNICAL

The Basics of Cleaning and Cleaning Validation


Jean-Michel Cardot and E. Beyssac
uring the past three years, cleaning or cleaningvalidation problems ranked in the top four of the 483s and warning letters issued by the FDA (1, 2), reflecting the agencys emphasis on cleaning. The main points mentioned are a lack of documented procedures, inadequate training of workers, and insufficient validation of analytical and/or cleaning methods. Manufacturers are concerned about the definition of clean. A clean product, surface, apparatus, fluid, or gas is one with an acceptable and predefined level of contamination (3). Contamination is defined as the presence of contaminant, and contaminant means a solid, liquid, biological or gaseous matter, a microorganism, or any combination of these likely to alter in any way the health or safety of workers, patients, product, or process. A contaminant is not always a dangerous substance, which is defined as any substance which, by reason of its characteristics, constitutes PRODUCT FOCUS: BIOPHARMACEUTICALS PROCESS FOCUS: BOTH UPSTREAM (PRODUCTION) AND DOWNSTREAM (PROCESSING) WHO SHOULD READ: MANUFACTURING, PROCESS DEVELOPMENT, QA/QC, AND
REGULATORY PERSONNEL

a danger to the health, safety or physical well-being of a worker. Those definitions underline two main points. First, absolute cleanliness does not exist; the definition above proposes an acceptable and predefined level, emphasizing the need to define acceptable limits. Second, contaminants are not always linked with an active ingredient but can come from other sources, including product residues, degradation products, preservatives, chemical intermediates, excipients, cleaning agents, microbiological products, process agents (fluids and compounds such as oil, unclean steam), and process equipment (metal, glass, or fibers generated during production). External contaminants such as pollens and pesticides are more difficult to monitor because they are generated outside the company and usually unexpected.

WWW.PHOTOS.COM

A MULTIPARAMETRIC PROCESS

KEYWORDS: CONTAMINATION, CLEANING, CIP, VALIDATION, TOC, OPLC LEVEL: BASIC


66 BioProcess International J UNE 2007

Contamination is a multiparametric process, as presented in Figure 1 (4). Three main parameters should be taken into account: the source of contamination (e.g., a previous API in the case of cross contamination); the vector (e.g., an unclean apparatus); and the receptor (the product, component, or apparatus that gets contaminated). Numerous solutions exist to break the links between these factors, such as good plant design, specific and well-designed equipment, and monitored flow of equipment and workers. But one of the most important elements is cleaning, itself.

All equipment zones should be cleaned efficiently, including difficultto-access areas that can influence overall production quality. Such critical points are defined as zones or spots in systems on which an absence of control will bring an unacceptable risk for security (3). Examples include the bottom of a tank, O-rings, U-bends, ceilings, and traps. Cleaning and cleanliness are linked to four categories of items that should be considered before implementing any cleaning strategy (Figure 2): products, equipment, production, and economics. Products: What are the product characteristics, their toxicities, their solubility properties, and the degradation products that might appear during cleaning? What is the risk of exposure for coworkers? What

ecological risks may be incurred (disposable of solvents, for example)? Equipment: Can a clean-in-place process be installed? Is equipment easy to dismantle? Is the nature of the material compatible with cleaning agents and solvents? Can all the parts of the equipment be accessed? Production: Does the process use specific apparatus? What are the planned uses of each apparatus and/or room? What is the duration of the cleaning process and cleaning control? Can (or must) production be stopped during the cleaning-control period? Economics: What are the cost comparisons between using dedicated or disposable equipment? What are the cleaning costs (manpower, energy, stopping of production)? Were all cleaning costs included in the price estimation? Once those four topics have been addressed, the choice of cleaning procedure or strategy (dedicated, for example) can be made and justified.

CHOOSING A CLEANING METHOD

The main factors of a cleaning processes are the cleaning agents (including solvent types), mechanical actions (including the type of cleaning clean-in-place, clean-out-of-place, or manual), duration of contact, temperature of the cleaning solvent, and the time that elapses between the end of production and cleaning. The choice of a cleaning agent depends on the nature of the apparatus (particularly its material of construction), the washing solvent used, and the substances to be removed. The consideration will be the physicochemical characteristics of the cleaning products, and there are three main classes to choose from, depending on the nature of the substances to be removed. Basic cleaning agents work better for organic compounds, acidic agents for mineral compounds, and neutral agents act as surfactants. A mixture of methods or sequential cleaning with different types of agents may be appropriate. Additives can also be added to cleaning agents, among them antideposit agents, sequestrates, complexing agents, antifoam agents,
68 BioProcess International J UNE 2007

corrosion inhibitors, oxidants, and chlorides (although their use on stainless steel should be strictly monitored). The best cleaning agent would be water soluble, would act at low concentrations, could be removed by rinsing without residuals, could be traced in samples, would not harm materials, would be fully documented (including composition and a safety data sheet), and would be GMP compliant. Usually the first cleaning agent to be tested is a neutral agent; then, if that fails, other types of agents are tested. One main goal is to know the full composition of the product from its supplier and have a guaranty of stable composition over time (5). Cleaning agents should also be considered as contaminants themselves, which can be neutralized and removed by rinsing. Absence of cleaning-agent traces after rinsing should be validated (using fixed trace limits) and controlled after each cleaning procedure (59). Procedures: The four main types of cleaning procedures fall into two main categories, automated and manual cleaning. Clean-in-place (CIP) cleaning and rinsing runs are monitored by an automated system and follow a repeatable cycle. As a term, CIP refers often to removing accessible contaminants, so it often could more accurately be described as wash-in-place (WIP). Clean-out-ofplace (COP) requires dismantling an apparatus, washing it in a central washing area using an automated system, and checking it at reassembly. Semiautomatic methods use both

Figure 1: Cleaning as a tool for quality (adapted from Reference 3)

manual and automatic processes (e.g., combining spray balls and manual cleaning). Pure manual cleaning is a second option. The mechanical action of automated cleaning is often represented by spray balls (stationary, rotary, rotary jet) and manual cleaning by sponge, brushes, single-use cloths, and so on. In addition to manual cleaning and COP, ultrasonic batch can be used for small pieces of equipment. Cleaning efficacy is achieved through a succession of steps and techniques (an example is shown in Figure 3). This implies knowing all the production steps; all the equipment used and where it is used; and the flow of consumables, products, materials, workers, and so on for the cleaning strategy. All equipment, rooms, and flows must be defined and introduced to the cleaning process associated with its contamination risk. Water and solvent quality is a main factor to consider. Tap water cannot be

Figure 2: Points to take into consideration (adapted from Reference 3)

Figure 3: Steps used during the cleaning of an equipment

PERFECT

A cleaning method uses water as a solvent with minimal concentration of the cleaning agent at an ambient temperature over the minimum period of time.

to a cleaning agent can corrode and damage the surface of equipment. That can place later batches at risk and create new critical points for later cleaning procedures. Often the cleaning process is developed at the laboratory stage using coupons (intact and damaged) of various materials. Based on overloaded contaminants and overdried coupons, the best cleaning agent or combination of agents are selected. Choices are made regarding the concentration of cleaning agent, type of solvent (water is preferred), parameters (time and temperature), and required mechanical actions (spray balls, high-pressure water, ultrasonic bath, and so on). A perfect cleaning method is one that uses water as a solvent, with a minimal concentration of the agent itself at an ambient temperature over the minimum amount of time.
Sampling: The first step in cleaningprocess control is recovery of a sample to be analyzed. As defined in a Canadian health guidance (5), the sampling methods used should be capable of quantitatively measuring levels of residues remaining on the equipment surfaces after cleaning. The sampling technique used should detect both insoluble and soluble residues. Many processes combine two main techniques (swabbing and rinse-water analysis), although some use only one. Table 1 sums up the main advantages and drawbacks of each method. Choice of swab materials and solvents for swabbing are important. Usually, swabs are made of polyester, which presents good solvent resistance and minimally interferes with analytical methods. Another advantage of such swabs over cotton is that they leave no particles behind on the surface. Solvents should exhibit strong ability to dissolve and remove compound traces without damaging equipment. In addition, they must not harm personnel or the environment and leave no traces. A swabbing procedure covers at minimum the surface to be swabbed, the volume of solvent, the process of swabbing itself, transport of

CONTROL OF CLEANING

used for cleaning and rinsing because its quality varies from one site to another. Water hardness can vary, for example, and large quantities of chloride, endotoxins, or microorganisms may be present, as well as trace amounts of total organic carbon. The quality of final rinse determines the cleanliness of equipment, so purified water or waterfor-injection (WFI) are used. Temperature and duration of a cleaning agents presence are also important factors. The maximum temperature that could be used is the boiling point of the solvent, and the time required depends on the tests performed. Neither the highest temperature nor longer duration is always the best choice. A high temperature can coagulate proteins, making them stickier and more difficult to remove, and overexposure
70 BioProcess International J UNE 2007

the swabs, and the numbering of samples. A rinsing procedure specifies at least the type, volume, and temperature of the solvent; the maximum time elapsing between cleaning and rinsing (to avoid overdrying contaminants); and the contact time as well as sample volume, time of sampling, transport of samples, and the maximum duration between sampling and analysis. One of the first tests performed on a sample is a recovery test, which can be performed on coupons. A main goal is to have consistent recovery (a percentage of recovered/theoretical quantities) over a range of concentrations (low, medium, high). A recovery of 95% is considered good. If recovery is poor, the reason for it must be found and a correction factor put in place. Calculating Limits: Trace amounts of contaminants can be calculated using various techniques. Acceptance criteria for residues should be defined and justified and are a function of the nature of product, the toxic and/or therapeutic dose of the current product, the concentration in the next product (depending on the size of the next batch size and the therapeutic dose), the type of equipment (dedicated or not), and contamination risks. The only clear rules presented in guidelines are those of the Canadian Guidelines (5) similar to PIC/S (10).
Carry-over of product residues should meet defined criteria, for example the most stringent of the following criteria (i, ii, iii):

(i) NMT (Not More Than) 0.1% of the normal therapeutic dose of any product to appear in the maximum daily dose of the following product; (ii) NMT 10 ppm of any product to appear in another product; (iii) No quantity of residue to be visible on the equipment after cleaning procedures are performed. Spiking studies should determine the concentration at which most active ingredients are visible. (iv) For certain allergenic ingredients, penicillins, cephalosporins or potent steroids

and cytotoxics, the limits should be below the limit of detection by best available analytical methods. In practice this may mean that dedicated plants are used for these products. (10)

Table 1: Main advantages and drawbacks of sampling methods Method Swabbing Advantage Level of contamination or residue per given surface area Soluble or insoluble contaminants sampled Easy to use with specific analytical methods Greater stability of samples Large surface area Sample otherwise inaccessible systems or parts Easy and on-line possible (e.g., TOC) Drawbacks Not all surfaces are easily accessible because of equipment design and/or process limitations Swab adsorption of contaminants can create recovery problems A manual approach (depends on operators) System is opened for swabbing (sometimes needs recleaning afterward) Insoluble residues or contaminants are not recovered Occluded contaminants in equipment are not recovered Conservation of samples is not good Large dilution

The third of the above guidelines calls attention to the type of traces: whether colored or not, visible in normal lightning or only under ultraviolet light, the distance (<2 m) and angle (normal or oblique light) at which they are detectable, the material on which they are found (type of surface, through a glass or not). The first and second guidelines depend on batch size. All equipment in contact with a product should be identified and its the total surface calculated. The calculation of the carryover quantity is based on Equation 1. The first and second rules above pose no equivalent challenges. Preventing carryover of product traces into later batches sometimes requires more stringent cleaning procedures. If the risk of contamination is not evenly distributed between apparatuses, then limits should be calculated independently for each part of an apparatus that contacts the product stream. In addition to incorporating swab results, the final value should also factor in recovery rates. Those limits are calculated for all products in contact with the equipment, factoring in the cleaning agents and the total surface area. This underlines another problem: Calculation of the exact surface area of equipment that contacts the product, especially the interiors of all the pipes. The same guidance note stipulates that For certain allergenic ingredients, penicillins, cephalosporins or potent steroids and cytotoxics, the limits should be below the limit of detection by best available analytical methods. In practice this may mean that dedicated plants are used for these products. That consideration implies, with the current evolution of analytical methods (such as LC-MSMS and atomic absorption), that

72

Rinse water

traces will always be detectable (11) and that equipment should be dedicated. The same option is mentioned in ICH Q7A (12): The use of dedicated production areas should also be considered when material of an infectious nature or high pharmacological activity or toxicity is involved (e.g., certain steroids or cytotoxic anticancer agents) unless validated inactivation and/or cleaning procedures are established and maintained. In case of new agents with unknown daily doses (a first clinical batch, for example), the first step is to consider NOEL (the no observable effect level) and calculate the first dose to be administered in humans. Based on this expectation, a 0.1% rule can be applied in a first attempt (13, 14). With formulations used both for adults and children, the worst case (children) is often taken into account to calculate limits. Analytical Methods: As stated in Annex 15 of the European GMP rules Validated analytical methods having sensitivity to detect residues or contaminants should be used. The detection limit for each analytical method should be sufficiently sensitive to detect the established acceptable level of the residue or contaminant (6). Validation of analytical methods might follow ICH Q2 (15). Before analytical development, certain information should be known. First,

the moieties that should be analyzed are chosen, and second, their limits are calculated. Various types of detection methods are used, from the simplest to the most sophisticated (Table 2).
Total organic carbon (TOC) is considered a good general method for analyzing rinse water and swab extracts, and it is described in the USP compendium (16). On the FDAs page, some comments have arisen: We think TOC or TC can be an acceptable method for monitoring residues routinely and for cleaning validation. But in order for TOC to be functionally suitable, it should first be established that a substantial amount of the contaminating material(s) is organic and contains carbon that can be oxidized under TOC test conditions. This is not a trivial exercise because we know that some organic compounds cannot be reliably detected using TOC.

TWO EXTREME CASES

Equation 1: Formula for calculating total surface area carryover quantity


J UNE 2007

TOC use may be justified for direct surface sample testing as well as indirect (rinse water) sample testing. In either case, because TOC does not identify or distinguish among different compounds containing oxidizable carbon, any detected carbon is to be attributed to the target compound(s) for comparing with the established limit. Thus, a firm should limit background carbon (i.e., carbon from sources other than the contaminant being

BioProcess International

Table 2: Summarizing some analytical methods used in cleaning validation Method Visual inspection Resistivity pH analysis Acid/base dose Loss in drying Spectrophotometry TLC HPLC GC Biochemical and/or biological analysis TOC OPLC Sensitivity + ++ ++ ++ ++ +++ ++ +++ +++ ++ +++ +++ Specificity + 0 0 0 0 ++ +++ +++ +++ +++ 0 +++ Identification N N N Y Y Y Y Y Y Y Y Y Rapid +++ +++ +++ ++ ++ ++ ++ ++ ++ ++ ++ ++ Simple +++ +++ +++ ++ ++ +++ ++ + + ++ ++ ++ Cost 0 + + ++ + ++ ++ +++ +++ ++ +++ ++

TLC = thin-layer chromatography GC = gas chromatography OPLC = overpressure layer chromatography

HPLC = high-performance liquid chromatography TOC = total organic carbon analysis

removed) as much as possible. The established limit, or the amount of residue detected for comparison to the specification, should correct for the target materials composition of carbon. As for any cleaning method, recovery studies are necessary (211.160(b)). If TOC samples are being held for long periods of time before analysis, a firm should verify the impact of sample holding time on accuracy and limit of quantitation. (9)

As with all the methods, TOC should be validated and its limits calculated.

used semiquantitative method that may replace TOC or other analytical methods with faster and better results (17). As an example, it can measure traces following estradiol production with a lower detection limit of 0.03 0.05 g/sample and good specificity because it differentiates between close compounds such as allylestrenol, estradiol, ethynodiol, levonorgestrel, and norethisterone. Users can run 3096 samples (1020 min) at a time. This test method can be acceptable if validated.
74 BioProcess International J UNE 2007

Overpressured layer chromatography (OPLC) is a rarely

Cleaning phases should be implemented through a validation master plan. ICH Q7A (12) states that The companys overall policy, intentions, and approach to validation, including the validation of production processes, cleaning procedures, analytical methods, in-process control test procedures, computerized systems, and persons responsible for design, review, approval, and documentation of each validation phase, should be documented. Normal qualification techniques precede validation. Qualification applies to apparatuses, probes, and sampling instruments. The objective of cleaning validation is to verify effectiveness of a cleaning procedure for removal of all moieties including product residues, degradation products, preservatives, excipients, and cleaning agents. The goal is to reduce the amount of routine analytical monitoring (5, 18). As presented in EU GMP annex 15 (6), validation of cleaning processes should be based on a worst-case scenario including challenging the cleaning process (using reduced cleaning parameters such as overloading contaminants, overdrying equipment surfaces,

VALIDATION OF CLEANING PROCESSES

minimal concentration of cleaning agents, and minimum contact time of detergents) to show that the challenge soil can be recovered in sufficient quantity to ensure that the cleaning process is removing soil to the required level. As for media fill testing, at least three consecutive applications of a cleaning procedure should be performed and shown to be successful to prove that the method is validated (6). Especially in cases of manual cleaning, personnel might also be considered as validated: All persons involved in cleaning must be trained and certified. To ensure reproducibility of cleaning methods, personnel training is key to successful validation and control, especially in cases of manual and semiautomated processes. If automated procedures are used, then qualification and validation of the cleaning equipment should be performed, with consideration given to monitoring critical control points and parameters with appropriate sensors and alarm points to ensure that the process is highly controlled and the computer system validated. For first trials, specific nontoxic compounds can be used, such as colorants or riboflavin, which can be visually detected at low traces. With highly potent drugs (substances either toxic or hazardous), products that simulate the physicochemical properties of a substance to be removed should be considered for validation purposes in a first step instead of the substances themselves (6). Raw materials sourced from different suppliers may have different physical properties and impurity profiles. When applicable, such differences should be considered when designing and validating cleaning procedures because such materials from different suppliers may behave differently. Cleaning and cleaning validation are continuous processes from the beginning of development through full production and include periodic revalidations. EU GMP annex 15 stresses that Facilities, systems and processes, including cleaning, should be periodically evaluated to confirm

the interval between manufacturing and cleaning as well as that between cleaning and beginning production on a new batch.

A cleaning

PROBLEM can

WITHOUT EFFECTIVE CLEANING . . .

have health, economic, environmental, and regulatory consequences.

that they remain valid. Where no significant changes have been made to the validated status, a review with evidence that facilities, systems, equipment and processes meet the prescribed requirements fulfils the need for revalidation (6). Revalidations are scheduled at a predefined time. In practice, revalidation is mandatory when equipment is replaced or significantly altered, in the cases of certain types of process changes, or when a products proportion or dose is modified. Cleaning procedures must be written procedures that assign responsibility for sanitation and describing the cleaning schedules, methods, equipment, and materials to be used in cleaning buildings and facilities (12). Phrases like test until clean should not be used. As for all other processes, a record is to be made of each operation. ICH Q7A requires that precise records of major equipment use, cleaning, sanitation, and/or sterilization and maintenance should show the date, time (if appropriate), product, and batch number of each batch processed in the equipment and the person who performed the cleaning and maintenance (12). Training and personnel experience are also factors to be recorded. Many other factors are checked during validation, such as adequate lighting in all areas to facilitate cleaning, maintenance schedules, and proper operations and critical points to check if sampling is justified. An important element to be validated is
76 BioProcess International J UNE 2007

Cleaning problems can lead to more than having to withdraw or destroy only one batch. A cleaning problem can have various consequences to health, economics, the environment, and regulatory approvals. Absence of good cleaning leads to contaminated drugs, which poses a risk for patients as well as manufacturing personnel. Ecological problems are linked to pollution risks that can negatively affect a companys image. Regulatory risks include possible withdrawal of authorization. Economic problems are not restricted to production delays and stock shortages or losses (increasing internal costs), but also manifest as public relations problems for a company, complicating negotiations with regulatory authorities as well as financial institutions.

REFERENCES

1 McCormick D. FDAs Evans Reviews Causes of Warnings and Recalls. Pharm. Technol. October 2005; www.pharmtech.com/ pharmtech/content/printContentPopup. jsp?id=190230 and ?id=383878. 2 McCormick D. Poor OOS Review Leads Causes of FDA Citations. Pharm. Technol. November 2006; www.pharmtech. com/pharmtech/content/printContentPopup. jsp??id=383878. 3 ISO 14644: Cleanrooms and Associated Controlled Environments. International Organization for Standardization: Geneva, Switzerland, 19992006; www.iso.org. 4 Miled M. Thse Pour le Diplme dtat de Docteur en Pharmacie. Clermont-Ferrand, 2002. 5 Health Canada GMP Committee. Canadian Health Products and Food Branch Inspectorate Guidance Document: Cleaning Validation Guidelines, June 2002; www.hc-sc.gc. ca/dhp-mps/compli-conform/gmp-bpf/ validation/cleaning-nettoyage_e.html. 6 Working Party on Control of Medicines and Inspections. Qualification and Validation. Annex 15 to the EU Guide to Good Manufacturing Practice. European Commission: Brussels, Belgium, July 2001. http://ec.europa.eu/enterprise/pharmaceuticals/ eudralex/vol-4/pdfs-en/v4an15.pdf. 7 Active Pharmaceutical Ingredients Committee. Cleaning Validation in Active Pharmaceutical Ingredient Plants. European Chemical Industry Council: Brussels,

Belgium, December 2000; http://apic.cefic. org/pub/4CleaningVal9909.pdf. 8 Guide to Inspections Validation of Cleaning Processes. US Food and Drug Administration: Rockville, MD, 2006; www.fda.gov/ora/ inspect_ref/igs/valid.html. 9 US Food and Drug Administration. Questions and Answers on Current Good Manufacturing Practices, Good Guidance Practices, Level 2 Guidance; www.fda.gov/cder/ guidance/cGMPs/equipment.htm; www.fda. gov/cder/guidance/cGMPs/production.htm. 10 Recommendations on Validation Master Plan, Installation and Operational Qualification, Non-Sterile Process Validation, Cleaning Validation (PI 006-2), July 2004; Pharmaceutical Inspection Convention, Pharmaceutical Inspection Cooperation Scheme: Geneva, Switzerland, July 2004; www.picscheme.org/publis/recommandations/ PI%20006-2%20Recommendat_121CF0.pdf. 11 Le Blanc D. Documents of Interest for Cleaning Validation. Cleaning Validation Technologies: Kodak, TN, July 2006; www. cleaningvalidation.com/SpecificDocuments. htm. 12 ICH Q7: Good Manufacturing Practice Guidance for Active Pharmaceutical Ingredients. Federal Register 66(186) 25 September 2001: 4902849029; www.ich.org/ LOB/media/MEDIA433.pdf. 13 David C, Naumann B, Hecher L. Setting Health-Based Residue Limits for Contaminants in Pharmaceuticals and Medical Devices. Quality Assurance: Good Practice, Regulation and Law 1(3) 1992: 171180. 14 Fourman G, Mullen M. Determining Cleaning Validation Acceptance Limits for Pharmaceutical Manufacturing Operations. Pharm. Technol. April 1993: 5460. 15 ICH Q2(R1): Validation of Analytical Procedures: Text and Methodology. Federal Register 62(96) 19 May 1997: 2746327467; www.ich.org/LOB/media/MEDIA417.pdf. 16 USP <643> Total Organic Carbon; USP <1112> Microbiological Attributes of NonSterile Pharmaceutical Products: Application of Water Activity Determination. US Pharmacopeia: Rockville, MD; www.usp.org. 17 Katona Z, et al. Cleaning Validation Procedure Eased By Using Overpressured Layer Chromatography. J. Pharmaceut. Biomed. Anal. 22, 2000: 349353. 18 ICH Q9: Quality Risk Management. Federal Register 71 (106) 2 June 2006: 32105 32106; www.ich.org/LOB/media/ MEDIA1957.pdf.

Corresponding author Jean-Michel Cardot and E. Beyssac are professors at the University Clermont 1, UFR Pharmacie, ERT-CIDAM, Biopharmaceutical Department, 28 Place H. Dunant, BP 38, Clermont-Ferrand, F-63001 France; j-michel.cardot@u-clermont1.fr.

You might also like